基于布朗模型和小波变换的图像分形维数计算  

Computation of Fractal Dimension for Image Based on Brownian Model and Wavelet Transform

作　　者：聂笃宪[1] 曾文曲[2] 易珺[3] 

机构地区：[1]华南农业大学理学院,广东广州510642 [2]广东工业大学应用数学学院,广东广州510090 [3]广东药学院医药商学院计算机系,广东广州510006

出　　处：《电脑知识与技术》2006年第8期124-124,133,共2页Computer Knowledge and Technology

基　　金：KZCX1-SW-18资助.

摘　　要：函像分形维数是反映图像纹理特征的重要因素，也是图像分割的主要依据；通常，图像的分形维数多数采用盒维数计算方法来得到．但是避免不了计算时阚值选择带来不精确的问题，本文结合小波变换和布朗模型，提出了一种新的计算方法，并且和盒维数方法计算结果进行比较，结果表明，通过本文的计算方法得到的图像分形维数较准确。Fractal dimension of image is an important factor that reflects the features of image texture and that it is also main foundation of image segmentation,Usually, the fractal dimension of image is obtained by using the box-counting method. However,it can not avoided by using the box-counting method that the inaccurate problem was caused by the chosen threshold when calculating the fractal dimension.Combined wavelet transform with Brownian model,a new counting method for fractal image is presented in the paper,Compared with the results of image fractal dimension by using the box-counting method, Simulation results show that the method in the paper is fine and accurate,

关 键 词：分形 分形维数 小波变换 布朗模型 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]