基于图像边界提取的果园作业机器人自主导航系统

摘要

针对果园中农业移动机器人的自主导航问题，提出了一种基于图像边界提取的自主导航系统。首先，通过摄像机采集果园场景图像，并利用均值偏移算法对图像中的像素点进行聚类；然后，采用基于图论的图像分割算法，根据预定义类别对图像进行分割；接着，提取出所需类的图像区域，并利用 Canny 算子对该区域图像的边缘进行检测和滤波；最后，利用 Hough 变换提取出该区域的边界线，从而获得机器人的行进路线。实验结果表明：该方案在不需要先验知识的情况下，能够快速、有效地获得最优路径。%For the issue that the autonomous navigation of agricultural mobile robots in the orchard , An autonomous navi-gation system based on image boundary extraction is proposed .Firstly, the scene images of the orchard were collected by the camera , and the pixels in the image were clustered by mean shift algorithm .Then , the image segmentation algorithm based on graph theory is used to segment the image according to the predefined categories .After that , the image area of the desired class is extracted , and the edge of the region image is detected and filtered by the Canny operator .Finally , Hough transform is used to extract the boundary line of the region , so as to obtain the robot's travel route .Experimental results show that the proposed scheme can obtain the optimal route quickly and efficiently without need of prior knowledge .