一种基于变精度粗糙集的图像分类方法

基于变精度粗糙集,引入近似区分矩阵的概念,提出一种基于颜色特征的图像分类模型及其分类算法.变精度粗糙集理论在数据分类应用中主要是将集合间精确的包含关系改为多数包含关系,既允许一定程度的错误分辨率存在.用该方法进行图像资源的分类,克服了经典粗糙集不宜处理带有噪声的数据和决策表不协调的分类问题的缺陷,同时又大大简化分类规则,且形成的规则集便于用户理解.完善近似空间的概念.实验结果表明：在处理决策表不协调的图像分类问题,变精度粗糙集方法性能良好,分类准确、高效.

Variable Precision Rough Set Based Approach for Image Classification

Based on variable precision rough sets,this article introduces a similar distinction Matrix, and raises a color based on the characteristics of the image classification model and its classification algorithm. Variable precision rough set theory in the data classification applies the main idea to amend the collection of precise. This includes a relationship to the majority of relations, not only to allow a certain degree of error resolution there, hut the method was also used to image the classification of resources and to overcome the classic rough set of noise which should not be handled with out the lack of coordination of data. The decision-making table not only determines/measures the issue of classification of defects, but also greatly simplifies the classification rules and makes it easy for users to understand. It also improves the concept of the space similarity. The results showed that lack problems of coordination in dealing with a decision-making table of image classification were changed to more precie in the variable precision rough set way for an accurate and efficient classification.