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 6 几种算法在U45数据集上的定量评估结果

Tab.6 Quantitative evaluation results of several algorithms in U45

算法 PSNR SSIM UIQM UCIQE CLAHE 16.8693 0.6877 2.6946 0.3805 UDCP 14.4794 0.7231 2.9653 0.3694 Retinex 13.2449 0.6444 2.8903 0.3594 IBLA 18.1902 0.7778 2.8492 0.4025 CycleGAN 21.1015 0.8083 3.2103 0.4188 FUnIE-GAN 20.5327 0.7818 3.2946 0.4245 LANet 22.6146 0.8193 3.3635 0.4454 SCEIR 23.5601 0.8231 3.3703 0.4429 MGCycleGAN 23.7336 0.8249 3.4167 0.4586