Movement Tracking in Real-time Hand Gesture Recognition

61
Movement Tracking in Real-time Hand Gesture Recognition Presenters Purohit Pankaj (W579074) Salagar Muaaz (W579080) Kulkarni Pranav (2010BCS203) Desale Ritesh (2010BCS211) Seminar Guide Mrs. S. S. Solapure

Transcript of Movement Tracking in Real-time Hand Gesture Recognition

Page 1: Movement Tracking in Real-time Hand Gesture Recognition

Movement Tracking in

Real-time Hand Gesture Recognition

PresentersPurohit Pankaj (W579074)Salagar Muaaz (W579080)Kulkarni Pranav (2010BCS203)Desale Ritesh (2010BCS211)

Seminar GuideMrs. S. S. Solapure

Page 2: Movement Tracking in Real-time Hand Gesture Recognition

Agenda IntroductionReference Paper and Its Contents

ApplicationConclusionQuestions

Page 3: Movement Tracking in Real-time Hand Gesture Recognition

Agenda IntroductionReferenced Paper and Its Contents

ApplicationConclusionQuestions

Page 4: Movement Tracking in Real-time Hand Gesture Recognition

Introduction

Page 5: Movement Tracking in Real-time Hand Gesture Recognition

Introduction

Gesture Recognition

Hand Gesture Recognition

Page 6: Movement Tracking in Real-time Hand Gesture Recognition
Page 7: Movement Tracking in Real-time Hand Gesture Recognition

Introduction (Contd.)

What is it?

How it works?

Page 8: Movement Tracking in Real-time Hand Gesture Recognition
Page 9: Movement Tracking in Real-time Hand Gesture Recognition

Introduction (Contd.)

What is its NEED?

Advantages Natural Interaction Builds a Richer Bridge Remote Interaction Wonderful Gaming Experience

Page 10: Movement Tracking in Real-time Hand Gesture Recognition

Agenda IntroductionReference Paper and Its Contents

ApplicationConclusionQuestions

Page 11: Movement Tracking in Real-time Hand Gesture Recognition

Reference Paper

Movement Tracking in Real-time Hand Gesture

RecognitionAuthored by

Hong-Min Zhu & Chi-Man PunDepartment of Computer and Information

ScienceUniversity of Macau, Macau SAR, China

{ma86560, cmpun} [at] umac.mo

9th IEEE/ACIS International Conference on Computer and Information Science

Page 12: Movement Tracking in Real-time Hand Gesture Recognition

Reference PaperWhat does it say?This paper deals with overcoming of SCHD technique for Hand Gesture Recognition using newly improved Algorithm, IFDHD

Page 13: Movement Tracking in Real-time Hand Gesture Recognition

Procedures in General Framework of Gesture Recognition

Page 14: Movement Tracking in Real-time Hand Gesture Recognition

Previous Work DoneTemporal Hand GestureAssumptions

Camera User Synchronization Uniform Lightening Condition Simple Background Features Frame Rate – Gesture Speed

Coordination

Page 15: Movement Tracking in Real-time Hand Gesture Recognition

S C H D Skin Color based Hand Detection

J. Kovac and P. Peer – Designed Skin Classifier

Rules Pixel is classified as a skin pixel if:

Value of Red > 95, Green > 40 and Blue > 20 & max{R, G, B} - min{R, G, B} > 15 & |R - G| > 15 and R > G and R > B

Page 16: Movement Tracking in Real-time Hand Gesture Recognition

Proposed Solution Problems with SCHD

Computationally Expensive Skin-like Object Ambiguity Illumination Parameters Skin Color Variation

Solution – Motivated from BSHD

IFDHDInter-Frame Difference based Hand Detection

Page 17: Movement Tracking in Real-time Hand Gesture Recognition

Proposed Solution

Hand DetectionModule

Motion TrackingModule

Page 18: Movement Tracking in Real-time Hand Gesture Recognition

Hand Detection Module

Figure 3.1 Zoomed Mode for Hand Detection Module

Page 19: Movement Tracking in Real-time Hand Gesture Recognition

Algorithm for Hand Detection

Input: Frames Fi = 1..N from video segmentSteps:

1. Convert frame F1 to grayscale2. Repeat (until end of video segment)

1. Convert frame Fi to grayscale2. Intensity difference image D0 = |Fi – F1|3. Binary image I = (D0 > T0)4. Do image opening on I followed by closing5. Splitting of Large regions into max size boundary box

as 60x806. Calculate center co-ordinate

Output: Center coordinate of each region in each frame

Page 20: Movement Tracking in Real-time Hand Gesture Recognition

Experimental ResultsSCHD based Hand DetectionLightening Condition IFDHD based Hand DetectionSCHD based Movement Tracking IFDHD based Movement TrackingEfficiency Measurement

Page 21: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 4. Result of SCHD(a) original frame

Page 22: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 4. Result of SCHD(b) Skin Pixel Classification

Page 23: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 4. Result of SCHD(c) De-noise & Region

Connection

Page 24: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 4. Result of SCHD(d) Region Splitting

Page 25: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 4. Result of SCHD(e) Centers of Each Region

Page 26: Movement Tracking in Real-time Hand Gesture Recognition

Experimental ResultsSCHD based Hand DetectionLightening Condition IFDHD based Hand DetectionSCHD based Movement Tracking IFDHD based Movement TrackingEfficiency Measurement

Page 27: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 5. Effect of Lightening Condition

(a) Original Frame

Page 28: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 5. Effect of Lightening Condition

(b) Skin Pixel Classification

Page 29: Movement Tracking in Real-time Hand Gesture Recognition

Experimental ResultsSCHD based Hand DetectionLightening Condition IFDHD based Hand DetectionSCHD based Movement Tracking IFDHD based Movement TrackingEfficiency Measurement

Page 30: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 6. Result of IFDHD

(a) 1st Frame

Page 31: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 6. Result of IFDHD

(b) 11th Frame

Page 32: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 6. Result of IFDHD

(c) Subtraction: (b) - (a)

Page 33: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 6. Result of IFDHD

(d) Threshold of (c)

Page 34: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 6. Result of IFDHD

(e) De-noise

Page 35: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 6. Result of IFDHD

(f) Region Splitting

Page 36: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 6. Result of IFDHD

(g) Centers of Regions

Page 37: Movement Tracking in Real-time Hand Gesture Recognition

System Domain (Contd.)

