摘要
图像去噪是图像处理领域的研究热点,数字图像去噪方法研究仍然是一项富有挑战性的工作.本文以性能卓越的曲波(Curvelet)变换理论为基础,提出了一种基于模糊支持向量机(FSVM)的曲波域图像去噪算法.该算法的基本工作原理为:首先,对原始噪声图像做曲波分解以获得变换系数;然后,结合噪声分布特点确定系数空间性,并构造出FSVM的训练特征;最后,对高频曲波系数进行模糊分类与自适应阈值去噪,并进一步对去噪后系数进行曲波重构以得到去噪图像.通过仿真实验结果,证明了本文算法在消除伪吉布斯(Gibbs)现象的同时,具有较强的抑制噪声能力和边缘保护能力.
Image denoising is of great importance in the field of image processing.In this paper,we present an effective image denoising approach using curvelet transform and fuzzy support vector machine（FSVM）.Firstly,the noisy image is decomposed into use curvelet transform frequency and orientation in response to different sub-bands.Secondly,the feature vector noisy pixels of the image is constructed space regularity curvelet domain and fuzzy support vector machine（FSVM）obtained through training model.Then the curvelet detail coefficients are divided into two categories（edge correlation coefficient and noise-related）by the FSVM training model.Finally,the detail subbands of curvelet coefficients are denoised by using the adaptive thresholds.The simulation results prove that the proposed algorithm can not only eliminate the pseudo-Gibbs（Gibbs）phenomena,but also has the strong ability to suppress noise and edge protection.
出处
《辽宁师范大学学报（自然科学版）》
CAS
2016年第1期44-49,共6页
Journal of Liaoning Normal University:Natural Science Edition
基金
国家自然科学基金项目(61472171
61272416)
辽宁省教育厅高等学校科学研究一般项目(L2015289)
辽宁师范大学青年科学研究项目(LS2014L016)
关键词
图像去噪
曲波变换
模糊支持向量机
自适应阈值
image denoising
curvelet transform
fuzzy support vector machine
adaptive threshold
