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 5 几种算法在EUVP和UIEB中绿色失真图像上的定量评估结果

Tab.5 Quantitative evaluation results of green distorted images on EUVP and UIEB using several algorithms

算法 PSNR SSIM UIQM UCIQE CLAHE 13.6256 0.7201 2.8658 0.3456 UDCP 18.8771 0.7924 3.0488 0.3817 Retinex 12.7404 0.6743 2.5165 0.3249 IBLA 20.0739 0.8097 3.3961 0.4194 CycleGAN 22.2757 0.8136 3.3433 0.4291 FUnIE-GAN 22.1953 0.8220 3.3651 0.4226 LANet 23.5396 0.8404 3.5889 0.4446 SCEIR 24.1619 0.8428 3.5358 0.4568 MGCycleGAN 24.6653 0.8594 3.6193 0.4774