Identification of Nutrient Deficiency in Rice Leaves using DenseNet-121

India is renowned for rice farming. Rice as a most important crop has a huge bang on people's health since many cultivate it for a living, it provides jobs, and many small industries are dependent on its cultivation. There is a risk of significant losses due to various diseases and pets that can affect this planting. With the help of modern science and technology a number of research projects and methods are proposed to improve rice yields. In the plant nutrient system, leaf uniqueness varies according to potassium (K) phosphorus (P), and nitrogen (N) dietary pressures. Here we imported the dataset and performed data augmentation to enhance the mass of dataset and to obtain more accuracy. Then we applied DenseNet-121 which is one of the DenseNet group of models designed to perform image classification. It is then evaluated with Adam, RMSprop, Gradient descent optimizers which has given 95.40%, 98.49%, 93.50% of accuracy. This entire approach is done to identify the type of fertilizer that should be used in order to increase the yield of crop.