基于SURE估计的图像块稀疏收缩去噪算法

Image Block Sparsity Shrinkage Denoising Algorithm Based on SURE Estimation

针对图像过完备稀疏收缩去噪的阈值选取问题,根据图像的常规稀疏模型,提出一种基于SURE无偏估计的自适应阈值选择算法。在一阶可导收缩函数的基础上,推导阈值选择的优化目标函数,证明该函数是关于阈值的凸函数,利用黄金分割法搜索其全局最小值。仿真结果表明,该算法选择的阈值接近峰值信噪比阈值曲线的极大值点,将该算法应用于图像的块稀疏模型,可取得比常规稀疏模型更好的去噪效果。

Aimming at the choice of threshold under over compeleted sparese shrinkage denoising of image,a new adaptive threshold selection algorithm is investigated over image normal sparse model based on SURE agonic estimation.Based on the one order derivable shrinkage function,the optimal objective function about threshold selection is derived and it is shown to be convex function on threshold,and then its global minimum is searched by golden section method.Simulation result shows that the choice of threshold is closer to the maxima of PSNR threshold curve.The new algorithm is extended over image block sparisity model,and a better denoising result than normal sparse model is gotten.