Development of a web-based land-use mapping

Abstract

Local Government Units (LGUs) in the Philippines are mandated to conduct land classification mapping of their areas which brings up an issue about a time-consuming and expensive data gathering. This paper aims at achieving a land use classification system that correlates with the traditional way of mapping of LGUs. This study introduces the development of a web application that can classify land use (Agricultural, Commercial, Residential, and Industrial) in different areas of Greater Manila. The system involves the utilization of freely accessible satellite images from Google Maps and benchmark dataset from the EuroSAT dataset for the categorization of the areas through deep learning using Convolutional Neural Network (CNN). Evaluation results show that the system in a web application could help with the mapping of the Greater Manila area with the web application being simple yet informative and efficient.