基于计算机视觉的全自动原木板材分检系统研究  

Raw Wood Material Sorting System Based on Computer Vision

作　　者：胡跃明[1] 戚其丰[1] 劳卫伦[1] 

机构地区：[1]华南理工大学自动化科学与工程学院,广东广州510640

出　　处：《林业机械与木工设备》2005年第8期12-15,共4页Forestry Machinery & Woodworking Equipment

摘　　要：原木板材的颜色纹理检测以及分类是木质产品生产过程中的一道重要工序。本文讨论分析了板材检测分类系统的硬件设计和系统涉及的算法,在分析不同聚类分析方法的基础上,提出了采用最大最小距离聚类法实现板材聚类,并对有关算法进行了软件设计和相关的实验研究,表明了算法的有效性。The detection and classification of the wood material is one of the important procedures in the manufacture process of the woody products. This paper not only introducesd the hardware design of the system, but also discussed the algorithms used for the detection and classification in detail. In addition, based on analyzing different clustering analysis methods, the biggest and lest distance clustering algorism was brought forward to realize the wood material clustering.

关 键 词：板材 颜色 纹理 特征提取与选择 聚类分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]