Extraction of spatial-temporal features for vision-based gesture recognition

One of the key problems in a vision-based gesture recognition system is the extraction of spatial-temporal features of gesturing. In this paper anapproach of motion-based segmentation is proposed to realize this task. The directmethod cooperated with the robust M-estimator to estimate the affine parametersof gesturing motion is used, and based on the dominant motion model the gesturingregion is extracted, i.e., the dominant object. So the spatial-temporal features ofgestures can be extracted. Finally, the dynamic time warping (DTW) method isdirectly used to perform matching of 12 control gestures (6 for "translation" orders,6 for "rotation" orders). A small demonstration system has been set up to verify themethod, in which a panorama image viewer can be controlled (set by mosaicing asequence of standard "Garden" images) with recognized gestures instead of the 3-Dmouse tool.