Kinematic Analysis of Depth Jump using Computer Vision

Abstract

Currently, the analysis of the kinematics of plyometric exercises, like the depth jump, is carried out indirectly and subjectively by coaches and professionals. However, this approach has a high degree of inaccuracy, compromising the identification of technical failures and increasing the risk of injury. This paper presents an algorithm to evaluate the kinematics of deep jumping. The methodology adopted is to use the combination of computer vision techniques. This approach was validated against the Kinovea software, and the results demonstrated high precision, achieving an average angular error of less than 1.0° for the knee, hip, and ankle joints. Additionally, performance metrics such as contact time and jump height showed satisfactory consistency, proving the tool's viability for assisting coaches and athletes in optimizing performance and preventing injuries.