Edge Strength Extraction using Orthogonal Vectors for Shot Boundary Detection

Content Based Video Retrieval mainly focuses on automatic indexing, retrieval and management of video data. The first step of the Content Based Video Retrieval is Shot Boundary Detection. In this paper, we propose a new method for detecting shot boundaries in video sequences by extracting edge strength using orthogonal vectors from blocks of the frames. Also, we propose transition (boundary) identification procedure which is capable of detecting the shot / non shot boundaries in the video sequences. The evaluation of the proposed method is performed by testing on TRECVid video test set containing various types of shot transition with lighting effects, object and camera movement within the shots. The results show that the proposed method yields better result compared to that of the existing shot boundary detection methods. Also, validation of the proposed method is performed using the performance indices like MSE, RMSE, R2.