<p>The aim of this study is to design a motion posture recognition and evaluation system based on motion sensing technology and machine learning, in order to achieve real-time monitoring and objective evaluation of athletes’ postures. A requirement analysis was conducted to determine the functional modules and overall architecture of the system. The system uses motion sensors to collect real-time motion data of athletes. In the data preprocessing stage, the collected signals are denoised and standardized, and motion postures are recognized and analyzed through machine learning algorithms. An athlete motion evaluation model is established, which mainly relies on bone keypoints and feature extraction methods, as well as optimization strategies for the algorithm model. After system testing, the results show that the system has high accuracy and robustness in recognizing and evaluating motion postures. Through in-depth analysis of data, personalized suggestions for optimizing exercise posture can be provided to athletes, which can help improve athletic performance and reduce the risk of injury, and effectively support the scientific and intelligent training of sports.</p>