期刊论文

中文

武汉轻工大学学报

2095-7386

2019

38

5

69-74+80

10.3969/j.issn.2095-7386.2019.05.014

国家重点研发计划(2016YFD0100202).

本校为第一机构

利用图像处理技术从甘蓝型油菜植株中识别出油菜根部及根部肿瘤等信息,进而分析甘蓝型油菜植株的根肿病程度。从原图像中提取出目标是解决实际问题的关键步骤,因此从复杂背景的甘蓝型油菜植株图像中分割出根部区域是准确识别其病害程度的关键。提出以EXG过绿颜色分割算法与高斯YCbCr颜色算法相结合的分割方法。首先通过EXG过绿颜色分割方法分离出绿色的茎叶部分,其次将RGB颜色空间转换到YCbCr颜色空间,利用高斯模型判断根的像素点,因此分割出甘蓝型油菜植株的根部区域,进一步利用数学形态学方法将得到的根部二值图进行膨胀与腐蚀的去燥操作,最终完成图像分割。利用该方法对不同拍摄环境且背景复...

Image processing technology was used to identify the roots and root tumors of Brassica napus L.plant,and to analyze the degree of clubroot disease in Brassica napus L. plants. Extracting the target from the original image is the key step ,therefore,segmenting the root region from the complex background image of the Brassica napus L. plant iscritical to accurately identifing the disease degree. Combining EXG over-green color segmentation algorithm and Gaussian YCbCr color algorithm,a segmentation method was developed in this paper. The green stem and leaf wasseparated by the EXG green color seg...