摘要
对水稻冠层图像中背景分割的方法进行了研究。应用杂草识别中常用的灰度化方法(超绿色法)进行实验,并提出了灰度化方法,然后运用迭代法获得最佳阈值将灰度图像转化为二值图像,最后对二值图像进行形态学腐蚀、膨胀的处理。并通过实验验证了该方法的优越性和可行性。
In this paper,the background segmentation method of rice used in weed identification to perform experiments.At the same transform gray image into binary image,use iterative method to dilation is used to the binary image.Then superiority and feasibi canopy is studied.Use the gray method（ExG） that is usually time,puts forward a new gray method.Afterwards,in order to obtain the optimal threshold.At last,morphological erode and ity of this method is verified by experiments.
出处
《工业控制计算机》
2013年第2期75-76,共2页
Industrial Control Computer
关键词
水稻冠层
计算机视觉
颜色特征
阈值分割
rice canopy
computer vision
colour feature
threshold segmentation
