基于图像处理的烟叶叶片结构分类算法

针对采用传统工艺方法进行烟叶叶片结构检测和分类的缺点,介绍了图像处理和分析技术在烟叶叶片结构分类识别中的应用,讨论彩色图像分割、连通区域标记、面积的检测、识别和分类等主要的图像处理算法,提出一种动态的烟叶叶片面积大小分类的算法,它是基于类的分离度思想而提出的.该算法对于不同品质的烟叶图像具有自适应性,且计算准确、速度快,能满足实时性要求.此算法已有效地应用于烟叶叶片结构自动分类系统,并取得了令人满意的效果.

A Classification Algorithm for a Tobacco Leaf Structure with Image Processing

To overcome the flaw of the conventional method of detecting and classifying the tobacco leaf structure, the application of image processing and analyzing in tobacco leaf structure classification and recognition are introduced. The details of some main image processing algorithms such as color image segmentation, connected component labeling, area detecting, recognizing and classifying are discussed. A algorithm for classifying tobacco leaf area dynamically is presented. It based on the concept of the scatter of classification. The algorithm is adaptive to various quality tobacco images. The algorithm is precious and the processing time is decreased. It al~ runs in real time. It works efficiently and stably with a tobacco leaf structure classifying automatically system. The practical results showed that it is feasible and satisfactory.