温室蔬菜病害智能识别图像预处理及其特征提取 - 吉林农业大学学报.pdf

吉林农业大学学报 2007, 29( 2) : 219~ 223 http:// xuebao1jlau1edu1cn Journal of JilinAgricultural University E-mail:jlndxb@vip1sina1com * ** 1 2 3 4 1 彭占武 , 王 雪 , 司秀丽 , 赵新子 , 袁洪印 ( 11 吉林农业大学工程技术学院, 长春 130118; 21 吉林农业大学菌物研究所, 长春 130118; 31 吉林农业大学信息技术学院, 长春 130118; 41 吉林省农机管理总站, 长春130062) : 采用计算机图像处理技术对温室蔬菜病 进行了智能化识别, 并以黄瓜霜霉病为例研究了温室蔬菜 病 智能识别图像预处理和特征提取的方法试验选择白色作为病 叶片的背景, 利用中值滤波法有效地去 除了噪声的干扰, 利用双峰法从背景中分离出病 图像, 再对图像进行边缘检测, 准确地提取了病斑的几何特 征该方法能够实现对病 图像的预处理, 并且能够准确地提取病 特征 : 蔬菜病 ; 图像预处理; 特征提取; 识别 : TP391 : A : 1000-5684( 2007) 02-0219-05 A Study on the Method of Image Pre-Processing and Feature Extraction for Intelligent Recognition of Greenhouse Vegetable Diseases 1 2 3 4 1 PENG Zhan-wu , WANG Xue , SI Xiu-li , ZHAO Xin-zi , YUAN Hong- in (11 College fo Engineering Technology, Jilin Agricultural University, Changchun 130118, China; 21Institute of ycology, JilinAgricultural University, Changchun 130118, China; 31 College of Inf or- mation Technology, JilinAgricultural University, Changchun 130118, China; 41JilinProvince Central anagement Stationof Agricultural achinery, Changchun 130062, China) Abstract : Using computer im ge processing technology, greenhouse veget ble dise ses were recognized in- telligently, nd t king the cucumber mildew s n ex mple, the method of im ge pre-processing nd fe- ture extr ction for intelligently recognizing greenhouse veget ble dise ses w s studied1Whitew s chosen s the b ckground of dise sed le f, medi n filter w s utilized to