Motion estimation in video sequences by edge detection

A fast approach to detecting and estimating multiple object motion in video sequences is presented. The method is based on transforming each image in the sequence of frames into vector projections formed by accumulating a pixel shift value along the rows and columns of the image. The projections corresponding to consecutive frames are in turn used to detect and estimate the individual horizontal and vertical components of the possible shifts. For that reason, block matching and maximum absolute difference (MAD) algorithms are used in detecting and tracking the moving objects in sequences. Sharp transitions or short edges in each image frame are also considered as an efficient method in the algorithm for decreasing the computational cost. Experimental results show that the usefulness of the method is presented not only in the context of motion analysis and ability to track objects, but also for a decrease in the processing time required for detection and tracking.