摘要
根据温室行间植物的位置特征,应用空域中值滤波对图像进行预处理,然后运用色度法和最大方差自动取阈值法对图像处理,最后应用种子填充法将杂草和作物进行分割。结果表明,机器视觉技术在对温室杂草的识别方面具有一定的优越性。
Acconding to the position feature of greenhouse plant, firstly, airspace median filter is applied to the image preprocessing. Secondly, the colorimetric method and the maximum variance auto threshold method are use to image pro- cessing. Finally, the seed filling algorithm is applied to segment crops and weeds. It is indicated that the machine vision has the superiority in recognition of greenhouse weeds.
出处
《信阳农业高等专科学校学报》
2012年第2期115-117,共3页
Journal of Xinyang Agricultural College
基金
信阳农业高等专科学校青年教师基金项目[201101003]
关键词
机器视觉
杂草
识别
算法
machine vision
weed
detection
algorithm
