Video Shot Boundary Detection: An efficient and an accurate approach based on Gabor Moments

In this paper, video shot boundary detection method based on Gabor moments is proposed. Each video frame is convolved with six 2-D Gabor filters corresponding to six different orientations, and the first and second moments of these convolved images are obtained, generating twelve dimensional feature vector. The frame dissimilarity values are compared against an adaptive threshold to localize the shot. Experimentation on the TRECVID video segments have been done to demonstrate the performance of the proposed approach for both cut and fade detection. The comparative study with other approaches is included to exhibit the superiority of the proposed approach in terms of efficiency and accuracy of shot detection.