An Approach to Monitor Urban Growth through Deep Learning based Change Detection Technique using Sentinel-2 Satellite Images

Urbanization, which refers to the growth of a city or a built-up region, is the process of transitioning from rural to urban regions. Urbanization is primarily brought on by natural population expansion, rural-to-urban migration, and annexation. Urban monitoring is the practice of observing and quantifying stability or change over a range of timescales inside cities. In addition to being a useful tool for studying the Earth's surface, satellite images may also be used to estimate different aspects behind rapid urbanization. Sentinel-2 images from 2017 to 2022 are utilised in this paper to assess the urban development in Lucknow, the Indian city of Uttar Pradesh's state. Hence, image indices like normalized difference built-up index (NDBI), normalized difference vegetation index (NDVI), and built-up index (BUI) extracted from Senitnel-2 images are utilized to classify the various land cover classes by applying the supervised classification technique. The classified images are further used to measure the urban areas and a deep learning-based change detection approach has been developed to detect the changes in urban areas that have been transformed into urban areas from agriculture or other areas. After conducting a critical analysis of change areas, it is observed that 52% of positive changes have been observed in urban areas, which shows that a large area has been transformed from non-urban to urban areas during the assessment period. In the future, an efficient and accurate urban monitoring system can be developed by using the proposed approach to establish sustainable and planned urban areas.