基于非局部Bayesian的单掩膜编码衍射成像算法

来源期刊:中南大学学报(自然科学版)2021年第2期

论文作者:司菁菁 党文伟 史姗姗 程银波

文章页码:450 - 458

关键词:相位恢复;编码衍射图案;图像自相似性;Bayesian估计

Key words:phase retrieval; coded diffraction pattern; image self-similarity; Bayesian estimation

摘    要:为了降低编码衍射成像系统的硬件复杂度,研究基于单个编码衍射图案(coded diffraction pattern, CDP)的成像算法。以图像本身具有的非局部区域自相似特性作为相位恢复中的信号约束,克服单个CDP携带信息过少的不足,提出一种基于非局部Bayesian估计与图像自相似性的相位恢复与图像重建算法。研究结果表明:当仅利用单个CDP重建图像时,本文算法获得的平均峰值信噪比分别比WF(Wirtinger flow),DOLPHIn(dictionary learning for phase retrieval)和ConPR(constrained phase retrieval)算法提高了15.00,5.00和0.23 dB;且其重建图像能够更好地保留细节信息,具有较高的主观视觉质量。此外,本文算法的运行时间约为DOLPHIn算法的1/4、ConPR算法的一半。

Abstract: To reduce the hardware complexity of coded diffraction imaging system, the coded diffraction imaging algorithm was studied based on a single coded diffraction pattern(CDP). Taking the non-local self-similarity characteristics of the image as the regularization in the process of phase retrieval, the lack of information carried by a single CDP was offset. A new phase recovery and image recovery algorithm was proposed based on the non-local Bayesian estimation and the self-similarity of the image. The results show that when only one CDP is involved, the peak signal to noise ratio achieved by our algorithm is in average 15.00, 5.00 and 0.23 dB higher than those achieved by WF(Wirtinger flow), DOLPHIn(dictionary learning for phase retrieval) and ConPR(constrained phase retrieval), respectively. Images reconstructed by our algorithm maintain more detailed information, and present higher subjective visual quality. Moreover, the runtime of our algorithm is approximately a quarter of that of DOLPHIn and half of that of ConPR.

相关论文

  • 暂无!

相关知识点

  • 暂无!

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号