Content Representation of Video Sequences Based on Salient Region

In order to provide more efficient content-based functionalities for video applications, it is necessary to extract meaningful regions from scenes as perceptual oriented representation of video content. We present a novel approach for salient region extraction, and simple visual features bound to salient regions will better represent the video content in perceptual manner. Since perceptual saliency for visual information is a subjective concept, a class-related fuzzy information granulation is constructed for mapping original feature space to concepts space. To detect salient regions, segmented homogenous regions are classified according to their salient importance. The experimental results using different types of video sequences are presented to demonstrate the efficiency and accuracy of our proposed algorithm.