Classification Method and Simulation Teaching Design for the Action Database Retrieval of Aerobics

In order to increase the accuracy and efficiency of calculation during the classification retrieval process in the calisthenics movement database, and relive the working pressure of calisthenics teachers, this paper proposes a classification retrieval algorithm for calisthenics movement database based on fuzzy correlation clustering in cluster classification mapping. First of all, it utilizes five modules including assisted introduction, information inquiry, system setting, information input and output, and information maintenance, to provide the design approach for the function modules in calisthenics teaching system; secondly, it puts forward a fuzzy correlation clustering algorithm based on cluster classification mapping, regarding cluster as a Gaussian spot which can be described by size, mean value and labels. What’s more, it optimizes the program by least square method and produces the output mapping of calisthenics movements by matrix multiplication. In the end, it verifies the efficiency of the approach being proposed by the simulation experiment in calisthenics movement database of Fudan University.