Table of Contents
[ ii ]
Chapter 2: Hand Gesture Recognition Using a Kinect
Depth Sensor 25
Planning the app 27
Setting up the app 27
Accessing the Kinect 3D sensor 28
Running the app 29
The Kinect GUI 30
Tracking hand gestures in real time 31
Hand region segmentation 32
Finding the most prominent depth of the image center region 33
Applying morphological closing to smoothen the segmentation mask 35
Finding connected components in a segmentation mask 36
Hand shape analysis 38
Determining the contour of the segmented hand region 38
Finding the convex hull of a contour area 39
Finding the convexity defects of a convex hull 40
Hand gesture recognition 40
Distinguishing between different causes of convexity defects 41
Classifying hand gestures based on the number of extended ngers 42
Summary 45
Chapter 3: Finding Objects via Feature Matching
and Perspective Transforms 47
Tasks performed by the app 48
Planning the app 50
Setting up the app 51
Running the app 51
The FeatureMatching GUI 52
The process ow 53
Feature extraction 55
Feature detection 55
Detecting features in an image with SURF 56
Feature matching 57
Matching features across images with FLANN 58
The ratio test for outlier removal 58
Visualizing feature matches 59
Homography estimation 61
Warping the image 64
Feature tracking 65
Early outlier detection and rejection 66
