Shot boundary detection algorithm based on color texture moments

Abstract

In this paper, a shot boundary detection method based on color texture moments is proposed. Here we apply the Local Feature Transform (LFT) on each frame of a video data for feature extraction purpose. Then the first and second moments are computed for each channel of the color space and the feature vector is formed. The dissimilarity measure higher than a predefined threshold indicates the presence of a shot cut. Experiments have been conducted on the standard TRECVID video database to evaluate the effectiveness of the proposed method. © 2011 Springer-Verlag Berlin Heidelberg.