Prajakta Kadukar
Computer science
Hometown: Tempe, Arizona, United States
Graduation date: Fall 2027
FURI | Fall 2025
Optimizing Hyperspectral Soil Data Analysis
This project uses artificial intelligence (AI) to improve how hyperspectral soil images are analyzed for environmental research. By developing encoder–decoder models, the study compresses large hyperspectral datasets and reconstructs them with minimal information loss, allowing faster and more accurate analysis. These models will be evaluated through reconstruction accuracy and visual feature preservation to identify meaningful soil patterns and carbon content. The results aim to enhance carbon mapping and support sustainable land management practices. Future work will explore integrating real-world soil data for further validation.
Mentor: Nakul Gopalan