Tomograph based concept detection of video data using PSO

Concept detection has always been an important area of research in image and video processing. Concept detection mainly is improved by precising detection results by using range of methods available. One such method is proposed in this work by using spatio-temporal video slices or called as video tomographs. The video tomographs can be a replacement for conventional key frames. Video tomographs are capable of capturing motion information which was not the case with the previous methods use. The another key novelty of this work is the basic methodology that is assembled upon a particle swarm optimization technique, for purpose of calculating subset of the existing base detectors should be united for developing a finest detector for every concept.