Fertilizer Recommendation with Stress Monitoring in Agriculture

Stress monitoring in agriculture is essential to increase farm production, encourage sustainability, and assist farmers while making choices that were ethical to food security and economic stability. The purpose of this study is to create a machine learning model to determine whether the plant is under stress or not. And if the plant experiences stress, what deficiency caused it, and recommend the appropriate fertilizer. We utilized the crop recommendation dataset and the fertilizer recommendation dataset as two datasets for this. Pair of classifiers to suggest fertilizer for the crop under stress, random forest and ANN are employed. As input parameters for classifiers, nitrogen, potassium, phosphorus, temperature, humidity, moisture, and crop kind are all employed. The Artificial Neural Networks (ANN) classifier has the highest accuracy (84%) when the overall performance of all the classifiers is considered. This study demonstrates algorithms can identify changes in the surroundings of the plant and forecast how the plant will react to these changes by analyzing the data. By comparing information produced using features and segments, our farmers can classify the stress and administer fertilizer. The automation of the stress detection process will boost agricultural productivity and benefit society.