基于生成式对抗网络和多级小波包卷积网络的水下图像增强算法
|
温佩芝,陈君谋,肖雁南,温雅媛,黄文明
|
Underwater image enhancement algorithm based on GAN and multi-level wavelet CNN
|
Pei-zhi WEN,Jun-mou CHEN,Yan-nan XIAO,Ya-yuan WEN,Wen-ming HUANG
|
|
表 2 本研究算法和其他算法在EUVP数据集上的各项指标对比 |
Tab.2 Comparison of indicators between proposed algorithm and other algorithms in EUVP dataset |
|
算法 | SSIM | PSNR | UIQM | UCIQE | NIQE | 原始图像 | 0.814 | 29.320 | 2.549 | 0.550 | 8.192 | IBLA | 0.464 | 15.040 | 1.045 | 0.698 | 11.930 | UDCP | 0.734 | 19.430 | 2.028 | 0.588 | 7.958 | ULAP | 0.694 | 23.190 | 2.213 | 0.577 | 8.073 | RGHS | 0.778 | 26.660 | 2.337 | 0.598 | 8.084 | Sea-thru | 0.772 | 21.380 | 2.593 | 0.591 | 8.123 | UWGAN | 0.880 | 29.020 | 3.345 | 0.554 | 7.294 | FunieGAN | 0.796 | 26.300 | 3.236 | 0.559 | 11.820 | FunieGAN-up | 0.768 | 26.200 | 3.029 | 0.557 | 8.259 | 本研究算法 | 0.932 | 30.610 | 3.196 | 0.570 | 7.212 | 本研究算法-无色彩校正 | 0.851 | 28.910 | 3.003 | 0.553 | 7.558 |
|
|
|