一种新的闭环校正策略研究

Research on a New Loop-closed Correction Strategy

视觉SLAM中闭环校正能有效减小相机位姿估计的累积误差。针对此问题,本文提出了一种新的闭环校正策略。为了验证算法的有效性,本文采用公开数据集KITTI来测试本文算法。实验结果表明本文算法能有效的减小累积误差。

Loop closed correction in visual SLAM can effectively reduce the cumulative error of camera pose estimation.To solve this problem,we propose a new closed-loop correction strategy.To verify the effectiveness of the algorithm,we use the public data set KITTI to test the algorithm.Experimental results show that the proposed algorithm can effectively reduce the cumulative error.