A Cost Effective and Field Deployable System for Soil Macronutrient Analysis Based on Near-Infrared Reflectance Spectroscopy

Abstract

Near-infrared reflectance (NIR) spectroscopy is a technique that shows many possibilities in the field of testing chemical and physical properties of soil. This paper is aimed to design a prototype of an integrated optoelectronic sensing system capable of estimating soil macronutrients- Nitrogen (N), Phosphorus (P), Potassium (K), pH and Organic Carbon (OC) which play major role in the process of plant growth. 24 soil samples have been collected from farming fields of Birbhum, West Bengal and calibrated using near infrared reflectance spectroscopy. Partial least square (PLS) model along with Standard Normal Variate (SNV) pre-treatment technique has been developed to extract features from the spectroscopic plot which is required for development of field deployable system. The handheld interface attempted in this paper comprises of a simple soil sieve, the output of which is fed in the optoelectronic system which exposes the sieved soil sample to selected IR wavelengths and processes the photodiode output using ARDUINO microcontroller. The trained PLS model embedded in the microcontroller provides estimation of the essential macronutrients with appreciable accuracy. This indigenous system is expected to enhance the scope of precision agriculture in rural India.