Page 38: Movement Tracking in Real-time Hand Gesture Recognition

Movement Tracking Module

Figure 3.2 Zoomed Mode for Movement Tracking Module

Page 39: Movement Tracking in Real-time Hand Gesture Recognition

Algorithm for Movement Tracking

Input: Region centers Detected in each frameSteps:

1. Initialize the start of frame2. Repeat (for each frame > 1)

1. Identify tail locations and store2. Calculate matrix of distances between centers and

tail locations3. Repeatedly select – min(Distance ( I ), Distance

( J )) 4. If Distance( I ) < Threshold then append Center to

Gesture and delete Distance( I ) Else initialize new Gesture start location

5. Select Gesture Frame that has the maximal standard deviation

6. Smooth movement track Gesture and interpolate it to Number of Center Coordinate falls coordinates

Output: Encoding of movement track

Page 40: Movement Tracking in Real-time Hand Gesture Recognition

Experimental ResultsSCHD based Hand DetectionLightening Condition IFDHD based Hand DetectionSCHD based Movement Tracking IFDHD based Movement TrackingEfficiency Measurement

Page 41: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 7. SCHD Based Movement TrackingFirst Row: Last Frame of Video Segment

Page 42: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 7. SCHD Based Movement TrackingSecond Row: Detected Digit Track

Page 43: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 7. SCHD Based Movement TrackingThird Row: Smoothed Track

Page 44: Movement Tracking in Real-time Hand Gesture Recognition

Experimental ResultsSCHD based Hand DetectionLightening Condition IFDHD based Hand DetectionSCHD based Movement Tracking IFDHD based Movement TrackingEfficiency Measurement

Page 45: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 8. IFDHD Based Movement TrackingFirst Row: Last Row of Video Segment

Page 46: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 8. IFDHD Based Movement TrackingSecond Row: Detected Digit Track

Page 47: Movement Tracking in Real-time Hand Gesture Recognition

Fig. 8. IFDHD Based Movement TrackingThird Row: Smoothed Track

Page 48: Movement Tracking in Real-time Hand Gesture Recognition

Experimental ResultsSCHD based Hand DetectionLightening Condition IFDHD based Hand DetectionSCHD based Movement Tracking IFDHD based Movement TrackingEfficiency Measurement

Page 49: Movement Tracking in Real-time Hand Gesture Recognition

Efficiency Measurement

Table 1. Comparing the Efficiencies of SCHD and IFDHD

Page 50: Movement Tracking in Real-time Hand Gesture Recognition

Outline IntroductionReferenced PaperApplication ScenarioConclusionQuestions

Page 51: Movement Tracking in Real-time Hand Gesture Recognition

Reference PaperAmerican Sign Language Recognition System for Hearing Impaired People Using Cartesian

Genetic ProgrammingAuthored ByFahad Ullah

Department of Computer Systems Engineering,University Of Engineering & Technology,

Peshawar, Pakistan

Proceeding of 5th International Conference on Automation, Robotics and Applications, New

Zealand

Page 52: Movement Tracking in Real-time Hand Gesture Recognition

Application Scenario

Why the interfaces are changing ?

How many Apps Out there? Have you tried X-box,PSP-2,Mac-OSX January 9, 2012, 66 million Xbox 360

consoles have been sold worldwide.

New era of Interfaces

Page 53: Movement Tracking in Real-time Hand Gesture Recognition

Application Scenario What if you can’t speak? ASL CGP (Cartesian Genetic Programming) How it works? Genetic programming an Overview:

Probabilistic search Darwinian principle of natural

selection Naturally occurring genetic operations

such as crossover and mutation.

Page 54: Movement Tracking in Real-time Hand Gesture Recognition

• Better individuals are preferred• Best is not always picked• Worst is not necessarily excluded• Nothing is guaranteed• Mixture of greedy exploitation and

adventurous exploration• Similarities to simulated annealing

(SA)

Probabilistic Selection Based On Fitness

Page 55: Movement Tracking in Real-time Hand Gesture Recognition

Workflow

Page 56: Movement Tracking in Real-time Hand Gesture Recognition
Page 57: Movement Tracking in Real-time Hand Gesture Recognition

ASL using CGP

26 English language alphabets are trained and Identified

The system uses 26 binary images representing the different alphabets

Mentioned system with a Dictionary correction ability in order to increase the overall accuracy of the system.

Page 58: Movement Tracking in Real-time Hand Gesture Recognition

Outline IntroductionReferenced PaperApplicationConclusionQuestions

Page 59: Movement Tracking in Real-time Hand Gesture Recognition

ConclusionProposed IFDHDServing Feature Extraction Stage

Overcoming the pitfalls of SCHD

Page 60: Movement Tracking in Real-time Hand Gesture Recognition

Outline IntroductionReferenced PaperApplicationConclusionQuestions

Page 61: Movement Tracking in Real-time Hand Gesture Recognition

Questions, IF ANY?

Q?Thank You