June 22, 2020
OSA Technical Digest

Potential for predicting soil properties using low cost near-infrared spectroscopy and machine learning

Wenbo Wang, Liming Hu, Benjamin K. Wilson, Matthew D. Keller

Using two public soil spectral libraries, we examined potential for low-cost nearinfrared devices for in-field testing. Results showed strong promise with conventional neural networks, comparable to previously published results with deep learning.

View Journal