Research on the method using Hopfield neural network to increase the resolution of the digital elevation model in grid form (Grid DEM)

https://tapchi.humg.edu.vn/en/archives?article=1002
  • Affiliations:

    Khoa Trắc địa - Bản đồ và Quản lý đất đai, Trường Đại học Mỏ - Địa chất, Việt Nam

  • *Corresponding:
    This email address is being protected from spambots. You need JavaScript enabled to view it.
  • Received: 10th-Jan-2019
  • Revised: 20th-Feb-2019
  • Accepted: 29th-Apr-2019
  • Online: 29th-Apr-2019
Views: 1070
Downloads: 462
Rating: 5.0, Total rating: 45
Yours rating

Abstract:

Nowadays, the digital elevation model in grid form (Grid DEM) has many applications in the fileds of economic and social life, especially in natural resource management and environmental protection. DEMs with higher spatial resolution will give more accurate and informative results. However, building them is expensive and takes a lot of time and effort. This paper proposes an algorithm that increases the spatial resolution of DEM in a new approach. In this approach, a model for smoothing and increasing the digital elevation model in grid form using minimum variogram value and a elevation constraint was proposed. The model was intergrated into a simple Hopfield Neural Network (HNN) model in which each pixel of a DEM are divided into m×m sub-pixels. The elevation of each sub-pixel are calculated based on minimum variogram value and an elevation constraint which can be stated that the everage of elevation of all subpixels located within a pixel must be equal to the elevation of the original pixel. The activation function used in this model of HNN is a simple linear function. Experimental results in Mai Pha, Lang Son, Vietnam showed the feasibility of this algorithm

How to Cite
Nguyen, H.Thu Thi 2019. Research on the method using Hopfield neural network to increase the resolution of the digital elevation model in grid form (Grid DEM) (in Vietnamese). Journal of Mining and Earth Sciences. 60, 2 (Apr, 2019).