Tracking object in video pictures based on background subtraction and image matching

Two challenge problems in tracking objects in video are object segmentation and matching. Most previous methods can work with some user-specific control. In this paper, we will propose an adaptive algorithm for colour object tracking in video sequences based on background subtraction and image matching by using multiresolution critical point filter (CPFs). Which the background subtraction segment the blob and identify all the objects in the image no matter what they are moving or not. Using results from segmentation phase, object tracking is performed using the multiresolution critical point filter algorithm. This is based on research from prof. Shinagawa's image interpolation. For now, our approach is suitable for tracking of the object through a sequence of 64*64 images.