摘要
针对传统迭代盲反卷积算法收敛速度慢、容易出现解模糊等问题,提出一种改进的图像迭代盲反卷积算法。利用动量矩求解图像的有限支持域,在支持域中使频率域和空间域交替迭代,从而实现图像的盲复原。仿真结果表明,与传统迭代盲反卷积算法和基于小波变换的盲反卷积算法相比,该算法的收敛速度较快,具有较好的图像恢复效果。
In order to overcome the problems such as slow convergence speed and fuzzy solution, this paper proposes an improved Iterative Blind Deconvolution（IBD） algorithm. Moment of momentum is utilized to obtain the limited support domain of image, and alternating iterative calculation is achieved in the frequency and space domain. Blind image restoration is completed. Compared with traditional iterative blind deeonvolution and IBD based on wavelet, simulation results show that the algorithm speeds up the convergence rate, and obtains a better restoration result.
出处
《计算机工程》
CAS
CSCD
2012年第8期204-206,共3页
Computer Engineering
基金
山西省科技攻关计划基金资助项目"医院器械物流管理平台的研究与开发"(20090321020)
中央高校基本科研业务费专项基金资助项目(ZY20120210)
天津市高等学校科技发展基金资助项目(20110709)
关键词
图像复原
盲反卷积
支持域
动量矩
image restoration
blind deconvolution
supporting domain
moment of momentum
