Infrared thermal image detection and facial expression recognition based on genetic algorithm in sports prediction simulation: Sports thermal modeling

Abstract

In the field of sports, accurate evaluation of athletes’ physiological and psychological states is the key to improve training efficiency and competition results. An infrared thermal image detection and facial expression recognition system based on genetic algorithm is developed for the prediction and simulation of sports. The thermal image data of athletes in training and competition were collected by infrared thermal imager. Then feature extraction and pattern recognition are carried out on these images using genetic algorithm to identify the key thermal image features reflecting the athlete’s state. The facial expression recognition part trains the model through deep learning techniques to recognize and analyze the athlete’s facial expressions to infer their emotional and psychological state. Finally, these data are integrated into a sports thermal model to predict athlete performance by simulating different training and competition scenarios. The experimental results show that the infrared thermal image detection system based on genetic algorithm can accurately identify the key physiological indicators of athletes, and the facial expression recognition system can successfully identify the emotional changes of athletes in different situations. By applying this data to sports thermal models, we are able to effectively simulate and predict athlete performance under specific training and competition conditions. The results of model prediction have high correlation with the actual performance, which verifies the effectiveness and practicability of the system.