NDQAT Seminar Report complete

Non-Destructive Techniques For Quality Evaluation Of Intact Fruits & Vegetables March 4, 2011

1

UNIVERSITY OF AGRICULTURAL SCIENCES, BANGALORE

DEPARTMENT OF HORTICULTURE

PHT 651 Seminar (0+1)

IV Semister 2010-11

SEMINAR REPORT ON

Non-Destructive Techniques for Quality Evaluation of intact

Fruits and Vegetables

Submitted To: Submitted By:

The Seminar Teacher, Vijay Rakesh Reddy, S.

Division of Horticulture, Sr. M.Sc (Hort) in PHT,

UAS, GKVK, Bangalore. PHK 935.


2

CONTENTS

Sl. No. Topic Page no.

1. Introduction

2. Quality components

3. Nondestructive Quality Evaluation Technology

4. Near-infrared spectroscopy

4.1. Principle of NIR Spectroscopy

4.2. Application of NIR Spectroscopy

4.3. Reflectance Measuring Method

4.4. Transmittance Measuring Method

4.5. Peach grading system using NIR Reflectance

4.6. Citrus grading system using NIR Full-Transmittance

5. X-ray and computerized tomography (CT)

6. Ultrasonics and Acoustic Technology

7. Nuclear magnetic resonance (NMR) techniques

8. E-nose Technology

9. Hyperspectral Imaging Technology

10. New Terminology

11. Case Studies

11.1. Non-destructive prediction of translucent flesh disorder in intact

mangosteen by short wavelength near infrared spectroscopy.

11.2. Rapid and non-destructive analysis of apricot fruit quality using FT-near-

infrared spectroscopy.

11.3. Non-destructive ultrasonic monitoring of tomato quality during shelf-life

storage.

11.4. Rapid and non-destructive analytical techniques for measurement of

apricot quality.

11.5. Hayward kiwifruits and plant growth regulators: Detection and effects in

post-harvest studied by Magnetic Resonance Imaging (MRI) and Scanning

Electron Microscopy (SEM).

11.6. Hyperspectral imaging for non-destructive determination of some quality

attributes for strawberry.

12. Conclusion

13. References


3

UNIVERSITY OF AGRICULTURAL SCIENCES, BANGALORE

DEPARTMENT OF HORTICULTURE

PHT 651 Seminar (0+1)

Name : Vijay Rakesh Reddy, S. Date : 04.03.2011

ID.No. : PHK 935 Time : 01:00 pm

Class : Sr.M.Sc.(Hort.) PHT Venue : Seminar Hall

Non-Destructive Techniques for Quality Evaluation of intact Fruits and Vegetables

Synopsis

Fruits and vegetables are increasing in popularity in the daily diets of people of both developed

and developing countries. This growing demand for high quality produce gives an immediate call for

more advanced post harvest technologies. Consumers are becoming more quality conscious and are

preferring more safer foods. Thus the produce quality and quality evaluation methods are becoming

extremely important. This increased the need for rapid and nondestructive quality evaluation of fresh

fruits and vegetables. Most of the existing methods are destructive and the recent developments in

science and information technology has led to the development of non-destructive techniques viz.,

visible or near infrared spectroscopy, ultrasonics, acoustics, electronic nose technology, hyperspectral

imaging to analyse the quality, chemical and physical parameters of fresh fruits and vegetables (Jha and

Matsuoka, 2000).

Sontisuk et al. (2007) used short wavelength near infrared (SW-NIR) transmittance

spectroscopy to predict an internal translucent flesh disorder in intact mangosteen fruits. They obtained

NIR absorption spectra for 193 mangosteen samples in the wavelength range of 640-980 nm on four

sides of each sample with a classification accuracy of 92.0 per cent.

Mizrach (2007) applied a non-destructive ultrasonic method to monitor the firmness and sugar

content of greenhouse tomatoes during their shelf life. He transferred the fruits from greenhouse to

controlled-temperature room and subjected them to non-destructive ultrasonic and destructive

penetration tests for firmness. A linear relationship between the attenuation and firmness till the end of

softening process was observed.

Hyperspectral imaging in the visible and near infrared (400 – 1000 nm) regions was tested for

non destructive determination of moisture content, TSS and acidity in strawberry by Elmasry et al.

(2007). Multiple linear regression (MLR) models for the moisture content (r = 0.91), TSS (r = 0.80) and

pH (r = 0.94) demonstrated good predication performance.

Bureau et al. (2009) performed diffuse reflectance measurements (800-2500 nm), physical,

physiological and bio-chemical measurements for 877 apricot fruits of eight cultivars harvested at

different ripening stages. They observed good prediction performance for soluble solids and titrable

acidity (TA) with correlation coefficients of 0.92 and 0.89 respectively and root mean square error of

prediction (RMSEP) of 0.98% Brix and 3.62 meq 100 g-1

fresh weight respectively.

Cristina et al. (2010) explored the feasibility of the two novel techniques (acoustic impulse

response technique for firmness and reflectance spectroscopy for fruit color) for rapid quality


4

measurement of apricot and determination of its optimal harvest date. They suggested that acoustic

firmness index and color parameters CIELAB, a* and ho could be used to distinguish between different

maturity stages of apricot fruit.

Taglienti et al. (2011) employed Magnetic Resonance Imaging (MRI) Spectroscopy to reveal

the role of two plant growth regulators on Hayward kiwi fruits, at any stage of cultivation and post-

harvest ripening and storage. They determined the internal morphology to depict structural features

related to hormone treatments and found that T2-weighted MRI images showed different internal tissue

organizations based on the use and type of PGR. They found that the auxins shortened the shelf life

while normal refrigeration exerted a stronger effect. In their investigation with Scanning Electron

Microscopy (SEM) they found significant differences at cellular level among different treatments in

terms of pore size and number of vacuoles.

Thus the non-destructive techniques of quality evaluation is gaining a good momentum in field

as well as processing industries. But rapid and accurate determination of internal quality posses

technical challenges because of complex structural, physical and chemical properties of fruits and

vegetables. As these methods are instantaneous and can simultaneously determine many parameters,

could be used for bulk handling of produce. Most of the developed countries are using various on farm

non-destructive methods. But in India due to technological hitches and lack of development of

equipments and their availability, has restricted their wider usage in quality evaluation of fruits and

vegetables. Therefore the development of an accurate, reliable and non destructive method for the

quality evaluation before harvest and at packaging site is critical to provide better quality produce.

REFERENCES:

BUREAU, S., RUIZ, D., REICH, M., GOUBLE, B., BERTRAND, D., AUDERGON, J. AND RENARD, C. M.

G. C., 2009, Rapid and non-destructive analysis of apricot fruit quality using FT-near-infrared

spectroscopy. Food Chemistry. 113: 1323-1328.

CRISTINA, P., GABRIEL, L. R., VIORICA, B. AND GHEORGHE, C., 2010, Rapid and non-destructive

analytical techniques for measurement of apricot quality. Romanian Biotechnological Letters.

15(2): 5213-5216.

ELMASRY, G., WANG, N., ELSAYED, A. AND NGADI, M., 2007, Hyperspectral imaging for non-

destructive determination of some quality attributes for strawberry. J. Food Eng., 81: 98-107.

JHA, S. N. AND MATSUOKA, T., 2000, Non-destructive techniques for quality evaluation of intact fruits

and vegetables. Food Sci. Technol. Res., 6(4): 248-251.

MIZRACH, A., 2007, Non-destructive ultrasonic monitoring of tomato quality during shelf-life storage.

Postharvest Biol. Technol., 46: 271-274.

SONTISUK, T., KWON, Y. K., ANUPUN, T., WARUNEE, T. AND YUTAKA, N., 2007, Non-destructive

prediction of translucent flesh disorder in intact mangosteen by short wavelength near infrared

spectroscopy. Postharvest Biol. Technol., 43: 202-206.

TAGLIENTI, A., SEQUI, P., CAFIERO, C., COZZOLINO, S., RITOTA, M., CEREDI, G. AND VALENTINI,

M., 2011, Hayward kiwifruits and plant growth regulators: Detection and effects in post-harvest

studied by Magnetic Resonance Imaging and Scanning Electron Microscopy. Food Chemistry.

126: 731-736.


5

Non-Destructive Techniques for Quality Evaluation of intact

Fruits and Vegetables

1. Introduction

Consumers are now more conscious about quality and source of their foods. Attempts made to

determine the quality of food materials are numerous, but most of them are destructive in nature. In

recent years, nondestructive methods of quality evaluation have gained momentum and considerable

attempts have been made to develop them. Fruits and vegetables are increasing in popularity in the

daily diets of people of both developed and developing countries. Product quality and quality

evaluation methods are naturally extremely important. The decisions concerning harvesting, maturity,

ripeness and quality are based mostly on subjective and visual inspection of the fruit‟s external

appearance. Several nondestructive techniques for quality evaluation have been developed based on the

detection of various physical properties that correlate well with certain factors of a product. The quality

of fruits and vegetables is mostly based on size, shape, color, gloss, flavor, firmness, texture, taste and

freedom from external as well as internal defects. Numerous techniques for evaluating the above

external quality factors are now available commercially. Internal quality factors such as maturity, sugar

content, acidity, oil content, and internal defects, however, are difficult to evaluate. Methods are needed

to better predict the internal quality of fruits and vegetables without destroying them. Recently, there

has been an increasing interest in non-destructive methods of quality evaluation, and a considerable

amount of effort has been made in that direction. But the real problem is how these methods are to be

exploited practically and what the difficulties are in implementing them. The objective of the present

paper is thus to review the application of the most recent non-destructive methods such as nuclear

magnetic resonance, x-ray computed tomography, near-infrared spectroscopy and some other important

methods and to evaluate their pros and cons for suitability in commercial application.

