In this paper, the initial parameters of C-V level set image recognition are optimized by using the global optimization characteristics of cultural algorithm, and a cultural algorithm C-V level set image recognition model is proposed, which is abbreviated as the CC-V model. The initial population space of the cultural algorithm is used to set the initial recognition parameters in a large range, and the population evolution is continuously optimized and guided by the situation knowledge and normative knowledge of the reliability space, so as to realize the global optimization of the recognition parameters and to timely terminate by judging the change of the image entropy fitness value. Through the analysis and comparison of the experimental results, the CC-V model has a better recognition effect than the C–V model. The partial differential equation image recognition model is applied to the video image sequence for moving target recognition. The background model is constructed by the block statistical histogram. The background difference method is used to locate the video moving target, and the minimum circumscribed rectangle of the multitarget positioning is used as the initial outline of the model recognition. The research results show that the nonintellectual factors of each group of ethnic traditional sports have an obvious effect on the application value of people, and the application effect of nonintellectual factors produced by different groups will also vary. All ethnic traditional sports groups show higher application value to nonintellectual character factors. The comprehensive application value of nonintelligence factors of the combat confrontation item group, accuracy item group, and endurance item group is relatively higher than the comprehensive application value of nonintelligence factors of other item groups.