INTELLIGENT PARKING LOT SYSTEM

29
DESIGN AND MANAGEMENT OF AN INTELLIGENT PARKING LOT SYSTEM BY MULTIPLE CAMERA PLATFORMS NEETHU.K.PHILIPS SAINTGITS COLLEGE OF ENGG.,KOTTAYAM 06/09/2022 1

Transcript of INTELLIGENT PARKING LOT SYSTEM

Page 1: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

DESIGN AND MANAGEMENT OF AN INTELLIGENT PARKING LOT SYSTEM BY MULTIPLE CAMERA PLATFORMS

NEETHU.K.PHILIPSSAINTGITS COLLEGE OF ENGG.,KOTTAYAM 1

Page 2: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• Traditional parking lots commonly use security camera, ultrasonic sensors or infrared ray

sensors to manage the parking lots.

• I present a hybrid intelligent parking system, which is able to:-

inform the drivers where is the empty parking space,

lend the drivers to easily record where they park,

provide remote monitoring, and

offer the parking spot leading service when drivers forget where they parked.

remote monitoring, detection and monitoring of parking in the personal sites, and fire

detection.

2

INTRODUCTION

Page 3: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

WHAT DO WE HAVE????

3

Page 4: INTELLIGENT PARKING LOT SYSTEM

05/03/2023 4

Page 5: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

WHAT DO WE NEED???

5

Page 6: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• The system contains four main parts: MAV system,parking spot marking system, exclusive

parking monitoring system and fire detection system.

• High mobility of the MAV and the on board camera -fix the blind spot.

• The on board camera also help MAV positioning by detecting the QR code at each parking spot.

• MAV detect every parking spot in site status-transfer the status to the central control computer in

real-time.

• Driver can use their smart phone application to call MAV, which can lend the driver to their car

when driver forget where they parked.

6

SYSTEM

Page 7: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• Implant both QR code and NFC tag in every parking spot.

• Scanning QR code might fail due to light condition.

• Latest NFC application to support the system when QR code is failed.

• QR code ,NFC tag -serial number of parking spot and the URL of the security camera.

• The driver can record the information through their smart phone by scanning the QR code

and NFC tag.

• Check their car condition by accessing the URL when they are away from their cars.

7

WHAT MAKES IT SMART??

Page 8: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• Webcam in every exclusive parking spot for license plate detection.

• Webcam connected to Raspberry Pi->cheap, simple, and easy to customize and expend.

• Some parking lot use IP camera to do the monitoring, however, it is too expensive and heavy and

it needs utility power.

• Thermal imager for our fire detection or abnormal heat and we set our fire nozzle on two-axis

platform, which can track the fire point and put out the fire directly.

• An intelligent bracelet for the security guard which they can use it to control the MAV for urgent

and special mission.

8

CONTINUES…

Page 9: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• We use Parrot AR. Drone 2.0 quipped with 2 ultrasound

sensors, a front camera, a belly camera, and an inertial

measurement sensor.

• MAV Control instructions were send from computer to MAV

through Wi-Fi and every data, which include those received

from sensors and the images from the camera also.

• We mainly use the belly camera to navigate and locate the

MAV and also in site detection.

9

A. MAV SYSTEM

Page 10: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• In order to patrol the MAV above the parking lot in certain order, we use parking gridlines, which is painted

in specific color, for MAV to track.

• First, the MAV capture the image from the bellycamera, which is a RGB image.

• And then, we transfer the RGB to HSV.

• Now we can separate the image to three matrix in Hue, Saturation, and Value.

• By using OpenCV, we can find the specific color we want. We need to detect the edge of the parking

gridlines by using OpenCV as well.

• After the edge detection, we need to do Hough Lines detection in order to connect all the pixel dots which we

found in edge detection.

10

Ground Line Detection and Navigation:

Page 11: INTELLIGENT PARKING LOT SYSTEM

05/03/202311

IMAGE PROCESSING

RGB TO HSV EDGE DETECTION

HOGH LINE DETECTION

Page 12: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• When we have the line, we still have to determine the direction of the straight line by using slope.

• When the slope is less than zero, the MAV needs to turn left. When the slope is greater than zero, the

MAV needs to turn right.