2. Quality Components

Consumers purchase fruits and vegetables on the basis of quality. Quality of horticultural

products is an important subject to those engaged in horticultural industries. Quality is defined as the

degree of excellence or superiority. Quality of fruits and vegetables is a combination of characteristics,

attributes, and properties that have significance and make for acceptability. Acceptability is depended

on highly subjective factors including sight, touch, smell, taste, and even hearing. The various

components of quality are used to evaluate fruits and vegetables. Quality is classified into external and

internal component as shown in Table below. Appearance, flavor, texture, nutritive value, and defect

factors are generally recognized as the five quality factors of fruits and vegetables.


6

3. Nondestructive Quality Evaluation Technology:

Quality evaluation methods can be the destructive and non-destructive. They include both

objective methods based on instrument reading and subjective methods on human judgement. Various

analytical methods are available for determination of quality related components using optics, X-ray,

mechanics, and electromagnetic (Table below).

Various techniques have been developed for evaluating internal quality of fruits and vegetables.

Specific examples of practical use are as follows: near infrared (NIR) reflectance/transmittance

technique to measure soluble solid content and acidity of peaches, pears, apples, citrus, melon, and

watermelon, acoustic response technique to evaluate the maturity and internal defects of melon and

watermelon, electrical capacitance technique to estimate soluble solid content and internal cavity of

watermelon.

On the other hand, magnetic resonance (MR) has been used to study the moisture and oil

content of agricultural materials. The magnetic resonance imaging (MRI) could prove to be an

extremely valuable technique to evaluate ripening, core breakdown, bruise, worm damage, chilling and

freezing. Currently, MR and MRI are not practical for routine quality testing. These equipments are

expensive and difficult to operate.

However, like all technologies, it is becoming cheaper, faster and more feasible for research and

specialized applications. MR and MRI techniques have great potential for evaluating the internal

quality of fruits and vegetables. Above all the techniques, NIR spectroscopy technique is very close to

practical use. It has been contributed to development and wide use of sorting and grading technology


7

during the last 10 years. This paper describes the theory, application, examples of practical use, and

some issues of near infrared spectroscopy.

4. Near-infrared spectroscopy

The use of near-infrared spectroscopy as a rapid and often nondestructive technique for

measuring the composition of biological materials has been demonstrated for many commodities. This

method is no longer new; as it started in early 1970 in Japan (Kawano, 1998), just after some reports

from America. Even an official method to determine the protein content of wheat is available (AACC,

1983). The National Food Research Institute (NFRI), Tsukuba has since become a leading institute in

NIR research in Japan and has played a pivotal role in expanding near-infrared spectroscopy

technology all over the country (Iwamoto et al., 1995). In Japan, NIR as a nondestructive method for

quality evaluation was started for the determination of sugar content in intact peaches, Satsuma orange

and similar other soluble solids (Kawano, 1994). To determine the solid content of cantaloupe Dull et

al. (1989) used NIR light at 884 nm and 913 nm. Initially the correlation of their findings was poor

mainly due to light losses. Later, Dull and Birth (1989) modified the earlier method and applied it to

honeydew melons; the improved method showed better correlation.

Similarly, a nondestructive optical method for determining the internal quality of intact peaches

and nectarines was investigated (Slaughter, 1995). Based upon visible and near-infrared

spectrophotometer techniques, the method was capable of simultaneously predicting the soluble solid

content, sucrose content, sorbitol content, etc. of intact peaches and nectarines, and required no sample

preparation. Now various NIR spectrometers are available and are being used commercially. Some

modifications in these available spectrometers, especially for holding the intact samples, are reported.

In the same sample holder a test tube for holding liquid food such as milk was also used to determine

fat content. Recently a low cost NIR spectrometer has been used to estimate the soluble solids and dry

matter content of kiwifruit. Errors are within the permissible limit and the time required for obtaining

data has been reduced. The influence of sample temperature on the NIR calibration equation was also

evaluated and a compensation curve for the sample temperature was developed to rectify the result.

4.1. Principle of NIR Spectroscopy

The electromagnetic spectrum covers a large range of photon energies which are commonly and

conveniently divided into the following bands: from longest to shortest wavelengths, radio wave,

microwave, infrared, near infrared, visible, ultraviolet, X-ray and gamma-ray (Fig. 1).

Fig. 1 Electromagnetic spectrum radiation

Optical properties indicate the response of matter to visible light wavelengths (400- 700nm) and


8

near infrared (700-2500nm). Optical properties are based on reflectance, transmittance, absorbance, or

scatter of light by the product.

When a light beam falls on an object, part of the incident beam is reflected by the surface. This

is called regular reflectance (or specular reflectance). The remaining radiation is transmitted through

the surface into the cellular structure of the object where it is scattered by the small interfaces within

the tissue or absorbed by cellular constituents. The absorbed portion of radiation can be transformed to

other forms of energy such as heat, chemical changes or luminescence. Part of the transmitted energy is

absorbed (absorption), some part is reflected back to the surface (body reflectance or diffuse

reflectance), and remain part is transmitted through the object (transmittance) as shown in Fig. 2.

Fig. 2 Distribution of incident light on an object

4.2. Application of NIR Spectroscopy

Near infrared (NIR) spectroscopy deals with irradiating the products with NIR light (i.e.

wavelength in the 700 nm to 2500 nm range) and with collecting and analyzing the absorbance

spectrum.

The advantages of NIR are numerous; this radiation is highly penetrative and thus can be

applied to the sample without any preparation, the spectra are not very discriminative but they

can give quantitative information on the major organic components of food products, the NIR

optical components are affordable and robust. For all these reasons, NIR spectroscopy has been

widely used in grains, beverages and sugar industry, milk and dairy products, fruits and

vegetables since Karl Norris first applied to the compositional analysis of cereal in 1963. It

reaches now on-line settings in the industry.

NIR spectrometers consist of light source, wavelength selector, and detector. Those are

classified to reflectance or transmittance measuring method according to an arrange of

components and detection methods.

4.3. Reflectance Measuring Method

In reflectance method, lamp and detector are located in the same direction as shown in Fig. 3.

When object is illuminated by lamps, the scattered reflected radiation is measured by the detector.

Diode array sensor used to measure the intensity of the radiation at each wavelength for laboratory and

industry applications. The concentrations of a component are calculated from the radiation intensity by

using a previously developed calibration equation. The sweetness grading system based on NIR

reflectance method can be used for the fruits having a thin peel such as peaches, apples, and pears.

However, reflectance method is limited that body reflectance means the diffused light from features at

depth to about 5 mm. The grading system is not available for the fruits with thick peel such as citrus,


9

melon, and watermelon.

Fig. 3 Reflectance measuring system

4.4. Transmittance Measuring Method

Transmittance method measures the transmitted light which passes through the tissues.

Transmitted light reflects the internal attributes of the object. Therefore, it gives us an overall

information of quality, and is able to make up for the weak points of the reflectance method.

Transmittance method can be divided into two; full transmittance and half transmittance

measuring method, according to the arrangement of light and detector. Full transmittance measuring

system makes light, object, and detector lined up and in order as shown in Fig. 4. This method applied

to the grain, milk, fruits of thin peel and small-sized fruits such as citrus.

In half transmittance method, many lamps arranged around the equator of the object to get the

high intensity of radiation. Lamp and detector are at right angles to each other as shown in Fig. 5. This

method is widely used for measuring sugar content, acidity, and internal defects of fruits and

vegetables; apple, pear, melon, and watermelon.

4.5. PEACH GRADING SYSTEM USING NIR REFLECTANCE

Most fruit sorters have used the mechanism which discharges fruits from trays tilted using

solenoid and spring. Since this mechanism causes damage to peaches, farmers evaded to use

conventional fruit graders. Grading operation of peaches depends on a manual way up to recently. In

order to grade peaches without any injuries and to sustain the quality, non-contact and real-time based

grader should be developed. To minimize damage caused on fruit during grading and handling peaches,


10

the prototype was devised to have a way of extracting each tray containing fruits from the moving path

of the grader (Fig. 6).

Fig. 6 Peach sugar content grading system

Sugar content within in a fruit varies according to a position; upper or lower, right or left, and

depth. To decrease the measurement error by position effects in detecting representative values of sugar

contents in peaches, two directions diverged fiber optics was developed to measure reflectance of both

sides at one time as shown in Fig. 7.

Fig. 7 Two directions diverged fiber optics

The operating program of the grader and the measuring algorithm of the quality index such as

size and sugar contents were developed using Visual Basic 6.0 language. All grading results were saved

in database and provided to farmers for helping them to produce the best quality peaches.

The multiple linear regression (MLR) model using 16 wavelengths for quality index was

developed to estimate sugar content of peaches with the determination coefficient of 0.71 and Standard

Error of Calibration (SEC) of 0.42Brix. The developed MLR model could predict also the sugar content

with the determination coefficient of 0.69 and Standard Error of Prediction (SEP) of 0.49Brix (Fig. 8).

