傅里叶谱和矩不变法结合的图像阈值分割  

Image thresholding method based on Fourier spectrum and moment-preserving principle

作　　者：张新明[1] 李双[1] 郑延斌[1] 张慧云[1] 

机构地区：[1]河南师范大学计算机与信息技术学院,河南新乡453007

出　　处：《计算机应用》2010年第8期2094-2097,2153,共5页journal of Computer Applications

基　　金：河南省重点科技攻关项目(102102210180);河南省教育厅科技攻关资助项目(2007520024;2008B520021)

摘　　要：针对矩不变法图像分割算法中存在的问题,提出了一种基于傅里叶谱和矩不变法相结合的图像分割方法。首先依据矩不变原理计算待分割图像的累计灰度分布函数值获取初始阈值,然后对图像直方图进行傅里叶变换,并得到傅里叶谱系数,最后利用该系数和累计灰度分布函数值对初始阈值进行调整,由于傅里叶谱能很好地"匹配"矩不变法获取的阈值,从而能自动调整初始阈值,使分割效果达到最佳。分割实验结果表明,该方法不仅运算速度快,而且有较好的分割效果和普适性。Aiming at the problems of the threshoding method based on the Moment-Preserving Principle MPP this paper presented an image thresholding method based on Fourier spectrum and MPP.Firstly the Cumulative Distribution Function Value CDFV was calculated according to the MPP to get the rough threshold value then Discrete Fourier Transform DFT was carried out to the histogram of the image to be segmented and compute its Fourier Spectrum FS finally the CDFV and the FS were used to change the rough threshold value.Owing to good adaptation of FS to the rough threshold this adjustment could get an optimal threshold.Experimental results show that the proposed method can get better segmentation result and better generalization with high speed.

关 键 词：图像分割 离散傅里叶变换 矩不变法 阈值化 傅里叶谱 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]