基于SIFT算法的移动机器人同时定位与地图创建

研究了基于尺度不变特征变换（SIFT）算法的移动机器人同时定位与地图创建（SLAM）方法,即在视角改变情况下,用SIFT算法对不同图像进行特征匹配,根据极线几何原理得到摄像头的旋转角度,将之与里程计的角度信息融合,从而实现较准确的自我定位与地图创建.实验表明：本方法利用电荷耦合器件（CCD）摄像头和里程计间内在的几何关系来实现SLAM,提高了低成本移动机器人的定位精度.

Mobile Robot Simultaneous Positioning and Mapping Based on SIFT Algorithm

The simultaneous localization and mapping （SLAM） method is investigated based on the Scale Invariant Feature Transform （SIFT） algorithm. In the algorithm the features from different images is matched with the variant view points. By combining the angle data from odometer and the rotate angle derived from epipolar geometry, the SLAM is thus implemented with higher precision. The algorithm mainly takes into consideration the geometry correlation between Charge Coupled Device（CCD） camera and odometer, and as a result the superior SLAM positioning accuracy is achieved.