This MLR model has better performance than the prediction model developed by using the detection of

single position in 1998.


11

Based on performance of the prototype, the non-destructive automatic peach sweetness grading

system was fabricated and evaluated. The peach sweetness grading system consists of photodiode-array

sensor, quartz-halogen lamp and one fiber optic diverged by two bundles for transmitting the lights to

two positions on fruit and re-transmitting the reflected lights from two position. The system could

classify peaches into 2 classes by soluble solid contents of peaches with the accuracy of 91 % and the

capacity of 7,200 peaches per hour

Fig. 8 Calibration and prediction results using MLR model

4.6. Citrus grading system using NIR Full-Transmittance

Conventional citrus grader grades citrus fruits based on size differences of the fruits. The fruits

are graded by transferring the fruits inside the series of rotating drums, which have different size of

holes. Fruits smaller than the hole-size are graded at the drum, remaining bigger size fruits are graded at

next drums. The capacity of the grader can be easily changed by adjusting width or rotating speed of

drums. The drum type citrus grader has good performance for grading citrus size, but it cannot be used

to grade the fruit based on its internal quality, such as maturity and internal defects. Immature or

obvious defected fruits are sorted manually before putting in the drum grader. Because of tremendous

labor requirement for this sorting operation, the automation of the operation is much demanded.

The on-line citrus grading system developed by NIAE (National Institute of Agricultural

Engineering) comprised an NIR transmittance measuring part and a grading part. The grading system

was designed so that citruses were fed one by one into NIR transmittance system. To reduce variability

of the spectra, fruits were always placed flat on conveying belt between the lighting unit and the NIR

detector as shown in Fig. 9. Design of feeding and conveying line is very important because the feeding

posture of citrus affects the accuracy of the internal quality evaluation using optical properties.

NIR full transmittance measuring part consisting of a specially designed light source, NIR

spectrophotometer having CCD array (Ocean optics, USA.), a controller, and photo sensors was

constructed. NIR transmittance spectra were acquired with the 2048 pixel CCD-array NIR detector. For

the strong illumination, a 300 W halogen lamp was used. Optical fibers and the lens assembly were

used to collect light and provide strong luminance enough to pass through the flesh of the citrus fruit as

shown in Fig. 9. NIR spectrum was acquired for 30 ms, and the wavelength of the spectrum was ranged

from 650 nm to 955 nm.


12

Fig. 9 Arrangement of light and detector

The Partial Least Square Regression (PLSR) model was developed to estimate sugar content

and acidity of citrus. The calibration model for sugar contents was developed with 463 spectrum

samples and its correlation coefficient and standard error of calibration (SEC) were 0.936 and 0.33Brix,

respectively (Fig. 11(a)). The validation result of model with 462 unknown samples is illustrated in Fig.

11(b). As shown in Fig. 11(b), the correlation coefficient and standard error of prediction (SEP) were

0.897 and 0.42Brix, respectively. Fig. 10 shows the prototype citrus grading system, and the capacity is

the 14,400 citrus per hour.

Fig. 10 Sugar content, acidity, and size grading system of citrus

Fig. 11 Calibration and prediction results using PLSR model


13

5. X-ray and computerized tomography (CT)

X-ray imaging is an established technique to detect strongly attenuating materials and has been

applied to a number of inspection applications within the agricultural and food industries. In particular,

there are many applications within the biological sciences where we wish to detect weakly attenuating

materials against similar background material. X-ray computed tomography (CT) has been used to

image interior regions of apples with varying moisture and, to a limited extent, density states (Tollner et

al., 1992). The images were actually maps of x-ray absorption of fruit cross sections. X-ray absorption

properties were evaluated using normal apples alternatively canned and sequentially freeze-dried, fruit

affected by water core disorder, and normal apples freeze-dried to varying levels. The results suggested

that internal differences in x-ray absorption within scans of fruit cross-sections are largely associated

with differences in volumetric water content. Similarly, the physiological constituents have been

monitored in peaches by CT methods in which x-ray absorbed by the peaches is expressed in CT

number and used as an index for measuring the changes in internal quality of the fruit (Barcelon et al.,

1999). Relationships between the CT number and the physiological contents were determined and it

was concluded that x-ray CT imaging could be an effective tool in the evaluation of peach internal

quality.

In another study, the potential for Compton scattered x-rays in food inspection was evaluated

by imaging the density variation across a food material by measuring the Compton scatter profile

across polystyrene spheres with internal voids (MacFarlane et al., 2000). In this study particular

attention was paid to simulate the obscuring influence of multiple scatter. The simulated result was

found to be in close agreement with the experimental observation. Some experimental test sample of a

Perspex block with various embedded soft materials showed that care should be taken to ensure that the

transmission image is taken with x-rays within an appropriate energy range (Zwiggelaar et al., 1997).

For low Z materials the contrasts between the materials became more pronounced at lower x-ray

energies. If more than one soft material has to be distinguished from the surrounding area it may be

advantageous to image over a range of x-ray energies.

6. Ultrasonics and Acoustic Technology

Mizrach et al., (1989) evaluated the use of high-power, low-frequency ultrasonic excitation for

determination of fruit tissue properties; they designed an experimental system for determination of

basic acoustic properties of some fruits and

vegetables, namely, wave propagation velocity and

attenuation. Further studies by Mizrach et al.,(1991),

Mizrach et al.,(1992) correlated ultrasonic properties with

some ripening parameters of the fruit tissue. The

strong interdependence of these properties and

parameters indicated the potential for the use of this

acoustic method for the evaluation of Firmness

properties in fruit tissue. Mizrach et al., (1996a)

measured local ultrasonic surface waves excited on the

peel of a whole fruit, and examined their connection to

the length of storage time. They indicated that

scattering of the experimental data of both destructive and nondestructive measurements formed a

problem in the interpretation of these measurements. Mizrach et al., (1996) applied ultrasonic testing to

the nondestructive quality evaluation of avocado fruits (cv. `Ettinger') and suggested a technique to

reduce the scattering of the results; they hypothesized that the rate of change in fruit Firmness and in

the associated acoustic measurement results formed a preferred predictor of the fruit's biological age at


14

full ripening. Therefore, they suggested a method to shift the data set along the time scale until all soft

fruits appeared in the same Firmness range. The authors used a nonlinear regression procedure for

determining models for relating variation in the ultrasound parameters and Firmness to storage time,

and concluded that models of this type may be used to evaluate ripening and shelf life of the avocado

fruit.

Acoustics technology

In the early 1970s, the potential of acoustics technology had been recognized in the food

industry. During the last 30 years, acoustics technology has developed quickly and has become a

primary method for fruit sorting and grading. When an acoustic wave reaches the agricultural products,

the reflected or transmitted acoustic wave depends on the acoustic characteristics of the agricultural

products. The reflected or transmitted acoustic wave can provide information on the interaction

between acoustic wave and agricultural products, and acoustic characteristics such as attenuation

coefficient, transmitting velocity, acoustic impedance, and natural frequency can be obtained from the

reflected or transmitted acoustic wave. Different agricultural products have various acoustic

characteristics because of their different internal tissue structures. Fig. 1 represents the visualization of

watermelon surface vibration caused by a pendulum at an angle of 45o. This figure indicates that

impulse response transmits along the surface of the sample with a uniform velocity in all directions

(Sugiyama et al., 1994). Mizrach et al. (1994) evaluated the internal physiochemical parameters of

melons such as firmness, dry weight, and total soluble solids content (TSS) by acoustic properties.

Acoustic impulse impedance technique was also used to detect hollow heart, soluble solids content

(SSC), flesh color, effe-gi firmness, and mass of different fruits. Schotte et al. (1999) stated that

acoustic properties could be used to evaluate internal qualities of all fruits and vegetables.

Watermelons/melons should be harvested when completely mature in order to achieve the best

taste and texture. The maturity of watermelons is always judged by human beings according to some

typical characteristics, such as withering of tendril, changes in belly‟s color, and a thumping test. When

thumped by fingers, the mature watermelons will give a dull sound as against a metallic sound of

unripe watermelons. Clark (1975) developed an accurate instrument to measure the maturity of

watermelons quantitatively and nondestructively, and this technique was based on the relationship

between the ripeness and the times that the sound reflected from and transmitted through a watermelon.

Yamamoto et al. (1980) used an acoustic impulse response method to measure the internal qualities of

watermelons. The natural frequency of watermelons was induced by impacting them with a wooden

ball pendulum, and the instrument used in this method was simpler than that of acoustic vibration

resonance method.


15

Hayashi et al. (1992) found that the impulse waveform of acoustic signals could be used to

assess the maturity of muskmelons, and the correlation coefficient (r) between the transmission velocity

and fruit hardness was 0.83. In order to judge the maturity and other internal qualities of watermelons

without damage, He et al. (1994) developed a simple pendulum hitting device to study the feature

curves of the sound waveform of watermelons. The power spectral density of the sound waveform of

watermelons was analyzed by a fast Fourier transform (FFT) method.

Sugiyama et al. (1994) studied the relationship between the transmission velocity and firmness

of muskmelons. They found that the transmission velocity became lower when muskmelons ripened.

