부산대학교 도서관

In current times, diseases in the crops are a leading cause of the decrease in agricultural yield. This affects the agricultural economy badly. The early detection of the diseases is required to ensure that the yield is good. With the increase in the population, the burden on farmers is also increasing. The model that is proposed in this paper solves the problem of early disease detection in the pomegranate fruit crop. This model is a combination of Convolutional Neural Network and Random Forest Classifier. The CNN extracts the features of the images and then it is fed to the Random Forest classifier to classify the images. The images are classified into healthy fruit and five diseased fruit categories namely, Aspergillus Fruit Rot, Bacterial Blight, Anthracnose, Cercospora, and Pseudocercospora Punicae. Image preprocessing techniques such as cropping, flipping, enhancement of contrast, zooming, etc. are used to improve the quality of the dataset. The images used for training the model are collected from both primary and secondary resources. The proposed model is a reliable solution for the early detection of the diseases in the pomegranate fruits.