| 基于概率神经网络板材纹理分类识别的研究 |
| |
| 引用本文： | 王辉,杨林,丁金华,李明颖.基于概率神经网络板材纹理分类识别的研究[J].大连轻工业学院学报,2009(5). |
| |
| 作者姓名： | 王辉 杨林 丁金华 李明颖 |
| |
| 作者单位： | 大连工业大学机械工程与自动化学院; |
| |
| 基金项目： | 黑龙江省自然科学基金资助项目(C2004-03) |
| |
| 摘 要： | 为了实现对板材纹理识别的自动化,提出了一种基于概率神经网络的板材纹理分类识别方法。首先,获取板材的灰度共生矩阵特征参数,并进行特征选择。然后,根据研究对象设计PNN分类器,进行分类实验,识别率为88.00%。结果表明,该方法是有效的,用其对板材纹理进行分类是基本可行的。
|
| 关 键 词： | 板材 概率神经网络 灰度共生矩阵 识别 |
Classification and recognition of wood block texture based on PNN |
| |
| WANG Hui,YANG Lin,DING Jin-hua,LI Ming-yingSchool of Mechanical Engineering & Automation,Dalian Polytechnic University,Dalian ,China.Classification and recognition of wood block texture based on PNN[J].Journal of Dalian Institute of Light Industry,2009(5). |
| |
| Authors: | WANG Hui YANG Lin DING Jin-hua LI Ming-yingSchool of Mechanical Engineering & Automation Dalian Polytechnic University Dalian China |
| |
| Affiliation: | WANG Hui,YANG Lin,DING Jin-hua,LI Ming-yingSchool of Mechanical Engineering & Automation,Dalian Polytechnic University,Dalian 116034,China |
| |
| Abstract: | To recognize automatically wood block texture,a method was proposed based on the probabilistic neural networks.The parameters of GLCM were obtained and feature selection was carried on,and PNN classifier was designed according to object of study.The recognition rate was up to 88.00%,which indicated that the method was effective,and the identification of wood block texture with this method was feasible. |
| |
| Keywords: | |
| 本文献已被 CNKI 等数据库收录！ |
|
