皮料自动分类中纹理分析方法的研究

阐述如何采用图像纹理分析方法对皮件生产中的皮料进行纹理分析识别．在对已有纹理分析方法对比研究以及大量实验的基础上，采用相隔一定距离的水平扫描线上的图象灰度数据作为研究对象，利用从中提取出的特征值，可以对皮料按精细度和深刻度进行识别分类，取得了较高的分类精度，这使皮料纹理的计算机实时自动识别分类成为可能，也为计算机图象处理提供了一种有效的分析方法．

TEXTURE ANALYSIS ON AUTOMATIC CLASSIFICATION OF LEATHER

In this paper, image texture analysis method is adopted in texture recognition of the leather used in the production of clothing. Based on many experiments and the trying of the present texture analysis methods,a new method is developed by using the image pixel grey value drawn from the scaning line sperated with a cetain distance. A high classification accuracy is achieved using the features extracted from the data mentioned above according to the fineness and sharpness of the leather texture. This makes the on-line automatic classification of leather possible,and a kind of effective method for image processing is provided also.