Video Motion Segmentation Based on Double Sliding Window
This paper puts forward a motion segmentation method based on double sliding window. In order to learn about the motions of a person in a certain period of time and do further research on the video segmentation, we should do some research on the motions people have every day. So the new method to segment video motions is presented. The new algorithm is to segment the different motions of the human body in the videos. Considering that the frequency of two different kinds of motion is also different, so in order to segment the motions, we mainly segment them according to the unique frequencies of different movements. Finally, this method has smaller computation and is easy to be implemented. Experimental results demonstrate the performance and feasibility of the new method.
double sliding window video motion segmentation waveform human motions
Jiaojiao Wu Bo Yin Wenjuan Qi
College of Information Science and Engineering Ocean University of China Qingdao, China
国际会议
杭州
英文
232-235
2011-10-28（万方平台首次上网日期，不代表论文的发表时间）