Video Object Tracking by Integrating Background Constructing and Motion Estimation

We introduce an algorithm of semiautomatic video segmentation and tracking by integrating background constructing and motion estimation. Firstly, we extract the video object of the initial frame using the artificial way, then, build the background of the initial frame, finally, the video object tracking is completed based on background updating, color and motion information of the video object. Experimental results are compared with the results of other algorithm, and demonstrate the performance of our algorithm when the video object rapidly motions or is still.