基于粗糙集阴影区域的检测与分类

Image Shadow Detection and Classification Based on Rough Sets

粗糙集理论是一种新的处理模糊和不确定性问题的数学工具。本文提出一种基于粗糙集阴影边缘分类方法。该算法根据粗糙集理论、梯度、最大邻域差及噪声的条件属性,将一幅图像划分为不同的子图像;然后对子图像分别进行处理,得到阴影边缘点;再利用边缘生长对边缘点进行细化和跟踪,删除那些假边缘点;然后根据阴影边缘构出假想的阴影区域,并统计这些区域的灰度直方图,求得阴影区域的灰度区间。根据该灰度区间可以得到阴影区域,再根据这些区域的灰度、形状、面积等特征对阴影进行分类。通过对所得结果进行分析可知,结合粗糙集理论的阴影图像边缘检测算法与其他的常规检测方法相比,无论从视觉效果还是检测精确度上都有所改善。

The theory of the Rough sets is a new mathematics tool which is used to process fuzzy and indetermination problem, this paper puts forward a new method of the shadow edge classification based on Rough sets, which according to the theories of the Rough sets and the condition attribute of the gradient, the biggest error of neighborhood and the noise. The method divides a picture into the different several sub-pictures, then respectively process the sub- picture and get the shadow edge points. Then these edge points are thinned and followed and those false edge points are deleted. Then a imaginable shadow district is formed according to the shadow edge. The method statistics the gray histogram of these districts and get a gray zone of the shadow district. Then the shadow district is got and the shadow can be classified according to some characteristics such as the gray and the shape and the area of zone etc. It draws the conclusion by analyzing the result, the detecting accuracy and visual effect are improved by comparing the edge detection method based on rough set with normal methods.