Video sequence processing based on spatiotemporal independent component analysis

Compression and content-based video retrieval (CBVR) are essential needs for efficient and intelligent utilizations of vast multimedia databases over the Internet. In video sequences, object based extraction techniques are gaining importance in achieving compression and performing content-based video retrieval. In this paper, we propose a novel technique to extract objects from video sequences based on spatiotemporal independent component analysis (stICA), -wavelet analysis and edge detection. The frame data obtained after stICA analysis is further processed using wavelet, and edge detection techniques to improve the accuracy of the extracted object. Initial studies demonstrate high potential for stICA based object extraction technique in comparison with existing techniques. Nine frames captured from a video sequence are used in our experiments and high extraction accuracy was achieved.