Analysis of Image Preprocessing Effects in a Landsat Image Simulation

• Published in: KSCE journal of civil engineering Vol. 24; no. 7; pp. 2186 - 2192
• Main Authors: Seo, Dae Kyo, Eo, Yang Dam, Paik, Geun Woo
• Format: Journal Article
• Language: English
• Published: Seoul Korean Society of Civil Engineers 01.07.2020; Springer Nature B.V; 대한토목학회
• Subjects: Civil Engineering; Engineering; Environmental effects; Geotechnical Engineering & Applied Earth Sciences; Image processing; Industrial Pollution Prevention; Landsat; Methods; Preprocessing; Radiation; Random variables; Regression analysis; Regression models; Remote sensing; Satellite imagery; Simulation; Surveying and Geo-Spatial Engineering; Vegetation; 토목공학; normalization; Random forest regression; Gaussian process regression; Image preprocessing; Multivariate linear regression; Phenological and radiometric; Image simulation
• ISSN: 1226-7988; 1976-3808
• DOI: 10.1007/s12205-020-0056-8

Optical remote sensing has limitations in obtaining images due to weather and environmental effects, so these limitations must be overcome to produce time-series image data. As an alternative to this, research are being conducted to simulate images at a specific time for which a specific image is needed. The purpose of this study is to improve the results of this process by preprocessing the input images of a multiple linear regression model alongside other remote sensing image simulation methods. Specifically, the input images, which are applied to a multi-linear regression equation, are preprocessed for phenological and radiometric normalization by a random forest regression model. The experimental results show that the proposed method is superior to the conventional methods both visually and quantitatively.