基于L1范数的图像超分辨率及差分统计模型  

L1-norm-based Image Super-resolution and Its Differential Statistical Model

作　　者：倚海伦[1] 王庆[1] 

机构地区：[1]西北工业大学计算机学院,西安710072

出　　处：《计算机工程》2007年第24期203-205,共3页Computer Engineering

基　　金：国家教育部新世纪优秀人才支持计划基金资助项目(2006-1801)

摘　　要：在L1范数图像超分辨率重建框架下,引入参数自适应估计,该方法对模型误差表现出良好的稳健性并且可以加速收敛。结合差分图像统计特性和概率先验模型,解释了L1范数形式的双边全变差正则项概念,利用Kullback-Leibler距离证明了该正则项的优越性,并分析了混合先验模型在超分辨率重建中应用的可行性等问题。A method for automatically estimating parameters is applied to the L1-norm-based image super-resolution reconstruction framework. The approach can get a stable result with model errors and can improve convergence. A novel explanation of L1-norm-based bilateral total variation（BTV） term is presented according to pixel differences statistics, combined with probability prior model. Superiority of BTV is validated using Kullback-Leibler distance. A theoretical analysis of the feasibility and the problems of super-resolution reconstruction are given using mixture prior distribution.

关 键 词：超分辨率 正则化 差分图像 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]