摘要
以全变分去噪模型为基础,详细阐述全变分模型的构造,利用图像的最优关联度作为停止准则的参考,获得一种新型的全变分图像去噪模型。该模型不仅能够获得一种较好的迭代准则,同时,图像去噪后的效果接近最优值。实验结果表明,该算法能够获得较好PSNR值,而且算法的复杂度也降低。
Taking the total variation denoising model as the basis,in this paper we expatiate on the structure of total variation model.By using the optimal relation of image as the reference for stop criterion,we gain a novel denoising model of total variation image.The model can obtain a quite good iteration criterion,and meanwhile its effect also approaches the optimal value after the image is noise removed.Experimental results show that algorithm can get preferable PSNR (peak signal-to-noise ratio)value,the computation complexity is decreased as well.
出处
《计算机应用与软件》
CSCD
北大核心
2014年第2期195-197,213,共4页
Computer Applications and Software
基金
国家自然科学基金项目(61171077)
江苏省高校自然科学研究项目(12KJB510025
12KJB510026)
交通部应用基础研究项目(2011-319-813-510)
南通市引进人才项目(03080415
03080416)
南通大学创新人才基金项目(2009)
关键词
全变分
停止准则
最优关联度
Total variation
Stop criterion
Optimal relation