A Novel Metric For Efficient Video Shot Boundary Detection

In recent years, with the current state of multimedia and network technology, large amount of digital video information are being generated each day, which is a unprecedented challenge to video retrieval and analysis. Traditional video retrieval methods need to annotate videos manually, and then retrieval annotation to achieve the goal of video retrieval. But for the rapid growth of video data, the workload of manual annotation is clearly too large and has become almost impossible task. In order to make better use of video data, content-based video retrieval technology is proposed. The technology automatically extract video feature and complete video classification and retrieval rapidly. Video shot boundary detection is usually the first step of content-based retrieval of video to segment a video clip into meaningful shots, get ready for keyframe extraction.We focus on video shot boundary detection in this paper. Taking the processing speed into account, most state-of-the-art methods make use of the frame histogram to extract shot boundary characteristics. In this paper we propose a different approach with a novel metric, which essentially captures the observation that within any shot, a pixel value in any frame usually has a pixel value very close to it within a small neighborhood in an adjacent frame. It turns out that the proposed approach can make better use of frame structural content than the histogram approach. In addition, the proposed metric has a low computational complexity. We propose a video shot boundary detection algorithm based on the proposed metric for detecting both cut transition (CT) and gradual transition (GT) boundary.Experimental results show that the proposed approach enjoys better detection rates over the state-of-the-art with competitive processing speed.