IoT based Soil Nutrients Analysis and Monitoring System for Smart Agriculture
M. Pyingkodi, K. Thenmozhi, M. Karthikeyan, T. Kalpana, Suresh Palarimath, G. Bala Ajith Kumar
Abstract
Soil fertility is an important factor in determining soil quality as it reflects how well the soil can support plant growth in agriculture. Soil sensor and Arduino can be used to quickly determine the nutrient content of the soil. Nitrogen, phosphorus, and potassium are all considered as important nutrient source components. These components should be measured in order to determine how much extra nutrient content should be added to the soil in order to increase the crop fertility. Soil fertility can be detected by using NPK sensors. Soil nutrient concentration data can help us to determine whether the soil used to support plant production is nutrition deficient or abundant. The nutrient content of the soil samples can be obtained in various ways by using sensing element or mass spectrogram. However, the spectral analysis method is inconvenient, where the records are only 60-70% accurate. By comparing the spectrum analysis method with classic wet chemistry methods, the accuracy of the products needs to be fully resolved due to a scarcity of data. Hence, to detect soil nitrogen, phosphorous, and potassium, a soil NPK sensor should be used. By utilizing a soil NPK sensor, which is of limited cost, fast and easy, elevated, and transportable. Its advantage over a standard detection approach is that it provides extremely fast measurements with accurate data. This paper analyzes and compares different nutrient levels in soil by using kernel density estimation algorithm and machine learning.