Performance Evaluation of Edge Based Motion Detection Algorithm for Differential Time Lapse Video in Real Time Application

The Time Lapse Technique is cinemography technique where by each film frame is captured at much slower rate than it will replay back. When replayed at normal rate, time appears to be moving faster and thus time lapsing. Time lapse technique is very powerful technique to document slowly changing/developing/moving phenomenon. The natural movements like sunrise, blooming of flowers periodic time lapse is used, while for uncertain activities like human activity, birds flying, animal movement etc aperiodic time lapse video is used. One of the classifications of aperiodic time lapse video is differential time lapse video. In differential time lapse video, each frame are compared with previously stored frame, if movement found in the frame than frame is stored and considered as last stored frame, else frame is discarded. The quality of differential time lapse video depends on the frame selection criterion. Edge based motion detection criterion is used here to overcome the limitation of pixel to pixel bit matching based motion detection. In this paper, MATLAB based algorithm is used to generate differential time lapse video with edge based motion detection is used as frame selection criterion.