一种基于粗糙集理论的灰度图像增强方法

Gray Image Enhancement based on Rough Sets Theory

介绍了粗糙集的基本理论,探讨并提出一种基于粗糙集理论的灰度图像增强方法。根据像素的灰度值属性和噪声属性,将像素划分成明暗区域和噪声区域,利用上下近似原理,求得无噪声区的粗糙集表示方式,并对较亮无噪声子图,采用直方图均衡变换,对较暗无噪声子图,采用指数变换,最后重叠,从而达到增强非边缘区域,减弱噪声的目的。实验证明:提出的算法对低灰度级别有丰富细节的图像具有明显增强作用。

The basic theory of rough sets is given and a method for gray enhancement is proposed.In accordance with the condition attributes of gray value and noise,the pixels are classified into light areas,shade areas and noise areas.By using upper and lower approximation of the rough sets theory,the rough sets expressions of the non-noise areas are acquired,and for the light non-noise sub-image,the histogram equalization transform,is adopted,while for the shade non-noise sub-image,the exponential transform is employed.Finally the tow sub-images are overlapped,thus to achieve the enhancement of non-edge areas and the reduction of noise.The experiments shows that this method is of certain significance in the enhancement of those low gray-level images with rich details.