• Now set the image center. The MAV needs to know the line which camera has detected so that it won’t

deviate the default route during the flight.

• We can know the relative position of the line and the MAV by calculating the center of the line.

12

MAV

Where Should I

Go??

Page 13: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

2) Locating and Positioning the MAV:

• For locating the MAV, we use QR code at each parking spot. The area will be boxed out when QR code is

detected.

• QR code serve three purposes in our system: to locate the MAV its current position, to update the latest record

of the parking spot, and let the driver easier to record their parking spot information.

3) In Spot Detection:

• Before the MAV determining each parking spot whether is parked or not, it needs to set the position and

detection range of each spot.

• First we find the corner of spot, and then use the corner points to set the position and size of the image.13

CONTINUES..

Page 14: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• Since the indoor parking lot is usually dusky,we binarize the image.

• The image will become a large group of white dots and few scatter dots.

• We count the dots in pixels with specific area, and set a threshold to determine whether there is a

car in spot or not.14

IMAGE IDENTIFICATION

Page 15: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• Every spot in our parking lot has a QR code and a NFC tag.

• Update the latest record of the parking spot.

• The main purpose of marking system is that let the MAV knows which data belongs to which spot.

• Scan the QR code- the MAV could know the data, including spot image ,whether there is a car in spot or not.

• Website linked to security room and human-machine interface at the entrance, after the driver scan the QR

code, they can check their car when they away from the car by using smart phone or computer.

• NFC tag mounted can be under QR code. It is quicker and more efficient then QR code.

15

B. PARKING SPOT MARKING SYSTEM

Page 16: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

Exclusive parking could be used for VIP or people who own the spot.

• We use a Raspberry Pi with multiple webcam to monitor every exclusive parking spot.

1) Multiple Webcams:

• A Raspberry Pi with multiple webcam could reduce the cost of hardware devices.

• However, the CPU in Raspberry Pi is not very good.So we use clock control since we don’t need real-time

image from every spot.

16

C. EXCLUSIVE PARKING MONITORING SYSTEM

Page 17: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• It updates all spot image every two minutes, which means if there are 10 webcams connect to Raspberry Pi, it can

update twenty spots in two minutes.

• After we capture the image, we use FTP for our Remote transmission to host-side.

• It is a simple and basic internet transmission and it can be easily used in Windows. For remote control, we use

SSH(Secure Shell).

• Secure Shell (SSH) is the standard for secure file transfer and remote logins over the internet. All network traffic

is encrypted and optionally compressed, providing strong authentication measures and secure communications

17

CONTINUES..

Page 18: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• The first thing to do for the host-side processing is to find the license plate location-Very important to do first

as the license plate is very small compare to the whole image.

• To find the candidate region, we change the RGB image to grayscale image.

• Then we use Gaussian filter for smoothing and Sobel for edge detection.

• After few image processing, we can use MSER(Maximally Stable Extremal Regions) to create possible region

of the license plate.

18

License Plate Location and recognition:

Page 19: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

(a) Change to grayscale image

(b) Image after Sobel operation

(c) Image after Gaussian smooth

(d) Use MSER to select biggest stable area19

CONTINUES..

Page 20: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• The next step right after we have a clear image is circle out the character.

• Therefore, we draw several horizontal lines across the plate and calculate the number changes between black and

white in every line.

• After the statistics calculation, the area with dramatic changes is the part we want.

• The next thing is character separation. Use edge detection to detect discontinuities in the image intensity values.

Image attributes usually reflect significant changes in important events and changes in properties.

• After image processing, license plate model must be created by the license plate recognition. We trained

Tesseract-OCR engine as our identification which will be link to the manager and the owner of the interface.

20

CONTINUES..

Page 21: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• For Managers and owners interface, we can use Dreamweaver to create web pages and we

also set up a site with Appserver so that we can write the website into NFC Tag.

• Managers interface can monitor all parking spot status and if vehicles parked in wrong place, the manager can

warn. There is also a interface for car owners.

• The biggest difference between manager interface and car owner interface is the owner can only check his own

personal spot.

• However, the car owner can use guide map to find their exclusive parking spaces and by using a smart phone,

they can monitoring their car status anytime, any where.

