Motion Matching Based on Pose Similarity Using Movement of Body Parts
Ryohei Osawa, Shohei Adachi, Hiroshi K. WATANABE · 2022 IEEE 11th Global Conference on Consumer Electronics (GCCE) · 2022
The use of high-performance equipment to analyze motion has become popular in the field of sports. However, high-performance systems are expensive and difficult to install for amateur players and teams. Therefore, a method of analyzing motion based on video captured by smartphones and cameras is attracting attention as an easy-to-introduce system. One way for amateur athletes to practice efficiently is to compare their motion videos with the expert’s one. However, since different players operate at different speeds, it is difficult to make an accurate comparison by simply playing back the two videos at the same time. Therefore, we propose a method to synchronize the timing of similar motion in two videos with high accuracy. To improve the accuracy of the mapping, we corrected the right and left inversion detection of the skeleton and considered the amount of time variation of the skeleton. Through evaluation experiments, we confirm that the proposed method is effective in improving the performance of motion matching.