基于改进CycleGAN的多失真类型水下图像增强

吕振鸣,董绍江,夏宗佑,牟小燕,王明权

Multi-distortion type underwater image enhancement based on improved CycleGAN

Zhenming LV,Shaojiang DONG,Zongyou XIA,Xiaoyan MOU,Mingquan WANG

表 3 几种算法在EUVP和UIEB中蓝色失真图像上的定量评估结果

Tab.3 Quantitative evaluation results of blue distorted images on EUVP and UIEB using several algorithms

算法 PSNR SSIM UIQM UCIQE CLAHE 14.8243 0.6868 2.9543 0.3337 UDCP 14.7749 0.6927 2.8836 0.3779 Retinex 16.7380 0.7565 3.1448 0.3863 IBLA 19.5472 0.7950 3.2923 0.4309 CycleGAN 21.4092 0.8199 3.3596 0.4313 FUnIE-GAN 22.5418 0.8022 3.3146 0.4647 LANet 24.6979 0.8427 3.7755 0.4809 SCEIR 23.1637 0.8399 3.4921 0.4679 MGCycleGAN 24.6334 0.8481 3.7208 0.4746