According to this, they developed an instrument to measure the transmission velocity of sound wave in

muskmelons. The results showed that the transmission velocity of sound wave in edible muskmelons

ranged from 37 m/s to 50 m/ s. Stone et al. (1996) developed a portable device by acoustic impulse

impedance techniques to determine the maturity of watermelons in the fields. The sensor and recording

system (shown in Fig. 2) consisted of an acoustic impulse cylinder probe, an amplifier, a filter, a data

acquisition unit, a PC, and a hand control. In this system, a ceramic piezoelectric element that bonded

to a brass disk was mounted at the end of the cylinder and held in contact with a fruit and a solenoid

was used to deliver a mechanical impulse to the flat face of the piezo ceramic. The impulse was

transferred through the ceramic to the fruit. The vibration of the fruit caused by the impulse was sensed

by the piezo element. Tissue color, effe-gi firmness, mass, and SSC were used as indicators of maturity.

The best correlation coefficients between tissue color, effe-gi firmness, mass, SSC, and acoustic

parameters were 0.66, 0.53, 0.62, and 0.64, respectively. Sugiyama et al. (1998) also developed a

portable firmness tester for melons by measuring the transmission velocity of an impulse waveform that

was induced by the impact of a plunger. The tester consisted of a light-touch impact mechanism and

two microphone sensors (shown in Fig. 3).


16

The correlation coefficient between the transmission velocity and apparent elasticity was 0.94,

so this tester could be used to evaluate physiological changes in ripening melons.

The acoustic device consisted of a microphone, structural elements and a mechanical impact

generator. They found that band magnitude (BM) parameter had a higher correlation with hollow

volume (r = 0.62–0.67) than resonant frequencies parameter, and was a better predictor for internal

voids. Diezma-Iglesias et al. (2004) also used acoustic impulse response to detect internal defects of

watermelons nondestructively, and they found that band magnitude (BM) parameters showed the best

ability to detect internal disorders. The results of correct classification for whole and hollow

watermelons by acoustic device and human experience were 78.14% and 71.58%, respectively, and the

results of correct classification for sound watermelons and those with internal disorders including

hollows, bruises, and over ripeness by acoustic device and human experience were 89.07% and

82.51%, respectively. AVA Company (Applied Vibro Acoustic, Japan. http://www.ava. co.jp) stated

that everything in the world holds its own special frequency. Fig. 4 shows a frequency response

spectrum of watermelon. According to this theory, they developed an instrument for measuring the

internal qualities (hollow heart and maturity) of watermelons (shown in Fig. 5).

Lü (2003) and Rao et al. (2004) developed a quality inspecting system using acoustic

technology. The sound wave was collected by microphones and transformed into electric signal. Then

http://www.ava/


17

this electric signal was amplified and filtered by a processing circuit, and sampled by a data acquisition

board (shown in Fig. 6). The correlation between the transmission velocity and SSC of watermelons

was set up. The best correlation coefficient for different striking positions and growth status of

watermelons reached 0.81–0.95.

7. Nuclear magnetic resonance (NMR) techniques

The nuclear magnetic resonance technique, often referred as magnetic resonance imaging

(MRI), involves resonant magnetic energy absorption by nuclei placed in an alternating magnetic field.

The amount of energy absorbed by the nuclei is directly proportional to the number of a particular

nucleus in the sample such as the protons in water or oil. The theory of NMR is presented in detail

elsewhere (Farrar & Becker, 1971). The basic concepts, types of pulsed experiments and the type of

information that can be extracted from these experiments are described. Information on

experimentation, assembling hardware, conducting laboratory tests and interpreting the results is also

available from Fukushima and Roeder (1981). These authors also provided detailed theory for better

understanding of what a scientist should seek and what he might expect to find out by using NMR.

There are many applications of NMR in agriculture. The simplest among them is the determination of

moisture and oil content. But the NMR response many times is not clear and poses problems especially

when constituents other than water are present in the material. Besides the established relationship

between the moisture and output of NMR experiments, various other facts helpful in determining the

quality of food materials without destroying them are available in the literature: selections of chocolate

confectionary products can be made non-invasively by three-dimensional magnetic resonance imaging

(Miquel et al., 1998); using a spin echo pulse sequence, 128×64×64 data sets were acquired with either

a 5- or 20-ms echo time, 500-ms repetition time and four signal averages, in total 2-h scan time. Such

images localize and distinguish between the constituents, and visualize both the internal and external

structure of matter.

Most perishable food products are now marketed in packaged form. To increase the

marketability longer shelf life is needed and this is achieved by freezing and secondary processing of

the food. During freezing it is natural that ice will form within the food that may change its

characteristics. Ice formation during food freezing can be examined using the NMRI method as the

formation of ice has been seen to reduce the spatially located NMR signal. The characteristics of a food

can be better controlled as MRI can serve to assess freezing times and the food structure during the

freezing process (Kerr et al., 1998). The secondary processing changes almost all characteristics of a

food, such as physical and aerodynamic, thermal and hygroscopic properties, which in turn, change its


18

key acceptability factors, i.e. sensory texture and taste. The sensory texture of cooked food such as

potatoes has been predicted using the NMRI technique. In addition, NMR image intensity, the ratio of

the oil and water resonance peaks of the one-dimensional NMR spectrum, and both the spin-lattice

relaxation time and spin-spin relaxation time of water in the fruit are correlated with maturity of a fruit

like avocado before harvesting. This important finding has desirable features for high speed sorting

using a surface-coil NMR probe that determine the oil/water resonance peak ratio of the signal from

one region in an intact fruit (Jha and Matsuoka, 2000)

An on-line nuclear magnetic resonance quality evaluation sensor has recently been designed,

constructed and tested. The device consists of a super-conducting magnet with a 20 mm diameter

surface coil and a 150 mm diameter imaging coil coupled to a conveyor system. The NMR spectra of

avocado fruits and one-dimensional magnetic resonance images of fresh cherries were acquired by this

system. These spectra were used to measure the oil/water ratio in avocados and this ratio correlated to

percent dry weight. One dimensional magnetic resonance images of cherries were later used to detect

the presence of pits inside.

8. E-nose Technology:

An Electronic Nose, also called an Artificial Olfactory System, is a machine that is designed to detect

and discriminate among complex odors using sensor arrays. The sensor array consists of broadly tuned sensors

that are treated with a variety of odor-sensitive biological or chemical materials. An odor stimulus generates a

characteristic fingerprint from the sensor-array. Patterns or smell prints from known odors are used to construct a

database and train a pattern recognition system so that unknown odors can subsequently be classified and

identified in a quantifiable and reproducible way. The sensors try to mimic the human olfactory receptors, and

processing of the data is conceptually analogous to the process that happens in the olfactory bulb. The final

classification is executed through soft-computing techniques (via neural networks or multivariate statistical

analysis) that are similar to the identification mechanisms of the human brain.

Shown above is a block diagram of the Electronic Nose. From the block diagram, it can be seen

that an Electronic Nose system primarily consists of four functional blocks, viz., Odor Handling and

Delivery System, Sensors and Interface Electronics, Signal Processing and Intelligent Pattern Analysis.

The most complicated component of the electronic olfaction process is the odor capture

mechanism and the sensor technology to be deployed for such capturing. Any sensor that responds

reversibly to a chemical in gas or vapor phase has the potential to be developed in an Electronic Nose

format. Some of the essential and desirable properties of the chemical micro-sensors to be used in the

Electronic Nose are Sensitivity, Speed of response, Reproducibility, Reversibility, Portability and so

on. Three technologies, viz., Metal Oxide Semiconductors (MOS), Conducting Polymers (CP) and

Bulk Acoustic Wave Devices (BAW), are generally used for making these chemo sensors for use in

Artificial Olfactory Systems.


19

The responses generated by an array of odor sensors may be processed using a variety of

techniques. The pattern recognition engine can be developed using both parametric and non-parametric

methods. Parametric methods are commonly referred to as statistical approaches and are based on the

assumption that the sensor data set may be represented as a probability distribution function. The non-

parametric methods do not assume any probability distribution function and deal with biologically

inspired techniques, viz., Artificial Neural Networks (ANN) and expert systems.

The human olfactory system can't detect many substances until their concentrations reach

several times their hazardous level, and it's insensitive to some substances altogether. AEMC's

Electronic Nose (ENose) has a much wider dynamic range than the human nose -- from less than one

part per million (ppm) to 10,000 ppm. Thus, it can detect chemicals in concentrations so small that

people could not smell them, or so large that they would overwhelm human noses (or a mass

spectrometer). It can detect chemicals that people find odorless. It doesn't suffer from "odor fatigue,"

the human tendency to grow accustomed and insensitive to smells that start small and increase

gradually. And of course, it doesn't catch colds.

How it works

A person's olfactory system employs millions of

receptor cells of an estimated 350 different types. Each

type responds differently to the molecules in each

inhaled breath. The brain compares the pattern of

responses with patterns stored in its memory to

determine whether an aroma represents, for example, a

rose or a chocolate chip cookie. In place of the human

nose's receptor cells, the ENose uses an array of 28 films

of 16 different types. The films are insulators, but are

impregnated with carbon particles that enable them to

conduct electricity. ENose's 32 sensors, consisting of 16 different polymers

Each film absorbs, to a greater or lesser degree, certain classes of chemical compounds --

alcohols or ketones, for example -- when a waft of air brings them into contact. Depending on the kind


20

and amount of the compound it absorbs, the film swells or shrinks by a characteristic amount. Swelling

drives the carbon particles apart, reducing the ability of the film to conduct electricity (or put another

way, increasing its resistance). Shrinking draws the carbon particles closer together, making it easier for

electric current to flow across the film (decreasing its resistance).

