论文部分内容阅读
采用动态跟踪方法实现自主移动机器人的特征跟踪和实时定位,首先利用Harris算子对t时刻与t+1时刻机器人周围场景的左右摄像机序列图像进行特征提取,并对提取出的图像特征点进行初步匹配,最后采用相邻帧间动态比较,获得左右相机前后序列的匹配特征,实现特征跟踪,进行机器人的实时定位。经过实验室内和实验室外环境的验证,该动态跟踪方法实时性较强、精度较高、有较好的定位能力,可以满足机器人的定位需求。
Firstly, the Harris operator is used to extract the features of the left and right camera sequence images of the surroundings of the robot at time t + 1 and the initial feature points of the extracted image are preliminarily obtained by using the dynamic tracking method to realize the feature tracking and real-time positioning of the autonomous mobile robot. Match, and finally use the dynamic comparisons between adjacent frames to obtain the matching features of the left and right cameras before and after the sequence, to achieve feature tracking, robot real-time positioning. After the laboratory and laboratory environment validation, the dynamic tracking method is more real-time, higher accuracy, better positioning ability to meet the robot’s positioning needs.