基于遗传算法的盲超分辨率图像的重建  

Blind super-resolution image reconstruct based on genetic algorithm

作　　者：吴伟巍[1] 曾庆山[1] 刚轶金[1] 

机构地区：[1]郑州大学电气工程学院,郑州450001

出　　处：《计算机工程与应用》2009年第35期199-201,共3页Computer Engineering and Applications

基　　金：河南省自然科学基金No.0611054100~~

摘　　要：针对盲超分辨率图像复原问题,提出了一种基于遗传算法的盲超分辨率图像重建方法。这种方法建立了一个由高分辨率图像和模糊项相关联的正则化函数,并利用遗传算法的全局寻优能力使得这个正则化函数最小化,从而由一组退化的低分辨率图像求得高分辨率图像,同时评估出导致模糊原因。仿真结果表明算法的可行性。In allusion to blind super-resolution images for recovery,an approach of blind super-resolution image reconstruct based on genetic algorithm is proposed.This approach builds a regularization energy function,where regularization is carried out in both the image and blur domains,minimizes the regularization by using global optimizing capacity of genetic algorithm,then gets the high resolution image from multiple frames degraded low resolution images,and also estimates the blurs.The results of simulation show that the approach is feasible.

关 键 词：盲超分辨率 遗传算法 正则化函数 模糊 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]