An object-based shot boundary detection using edge tracing and tracking

With the upcoming development of MPEG-7, research in shot boundary detection has become more popular. Traditionally, shot boundary has been detected by measuring the differences between the features of adjacent frames, and the threshold was constructed based on a trial and error basis, which usually achieved an accuracy of nearly 90%. However, the remaining errors were left unresolved. In this paper, we will tackle the problem by using object-based detection, which utilizes information from multiple frames through a time stamp transferring mechanism. Unlike classical techniques, our method can handle a gradual transition of any length. The concept of edge object tracking is developed to track moving objects as well as unbounded objects across the gradual transition scenes. We will also prove that object-based detection is not merely an algorithm that can detect shot boundary, but is superior to traditional indicators for detecting gradual transition. Such a technique works very well with frames containing skipped blocks and is independent of the transition period and the amount of disturbances within the sequence, which cannot be handled by indicators based on adjacent frame comparison. (C) 2001 Academic Press.