一种适于数字灰度图像的分数维计算方法  

A Fractal Dimension Calculation Method for Gray Level Images

作　　者：姬光荣[1] 王硕[1] 年睿[1] 冯晨[1] 赵静[1] 

机构地区：[1]中国海洋大学电子工程系,山东青岛266071

出　　处：《中国海洋大学学报（自然科学版）》2006年第5期709-711,共3页Periodical of Ocean University of China

基　　金：国家自然科学基金项目(60572064)资助

摘　　要：分数维是分形分析的重要特征量。目前普遍采用的分数维计算方法是盒维数算法。但是对于数字灰度图像,该算法所得到的分数维非常粗糙,以至于难以作为图像分析特征。基于面积度量精确性和计算的相对简单性,提出1种利用面积覆盖的分数维计算方法。该方法的主要特点是对于灰度图像计算所得的分数维几乎是连续的,为分数维作为图像分析特征开拓了广泛的应用前景。Fractal dimension is a very important feature in fractal analysis. The popular method for fractal dimension estimation is the box-counting method, but for gray level images, the fractal dimensions are so coarse that they could hardly be used as the features for image analysis. Because of the accuracy and relative simplicity of area measurement, a new fractal dimension estimating method is proposed based on the area coverage. The main characteristic of the method is that the calculated fractal dimensions are almost continuous for a gray level image. The method may have broad application prospects in image analysis in terms of fractal dimention.

关 键 词：分数维 盒维数 体积覆盖 面积覆盖 数字灰度图像 

分 类 号：TN911.73[电子电信—通信与信息系统]