Low-complexity range tree for video synopsis system

This work proposes an efficient video retrieval technique for video synopsis. In a video system, the Region of Interest (ROI) should be extracted in a long video effectively such that users can browse it quickly and easily. Focusing on the characteristics of objects in the foreground of real-world video sequences, this work employs the Gaussian Mixture Model (GMM) and color-histograms for object detection. In order to reduce the search time, a new video synopsis search approach, a low-complexity range tree algorithm, is proposed to improve the effectiveness of searches for objects of interest matching pre-set conditions. With the time and space redundancy-reducing techniques of video synopsis, the objects of interest can be displayed within a short time. Objects and events can be found and displayed quickly without allocating time to watching non-ROIs. For the test video sequences, the results show an accuracy rate of 97 % and a processing speed of 32 FPS (frames per second) in the online phase, and the time complexity of object searching is reduced from O( N ) to O(log D -1 N ).