Automized Mango Leaf Disease Detection Using Various Optimizer of CNN

Mango is one of the most grown fruits and is considered to be one of the most consumed around the world. Despite being so crucial economically and nutritionally the farmers are not aware about the different diseases that mango plants can adhere to. The research is intended to find out an efficient way for the farmers to determine the disease at an early stage and cure the plant accordingly so that it does not affect it for the whole life span. To achieve the objective deep CNN was chosen as it is the best in image classification, furthermore tuning of the optimizer was done to find the best fit for the model. Optimizers used were Adam, Adamax, SGD, RMSprop, Adadelta. After many experiments we determined that the best fit optimizer for the deep CNN is “Adamax” with an accuracy of 94%. This is an ideal accuracy for a classification model over a real life dataset.