A computer program reads the pattern of resistance

changes across the array, compares it to the patterns stored in its

memory from laboratory testing, and identifies the chemical

that has been "smelled." If the chemical is in a concentration

deemed to be dangerous, the ENose can sound an alarm to

notify the crew or possibly, sometime in the future, activate an

automatic system to remove the pollutant from the air via fans

and filters.

Sixteen types of film allow for more than 60-thousand possible combinations, each of which

could potentially signal a different chemical. In order to develop the ability to recognize each chemical,

the ENose must be trained. It is exposed repeatedly to varying concentrations of known compounds in

varied order, and the resulting patterns of resistance changes across the array of films are turned into a

computer algorithm.

The ENose can also be trained to recognize partial patterns caused by substances for which it

has not been specifically trained -- enough to identify the chemical's functional group. Currently it can

identify alcohols and aromatics, the two groups that include most of what is likely to present a problem

aboard the ISS, but it could be educated enough to identify any group.

Application areas of the Electronic Nose are:

Environmental monitoring

Monitoring of air, water and land.

Medical Diagnostics and Health Monitoring

Breath Monitoring

Eye Infection

Medical Environmental Monitoring

Leg Ulcers

Cultured Bacteria

Food and Beverage Applications

Quality and process monitoring of fruits, vegetables, meat, fish, brewery, tea, coffee and so on.

Automotive and Aerospace Applications

Detection of hazardous gas within automobiles, spacecrafts.

Narcotic Detection.

Application in Cosmetics and Fragrance Industry

Detection of Explosives

One of the most immediate targets for Electronic Nose application is the Food and Beverages

sector. Traditionally, the human panel plays the most important role in quality estimation of food and

agro products. But human panel testing is highly subjective and scarcely repeatable as well as

susceptible to individual psychophysical conditions of the testers. In order to optimize the evaluation of

quality and to enhance marketability of food and agro products, there is an increasing interest in non-

destructive methods to assist in the classification of different grades. And deployment of Electronic

Noses for this purpose appears to be the optimal solution. For example, the quality of Tea is principally

Illustration of how the ENose works


21

dependent on aroma and flavor. This quality is conventionally estimated by a sensory panel (called tea

tasters). To get rid of the inherent subjectivity and human-dependence in this kind of sensory panel

estimation, the Electronic Nose may be employed for the discrimination of tea aroma and flavor. For

tea-nose, the organoleptic scores available from tea flavor wheel for various samples are co-related with

the sensor data to effect classification and identification.

9. Hyperspectral Imaging Technology:

Over the years, optical sensing technologies have been investigated as potential tools for non-

destructive evaluation and inspection for food quality and safety. In particularly, methods based on two

mature technologies of imaging and spectroscopy have been widely studied and developed, resulting in

many successful applications in the food industry. However further development of these conventional

imaging and spectroscopy techniques is limited by their inabilities to obtain sufficient information from

individual food items. Therefore by taking the most useful characteristics of these two mature

technologies: imaging and spectroscopy, hyperspectral imaging (or imaging spectroscopy) has emerged

as a technology with great potential for effective and nondestructive quality and safety evaluation and

inspection in the area of food processing.

A conventional imaging system or more specifically computer vision is a common technique for

detecting surface features. The system normally consists of lighting and an area detector, with the light

source providing illumination to the sample and the area detector capturing mixed spectral contents

from the sample. Spatial information of the sample is obtained in the forms of monochromatic or colour

images, therefore conventional imaging system is used for colour, shape, size, surface texture

evaluation of food products and for surface defects detection in food inspection, however it cannot

identify or detect chemical properties or characteristics from a food product.

On the other hand, conventional spectroscopy system is a technique for evaluating chemical

properties or characteristics of food products. Such a system generally includes a light source, a

wavelength dispersion device, and a point detector. In this system, light is dispersed into different

wavelengths after interaction with the sample in order for the point detector to collect the dispersed

light to obtain spectral information from the sample. As the point detector has its size limitation,

conventional spectroscopy system cannot cover a large area or a small area with high spatial resolution.

Therefore the technique does not provide the spatial information which is regularly required for and is

critical in food inspection.

With the integration of the main features of imaging and spectroscopy, hyperspectral imaging

can simultaneously acquire both spatial and spectral information that is critical to the detection of food

safety and evaluation of food quality attributes. A typical hyperspectral system consists of a light

source, a wavelength dispersion device, and an area detector. The images are acquired over the visible

and near-infrared (or infrared) wavelengths to specify the complete wavelength spectrum of a sample at

each point in the imaging plane. These images are then combined and form a three dimensional

hyperspectral cube, with two dimensions for describing spatial information and the third one for

spectral information. In this hypercube, each spectral pixel corresponds to a spectral signature (or

spectrum) of the corresponding spatial region, recording the entire measured spectrum of the imaged

spatial point. Therefore the measured spectrum indicates the ability of the sample in absorbing or

scattering the exciting light, representing the inherent chemical properties of a sample. As a result, the

technology provides us with unprecedented detection capabilities, which otherwise cannot be achieved

with either imaging or spectroscopy alone. If conventional imaging is to provide the answer to the

question of where and conventional spectroscopy is to provide the answer to the question of what,

hyperspectral imaging is a technique to provide the answer to the question of where is what.


22

Hyperspectral imaging techniques have received much attention for food quality and safety

evaluation and inspection. Many approaches and applications have shown the usefulness of

hyperspectral imaging in the food industry. These applications include meat quality assessment,

automated poultry carcass inspection, quality evaluation of fish, bruise detection of apples, quality

analysis and grading of citrus fruits, bruise detection of strawberry, visualization of sugar distribution

of melons, measuring ripening of tomatoes, defect detection of pickling cucumber, and classification of

wheat kernels. Interested readers are advised to refer to the following list of publications for further

information (Sun, 2009).

10. NEW Terminology

Calibration is a comparison between measurements – one of known magnitude or correctness made or

set with one device and another measurement made in as similar a way as possible with a second

device.

Validation: Validation of Calibration model is the most essential step because it proves its ability to

predict the desired component concentration/ parameter.

Ultrasonic velocity: term used for description of wave propagation. i.e distance travelled per unit time.

Attenuation (in some contexts also called extinction) is the gradual loss in intensity of any kind of

flux through a medium.

Attenuation coefficient : Logarithmic loss of wave energy per unit distance.

11. CASE STUDIES:

11.1. Non-destructive prediction of translucent flesh disorder in intact mangosteen by short wavelength

near infrared spectroscopy.

Sontisuk, T., Kwon, Y. K., Anupun, T., Warunee, T. and Yutaka, N (2007).

11.1.1. Introduction

The quality of mangosteen (Garcinia mangostana L.) fruit is measured not only by external

factors such as color, shape, size, skin blemishes, latex staining and insect damage, but also by internal

factors such as translucent flesh, yellow gummy latex and hardening pericarp which are also very

important for consumer acceptance. Translucent flesh disorder is one of the most important internal

factors determining the quality of mangosteens. A floating technique using differences in specific

gravity is currently undertaken for the non-destructive detection of translucent flesh in mangosteens,

but it is still not reliable.

In recent years, NIR spectroscopy has been reported to be a very useful technique for sorting

various intact fruit. However, it is difficult for the light emitted from the light source to pass through

the intact fruit in the long wavelength near infrared (LW-NIR) range of 1100–2500 nm because of the

fruit‟s high moisture content.

11.1.2. Materials and Methods:

Commercially available mangosteens were purchased from a local fruit auction in Bangkok,

Thailand. The samples were delivered to the laboratory of Kasetsart Agricultural and Agro-Industrial

Product Improvement Institute (KAPI) and then kept in a room at a constant temperature of 25oC prior

to the recording of spectral data on the following day by a commercially available SW-NIR instrument

(PureSpect, Saika TIF., Japan) as shown in Fig. 1. About 1000 intact Thai mangosteens were used to

http://en.wikipedia.org/wiki/Flux


23

study the optimum conditions of SW-NIR measurement and 193 intact mangosteens were used for

evaluating the accuracy of translucent flesh disorder detection.

11.1.3. Results and Discussion:

a) Investigation of the instrument setup for optimal spectral acquisition

Due to each mangosteen sample having different internal factors such as translucent flesh,

hardening pericarp and seeds, the dynamic ranges of the SW-NIR intensity spectra were very different,

especially, between translucent flesh and hardening pericarp. The hardening pericarp, which is usually

produced after a physical shock during harvesting and transporting, caused low signal-to-noise ratio

spectra because its tissue has low light-penetration.

b) Investigation of light sources and SW-NIR spectra

Different light source levels of 100, 200 and 400W were tested for this experiment. The SW-

NIR spectra of mangosteens featured noise and a low-level intensity when using a low-level light

source compared to using a higher level light source. The spectra intensity of translucent flesh was

excessive when using a 400W light source. However, at 100W mangosteens with hardening pericarp

yielded very low-level spectra and signal-to-noise ratios while the spectra of mangosteens with

translucent flesh were more strongly produced. The results showed that the light level is one of the

important measurement conditions in order to correctly obtain spectra of translucent flesh and internal

disorders of mangosteens.

c) Investigation of measurement speed effect on SW-NIR spectra

The SW-NIR spectra of samples were collected at different measurement speeds (3, 6 and 12

m/min). We found that the spectra of mangosteen samples were better at the lower speeds in terms of

the signal-to-noise ratio. Therefore, the optimum measurement speed for this study was selected at 3

m/min with an integration time of 78 ms.


