Motion Video Image Based on Sensor and Multiprocessing Technology in Track and Field Sports

Abstract

In order to use motion video image processing technology to better guide and evaluate athletes’ technical and tactical ability, this paper first constructed a motion video image acquisition system. Secondly, a fuzzy kernel estimation method for uniform linear motion based on cepstrum property is proposed. Then, the motion blur areas to be processed are selected, and only these areas are deblurred. Finally, the effective removal of local motion blur is realized, and a clear scene image is obtained. Then, the maximum value of the total entropy of the image is calculated by using the information entropy theory, and the particle swarm optimization algorithm is introduced to find the maximum threshold of image segmentation. Finally, small wave optical flow estimation algorithm and rectangular window scanning algorithm across scales of motion image target detection algorithm are proposed, to not only solve the traditional optical flow estimation for fast moving object detection accuracy but also improve the efficiency of the optical flow computation. Compared to many kinds of algorithms, this paper proposed an algorithm that can improve the accuracy of moving target detection and measurement accuracy. And, the detection accuracy of the proposed algorithm is up to 86.5%. The estimated accuracy was as high as 65%. The segmentation accuracy is up to 95%.