Automated Plant Pathological Foliar Disease Identification using Deep Learning Approach

Abstract

Agriculture is considered as one of the backbones to the nation. Food is the essential item for all living being. Farmers plays an important role in cultivating the crops, plants and taking good care of them. Disease found in plant is very common during farming. Many farmers could able to recognize the disease visually and treat them accordingly. But if the plant farming is very big, then the manual inspection of diagnosing the disease in plant is a time-consuming process and could not able to treat them at the initial stage. Hence, we must automate the diagnosing procedure of predicting the disease which helps the farmers to treat the plant with disease at earlier stage. We are proposing deep learning algorithm which can able to predict various kind of diseases found in the apple tree efficiently. We have preprocessed the input images and different augmentation is been carried out to avoid overfitting of the model. Trained the model using Residual neural network and could able to achieve 96.6% accuracy in the test dataset.