24

d) Discriminant analysis

The averaged absorbance spectra from the four

scans of 193 intact mangosteens are shown in Fig. 4. There

were four absorbance peaks, one influenced by chlorophyll

around 680 nm and three by water around 760, 840 and

970 nm. The absorbance SW-NIR spectra of the

translucent flesh mangosteen were lower or the

transmission ratio of SW-NIR in translucent flesh

mangosteen was higher. Mangosteen fruit with hardening

pericarp showed a different downward peak on the spectra

in the range around 825–875 nm. Hardening pericarp may

be found in a fewparts or around the whole mangosteen.

There were four samples in this experiment that had

hardening pericarp on all four measurement points around

the mangosteen. It may be that exclusion of the hardening pericarp spectra (or the spectra with high

second peak in intensity) from the training set may improve the accuracy. Four rawspectra of each fruit,

acquired from the four positions described earlier, were used to create four training sets. They were the

averaged spectra set (one averaged spectra per sample), the amended averaged spectra set (averaged

spectra omitting the spectra of the hardening pericarp fruit), the individual spectra set (four spectra per

sample) and the amended individual spectra (leaving out spectra from the hardening pericarp point of

measurement).

Table 1:SW-NIR classification for normal flesh and translucent flesh mangosteen using

discriminant analysis and leave-one-out cross-validation

The discriminant analysis results are shown in Table 1. The classification error was the lowest

in the amended individual spectra set where 632 samples out of 687 were correctly classified (92.0%

accuracy). The classification accuracy was slightly improved for those training sets containing each


25

individual spectra (92.0 and 91.2% compared with 89.9 and 89.1%). Interestingly, the improvement in

accuracy may be due to the better classification of the translucent flesh fruit. For instance, for the

averaged spectra set, 7 samples of 23 translucent flesh mangosteen (30.4%) were correctly identified

whereas 32 samples out of 92 (34.8%) were correctly classified for the individual spectra set. The

higher classification accuracy may be due to each individual spectra containing more accurate

translucence information than the averaged spectra.

They suggested the use of 3- or 6-position spectra to obtain a highly efficient and stable

calibration equation. The hardening pericarp information in the spectra seemed to have an influence on

the classification accuracy. Table 1 shows accuracy of 89.1% for the averaged spectra set against

89.9% for the amended averaged spectra set and 91.2% for the individual spectra set in comparison to

92.0% for amended individual spectra set. In this case, the translucent flesh may cause a greater change

in the intensity than the hardening pericarp. The Fisher‟s linear discriminant functions were studied in

order to search for the most informative wavelengths. The coefficients with the highest values were

found in the range 673–762 nm, which corresponded to the chlorophyll peak and the first water peak in

the absorbance spectra. The chemical bands associated with this wavelength range are not understood

and require further investigation.

11.1.4. Conclusion:

The absorbance of mangosteen fruit at the short wavelength was sufficient to use for translucent

flesh disorder detection in mangosteens. The accuracy can be increased by rejection of hardening

pericarp mangosteen spectra before evaluation. However, the accuracy of detection by using SW-NIR

transmittance spectroscopy for translucent flesh mangosteens could be improved in the future by

systematic consideration of other effects such as fruit size, pericarp thickness, seeds, gummy latex and

skin color.

11.2. Rapid and non-destructive analysis of apricot fruit quality using FT-near-infrared spectroscopy.

Bureau, S., Ruiz, D., Reich, M., Gouble, B., Bertrand, D., Audergon, J. and Renard, C. M. G. C.

11.2.1. Introduction:

The objective of this study was to evaluate the potential of near-infrared spectroscopy as a non-

destructive method to predict apricot quality traits such as firmness, soluble solids, titratable acidity,

ethylene production, sugar and acid contents through the comparison with standard techniques. The

evaluation was performed on a large diversity of fruits covering genetic factor and physiological stages

which was expected to be representative of the variability observed on the market. Samples belonging

to eight cultivars harvested at different stages of ripening were used as a calibration set. The prediction

models for each quality parameters were developed with partial least square technique.


FT-NIR spectra were recorded on a multi-purpose analyser (MPA) spectrometer (Bruker Optics,

Wissembourg, France) equipped with an integrating sphere to provide diffuse reflectance

measurements and a TE-InGaAs detector. The MPA was completely software-controlled by the OPUS

software Version 5.0 which was provided by Bruker Optics. The NIR spectrum of each sample was

obtained by taking the average of 32 scans. It was acquired between 800 and 2700 nm at 2 nm spectral

resolution, with scanner velocity of 10 kHz and a background of 32 scans. The time required to achieve

a spectral measurement was 30 s. The intact apricots were placed on an automated 30-position sample

wheel, each position corresponding to an 18 mm diameter hole. Apricots were placed at each position

with their stem–calyx axis horizontal. On each apricot, a diffuse reflectance spectrum was measured on

two opposite sides, the first on the blushed side and the second on the un-blushed side.


26


The general profile of the absorption spectra for apricot (Fig. 1) is very similar to that of other

plant materials as mandarin, apple or tomato. These spectra are in fact dominated by water absorption

bands with overtone bands of the OH-bonds at 970, 1190 and 1450 nm and a combination band at 1940

nm.

11.2.4. Conclusion

On intact apricot, near-infrared wavelength (800–2500 nm) could be used to accurately predict

soluble solids and titratable acidity. But, the prediction of the other quality traits such as firmness,

ethylene production, individual sugars and organic acids seemed not to be sufficiently accurate. In the

future we plan to check the robustness of these established models on other apricot fruits, including

different cultivars, harvested over different years and cultivated in different orchards. According to its

non-destructive and rapid characteristics, the near-infrared spectroscopy appeared already as a suitable

technique for screening quality parameters of a large number of apricots.


27

11.3. Non-destructive ultrasonic monitoring of tomato quality during shelf-life storage.

Amos Mizrach (2007)

11.3.1. Introduction

Ripening indicators in greenhouse tomatoes (Lycopersicon esculentum Mill) are softening of the

flesh, decreasing acidity, and change in color. The change in sugar content during maturation is not

considerable, but is important to consumers. Tissue firmness is the property most relevant to consumer

perceptions of ripening, and it is the factor most closely related to the stage of maturity in tomatoes.

Atrained person can assess this parameter, but at present a destructive measurement method is required.

Consumer demand for high-quality products raises the need for a reliable, rapid, nondestructive, non-

invasive technique for maturity determination, especially during harvest and at the packing site. There

are several techniques for nondestructive measurement of quality parameters in various fruit and

vegetables, e.g., acoustic impulse response, and ultrasound. The ultrasonic technique was suggested

previously to determine some quality parameters of fruit and vegetables but not for fresh greenhouse

tomatoes. Scientists used a similar ultrasonic technique to evaluate chilling injury in tomatoes, but

found difficulties with contact pressure between the probes and the tomato flesh. They applied

destructive penetration of the probes to record their results and suggested the development of a

nondestructive ultrasonic technique to make this method applicable.


The mechanical structure of fruit tissue, its physicochemical quality indices, and each change in

quality attributes of the fruit, affect the amplitude of the ultrasound signal passing through the fruit

tissue, and the attenuation of the ultrasonic wave can be evaluated. Previous studies with avocado and

mango fruit suggested that there is a good correlation between the attenuation of the ultrasonic signal

and the mechanical and physiological properties of the fruit tissue and this correlation was examined

for tomatoes in the present study.


The attenuation of the ultrasonic waves passing through the tissue of the fruit was calculated

during the course of storage for the fruit in groups N1–N10 and 1–70. The calculated attenuation was

plotted against storage time for the N1–N10 group (Fig. 1). These fruit were subjected to a


28

nondestructive test only; each of the 10 fruit was tested at the same point each time during the course of

the experiment. Each data point in Fig. 1 represents the mean value for the same 10 fruit in the batch.

The attenuation measurements revealed a decreasing trend over the course of time, and a parabolic

expression yielded the best-fitting curve for the plot of the mean attenuation values against time (R2 =

0.965). The results of the attenuation measurements of the ultrasonic waves passing through each fruit

of the 1–70 group, showed wide scattering, and no clear tendency over the course of time could be

seen. This might be because the measurement locations for the fruit of group 1–70 were arbitrarily

chosen, whereas the attenuation calculated for group N1–N10 was based on daily measurements at the

same location on the fruit peel.

The physiological behavior of the tomatoes during storage is illustrated in Fig. 2. The results

show the expected trend in firmness: the mean value decreased in the course of time from about 8.9

down to 2.2 N. The sugar contentwas found to be quite constant, as expected, in the range of 2–2.5%

Brix. The average values of attenuation in group N1–N10 and of firmness in group 1–70 showed

similar trends when plotted against time (Figs. 1 and 2, respectively); parabolic expressions.

