Hand Gesture Extraction by Active Shape Models

The paper applied active statistical model for hand gesture extraction and recognition. After the hand contours are found out by a real-time segmenting and tracking system, a set of feature points (Landmarks) are marked out automatically and manually along the contour. A set of feature vectors will be normalized and aligned and then trained by Principle Component Analysis (PCA). Mean shape, eigen-values and eigenvectors are computed out and composed of active shape model. When the model parameter is adjusted continually, various shape contours are generated to match the hand edges extracted from the original images. The gesture is finally recognized after well matching.