改进全变分的图像去噪  

Image Denoising on Improved Total Variation

作　　者：何坤[1] 郑秀清[2] 琚生根[1] 张永来[2] 

机构地区：[1]四川大学计算机学院,成都610065 [2]四川师范大学信息技术学院,成都618300

出　　处：《电子科技大学学报》2016年第3期463-468,共6页Journal of University of Electronic Science and Technology of China

基　　金：四川省科技支撑计划(2013SZ0157)

摘　　要：为了弥补各向同性扩散去噪的非保边性、异性扩散的耗时性,分析了各向同性和异性扩散的机理,根据噪声对像素变化的影响,设计了新的扩散函数,理论上分析了该函数的扩散性能:对平滑区各向同性扩散,边缘区实现各向异性扩散。在传统全变分去噪的基础上,提出了改进全变分的图像去噪模型,运用固定点代算法设计了相应的离散迭代函数。实验结果表明,该算法在图像平滑区进行各向同性扩散,继承了各向同性的优点,降低了传统全变分的运行时间;在边缘区实现了各向异性扩散保护了图像边缘,提高了图像的峰值信噪比和视觉效果。By analyzing the mechanism of isotropic and anisotropic diffusion in image denoising, a new diffusion function based on the effect of noise on the image pixel variation is designed in this paper. The diffusion performance of this function is isotropic diffusion in the smoothing sub-region and anisotropic diffusion on the edge. Then, an improved total variation denoising model is proposed based on the traditional total variation （TV）, and the corresponding discrete iterative function is designed by using the fixed point iteration algorithm. The algorithm uses isotropic diffusion on the smooth region, inheriting the advantages of the isotropic diffusion, reducing the running time of the traditional TV, and protecting the edge by using anisotropic diffusion. Experimental results show that this algorithm can improve image peak signal to noise ratio （PSNR） and visual effects.

关 键 词：各向异性 扩散函数 图像去噪 各向同性 全变分 

分 类 号：TP394.1[自动化与计算机技术—计算机应用技术]