粗糙集理论在图像增强处理中的应用

Application of Rough Sets Theory to Image Enhancement Processing

摘要:

对粗糙集理论在图像主要特征提取、图像自动识别与边界检测等方面的应用方法进行研究.采用地学实地野外调查中获取的彩色图像作为实验对象,选用表征颜色、纹理的17个特征属性作为条件属性构建决策表,通过属性约简、分类和规则提取指导图像处理,最终实现了对目标地物的分类分区增强处理及区域边界划分.实验结果表明:该方法可以较好地对图像进行分区处理和边界划分,结果较令人满意;并且得出影响实验图像分区的主要条件属性为蓝色标准偏差、红色方差、红色与蓝色相关性、彩色特征、对比度及熵.

Abstract:

This paper presents a study of the application methods of rough sets theory to the main image features obtaining, image automatically recognizing and edges detecting. The color image from the field research was processed. The color characters and texture characters were used to build a decision table including seventeen features. The objects were classified and enhanced by cutting attributes, classifying and rule-obtaining. At the same time,the region edge image was made. The test result shows that this method can better segment the image, enhance the regional image and detect the edge of the image. And the blue standard deviation, red variation, correlation index between red and blue, color feature, contrast and entropy are regarded as the main condition attributes.