Research on Optimization of Multi-Camera Placement Based on Environment Model
Liming Tao, Renbo Xia, Jibin Zhao, Fangyuan Wang, Shengpeng Fu · 2023
Aiming at the model distortion, constraint redundancy, and lack of occlusion detection, an optimization algorithm for multi-camera placement based on environment model is proposed to minimize escape rate and measurement error. First, the measurement error model is established by the reverse modeling method, which reveals the relationship between the pixel quantization error and the measurement error. Second, the environment model is split into camera candidate position model, feature model and obstacle model. Based on the division of field of view and the signedness of tetrahedron volume, a method for judging the relationship between the point and the field of view is proposed by calculating the determinant. Thirdly, the Bresenham algorithm is extended to three-dimensional space for occlusion detection. Finally, a fitness function is built to minimize the escape rate and measurement error. The particle swarm optimization algorithm is used to optimize the multi-camera placement. Extensive experiments demonstrate that the proposed method is effective and robust.