Application of sports video image analysis based on fuzzy clustering algorithm
Since the current sports video image segmentation methods have the defects of rough image segmentation result and high space distortion rate,a sports video image segmentation method based on fuzzy clustering algorithm is proposed. The basic theory of fuzzy clustering algorithm is introduced. The two-order fuzzy attribute containing the normal distribution and gray value is established by virtue of the time-domain difference image to give the fuzzy attribute S-membership function,so as to per-form the fuzzy clustering to the time-domain difference image. The sports video image segmentation result is obtained with edge detection. The experimental results show that the proposed method has high space accuracy,perfect noise iteration performance and low space distortion rate,and can segment the complex sports video image accurately to obtain the high-definition image.
国家科技期刊平台
NSTL
万方数据
维普
