摘要
基于图像识别是小麦腥黑穗病分类鉴定中正在研究的一种新技术。文章研究了提取小麦腥黑穗病病菌冬孢子形态学特征的图像处理方法。通过获取小麦腥黑穗病显微图像,对显微图像进行图像分割,中值滤波以及试探性扩张矩形等方法获得了病菌显微图像冬孢子区域的内接矩形,提取了基于灰度共生矩阵的纹理特征与形状特征。实验数据显示该方法能有效地提取出小麦腥黑穗病的形态学特征,对后续基于模式识别鉴定小麦腥黑穗病类别具有参考意义。
It is a new technology that identifies species of the Tilletia diseases based on image recognition under study.And Introduce the method of image processing extracts the morphological characters of the Teliospore of the Tilletia diseases. After obtain the Tilletia diseases micrographs, get the inscribed rectangle of the Teliospore in the micrographs of germ by using Image Segmentation, median filter and expanding rectangular tentatively. Extract the textural features and shape feature based on Gray-level Co-occurrence Matrix. The experimental data show that the method could extract morphological characters of the Tilletia diseases effectively, which has reference meaning for identifying the species of the Tilletia diseases based on pattern recognition later.
出处
《电脑与信息技术》
2015年第2期1-3,16,共4页
Computer and Information Technology
基金
2013年质检总局科技计划项目(项目编号:2013IK252)
关键词
小麦腥黑穗病
分类
图像处理
形态学特征
Tilletia diseases
classification
image processing
morphological character
