【摘要】 针对红外遥感图像的超分辨率重建问题,提出了一种深度卷积残差学习网络。通过加深网络来扩大感受野,使在重建过程中能够利用到更多的邻域信息,提高了重建图像质量。使用残差网络,使模型能够更好的学习到先前信息,在提高数据稀疏性的同时,提高了算法效率,在亚像素卷积层提高红外遥感图像的尺度。实验结果表明,与常用深度学习网络相比,该网络得到的重建图像具有更好的重建效果。更多还原

【Abstract】 Aiming at the problem of super-resolution reconstruction of infrared remote sensing images,a deep convolution residual learning network is proposed. More neighborhood information can be used in the reconstruction process by deepening the network to expand the receptive field so as to improve the quality of the reconstructed image. Using residual network,the model can learn from the previous information better and improve the data sparsity to improve the efficiency of the algorithm. And the scale of the infrared remote sensing image is improved in the subpixel convolution layer. The experimental results show that the reconstructed images obtained by the proposed network have better reconstruction effect compared with that by general deep learning networks.更多还原