基于立体视觉的移动机器人障碍检测算法研究

针对室内导航环境的特点以及匹配精度和匹配算法的实时性问题,本文提出了一种基于区域分割的障碍检测方法。通过区域分割,把图像对应点的匹配问题简化为区域边界点的匹配,提高了匹配的精度。分割后的区域作为一个独立的对象进行处理,障碍物三维信息计算转化为视差的判断,提高了匹配的速度。该算法能够实现在室内环境下的实时障碍检测,不需要利用环境的先验知识,具有较强的自适应性。实验结果表明了该方法的有效性。

A Study on Obstacle Detecting Algorithm of Mobile Robot Based on Stereovision

Aiming at the features of indoor navigation environment,the paper presents an obstacle detection algorithm based on region segmentation.The kind of region can be handled as a independent object.The algorithm can realize real time obstacle detection in indoor environment without using the priori knowledge of environment,and it has strong suitability.The experimental results show that the algorithm is effective.