Intelligent Action Recognition and Dance Motion Optimization Based on Multi-Threshold Image Segmentation

Abstract

In order to improve the recognition and optimization effect of dance movements, this paper combines multi-threshold image segmentation technology to perform intelligent recognition of dance movements. Moreover, this paper analyzes the window focus calibration process in the cascade multi-threshold expansion method and focuses the window on the object body to expand from small to large. Then, due to the local optimum problem of single-layer expansion, this paper proposes a cascade expansion method, and comparative experiments analyze the effect of this method on window optimization. In addition, this paper combines the actual situation of dance movement training to construct an intelligent action recognition and dance movement optimization system based on multi-threshold image segmentation. The research shows that the intelligent action recognition and dance motion optimization method based on multi-threshold image segmentation proposed in this paper has a good effect.