Disease Detection Using Transfer Learning In Coffee Plants

Every year, various biotic and abiotic factors cause a lot of damage to plants leaves. Various methods have been introduced in the past to tackle these problems and among them, Convolutional Neural Networks(CNN) have proven to be very efficient and accurate. This research aims to timely detect leaf diseases in coffee plants and hence help in increasing the production of coffee. Coffee plants suffer from various leaf diseases like Leaf Miner, Phoma which have visible symptoms and can be classified. CNN can detect the different patterns present on the leaf and help classify them. We have proposed a model which uses CNN along with transfer learning. Also, Data augmentation has been used to enlarge the data set. The accuracy achieved is 98.51%.