Gaze tracking means to detect and follow the direction in which a person looks. This can be used in for instance human-computer interaction. Most existing systems illuminate the eye with IR-light, possibly damaging the eye. The motivation of this thesis is to develop a truly non-intrusive gaze tracking system, using only a digital camera, e.g. a web camera. The approach is to detect and track different facial features, using varying image analysis techniques. These features will serve as inputs to a neural net, which will be trained with a set of predetermined gaze tracking series…
Contents
1 INTRODUCTION
1.1 Background
1.2 Problem Description
1.3 Method
1.4 Thesis Outline
2 PREVIOUS WORK
2.1 Existing Gaze Tracking Systems
2.1.1 Infrared Light
2.1.2 Electro-Oculogram
2.1.3 Truly Non-Intrusive Gaze Trackers
2.2 Applications
2.2.1 Human Computer Interaction
2.2.2 Usability and Advertising Studies
2.2.3 Video Compression
2.2.4 More
3 FRAMEWORK
3.1 My Concept of Tracking Gaze
3.1.1 Setup Restrictions
3.1.2 Linear Regression
3.2 Interesting Facial Features
3.3 Provided Resources
4 HUMAN VISUAL SYSTEM
4.1 Functional Description
4.2 Eye Movements
5 ROTATIONAL SYMMETRIES
5.1 Introduction
5.2 Local Orientation in Double Angle Representation
5.3 2:nd Order Rotational Symmetries
6 DETECTION AND TRACKING OF FACIAL FEATURES
6.1 Color Space Transformation to Find the Face
6.1.1 Re-sampling and Low Pass Filtering
6.1.2 Color Spaces – RGB to YCrCb
6.1.3 Morphological Operations
6.1.4 Face Finder Implementation
6.2 Coarse to Fine Eye Detection
6.2.1 Rotational Symmetries
6.3 Detecting Corner of Eye
6.4 Detecting Nostrils
6.5 Geometrical Features
7 NEURAL NETWORK LEARNING
7.1 Introduction
7.1.1 Imitating the Brain
7.2 Multilayer Perceptrons
7.2.1 Perceptron
7.2.2 Two Layer Perceptron
7.3 Error Back-propagation Algorithm
7.3.1 Learning by Error-Correction
7.3.2 Back-propagation
7.3.3 Derivation of Delta Rule
7.3.4 Parameters
8 IMPLEMENTATION
8.1 Matrices in MATLAB
8.2 Training the Neural Net
8.3 and Testing it
8.3.1 Accuracy
9 RESULT AND DISCUSSION
9.1 Facial Feature Detection
9.1.1 Face Finder
9.1.2 Eye Detection
9.1.3 Detecting Corner of Eye
9.1.4 In Search for Nostrils
9.1.5 The Geometrical Features
9.2 Back-Propagation
9.2.1 Learning Parameters
9.3 Gaze Tracking
9.3.1 Accuracy
9.4 The Features Effect on the Outcome
9.4.1 Correlation
9.4.2 Principal Component Analysis
9.4.3 Salience
10 CONCLUSION AND FUTURE WORK
10.1 Conclusion
10.2 Future work
Author: Bäck, David
Source: Linköping University
Download URL 2: Visit Now