手术机器人多视点光学定位遮挡动态补偿算法
This paper proposes a dynamic compensation algorithm to address the problem of reduced positioning accuracy caused by local occlusion in a multi-view optical positioning system. The algorithm monitors the three-dimensional data of the positioning points in real time to quickly detect occlusion and accurately locate the occluded area. During the occlusion period, the unaffected positioning point data is combined with the Kalman filter for triangulation, and a dynamic weight adjustment mechanism is introduced to assign weights according to the credibility of each positioning point to optimize the position estimation accuracy. After the occlusion is removed, the algorithm fuses all the positioning point data to correct the target position, and adjusts the prediction model parameters through error analysis to further optimize the performance. Experimental results show that the algorithm significantly improves the robustness and positioning accuracy of the system, and provides an effective solution for the application of multi-view optical positioning systems in complex occlusion environments.
本文提出了一种动态补偿算法，旨在解决多视角光学定位系统中因局部遮挡导致的定位精度下降问题。该算法实时监测定位点的三维数据，以快速检测遮挡现象并准确定位遮挡区域。在遮挡期间，算法将未受影响的定位点数据与卡尔曼滤波相结合进行三角测量；同时引入动态权重调整机制，根据各定位点的可信度分配权重，从而优化位置估计精度。遮挡消除后，算法融合所有定位点数据以修正目标位置，并通过误差分析调整预测模型参数，进一步优化系统性能。实验结果表明，该算法显著提升了系统的鲁棒性和定位精度，为多视角光学定位系统在复杂遮挡环境下的应用提供了有效的解决方案。