摘要
针对微光环境下的光子计数集成成像系统,提出了一种基于复合光子计数模型和Lucy-Richardson算法的元素图像增强算法。首先,根据集成成像原理,用3DSMAX软件仿真合成孔径集成成像系统采集的目标元素图像,对其进行归一化后根据光子计数模型获得光子计数元素图像。然后,根据复合光子计数模型对元素图像进行处理,利用泊松分布的可加性提高元素图像的质量。最后,利用Lucy-Richardson算法迭代增强元素图像,并根据重构公式得到重构图像。实验结果表明,增强的元素图像峰值信噪比随算法迭代次数的增加而提高并逐渐趋于稳定,且用增强元素图像重构的图像效果优于传统的极大似然法。
For the photon counting integral imaging system in a low-light environment,this paper proposes an element image enhancement algorithm based on the composite photon counting model and the Lucy-Richardson algorithm.First,in compliance with the integral imaging principle,3 DSMAX software is used to simulate the target element image collected by the synthetic aperture integral imaging system.After normalization of the target element image,the photon counting element image is obtained via the photon counting model.Then,it is processed in light of the composite photon counting model and its quality is enhanced in view of the additivity of the Poisson distribution.Finally,the Lucy-Richardson algorithm is iterated to enhance the element image and the reconstruction formula is employed to generate the reconstructed image.Experimental results show that the peak signal-to-noise ratio of the enhanced element image increases with the algorithm iteration number and gradually stabilizes.In addition,enhancing the element image delivers a better reconstruction result than the conventional maximum likelihood estimation method.
作者
李尚
儒葛鹏
Li Shangru;Ge Peng(School of Physics and Optoelectronics,South China University of Technology,Guangzhou,Guangdong 510641,China)
出处
《光学学报》
EI
CAS
CSCD
北大核心
2021年第17期57-63,共7页
Acta Optica Sinica
基金
国家自然科学基金(61805082)
中央高校基本科研业务费专项(2019MS125)。
