摘要
全局增强未过多考虑图像局部细节,单独使用局部增强会损失图像的亮度。多尺度Retinex算法既能保持动态范围压缩又能实现整体再现,但易出现光晕现象,色彩失真,图像细节信息容易丢失,且运算时间较长。针对上述不足,提出了图像融合增强算法:采用离散小波变换实现对局部增强、全局增强图像及改进多尺度Retinex增强图像的融合。改进多尺度Retinex算法使用引导滤波替换高斯滤波估计反射分量。通过对仿真结果定性以及在峰值信噪比,结构相似度等方面进行定量对比,实验结果证明较原图和MSR(multi-scale-retinex)算法处理图像,该算法针对图像昏暗、提取信息不全问题具有更好的显示效果。
Global enhancement does not give much consideration to local detail in the image,and local enhancement alone loses the brightness of the image.The multi-scale Retinex algorithm can not only maintain dynamic range compression but also achieve overall reproduction,but it is prone to halo phenomenon,color distortion,easy loss of image detail information,and long operation time.In view of the above shortcomings,this paper proposes an image fusion enhancement algorithm:discrete wavelet transform is used to achieve the fusion of local enhancement,global enhancement image and improved multi-scale Retinex enhanced image.The improved multiscale Retinex algorithm uses guided filtering instead of Gaussian filtering to estimate the reflection component.By qualitatively comparing the simulation results and the peak signal-to-noise ratio and structural similarity,the experimental results show that compared with the original image and MSR(multi-scale retinex)algorithm processing images,the algorithm has better display effect when processing images,ensuring contrast enhancement and complete image detail display.
作者
单永奇
刘文波
滕子煜
王倩岚
SHAN Yong-qi;LIU Wen-bo;TENG Zi-yu;WANG Qian-lan(Nanjing University of Aeronautics and Astronautics,Nanjing 211000,China;Non Destructive Testing and Monitoring Technology for High speed Transport Facilities Key Laboratory of Ministry of Industry and Information Technology of the People′s Republic of China,Nanjing 211000,China)
出处
《航空计算技术》
2023年第5期81-84,88,共5页
Aeronautical Computing Technique
基金
国家自然科学基金项目资助(61871218)
国家重点研发计划项目资助(2018YFB2003304)
中央高校基本科研业务费项目资助(NJ2019007
NJ2020014)。
