Soil nutrient information extraction model based on transfer learning and near infrared spectroscopy

Abstract

This study aims to combine the transfer learning algorithm and near-infrared spectroscopy technology to build a soil nutrient information extraction model. We determine the soil nutrients with near-infrared spectroscopy technology; moreover, the characteristic bands of soil nutrients were also determined for preprocessing the soil nutrient information. Then, combined with the transfer learning algorithm, soil nutrient information extraction model was constructed. Simulation results showed that the proposed model could improve the efficiency of soil nutrient information extraction, and also obtain high-confidence feature extraction results. This will solve the problem of low-accuracy and efficiency of soil nutrient information extraction through traditional models.

Keywords

• Transfer learning
• Near-infrared spectroscopy
• Soil nutrient
• Information extraction model
• Straw method
• Principal component analysis