基于分级关键帧筛选的RGB-D SLAM算法研究

Research on RGB-D SLAM algorithm based on hierarchical keyframe filtering

利用关键帧求解SLAM算法(simultaneous localization and mapping)能够提高SLAM系统的实时性与精确度。针对现存关键帧筛选算法中存在的计算复杂度高、图像帧冗余以及鲁棒性较差等问题,提出一种分级关键帧筛选方法。该算法考虑了SLAM系统在不同运行阶段时对关键帧的要求,首先结合旋转度指数与地图点跟踪筛选出一级关键帧用于后端优化与回环检测,再利用相邻帧在空间上的相对运动距离筛选出二级关键帧用于三维地图构建,最后,实现了基于此二级筛选算法的RGB-D SLAM系统。实验表明,一级关键帧算法能提高SLAM系统的定位和建图精度,二级关键帧算法则有效减少了数据冗余,提高了建图效率。

Using keyframe group to solve the SLAM algorithm can improve the real-time performance and accuracy of the SLAM system.Due to the high computational complexity,data redundancy and low robustness in the existing keyframe filtering algorithm,this paper proposed a hierarchical filtering method.The method combined rotation index and map-points tracking to select first-group keyframes for back-end optimization and loop-detection,and then used the relative motion distance to select second-group keyframes for 3D reconstruction.Experiments demonstrate that using first-group keyframe selection algorithm can improve the localization and mapping accuracy,and using second-group keyframe selection algorithm can effectively reduce redundant frames and increase 3D reconstruction efficiency.