The combination of the NDBI and machine learning algorithms to classification the development of urban areas in Surabaya uses Landsat 8 Imagery.

Development is a transformation that occurs in order to better people’s lives. In this instance, efforts are required to monitor development progress swiftly and efficiently so that outcomes that support the government’s new policies are obtained. With a population of 2,848,583 people, Surabaya is the largest metropolitan city in Eastern Indonesia. An increase in population can impact Surabaya’s development and generate potential issues with urban planning development. Through these issues, remote sensing technology can efficiently and swiftly detect advances in Surabaya. This research study employs Landsat 8 image data with a combination of Normalized Build-Up Index (NDBI) as pre-processed pixels to segment into urban areas. It employs machine learning to classify urban areas to separate class types in depth. The findings show that the method successfully classified subclasses in residential and industrial areas. Random Forest (RF) outputs have an overall accuracy of 85%, whereas Decision Tree (DT) outputs have an overall accuracy of 84%. This study’s findings also indicate the possibility for practical applicability to various case studies, such as conducting detections in forests, plantations, mining, and coastal environments.