Abstract:
Nitrogen (N), Phosphorous (P), and Potassium (K) are considered the most important nutrients and are essential components in the soil affecting the growth and yield of crops. For optimal growth of the plant, the nutrients N, P, and K present in the soil must be in a balanced proportion. However, based on the parent material (like sand, peat, and clay), climatic conditions, and the differences in the past management of the crop residues and the use of fertilizers, the farmers need to know the accurate proportions of N, P, and K to maximize the crop growth, production, and yield. Therefore, its measurement to maintain an accurate balance is crucial. In this paper, two methods to estimate N, P, and K in the soil are proposed which can provide results in real-time without the need for any chemicals. The first method makes use of electrical conductivity and pH sensors to measure these parameters from the soil and uses machine learning techniques to estimate the N, P, and K values. The second method makes use of optical sensors to measure the amount of light absorbed and reflected by the soil and uses regression techniques to estimate N, P, and K. In both cases, the N, P, and K values are classified into different classes. We obtain more than 75% accuracy in both cases. A hand-held electronic device to measure N, P and K can be easily designed using these techniques. The proposed schemes can optimize fertilizer usage as well as assist farmers in an economical and efficient crop yield.
