摘要
提出一种基于多频段区域互信息(MRMI)的光学成像系统性能评价方法。由信息论出发,结合人类视觉系统(HVS)分频感知原理,利用小波变换将物面与像面图像分解为多个频段,分别计算各频段区域互信息(RMI),再由小波滤波器和MTF的空间频段积分比计算各频段RMI的权重系数,加权综合得到多频段区域互信息(MRMI)。实验分析了具有不同像散度的光学系统点光源成像,最终得到的评价参数MRMI随着光学系统像散度的不断增大而趋于减小,曲线拟合结果表明MRMI值与像散度呈幂函数关系,确定系数达0.981以上。与MTF等传统光学系统性能评价方法相比,MRMI值量化反映了光学系统的成像性能,且与人眼视觉感知相一致。
An optical imaging system performance evaluation method based on multi-band regional mutual information(MRMI)is proposed.Starting from information theory,and combining with the human visual system(HVS)frequency sensing principle of using wavelet transform to the surface and the image plane image is decomposed into multiple frequency band,calculation of regional mutual information(RMI)for each frequency band.Then the weighted coefficients of RMI in each frequency band are calculated by the spatial frequency band integral ratio of wavelet filter and MTF,and the weighted synthesis results in multi-band regional mutual information(MRMI).The experimental analysis of point-light source imaging of optical systems with different astigmatism shows that the final evaluation parameter MRMI tends to decrease with the increasing astigmatism of optical systems.The curve fitting results show that MRMI value is a power function with astigmatism,and the determination coefficient is above 0.981.Compared with traditional optical system performance evaluation methods such as MTF,The MRMI value quantifies the imaging performance of the optical system and is consistent with human visual perception.
作者
施柯宇
付东翔
金涛
项华中
武杰
陈家璧
SHI Keyu;FU Dongxiang;JIN Tao;XIANG Huazhong;WU Jie;CHEN Jiabi(School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China;School of Medical Instrument and Food Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)
出处
《光学技术》
CAS
CSCD
北大核心
2020年第3期269-276,共8页
Optical Technique
基金
国家自然科学基金青年基金项目(61605114)
国家自然科学基金青年基金项目(61703277)。
关键词
信息光学
光学成像系统性能评价
区域互信息
小波分解
人类视觉系统
information optics
performance evaluation of optical imaging system
regional mutual information
wavelet decomposition
human visual system
