An algorithm to detect the moving object on MPEG-4 compressed streams is presented. The first step of the algorithm is to mark the moving regions based on the motion vectors
that have been filtered by cascade filter. Then a local object detecting is implemented by using DC coefficients of each marked block and calculating the difference between the central block and its neighbors'. Finally
the border is rectified by extending the objects' region and comparing border blocks' DC coefficients with the average background's value of the new region. Testing results show that the minimal decoding on compressed bit stream is performed and the detecting effect is dramatically improved. The algorithm performs well on different videos and keeps its robust character despite of other influence
such as light intensity.
关键词
Keywords
references
江华俊.视频图像运动特征的分析与提取[D].长春:吉林大学计算机学院,2004.
YU Xiaodong, DUAN Lingyu, TIAN Qi. Robust moving video object segmentation in the MPEG compressed domain [C]∥IEEE International Conference on Image Processing. Piscataway, USA: IEEE, 2003: 933-936.
AHMAD A M A, CHEN Duanyu, LEE Suh-Yin. Robust object detection using cascade filter in MPEG videos [C]∥Proceedings of the IEEE 5th International Symposium on Multimedia Software Engineering. Piscataway, USA: IEEE, 2003:196-203.
WANG Da-Jinn, CHEN Thou-Ho, LIAU Hung-Shiuan, et al. A DCT-based video object segmentation algorithm for rainy situation using change detection [C]∥Proceedings of the IEEE 1st International Conference on Innovative Computing, Information and Control. Piscataway, USA: IEEE, 2006:261-264.
ZENG Wei, GAO Wen, ZHAO Debin. Automatic moving object extraction in MPEG video [C]∥Proceedings of the IEEE International Symposium on Circuits and Systems. Piscataway, USA: IEEE, 2003:524-527.