基于立体视觉的移动机器人楼梯识别

本文提出了一种结构化环境下,基于立体视觉的机器人楼梯识别算法,并将算法该应到自主移动机器人上。该算法首先利用二维图像分析的方法搜索楼梯的疑似区域;进而利用立体视觉对各个疑似区域进行精确三维重建,结合三维信息重构楼梯平面,排除虚假疑似楼梯区域;最后判定机器人和楼梯的相对位姿关系,引导机器人爬楼梯。最终我们将该算法应用到了自主移动机器人上,通过在各种光照条件下的实验,进一步验证了该算法的准确性和快速性。

This paper proposes a stairway detection algorithm under the structured environment based on stereo vision,and this algorithm will be used for the Unmanned Ground Vehicle (UGV).The algorithm first is used two-dimensional image analysis method to search suspected staircase regions;further the stereo vision is applied to reconstruction the suspected region accurately.With 3-D information remodeling staircase plane,eliminate false suspected staircase regional ultimately determine robot and staircase relative position and attitude,and in stair climbing robot guidance.We eventually applied the algorithm to the UGV in order to verify the accuracy and speed of our method under various light conditions.