Modeling the development trends of Land use/Land cover in urban areas using remote sensing data and artificial intelligence algorithms

Abstract

Research on Land use/Land cover (LULC) change is one of the important contents serving the monitoring, management, and planning of natural resources, especially for urban areas. LULC change is the result of the synthetic relationship between natural and social factors. In this study, Landsat satellite image data in 2010, 2015 and additional information layers are used to predict LULC change in the Ho Chi Minh City area for 2020, then compared with the LULC classification results in 2020 to evaluate the accuracy and calibrate the model. Machine learning techniques including Random Forest (RF), Support Vector Machine (SVM), Classification and Regression Tree (CART), and maximum likelihood classification method are tested to classify LULC from Landsat satellite images, then select the method with the highest accuracy. To predict the future development trend of LULC, the study combines the use of the Cellular Automata (CA) mathematical model and artificial intelligence models, including logistic regression (LR) and artificial neural network (ANN), then selects the method with higher accuracy to predict the development trend of LULC in 2025 and 2030.