Implementation and Comparative Analysis of Semi-Automated Surveillance Algorithms in Real Time using Fast-NCC

Chaotic environment, irregular motion of objects creates challenging environment in the field of computer vision. Advance target tracking techniques are used to overcome these problems, but few parameters are considered ideal in those scenarios, or those parameters are ignored. In this research, cross correlation technique is applied for target tracking which is famous for feature extraction in image processing. Further, normalized cross correlation and fast-normalized cross correlation are implemented and results are compared. As enormous computation is required in these techniques, real time target tracking is factual challenge faced by this technique. High performance embedded hardware is required to implement these techniques. In this research, digital signal processor embedded board TMS320DM642 is carefully chosen for this purpose. This board is equipped with evaluation module and video decoder. Above mentioned techniques are implemented on TMS320DM642 and their results are carefully analyzed in this research.