An algorithm based on image processing technology was proposed for generating navigation directrix in orchard of complex circumstance. Image of R - B component was binarized by 2-D maximum between-cluster variance. A region of interest was detected by horizontal projection method to extract tree trunks. Crossing points between tree and ground were obtained by scanning the trunks and classified into different clusters in terms of their neighboring relation. Least square method was used for detecting two border lines, center points of crossing points between two border lines were extracted to generate navigation directrix for agricultural robot. The experimental results under various scene conditions showed that the algorithm could generate navigation directrix accurately, the accuracy was more than 90.7% which showed a good noise robust;the average processing time of a 640×480 pixels image was less than 119 ms and met the need of real-time.%针对果园导航环境的复杂性,提出了一种基于图像处理的果园导航基准线生成算法.采用二维Otsu算法获取最优分割阈值,对色差R-B分量图进行二值化处理;根据水平投影曲线的一阶导数变化规律,提取树的主干区域(ROI);根据区域中邻近像素灰度值的变化规律,提取主干与地面的交点作为特征点;根据点的分布特性对其进行归类,并利用最小二乘法拟合左右边界线,提取边界线上各行中心点生成果园导航基准线.在不同场景中的实验结果表明,该算法对噪声的干扰有较强的鲁棒性,使导航基准线的生成准确率高于90.7％;处理一幅640×480像素的图像平均耗时小于119ms,具有较好的实时性.
展开▼
