Research on Badminton Technique Movement Diagnosis and Instructional Feedback Optimization Based on Video Analysis

Accurate diagnosis of sports techniques is essential for improving athletic performance and reducing training inefficiencies. This study proposes a video-analysis-based framework for badminton technical movement diagnosis and instructional feedback optimization. A standardized video acquisition process is established to collect multi-view motion data, while pose-estimation algorithms are employed to extract spatiotemporal and kinematic features. A hybrid diagnostic model integrating threshold rules and machine-learning methods is developed to identify common technical deficiencies in high clear, smash, and net-shot movements. An eight-week teaching experiment involving 60 university students demonstrates that learners supported by the proposed system achieve significantly higher scores in movement standardization and technical execution than those receiving conventional instruction. The study validates the effectiveness of automated motion diagnosis and provides methodological references for computer vision, intelligent sensing, and human-motion analysis systems.