Detection of Jump Landing Performance of Basketball Player Using Image Processing and Support Vector Machine

This study is dedicated to developing an analytical system aimed at evaluating the jump-landing technique employed by basketball players. This innovative system relies on applying the Support Vector Machine (SVM) image processing algorithm in conjunction with the Kinect sensor to recognize and assess the landing postures of the players. The primary objective of this research is to establish a detection system that can provide precise assessments of landing techniques by comparing them with evaluations made by experienced basketball trainers or coaches. The study mainly concentrates on jump landings in basketball, specifically on maneuvers like layups and rebound jumps. The research outcomes have demonstrated a robust correlation of 95.07 percent between the evaluations generated by the program and those provided by the coach. Rigorous statistical analysis has validated the accuracy of these assessments, which were gathered exclusively from the coaching staff of a single professional basketball team. The resulting system offers valuable insights for enhancing the jump-landing technique of basketball players, making notable contributions to injury prevention and overall performance improvement in the sport.