光学学报, 2014, 34 (4): 0415003, 网络出版: 2014-03-25
多摄像机空间交会动态测量中遮挡问题的研究
Study on the Occlusion Problem in Dynamic Space Intersection Measurement with Multi-Camera Systems
机器视觉 多像机测量网络 遮挡处理 空间交会测量 machine vision multi-camera measuring net occlusion processing space intersection
摘要
遮挡问题是现有多摄像机测量网络在进行空间交会测量时存在的一个重要问题。当测量过程中有遮挡发生时,某些特征点可能仅能被一个像机观测到,传统的前方交会测量方法将不再可行。为了解决这一问题,通过特征点的几何约束来求解特征点位在全局坐标系下的粗略位置作为进一步迭代优化的初值,并根据可用的特征点(可以是能被所有像机观测到,也可以是仅被部分像机观测到)的多光线约束来建立最优化目标函数,通过迭代算法最终解算得出被测目标可靠的位置姿态。实验结果表明,该方法在解决实际应用中遮挡问题的同时,也能提供较高的测量精度。
Abstract
Occlusion is a major problem in dynamic space intersection measurement with multi-camera systems. When occlusion occurs in measuring process, certain control points can only be observed by one camera, so the typical method of space intersection can′t be useable. To process the occlusion problem, the locations of the control points can be roughly determined, the optimal results can be obtained through iterative optimization algorithms, and the objective function can be established by multi-plane constraints of the available control points (the points are observed not only by all the cameras, but also by parts of the cameras). The experimental results show that the method can solve the occlusion problem well, and provide reliable measuring accuracy at the same time.
王伟, 于之靖, 郭寅, 杨学友. 多摄像机空间交会动态测量中遮挡问题的研究[J]. 光学学报, 2014, 34(4): 0415003. Wang Wei, Yu Zhijing, Guo Yin, Yang Xueyou. Study on the Occlusion Problem in Dynamic Space Intersection Measurement with Multi-Camera Systems[J]. Acta Optica Sinica, 2014, 34(4): 0415003.