The poverty and vulnerability index assesses livelihood vulnerability on a large scale, crucial for improving living standards. Poverty reflects limited access to resources and income opportunities. To address this, we're leveraging GIS software and deep learning to map poverty levels using satellite imagery and socio-economic data in India.
Current solutions often rely on outdated or limited datasets, hindering accurate socio-economic assessments. Additionally, low spatial resolution in poverty mapping satellite imagery often overlooks localized poverty pockets within vast regions.
Our approach enhances accuracy by integrating GIS software and deep learning. By employing convolutional neural networks (CNNs) and recurrent neural networks (RNNs), we create a hybrid model that integrates satellite imagery and socio-economic data for precise poverty mapping and prediction. This integration allows us to aggregate spatial and temporal data at various scales and resolutions, ensuring the relevance and effectiveness of our model.