基于改进关键帧选择的RGB-D SLAM算法
发布时间：2018-08-31 18:38
【摘要】：关键帧选择是提高视觉SLAM(simultaneous localization and mapping)算法精度及实时性的重要因素.关键帧常以图像的帧间相对运动距离为选择依据.该方法虽简单有效,但实时性、鲁棒性较差且容易产生大量冗余关键帧.针对上述问题,提出一种改进的关键帧选择算法.该算法整合了帧间相对运动距离、帧间特征点跟踪以及最小视觉变化来选择关键帧并删除冗余关键帧.基于该算法,结合具有较好方向和光照不变性的ORB(oriented FAST and rotated BRIEF)特征,实现了RGB-D SLAM算法.在RGB-D数据集上的实验表明,改进的关键帧选择算法能够更精准、及时地选择关键帧,并在减少RGB-D SLAM中冗余关键帧的同时提高算法的实时性、建图和定位精度.
[Abstract]:Key frame selection is an important factor to improve the accuracy and real time of visual SLAM (simultaneous localization and mapping) algorithm. Key frames are often selected on the basis of the relative motion distance between frames. This method is simple and effective, but real-time, robust and easy to produce a large number of redundant key frames. To solve the above problems, an improved key-frame selection algorithm is proposed. The algorithm integrates the relative motion distance between frames, the tracking of feature points between frames and the minimum visual change to select key frames and delete redundant key frames. Based on this algorithm, the RGB-D SLAM algorithm is implemented by combining the ORB (oriented FAST and rotated BRIEF) features with good orientation and illumination invariance. Experiments on RGB-D dataset show that the improved key-frame selection algorithm can select key-frame more accurately and timely, and improve the real-time, mapping and positioning accuracy of the algorithm while reducing redundant key-frames in RGB-D SLAM.
【作者单位】： 杭州电子科技大学自动化学院;
【基金】：国家自然科学基金资助项目(61175093)
【分类号】：TP391.41
，
本文编号：2215831
[Abstract]:Key frame selection is an important factor to improve the accuracy and real time of visual SLAM (simultaneous localization and mapping) algorithm. Key frames are often selected on the basis of the relative motion distance between frames. This method is simple and effective, but real-time, robust and easy to produce a large number of redundant key frames. To solve the above problems, an improved key-frame selection algorithm is proposed. The algorithm integrates the relative motion distance between frames, the tracking of feature points between frames and the minimum visual change to select key frames and delete redundant key frames. Based on this algorithm, the RGB-D SLAM algorithm is implemented by combining the ORB (oriented FAST and rotated BRIEF) features with good orientation and illumination invariance. Experiments on RGB-D dataset show that the improved key-frame selection algorithm can select key-frame more accurately and timely, and improve the real-time, mapping and positioning accuracy of the algorithm while reducing redundant key-frames in RGB-D SLAM.
【作者单位】： 杭州电子科技大学自动化学院;
【基金】：国家自然科学基金资助项目(61175093)
【分类号】：TP391.41
，
本文编号：2215831
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2215831.html
