FURI | Fall 2025

Optimizing Hyperspectral Soil Data Analysis

Data icon, disabled. Four grey bars arranged like a vertical bar chart.

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.