基于CNN和Transformer聚合的遥感图像超分辨率重建
胡明志,孙俊,杨彪,常开荣,杨俊龙

Super-resolution reconstruction of remote sensing image based on CNN and Transformer aggregation
Mingzhi HU,Jun SUN,Biao YANG,Kairong CHANG,Junlong YANG
表 1 AID测试数据集6个随机场景下不同模型的PSNR和SSIM指标
Tab.1 PSNR and SSIM metrics of different models on six randomly selected scenes from AID test dataset
模型飞机场城市农田停车场运动场港口
PSNR/dBSSIMPSNR/dBSSIMPSNR/dBSSIMPSNR/dBSSIMPSNR/dBSSIMPSNR/dBSSIM
Bicubic26.220.687123.730.618229.490.734719.730.545125.450.690522.230.6699
Swinir24.430.659822.190.586728.150.714217.790.512624.200.666620.090.6401
CDC24.820.627222.500.586726.870.660120.040.554724.110.655021.710.6691
DAN25.700.692223.600.627728.750.730719.720.571125.310.696021.450.6682
real-Esrgan27.810.729624.820.676830.200.766421.060.646826.330.733822.840.7316
BSRGAN27.740.731825.240.675430.800.770421.790.641327.100.735123.570.7328
MM-realSR27.830.764925.640.722530.440.785222.420.694927.640.782623.950.7699
realHAT-TG27.760.747025.340.693330.530.773921.960.662226.930.749523.540.7426
本文模型29.500.785727.270.746232.430.806823.450.729229.570.806025.390.7860