<title>Detection of gradual scene changes for parsing of video data</title>

Abstract

The automatic video parser, a necessary tool for the development and maintenance of a video library, must accurately detect video scene changes so that the resulting video clips can be indexed in some fashion and stored in a video database. With the current existing algorithms, abrupt scene changes are detected fairly well; however, gradual scene changes, including fade-ins, fade-outs, and dissolves, are often missed. In this paper, we propose a new gradual scene change detection algorithm. In particular, we focus on fade-ins, fade-outs, and dissolves. The proposed algorithm is based on the chromatic video edit model. The video edit model indicates that, for sequences without motion, the second partial derivative with respect to time is zero during fade-ins, fade-outs, and dissolves. However, it is also zero for static scenes. Thus, the proposed algorithm computes the first (to disregard static scenes) and second partial derivatives, and if the norm of the second derivative is "small" relative to the norm of the first derivative, the algorithm declares a gradual scene change. The efficacy of our algorithm is demonstrated using a number of video clips and some performance comparisons are made with other existing approaches.

Scene-Based Shot Change Detection and Comparative Evaluation

“…Chromatic edit model [18]. Finally, in the chromatic edit model approach, the first partial derivative of the intensity with respect to time is computed.…”

“…These four algorithms were: (1) the plateau detection algorithm of Yeo and Liu [20]; (2) the variance curve approach of Meng et al [13]; (3) the twin comparison approach of Zhang et al [21]; and (4) the chromatic edit model approach of Song et al [18]. The variance curve approach and the chromatic edit model approach were reported with poor recall performance, whereas the twin comparison approach and the plateau detection approach scored high recall but low precision.…”

Shot-boundary detection: unraveled and resolved?

“…In [24], a chromatic video edit model for gradual transitions is built based on the assumption that discontinuity values belonging to such a transition form a pattern consisting of two piece-wise linear functions of time, one decreasing and one increasing. Such linearity does not apply outside the transition area.…”

“…Therefore, we compute the function as the relative change of frame variance in the middle of the sliding window, i.e., at the frame compared to frames close to window edges. Analytically, this can be written as in (24), shown at the bottom of the next page. Here, is the variance of the frame in the middle of the window with , while and are the variances of frames at both window edges.…”

Statistical sequential analysis for real-time video scene change detection on compressed multimedia bitstream

“…There have been a number of video segmentation approaches that appeared in the literature [2]- [6], which operate on the decompressed data, in the pixel domain. They may use methods based on color histogram differences [2], changes in the edge characteristics in the image [3], characteristic patterns in the standard deviation of pixel intensities for detection of fades [5], or linear modeling of transitions in gradual changes [4]. The color histogram-based methods are the most reliable for detection of abrupt changes.…”