Comparison and Analysis of Algorithms used for Detecting Slums in Satellite Images

Pallavi Saindane, Gayatri Ganapathy, Neha Prabhavalkar, Nilesh Bhatia, Aishwarya Vaidya

This paper presents a comparison of some neural network based approaches for analyzing satellite images. The motivation for this study comes from the potential use of such analysis for identification of slums from satellite images. Slums, also formally termed as informal settlements, can be identified from satellite images using image segmentation as well as object detection techniques. Policymakers often spend substantial time and resources to discover certain regions of interest. This paper presents a brief study and comparison of some known algorithms that can be helpful for classifying residential areas as developed and underdeveloped using satellite images as input. The approach can be further improved to identify the slums from satellite images. Such a process is then readily scalable and can reduce the time and resources spent by policymakers for this analysis. This can ultimately help policymakers in developing appropriate policies that lead to economic progress.

https://doi.org/10.31871/IJNTR.5.2.14