PROJECT TITLE :

Parallel algorithm implementation for multi-object tracking and surveillance

ABSTRACT:

A recently developed sparse illustration algorithm, has been proved to be helpful for multi-object tracking and this study is a proposal for developing its parallelisation. An online dictionary learning is employed for object recognition. After detection, each moving object is represented by a descriptor containing its appearance options and its position feature. Any detected object is classed and indexed consistent with the sparse answer obtained by an orthogonal matching pursuit (OMP) algorithm. For a real-time tracking, the visual info wants to be processed terribly fast while not reducing the results accuracy. However, each the big size of the descriptor and the expansion of the dictionary after each detection, curtail the system process. In this work, a unique accelerating OMP algorithm implementation on a graphics processing unit is proposed. Experimental results demonstrate the efficiency of the parallel implementation of the used algorithm by considerably reducing the computation time.