现代电子技术2016，Vol.39Issue(20)：159-162,4.DOI:10.16652/j.issn.1004-373x.2016.20.040
基于字典训练和高频增强的图像降噪研究
Research on image denoising based on dictionary training and high-frequency enhancement
摘要
Abstract
The noise image,especially for the image with high⁃density noise,could lose its many details after denoising. In order to solve this problem,a method based on dictionary learning and high⁃frequency enhancement is proposed in this paper. In this method,the noise image is denoised at first;the adding noise and denoising processes are simulated respectively with the sample image to obtain the denoising sample image , and then the sample image is subtracted from the denoising sample image to get the sample difference image;the sample difference image and the denoising sample image are trained respectively to get a pair of high and low resolution dictionaries,which will be used for rebuilding the high frequency lost when the image is denoised. The simulation results of the experiments show that the proposed method is superior to the BM3D method in the subjec⁃tive human vision and objective evaluation.关键词
图像降噪/字典训练/稀疏表示/K-SVD算法Key words
image denoising/dictionary training/sparse representation/K-SVD algorithm分类
信息技术与安全科学引用本文复制引用
宁煌,曾儿孟,黄智昌,靳寒阳..基于字典训练和高频增强的图像降噪研究[J].现代电子技术,2016,39(20):159-162,4.基金项目
国家自然科学基金（61362006）；广西自然科学基金（2013GXNSFAA019334）；广西无线宽带通信与信号处理重点实验室（GXKL0614202；GXKL0614101）；认知无线电重点实验室（2013ZR08）；桂林电子科技大学研究生教育创新计划项目 （）
