A novel video shot boundary detection framework employing DCT and pattern matching

The video Shot Boundary Detection (SBD) is an elementary step in realising a system capability to perform content based video search, structural analysis, data retrieval and video summation. Myriad research works in the past have been reported to construct SBD algorithms. However, the need of an error-free, meticulous and cost-effective SBD technique still persists; for applications viz. apt management, storage, browsing, video indexing and retrieval of multimedia data. This paper is an effort in the same direction with the aim of achieving high execution speed and greater accuracy. The proposed SBD technique in this paper incorporates three steps: (i) Candidate Segment Selection (ii) Cut Transition detection (iii) Gradual Transition detection. This paper adopts pixel based technique with candidate segment selection to speed up the SBD. For Cut Transition detection, the proposed method employs Discrete Cosine Transform (DCT) and for Gradual Transition detection, it employs Image Histogram and Pattern Matching. The comparison of MATLAB simulation results of the proposed SBD technique with those in literature manifest better results in terms of execution speed and accuracy.