Fertilizer Recommendation for Rice Crop based on NPK Nutrient deficiency using Deep Neural Networks and Random Forest Algorithm

Nutrient deficiency has a noteworthy impact on agriculture which results in reduced plant quality in turn reduced crop yield. A plant can have Multiple deficiencies at the same time so there is need of suggesting an appropriate Fertilizer considering all nutrient deficiencies in it. In this paper we suggested suitable fertilizer using random forest for detected nutrient deficiency from leaf images using different Neural Networks. We trained our dataset using four different Convolution Neural Networks and selected the network with highest accuracy and then the output of this network is sent to the trained model of Random Forest which suggests the suitable fertilizer. Since Potassium(K), Phosphorous(P) and Nitrogen(N) are three key nutrients required for Rice plant we considered the image dataset having leaves with N, P, K deficiencies.