基于支持向量机技术的粮虫图像分类技术研究  

Study on Stored-Product Pests Recognition Base On Support Vector Machine

作　　者：党培[1] 谭联[2] 

机构地区：[1]河南工业大学机电工程学院,河南郑州450007 [2]华北水利水电学院电力学院,河南郑州450011

出　　处：《粮食流通技术》2009年第6期33-37,共5页Grain Distribution Technology

基　　金：校级青年科研基金资助项目

摘　　要：介绍了支持向量机的基本思想,提出了一个基于支持向量机的粮虫模式识别系统。该系统先对粮虫图像进行小波边缘提取,根据灰度共生矩阵和局部统计方法提取小波分割后的图像纹理特征。最后利用支持向量机对粮虫图像进行分类。The foundations of support vector machines are introduced. An evaluation model based on SVM is made, and the model is tested to obtain better results. Edge detction based on wavelet multi-scale identity is made.The statistics features based on regional gray and the co-occurrence matrix of gray level are taken as performing image segmentation. It is expected that the SVM method would be further applied to the field.

关 键 词：特征提取 小波变换 支持向量机 图像识别 

分 类 号：S379.2[农业科学—农产品加工]