Downscaling Gridded DEMs Using the Hopfield Neural Network

Abstract

In this paper, a new method was investigated to enhance remote sensing images by alleviating the point spread function (PSF) effect. The PSF effect exists ubiquitously in remotely sensed imagery. As a result, image quality is greatly affected, and this imposes a fundamental limit on the amount of information captured in remotely sensed images. A geostatistical filter was proposed to enhance image quality based on a downscaling-then-upscaling scheme. The difference between this method and previous methods is that the PSF is represented by breaking the pixel down into a series of sub-pixels, facilitating downscaling using the PSF and then upscaling using a square wave response. Thus, the sub-pixels allow disaggregation as an attempt to remove the PSF effect. Experimental results on simualted and real datasets both suggest that the proposed filter can enhance the original images by reducing the PSF effect and quantify the extent to which this is possible. The predictions using the new method outperform the original coarse PSF-contaminated imagery as well as a benchmark method. The proposed method represents a new solution to compensate for the limitations introduced by remote sensors (i.e., hardware) using computer techniques (i.e., software). The method has widespread application value, particularly for applications based on remote sensing image analysis.

Digital Surface Model Super-Resolution by Integrating High-Resolution Remote Sensing Imagery Using Generative Adversarial Networks

“…Generally, DTM is a representation of various topographic features, such as slope, gradient, valley lines and ridges, organized in a onedimensional or multi-dimensional feature vector space overlaid on a two-dimensional geographic space [1]. DEM only characterizes the elevation of bare-Earth surface by removing all natural and built objects [9,10]. In contrast, DSM not only provides the elevation of bare-Earth surface but also includes the elevation of natural and built objects that exist above the terrain, such as buildings and vegetation [7,8].…”

Impacts of Resampling and Downscaling Digital Elevation Model and Its Morphometric Factors: A Comparison of Hopfield Neural Network, Bilinear, Bicubic, and Kriging Interpolations

“…Numerous methods and techniques have been proposed to increase the DEM accuracy for natural hazard predictions [14,19]. Some of these are based on the down-sampling of DEM to different resolutions, and they use these DEMs to calculate the input factors for natural hazard models in different regions [20]. Thus, the accuracy and grid DEM resolution can be increased using resampling and downscaling methods [19,20].…”

“…Some of these are based on the down-sampling of DEM to different resolutions, and they use these DEMs to calculate the input factors for natural hazard models in different regions [20]. Thus, the accuracy and grid DEM resolution can be increased using resampling and downscaling methods [19,20]. The common approaches for downscale resampling are bilinear, bi-cubic, and Kriging [21].…”

“…These methods are mostly based on approaches proposed for image super-resolution [22][23][24][25][26]. The result shows that the DEM accuracy has improved regarding root mean square error (RMSE) and the closeness to the reference data [19,20]. However, although showing potential for improvement of super-resolved DEMs' accuracy, applying these methods still required large amounts of training data and very high-capacity hardware to build complex neural network models.…”

“…In recent research, we proposed a method for downscaling DEM using Hopfield Neural Network (HNN) [20] with promising results. This method not only improves the accuracy of DEMs but also is computationally efficient and can be performed on personal computers without incurring high costs.…”