Vehicle Detection and Tracking Using the Block Matching Algorithm

The paper describes an approach to vehicle detection and tracking fully based on the Block Matching Algorithm (BMA), which is the motion estimation algorithm employed in the MPEG compression standard. BMA partitions the current frame in small, fixed size blocks and matches them in the previous frame in order to estimate blocks displacement (referred to as motion vectors) between two successive frames. The detection and tracking approach is as follows. BMA provides motion vectors, which are then regularised using a Vector Median Filter. After the regularisation step, motion vectors are grouped based on their adjacency and similarity, and a set of vehicles is built per singular frame. Finally, the tracking algorithm establishes the correspondences between the vehicles detected in each frames of the sequence, allowing the estimation of their trajectories as well as the detection of new entries and exits. The tracking algorithm is strongly based on the BMA. We consider the BMA output as the basic tracking information associated with each block and combine this already available block-level tracking with the grouping output so as to achieve the tracking of vehicles. Key-Words: Computer vision, traffic monitoring, vehicle tracking