可提高相机动态范围的图像融合方法研究  

Image Fusion Method for Improving Dynamic Range of Camera

作　　者：李洪博 董岩[1] 刘云清[1] 赵馨[1] 宋延嵩[2] LI Hongbo;DONG Yan;LIU Yunqing;ZHAO Xin;SONG Yansong(Changchun University of Science and Technology,College of Electronic Information Engineering,Changchun 130022,China;Institute of Space Photo-Electronic Technology,Changchun University of Science and Technology,Changchun 130022,China)

机构地区：[1]长春理工大学电子信息工程学院,吉林长春130022 [2]长春理工大学空间光电技术研究所,吉林长春130022

出　　处：《红外技术》2018年第9期887-893,共7页Infrared Technology

基　　金：吉林省科技厅重点科技研发项目(20180201002SF)

摘　　要：为提高相机动态范围,提出了一种基于相机光电参数的图像融合方法。该方法首先通过对相机的转换增益以及黑电平偏移进行测试确定融合系数,得到高动态范围图像,经过灰度拉伸后得到可显示的图像。从像质评价方面选取3组不同光强下融合图像,分别从主观和客观两方面与2种传统的融合算法进行比较同时将相机融合前后的动态范围进行了比较。选用辰芯G2020型互补性氧化金属半导体(Complementary Metal Oxide Semiconductor,CMOS)图像传感器完成实验测试,融合前高增益模式以及低增益模式下相机动态范围分别为59.2 dB和66.1 d B,经该方法处理后相机动态范围达到87.5dB,且成像质较好,同时本文方法计算量小,易于硬件实现。To improve the dynamic range of a camera,an image fusion method based on photon transfer curve is proposed in this paper.First,the method determines the fusion coefficient by testing the conversion gain and the black level shift of the camera to obtain a high dynamic range image,the image that can be displayed after a grayscale extension.Three groups of fusion images with different light intensities were selected for image quality evaluation.Compared with two traditional fusion algorithms from subjective and objective aspects,the dynamic range before and after the camera fusion was compared too.The Chen Xin G2020 CMOS image sensor was used to complete the experimental test.While the dynamic range of the camera is 59.2 dB and 66.1 dB in the high-gain and low-gain modes,respectively,it is 87.5 dB using the image fusion method with good imaging quality.At the same time,the computational cost of the method is low and the hardware is easy to implement.

关 键 词：图像动态范围 像素级图像融合 光子转移曲线 像质评价 

分 类 号：TN911.73[电子电信—通信与信息系统]