11.3.4. Conclusion:

This study examined the potential use of a nondestructive ultrasonic technique to monitor the

ripening of tomatoes during storage; the ultrasonic measurements were analyzed in parallel with the

destructive measurements of the firmness of the fruit. The ultrasonic measuring technique was

successfully applied by means of ultrasonic probes in contact with the fruit peel. The measured

attenuation was found to be linearly related to the firmness of the fruit during 8 days following entry

into storage. This suggests that this ultrasonic method might be used for nondestructive firmness

monitoring of tomatoes during their shelf-life. However, in the second group of tomatoes (1–70), in

which the measuring locations were arbitrary chosen, wide scattering in the attenuation measurements

as well as lack of a clear trend in the relationship between attenuation and firmness during the course of

shelf-life was found. The reason for this is not yet clear, but we suspect that the irregular flesh structure

of the tomato is a relevant factor, and this needs to be investigated.


29

11.4. Rapid and non-destructive analytical techniques for measurement of apricot quality.

Cristina, P., Gabriel, L. R., Viorica, B. and Gheorghe, C


Application of acoustic impulse response method is reported for evaluation of apple, kiwi,

mango, peach texture but not of apricot. There is a little information about background color

modification of apricot fruits during maturation on tree.The objective of this study was thus to define a

maturity and quality index for two apricot cultivars (Excelsior and Nicusor), through exploring the

potential of acoustic impulse response method and by measuring color parameters.


Objective color measurements were assessed using a HunterLab Mini Scan XE Plus

spectrophotometer. In the Hunter scale, L measure lightness and varies from 100 for perfect white to

zero for black approximately as the eye would evaluate it. The chromaticity dimension “a*” measures

redness when positive, gray when zero and greenness when negative, and “b*” measures yellowness

when positive, gray when zero, and blueness when negative.

Acoustic firmness was measured at the equator of the unpeeled fruit in two repetitions by using

an AWETA Acoustic Firmness Sensor. In this device, an acoustic signal is generated by means of

gentle impact on the equator of the fruit. This signal is processed and transformed to obtain a peak of

natural frequency, which is used to calculate the stiffness factor as f2x m2/3, here f represents this

frequency and m is the fruit mass.


Table 1. Variation of physico-chemical attributes of apricot

From the acoustic measurement changes in resonant frequency during maturation period of

apricot were observed, because of tissue softening. Acoustic firmness index decreased during fruit

maturation to value 3 at ripe stage. Both cultivars have fruits with high firmness at this stage what show

compact and high density fruit (table1).

Dry matter increases during the ripening of the apricot. Dry matter gain is more important

during the last two maturity stage and the less significant beyond first maturity stage (table1). As it can


30

be seen in table 1 the total carotenoids content increase and varied from 0.7 to 4.79 mg/ 100g f.w for

Excelsior cultivar and from 0.8 to 3.78 mg/100g f.w for Nicusor cultivar during maturation of fruit.

11.4.4. Conclusion:

Determination of color parameters is an easy, convenient and non -destructive method that provides

fruit color characterization, quality control and a good estimation of the carotenoid content of apricot.

Based on these results, most of the apricot breeding programs could use h° and a* as a quality criterion

to select new varieties with higher carotenoid content. This color parameter can be used successfully

for distinguishing different maturity stages of apricot and for establishing the optimum harvest time.

11.5. Hayward kiwifruits and plant growth regulators: Detection and effects in post-harvest studied by

Magnetic Resonance Imaging (MRI) and Scanning Electron Microscopy (SEM).

Taglienti, A., Sequi, P., Cafiero, C., Cozzolino, S., Ritota, M., Ceredi, G. &Valentini, M. (2011)


Kiwifruit, Actinidia chinensis, Planch and Actinidia deliciosa, A. Chev., is of Chinese stock, its

commercial development took place in New Zealand, later was successfully introduced in Italy and in

Mediterranean region and now is one of most appreciated fruits amongst consumers. The effects of

synthetic hormones on kiwifruit quality during post-harvest ripening have been investigated only by

evaluating few parameters, and in some cases data are in not in complete agreement.


Scanning Electron Microscopy (SEM) was performed using a SEM Zeiss EVO (Carl Zeiss

SMT Ltd., Cambridge, England) under variable pressure conditions (50 Pa), using secondary electron

detector and with an acceleration voltage of 20 kV; a slice of the outer pulp of the fruit with dimensions

of ca. 0.5 _ 0.5 mm2 was cut and examined as fresh hydrated material. SEM experiments were

performed only on kiwifruits at harvesting stage.


Table 1 summarized the firmness and Brix values measured for kiwifruit at harvest, and at the end

of storage in both considered conditions: i.e. normal refrigeration (NR) and controlled atmosphere


31

(CA). At harvesting stage firmness is quite similar for the three different plots, whilst the Brix value is

about 5–10% higher for untreated samples. Auxin appears to be the most affecting Brix level, with

cytokinin exerting a slighter reduction. When considering the NR and CA storage the untreated

kiwifruits have the highest firmness and sugar content. In both cases, fruits treated with cytokinin are

characterised by values similar to untreated samples, whilst those treated with auxin are largely lower.

This indicates that the use of cytokinin does not affect significantly the overall quality during post-

harvest storage; on the other hand, auxin plays an important role on kiwifruit quality, with CA and NR

being very similar.

11.5.4. Conclusion:

MRI and SEM have been used to study the effects of PGRs, i.e. auxin and cytokinin, in

kiwifruit; structural modification of the internal morphology, quality alteration during post-harvest and

comparison between different storage conditions have been addressed. MRI offers a reliable method for

revealing the use of PGRs when forbidden, i.e. organic farming, even at very long delay from

treatments, and the use of transverse relaxation times of the outer zones and the greyscale analysis of

T2-weighted images are useful parameters for recognizing the type of hormone. SEM images confirms

that modification of tissues at a cellular level occurred; pore size and vacuoles are the most affected,

further investigation is elucidate the physiological mechanisms responsible for the observed variations.

11.6. Hyperspectral imaging for non-destructive determination of some quality attributes for strawberry.

Elmasry, G., Wang, N., Elsayed, A. and Ngadi (2007)


Strawberry (Fragaria sp.) is one of the economically important fruits which are more popularly

fresh, used for garnishing cakes and pastries, flavored for juices and milk products, and processed into

jams and others. Thus, together with the recent concern for food quality and safety, automatic

technologies for judging the fresh quality of strawberry are being sought.


The hyperspectral imaging technique has been implemented in several applications such as

inspection of poultry carcasses, defect detection

or quality determination on apples, eggplants,

pears, cucumber and tomatoes as well as

physical, chemical and mechanical properties

estimation in various commodities. In addition,

a significant amount of research has been done

in the area of spectroscopy and hyperspectral

imaging applied specially to fruit analysis. The

successful attempts to evaluate internal

properties nondestructively were accomplished

using spectral technology for predicting sugar

content, soluble solids, firmness, moisture

content, acidity and so many other applications.

In strawberries, however, there had been very

limited literature on the use of spectroscopic

technique for quality estimation.


32


Fig. 3 shows the average reflectance spectra in a range of 400–1000 nm of strawberries

collected at different ripeness stages. The presence of water in the fruit gave a rise to the characteristic

absorption bands that appear as localized minima. The samples containing higher moisture contents

had lower reflectivity across their spectra. In spite of ripeness stage, the reflectance curves of

strawberry were rather smooth across the entire spectral region. In case of unripe fruit, the reflectance

curve had three broadband absorption regions around 500, 680, and 960 nm in addition to small

absorption region at 840 nm. The regions around 500 and 680 represent anthocyanin and chlorophyll

pigments which represent the colour characteristics in the fruit. The absorption regions in the NIR at

840 and 960 nm represent sugars and water absorption bands.

Table 1 shows that the model was very accurate for predicting moisture content with r of 0.90

and 0.96 for training and validation sets, respectively. The SEC and SEP were 6.085 and 3.874 for

calibration and validation sets, respectively. TSS was predicted with r of 0.8 and SEC of 0.233 for the

training set. The accuracy of the model in the validation set for predicting TSS was with r of 0.85 and


33

SEP of 0.184. The pH was predicted with r of 0.87 and SEC of 0.105 for the training set and r of 0.87

and SEP of 0.129 resulted from the validation sets.

11.6.4. Conclusion:

The PLS models were established between reflectance spectra and the quality parameters. The

models for the moisture content (r = 0.97, SEC = 6.085, SEP = 3.874, and nine latent factors), the total

soluble solids (r = 0.85, SEC = 0.233, SEP = 0.184, and four latent factors) and pH (r = 0.87, SEC =

0.105, SEP = 0.129, and six latent factors) show good prediction performance. Texture measures were

derived from the grey-level co-occurrence matrix (GLCM) of strawberry images to identify its ripeness

stage. High classification accuracy of 89.61% for correctly identifying strawberry ripeness stage was

achieved using the GLCM parameters at horizontal direction at angle of 0o.

12. CONCLUSION:

Determination of quality of any food material is actually a complex problem that requires a

variety of specific sensors, more than an accumulation of simple sensors. Various techniques are being

tried. NMR, x-ray CT, Ultrasonic‟s, Acoustics, E-nose, Hyperspectral imaging and NIR techniques

may be useful for a large volume of work in agriculture, especially for evaluation of qualities such as

maturity, internal quality of fruit and conditions of food materials after processing. Thus the non-

destructive techniques of quality evaluation is gaining a good momentum in field as well as processing

industries. But rapid and accurate determination of internal quality posses technical challenges because

of complex structural, physical and chemical properties of fruits and vegetables. As these methods are

instantaneous and can simultaneously determine many parameters, could be used for bulk handling of

produce. Most of the developed countries are using various on farm non-destructive methods. But in

India due to technological hitches and lack of development of equipments and their availability, has

restricted their wider usage in quality evaluation of fruits and vegetables. Research must continue to

focus on making their use easier and lowering the cost to make it within the reach of common

businesses /growers of both developed and developing countries. Therefore the development of an

accurate, reliable and non destructive method for the quality evaluation before harvest and at packaging

site is critical to provide better quality produce.

