基于ARMCortexA8平台喷药机器人路径检测与仿真.docVIP

基于ARMCortexA8平台喷药机器人路径检测与仿真 　　摘 要： 为了实现机器视觉在农业中的应用，研究了喷药机器人在植物道路中进行识别及路径检测的方法。以植物栽培中的道路为研究对象，采用HSV空间的S分量来实现道路图像灰度化，用Otsu算法进行二值化处理，将道路从绿色植物背景中分离出来，采用形态学滤波器对图像进行滤波操作。由于图像视野尽头道路弯曲的不确定性和道路尽头树木干扰等问题，提出将图像的前1/4行以黑色填充的方法和算法，使图像处理时减少不确定因素的干扰，提高导航线提取精度，最后利用霍夫变换得出喷药机器人在植物道路中的行走路径。结果表明，喷药农业机器人在植物道路中可准确识别到行走路径；道路识别时，处理一副像素450×310的图像平均耗时0.053 s；路径检测中，处理一副像素450×310的图像平均耗时0.062 s，可快速采集并完成图像处理，满足农业机器人导航的实时性要求。 　　关键词： 喷药机器人； 图像处理； 道路识别； 路径检测； 形态学滤波器 　　中图分类号： TN830.1?34； TP391.9 文献标识码： A 文章编号： 1004?373X（2017）22?0156?03 　　Abstract： In order to realize the application of machine vision in agriculture， the methods of spraying robot’s road recognition and path detection in plants are studied in this paper. Taking the road in plants as the research object， the component S of HSV space is adopted to realize the graying of road image. The graying image collected is binarized with Otsu algorithm， the road is separated from the green plant background， and then the image is filtered by morphological filter. Because of the problems such as the uncertainty of the curved road at the end of the image vision and the interference of the trees at the end of the road， the method that the first 1/4 lines of the image are filled with black is proposed to reduce the interference of the uncertain factors in the image processing and improve the extraction precision of the navigation line. The spraying robot′s walking path in plants is obtained with Hough transformation. The results show that the spraying robot can accurately identify the walking path in plants； when identifying the road， it takes the average time of 0.053 s to process an image with 430 × 310 pixel； in the path detection， it takes the average time of 0.062 s to process an image with 430 × 310 pixel. That means it can quickly capture and complete the image processing， and meet the requirement of real?time navigation for agricultural robots. 　　Keywords： spraying robot； image processing； road recognition； path detection； morphological fi