ビデオカメラとLiDARを用いたバレーボール選手の位置の取得に関する一考察

抄録

In the field of sports, 3D sensing technology, which uses video cameras and sensors to acquire the three-dimensional movements of athletes in a time series, has been attracting attention. However, when acquiring 3D positions with a video camera, it is necessary to capture the same object from multiple viewpoints. Therefore, we compared the coordinate positions of single view point cloud data acquired by inexpensive LiDAR with the 3D position of an athlete from images generated by a video camera, and confirmed that the error in the 3D position of the athlete was small. As a result, we confirmed that there was a low error in the athlete's positional accuracy. This demonstrates the feasibility of athlete position estimation from a single viewpoint and from a long distance, and also clarifies the challenges of using LiDAR and a camera.