13. REFERENCES:

American Association of Cereal Chemists. (1983). Approved methods of the Am. Assoc. of Cereal

Chemists, 8th Ed. (March). St. Paul, Minn.

BARCELON, E.G., TOJO, S. AND WATANABE, K.,1999, X-ray computed tomography for internal

quality evaluation of peaches. J. Agric. Eng. Res., 73: 323–330.

BUREAU, S., RUIZ, D., REICH, M., GOUBLE, B., BERTRAND, D., AUDERGON, J. AND RENARD, C. M.

G. C., 2009, Rapid and non-destructive analysis of apricot fruit quality using FT-near-infrared

spectroscopy. Food Chemistry. 113: 1323-1328.

CLARK, R.L., 1975, An investigation of the acoustic properties of watermelon as related to maturity.

ASAE Meeting, Paper Number: 75-6004.

CRISTINA, P., GABRIEL, L. R., VIORICA, B. AND GHEORGHE, C., 2010, Rapid and non-destructive

analytical techniques for measurement of apricot quality. Romanian Biotechnological Letters.

15(2): 5213-5216.

DIEZMA-I, B., RUIZ-A, M., BARREIRO, P., 2004, Detection of internal quality in seedless watermelon

by acoustic impulse response. Biosystems Engineering 88 (2): 221–230.


34

DULL, G.G. AND BIRTH, G.S., 1989, Nondestructive evaluation of fruit quality: Use of near infrared

spectrophotometry to measure soluble solids in intact honeydew melons. Hortscience, 24 :754.

DULL, G.G., BIRTH, G.S, SMITTLE, D.A. AND LEFFLER, R.G., 1989, Near infrared analysis of soluble

solids in intact cantaloupe. J. Food Sci., 54: 393–395.

ELMASRY, G., WANG, N., ELSAYED, A. AND NGADI, M., 2007, Hyperspectral imaging for non-

destructive determination of some quality attributes for strawberry. J. Food Eng., 81: 98-107.

FARRAR, T.C. AND BECKER, E.D., 1971, Pulse and Fourier Transform NMR: Introduction to Theory

and Methods. Academic Press, New York.

FUKUSHIMA, E. AND ROEDER, S.B.W., 1981, Experimental Pulse NMR. Addison-Wesley Publishing

Company, Reading, MA.

HAYASHI, S., SUGIYAMA, J., OTOBE, K., KIKUCHI, Y., USUI, S., 1992, Nondestructive measurement

for maturity of muskmelons by analysis of acoustic-signals. J. Jap. Soc. Food Sci. and Technol.

– Nippon Shokuhin Kagaku Kogaku Kaishi 39 (6): 465–470.

HE, D., LI, Z., WANG, H., 1994, On the characteristics of sound wave forms of watermelons. Acta

Universitatis Agriculturalis Boreali-Occidentalis 22 (3): 105– 107.

IWAMOTO, M., KAWANO, S. AND YUKIHIRO, O., 1995, An overview of research and development of

near infrared spectroscopy in Japan. J. Near Infrared Spectrosc., 3: 179–189.

JHA, S. N. AND MATSUOKA, T., 2000, Non-destructive techniques for quality evaluation of intact fruits

and vegetables. Food Sci. Technol. Res., 6(4): 248-251.

KAWANO, S., 1998, New application of nondestructive methods for quality evaluation of fruits and

vegetables in Japan. J. Jpn. Soc. Hort. Sci., 67: 1176–1179.

KAWANO, S.,1994, Non-destructive near infrared quality evaluation of fruits and vegetables in Japan.

NIR News, 5: 10–12.

KERR, W.L., KAUTEN, R.J., MCCARTHY, M.J. AND REID, D.S., 1998, Monitoring the formation of ice

during food freezing by magnetic resonance imaging. Food Sci. Technol., 31: 215–220.

LÜ, F., 2003, Non-destructive Quality Evaluation of Watermelon based on its Acoustic Property.

Zhejiang University, Hangzhou, China.

MACFARLANE, N.J.B., BULL, C.R., TILLETT, R.D., SPELLER, R.D., ROYLE, G.J. AND JOHNSON,

K.R.A., 2000, The potential for Compton scattered x-rays in homogeneity. J. Agric. Eng. Res.,

75: 265–274.

MIQUEL, M.E., EVANS, S.D. AND HALL, L.D., 1998, Three dimensional imaging of chocolate

confectionary by magnetic resonance methods. Food Sci. Technol., 31: 339–343.

MIZRACH, A., 2007, Non-destructive ultrasonic monitoring of tomato quality during shelf-life storage.

Postharvest Biol. Technol., 46: 271-274.

MIZRACH, A., GALILI, N., AND ROSENHOUSE, G., 1989, Determination of fruit and vegetable

properties by ultrasonic excitation. Transactions of the ASAE, 32: 2053-2058.

MIZRACH, A., GALILI, N., AND ROSENHOUSE, G., 1992, Half-cut fruit response to ultrasonic

excitation. ASAE Paper No. 923017. American Society of Agricultural Engineers, St. Joseph, MI,

USA.


35

MIZRACH, A., GALILI, N., AND ROSENHOUSE, G.,1996a, Method and apparatus of ultrasonically

determining of quality parameters of fresh products. US Patent No. 5, 589, 209.

MIZRACH, A., GALILI, N., GAN-MOR, S., FLITSANOV, U., & PRIGOZIN, I., 1996, Models of ultrasonic

parameters to assess avocado properties and shelf life. J. Ag. Eng. Res., 65: 261-267.

MIZRACH, A., GALILI, N., ROSENHOUSE, G., AND TEITEL, D.C., 1991, Acoustical, mechanical and

quality parameters of winter grown melon tissue. Transactions of the ASAE, 34: 2135-2138.

MIZRACH, A., GALILI, N., TEITEL, D.C., ROSENHOUSE, G., 1994, Ultrasonic evaluation of some

ripening parameters of autumn and winter-grown „Galia‟ melons. Sci. Hort. 56 (4): 291–297.

RAO, X., YING, Y., L, F., JIN, B., 2004, Development of a fruit quality inspecting system based on

acoustic properties. Transactions of the CSAM , 35 (2): 69– 71.

SCHOTTE, S., BELIE, N.D., BAERDEMAEKER, J.D., 1999, Acoustic impulse-response technique for

evaluation and modeling of firmness of tomato fruit. Postharvest Biol. and Technol. 17 (2): 105–

115.

SLAUGHTER, D.C., 1995, Nondestructive determination of internal quality in peaches and nectarines.

Trans. ASAE, 38: 617–623.

SONTISUK, T., KWON, Y. K., ANUPUN, T., WARUNEE, T. AND YUTAKA, N., 2007, Non-destructive

prediction of translucent flesh disorder in intact mangosteen by short wavelength near infrared

spectroscopy. Postharvest Biol. Technol., 43: 202-206.

STONE, M.L., ARMSTRONG, P.R., ZHANG, X., BRUSEWITZ, G.H., CHEN, D.D., 1996, Watermelon

maturity determination in the field using acoustic impulse impedance techniques. Transactions of

the ASAE, 39 (6): 2325–2330.

SUGIYAMA, J., KATSURAL, T., HONG, J., KOYAMA, H., MIKURIYA, K., 1998, Melon ripeness

monitoring by a portable firmness tester. Transactions of the ASAE , 41 (1): 121– 127.

SUGIYAMA, J., OTOBE, K., HAYASHI, S., USUI, S., 1994, Firmness measurement of muskmelons by

acoustic impulse transmission. Transactions of the ASAE, 37 (4): 1235–1241.

SUN, D.-W., 2009, Hyperspectral imaging for food Quality Analysis and Control, Academic Press/

Elsevier, San Diego, California, USA, 15 Chapters.

TAGLIENTI, A., SEQUI, P., CAFIERO, C., COZZOLINO, S., RITOTA, M., CEREDI, G. AND VALENTINI,

M., 2011, Hayward kiwifruits and plant growth regulators: Detection and effects in post-harvest

studied by Magnetic Resonance Imaging and Scanning Electron Microscopy. Food Chemistry.

126: 731-736.

TOLLNER, E.W., HUNG, Y.-C., UPCHURCH, B.L. AND PRUSSIA, S.E., 1992, Relating x-ray absorption

to density and water content in apples. Trans. ASAE, 35: 1921–1928.

YAMAMOTO, H., IWAMOTO, M., HAGINUMA, S., 1980, Acoustic impulse response method for

measuring natural frequency of intact fruits and preliminary applications to internal quality

evaluation of apples and watermelons. J. of Texture Studies, 11 (2): 117–136.

ZWIGGELAAR, R., BULL, C.R., MOONEY, M.J. AND CZARNES, S., 1997, The detection of soft materials

by selective energy x-ray transmission imaging and computer tomography. J. Agric. Eng. Res.,

66: 203–212.

NDQAT Seminar Report complete

Documents

Transcript of NDQAT Seminar Report complete