% SPGL1: A solver for large-scale sparse reconstruction % Version 1.7 (May 20, 2009) % % Files % spg_bp - Solve the basis pursuit (BP) problem % spg_bpdn - Solve the basis pursuit denoise (BPDN) problem % spg_lasso

基于双边滤波的retinex算法，暗光图像增强，有效去噪、提高对比度、去光晕；含论文和讲解PPT、实例代码(Retinex algorithm based on bilateral filtering can enhance dark image, effectively denoise, improve contrast and eliminate halo)

非局部均值（NL-means）是近年来提出的一项新型的去噪技术。该方法充分利用了图像中的冗余信息，在去噪的同时能最大程度地保持图像的细节特征。基本思想是：当前像素的估计值由图像中与它具有相似邻域结构的像素加权平均得到。(NL-means is a new denoise method. It utilized image's information, preserving detail while denoise.)

利用偏微分方程原理，建立二阶、四阶及TV模型对指静脉图像进行去噪处理，提升图像清晰度，以便于后续特征提取及识别。(Based on the principle of partial differential equation, the second-order, fourth-order and TV models are established to denoise the digital vein image, so as to improve the image clarity and f