摘要
本文提出了基于峰高-峰宽共生矩阵模型的纹理分析方法。利用从该模型中提取出的五个纹理特征参数,可以对皮料按其纹路进行分类识别,在实验室中取得了较高的分类精度。
In this paper, a kind of texture analysis method, based on peek height-peek width co-occurrences model, is presented. Using the five features extracted from this model, a high classification accuracy is obtained according to the texture of leather.
出处
《信号处理》
CSCD
北大核心
1996年第4期369-372,共4页
Journal of Signal Processing
关键词
特征提取
共生矩阵分析法
皮料
纹理识别
texture,feature extraction,model,extrema,leather classification