摘要
针对木材表面颜色自动分类的难题,在RGB颜色空间,将R、G、B三个颜色矩阵融合成一个特征矩阵,再对这个特征矩阵提取颜色三阶矩参数作为木材表面颜色分类的特征参数,设计了适合木材表面颜色分类的BP神经网络分类器,分类识别率达到98.67%,验证了提取特征参数的有效性。
Aimming at the classification of wood surface color, R, G and B was fused to 1 feature matrix in RGB color space, and extracted feature parameters of color moment. BP neural network, which was suitable to wood surface color, was designed. The correct rate of classification reached 98.67%, and accounted for the validity of the parameter that this paper extracted.
出处
《林业机械与木工设备》
2006年第8期27-29,共3页
Forestry Machinery & Woodworking Equipment
基金
黑龙江省自然科学基金(C2004-03)
哈尔滨市自然科学基金(2004AFXJ020)
关键词
木材
特征提取
分类
BP神经网络
wood
feature extraction
classification
BP neural network
