摘要
水稻病害图像分割是提高水稻病害识别准确率的前提。论文研究了HSV、Lab、YCb Cr和RGB颜色空间下水稻稻瘟病、细菌性条斑病和稻曲病等3种病害图像分割方法。实验结果表明,HSV颜色空间下H色调分量病斑和健康区域对比度不强,Lab颜色空间下A色调分量和YCb Cr颜色空间下Cr色差分量采用最大类间方差法存在漏分割和冗余分割的现象,论文利用在RGB颜色空间下R分量高于G分量和B分量的特征,提取了水稻病斑,取得了较好的分割效果。
Image segmentation of rice disease is the prerequisite to improve the accuracy of rice disease identification. In this paper,the image segmentation methods of 3 diseases are studied,namely rice blast,bacterial leaf streak and rice false smut,in HSV,Lab,YCb Cr and RGB color spaces. The experimental results show that the contrast of the disease spot and the health area isnot strong in the H color component of the HSV color space. The A hue component under the Lab color space and the Cr color difference component under the YCb Cr color space have the phenomenon of leakage segmentation and redundancy segmentation using the maximum variance method. The method used in this paper takes advantage of the characteristics that the R component is higher than the G component and the B component. The disease spots are extracted and good segmentation results are obtained.
作者
苏博妮
化希耀
范振岐
SU Boni;HUA Xiyao;FAN Zhenqi(School of Intelligent Manufacturing,Sichuan University of Arts and Science,Dazhou 635000;Dazhou Industrial Technology Institute of Intelligent Manufacturing,Dazhou 635000;College of Information Engineering,Tarim University,Alar 843300)
出处
《计算机与数字工程》
2018年第8期1638-1642,共5页
Computer & Digital Engineering
基金
国家自然科学基金项目"新疆阿拉尔垦区天然彩色棉花高光效株型数字化构建研究"(编号:61662064)
四川省教育厅科研项目(编号:17ZB0373
18ZA0419)
四川文理学院智能制造产业技术开发研究专项项目(编号:2017ZZ005Y
2017ZZ002Z)
四川文理学院科研项目(编号:2017KZ012Y)资助
关键词
水稻病害
颜色特征
图像处理
图像分割
flee disease
eolor features
image proeessing
image segmentation
