Cut scene change detection using spatio temporal video frame

Abstract

In this paper, we addressed the problem of cut or abrupt scene change detection in video sequence. Video is defined as a sequence of XY frame with flow direction T. Proposed method considered spatio-temporal frames TY and TX for detection. In XY frame, large pixel intensity variation occurred at the location of cut scene change and such phenomena produced a boundary line in spatio-temporal frames. Proposed boundary appeared as a vertical line for each scene change. Each scene categorizes as an uniform region and their transition produces a boundary line. These boundary has been detected with the help of edge detection method in both TY and TX frames simultaneously. A two way detection in TY and TX frames confirms the accuracy of detection algorithm and eliminates the false detection.