基于深层残差网络的山区DEM超分辨率重构

Super-resolution Reconstruction of DEM in Mountain Area Based on Deep Residual Network

针对大区域高分辨率数字高程模型(DEM)数据较难获取、超分辨率重构(降尺度)较低分辨率的DEM精度不高、难以满足实际需要的问题,提出一种对起伏特征较明显的山区DEM超分辨率重构的方法。利用较深层的神经网络充分学习高低分辨率DEM之间的非线性映射关系;为了降低训练难度,结合残差学习的方法进行数据训练。将双立方插值法、稀疏混合估计法重构的DEM及提取的坡度结果分别同深层残差网络法的结果进行对比,结果表明,3种方法DEM结果的差值平均值分别为0.41、0.34、0.34 m,RMSE分别为0.5945、0.5715、0.4869 m;坡度结果的差值平均值分别为3.02°、2.04°、1.99°,RMSE分别为3.6498°、3.1360°、2.7387°;处理时间分别为0.052、663.39、2.16 s。研究表明,对于10、20、40 m的DEM,本文方法在空间分布和误差方面优于其他方法,在耗时效率上也优于稀疏混合估计法,适合应用于梯田等地形复杂的区域进行超分辨率重构。

High-resolution digital elevation model(DEM)in large districts is difficult to be acquired due to the limitation of cost and technology.Usually,it can be obtained by super-resolution reconstruction(downscale)from low-resolution DEM.However,the accuracy of the DEM generated by conventional downscale methods is insufficient.With the development of image downscale,convolutional neural network(CNN)has achieved success.To improve DEM accuracy,a very deep convolutional networks super-resolution method(VDSR)was designed to reconstruct the terrace DEM with obvious undulation characteristics.The deep neural network was used to learn nonlinear mapping between high-resolution DEM and low-resolution DEM,at the same time,residual learning method were used to reduce training difficulty.In order to compare,bicubic interpolation method,sparse mixed estimation method and VDSR method were used to reconstruct the DEM and slope.The slope data were extracted from the DEM results.The mean value of DEM difference of three methods were 0.41 m,0.34 m and 0.34 m,respectively.The RMSE of DEM were 0.5945 m,0.5715 m and 0.4869 m,respectively.The mean value of slope difference of three methods were 3.02°,2.04°and 1.99°,respectively.The RMSE of slope were 3.6498°,3.1360°and 2.7387°,respectively.The running time were 0.052 s,663.39 s and 2.16 s,respectively.By comprehensive comparison,for 10 m,20 m and 40 m DEM,the result showed that VDSR method had great advantage in spatial distribution,error and running time,and it was suitable for super-resolution reconstruction in areas with complex terrain such as terrace.