21

In site system interface

Page 22: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• Most of the fire extinguisher and automatic sprinklers are used after the fire spread out and it may affect

many things far away from fire.

• We want to develop a fire system which can put out fire in early stage.

• We can combine thermal imager and computer vision processing together to solve this problem.

1) Fire Detection:

• Fire source detection is a key part.

• Thermal imager can instantly alert when abnormal temperature rise.

22

D. FIRE DETECTION SYSTEM

Page 23: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

23

• Through the color of thermal image, we can determine the temperature level and preset a

threshold to capture high temperature.

• After a region color analysis, the results were obtained by the use of back projection to draw

the outline of the fire source.

CONTINUES..

Page 24: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• Two-axis platform is a machine agency which can rotate up and down.

• Calculate the distance of the object by measuring motor rotation and pitch angle

obtained from the encoder of the two-axis platform.

• Under normal circumstances, it patrol automatically. If there is a specific place need to monitor, it can

also be switched to manual control.

• If the target was found, it will lock the target and rotate the motor control platform until the target is

located in the center of the horizontal axis of the screen.

24

Two-axis platform control and fire source location estimation:

Page 25: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• Rotate the pitch axis, let the target locate in the center of the screen.

• As the instrument is at fixed height, we can calculate the distance between the platform and the fire by current tilt

angle of the platform:

d = h * tan( 90 - δ ); δ is Two-axis platform rotation degree, h is the height of the platform.

3) Extinguishing device control:

• After calculating the position of the fire source, an alarm signal will be transmitted to the Arduino making a

warning buzzer sounds through serial port.

• Then, computer terminal sent coordinates of the fire sprinkler system, which convert from thermal imager to

Arduino system.

• Arduino system send signal to control platform to aim fire source and put out fire. 25

CONTINUES..

Page 26: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• Most of the MAV patrolling and detection are automatic. However, the security guards need to operate

MAV manually in special conditions.

• Thus, we design an intelligent bracelet to support special occasion. Manager is able to control MAV using

the bracelet.

• It contains built-in three-axis gyroscope and accelerometer, and through xbee, we can send instructions

from server end.

26

E. INTELLIGENT BRACELET SYSTEM

Page 27: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

• Intelligent parking system’s easy and simple interface could help the people save their time and money.

• Webcam with Raspberry pi can be expanded for any size parking lot.

• The vehicle detection and patrol route guidance can still be affected by many random reasons, resulting

in MAV possibly out of control.

• Update rate of the parking spot may not fast enough in high traffic load.

• In the future,there can be extensions for Raspberry Pi, including voice guidance, coordinate with fire

detection, to have more practical and efficient feature.

27

CONCLUSION

Page 28: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

[1] Chieh-Hsun Huang, Han-Sheng Hsu,Hong-Ren Wang, Ting-Yi Yang, Cheng-Ming Huang,on “Design and Management of an

Intelligent Parking Lot System by Multiple Camera Platforms”, Proceedings of 2015 IEEE 12th International Conference on

Networking, Sensing and Control Howard Civil Service International House, Taipei, Taiwan, April 9-11, 2015

[2] Hilal Al-Kharusi, Ibrahim Al-Bahadly on World Journal of Engineering and Technology on “Intelligent Parking Management

System Based on Image Processing”,2014, 2, 55-67

[3] G. Nagy, Y. Xu, “Automatic Prototype Extraction for Adaptive OCR”, Proc. of the 4th Int. Conf. on Document Analysis and

Recognition, IEEE, Aug 1997, pp 278-282.

[4] R. Yusnita, Fariza Norbaya, and Norazwinawati Basharuddin,on “Intelligent Parking Space Detection System Based on Image

Processing”, International Journal of Innovation, Management and Technology, Vol. 3, No. 3, June 2012

[5] Y. Yorozu, M. Hirano, K. Oka, and Y. Tagawa, “Electron spectroscopy studies on magneto-optical media and plastic substrate

interface,” IEEE Transl. J. Magn. Japan

REFERENCES

28

Page 29: INTELLIGENT PARKING LOT SYSTEM

05/03/2023

Thank You ! 29