摘要
对利用植物的位置来识别作物苗期田间杂草的方法进行了研究。根据苗期田间植物的位置特征 ,建立了基于机器视觉的分割苗期田间杂草的算法 DBW。通过比较分析各种算法的分割效果图和所耗费的时间 ,运用超绿色法灰度化原始图像 ,然后应用最大方差自动取阈法二值化图像 ,最后运用种子填充算法分割作物和杂草。研究表明 ,算法 DBW在实时性方面表现出一定的优越性 ,处理一幅 5 4 4× 117像素的图像只需大约 6 0
A method for detecting weeds using features of plant location at seedling had been studied. According to features of plant location at seedling, the DBW algorithm based on machine vision to segment weeds at seedling is found. Firstly, the extra-green algorithm is adopted to gray the source image according to the segmentation results and the expended time of those algorithms. Secondly, the algorithm of the threshold automatically extracted according to the maximum deviation is used to transform gray image to binary image. Finally, the seed-fill algorithm is applied to segment crop and weeds. It is indicated that the DBW algorithm has the superiority in real-time. It spent about 60 ms to process a 544×117 pixels image using the DBW algorithm.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2005年第1期83-86,共4页
Transactions of the Chinese Society for Agricultural Machinery
基金
国家"8 63"高技术研究发展计划资助项目 (项目编号 :2 0 0 1AA2 45 0 12 )