InfraTech Journal of Sustainable Architecture and Civil Engineering
Urban Growth Monitoring in Kano State Using Normalized Difference Built-Up Index and Convolutional Neural Network Based Change Detection
Abstract
Kabiru Abdullahi Abdulhamid and Parluhutan Manurung
The Kano State, Nigeria urban expansion is of significant relevance to environmental sustainability, economic growth, and infrastructure planning. The Normalized Difference Built-up Index (NDBI) in this study is paired with a patch-based Convolutional Neural Network (CNN) to detect the urban expansion patterns between 2010 and 2015. Multitemporal Landsat data was processed with atmospheric correction, geometric correction, cloud masking, and NDBI derivation in Google Earth Engine. NDBI imagery that was pulled was categorized employing a CNN model trained on square image patches to enable precise builtup area change detection. Consequence was that the urban sprawl in Kano grew extremely rapidly, particularly between 2013-2015, because of increased population, economic expansion, and investment in infrastructure. Urbanization corridors and hotspots of growth were extremely well mapped by the CNN model that was also cofirmed by visual overlays and classification metrics. The study highlights the promise of integrating spectral indices and deep learning for change detection in rapidly transforming sub - Saharan African cities. Although valuable, the NDBI - only approach has limited capacity to capture mixed land uses and zones of transition, and future research is required that utilizes multiple indices and state - of - the - art machine learning techniques. The findings provide actionable insights for sustainable urban planning and regional policy interventions.