使用改进的POCS算法的超分辨率图像复原  

Super-resolution image restoration via improved POCS algorithm

作　　者：谢甜[1] 

机构地区：[1]武汉科技大学,湖北武汉430081

出　　处：《电子设计工程》2013年第18期142-144,共3页Electronic Design Engineering

摘　　要：超分辨率复原技术的基本思想就是采用信号处理的方法,在改善图像质量的同时,重建成像系统截至频率外的信息。POCS(凸集投影)算法是一种广泛应用于图像超分辨率复原的方法。针对传统的POCS算法的边缘振荡效应,在分析其产生的原因,造成的影响的基础上,采用改进的POCS算法,以减少边缘振荡。采用基于小波变换模极大值的改进POCS算法进行图像超分辨率复原。实验结果表明,该方法有效的较少了复原图像的边缘振荡效应,是一种有效的图像超分辨率复原方法。Super-resolution image restoration may restore the high frequency information beyond the cut-off frequency of the imaging system by using signal processing method. POCS （projection onto convex sets） algorithm is a widely used method of image super-resolution recovery. Aiming at the edge ringing effect of traditional POCS algorithm, vis analyses the basic reason causing the effect, and adopts an improved POCS algorithm to reduce it. Based on wavelet transform modulus maxima improved pocs algorithm image super-resolution recovery. The experimental results show that the method is effective less the recovery image edge oscillation effect, is an effective super-resolution image restoration method.

关 键 词：凸集投影算法(POCS) 超分辨率图像复原 小波变换模极大值 边缘振荡 

分 类 号：TN911.73[电子电信—通信与信息系统]