摘要
针对现有双边滤波算法参数设定复杂、不具通用性等问题,提出一种基于参数估计的双边滤波算法。该算法从计算图像噪声水平和边缘强度出发,在弱纹理区域采用PCA对图像的噪声水平进行估计;利用Sobel算子检测出图像的有效边缘,并对其强度水平进行定义,进而在最佳峰值信噪比准则下自适应设置滤波参数的空间标准偏差和亮度标准偏差,从而得到自适应双边滤波器。最后,对本文滤波结果进行主客观质量评价。研究结果表明:本文算法是有效的。
Due to complex parameters and lack of adaptability to the existing bilateral filtering methods, an adaptive bilateral filtering method based on parameter estimation was proposed. Based on estimating noise level and edge intensity of image, the noise level was estimated from weak textured patches of image using principal component analysis, the edge intensity was defined by the detail performances of image, which was denoted by the Sobel edge detector, and the standard of space and intensity was acquired at the rule of optimal peak signal to noise ratio. Finally, the adaptive bilateral filtering was obtained and then evaluated by subjective effects and objective image quality assessment. The experiment results show that the algorithm is effective.
出处
《中南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2014年第11期3840-3845,共6页
Journal of Central South University:Science and Technology
基金
国家自然科学基金资助项目(61372167
61379104
61175029)资助项目
国防科技重点实验室基金资助项目(9140c610301080c6106
9140c6001070801)~~
关键词
双边滤波
主成分分析
SOBEL算子
峰值信噪比
bilateral filtering
principal component analysis
Sobel
peak signal to noise ratio
作者简介
南栋（1987-），男，陕西西安人，博士研究生，从事计算机视觉与图像处理研究；电话：18066620881；E-mail：nd．tian_53@163．com
