Stubble (Crop Residue) Burning Detection Through Satellite Images Using Geospatial Foundation Model: A Case Study in Punjab, India (Proposals Track)

Rajiv Ranjan (Plaksha University); Ying-Jung Chen (Georgia Institute of Technology); Shashank Tamaskar (Plaksha University); Anupam Sobti (IIT Delhi)

Paper PDF NeurIPS 2024 Recorded Talk Cite
Computer Vision & Remote Sensing Agriculture & Food

Abstract

Stubble burning is a significant environmental challenge globally, with widespread implications for air quality, greenhouse gases emission, soil degradation and health issues. This practice is particularly prevalent in agricultural regions across the world, though its impacts are notably severe in the northern India. This proposed work focuses on improving the detection of stubble (crop residue) burning in Punjab (India), through using the geospatial foundation model. This study leverages a series of satellite images where stubble burning incidents have been documented. By refining the model to incorporate local environmental factors, this study aims to improve the accuracy of stubble burning detection, thereby contributing to a scalable solution for real-time monitoring and intervention in crop residue burning practices worldwide.