基于机器视觉的作物对行喷药控制的研究

Study on spray control to aim toward crops rows based on machine vision

设计了一套基于机器视觉的作物对行喷药控制系统。首先由CCD摄像机拍摄成行作物图像,提取其色调值图像并用最大类间方差法二值化后,经数学形态学腐蚀后用Hough变换方法拟合作物行中心线,最后步进电机移动其喷头标识来对准此作物行。实验验证了该方法的正确性。

A row-following system based on machine vision was put forward. At first, the image of band crops captured by a CCD camera was passed on computer, then crops were segmented from background by obtaining H value and its binary image was obtained by means of OSTU. Finally, the center line of crops row was regressed by Hough transform after the binary image was morphologically erode, which was moved accurately by a nozzle mark through a step motor. Experimental results show that the method is effective.