Pomegranates Fruit Disease Classification Using EfficientNet Deep Learning Model

Abstract

Pomegranate stands among the most important valuable fruits globally because it contributes significantly to food security and economic development throughout agricultural communities. The combination of diseases including alternaria alongside anthracnose bacterial blight and cercospora creates severe problems that lead to 75% yield reduction and deteriorated quality and substantial financial loss for growers. Manual inspections along with expert consultations currently fail to detect diseases effectively because they consume significant time while being subjective and their responses are usually delayed for managing the condition. The proposed research introduces the EfficientNet deep learning model for pomegranate disease classification because it exhibits high accuracy alongside efficient computing capabilities. A model based on the EfficientNet has been devised to classify disease affected fruits. The model was train and tested using Pomegranates fruits diseases dataset for deep learning model. The study's experimental results indicate that our proposed work attains an accuracy of 98.73%. The Proposed Solution functions as one of many agricultural technology developments through its deep learning system that delivers scalable accessible classification of pomegranate diseases while achieving high performance. Additionally the models proposed in this research can be now expanded to enable predict in real-time the quality of vegetable and fruits, by use of IoT devices. Moreover, the proposed model could be progressed into an android app where smartphone would enable farmers to take images of their plants or fruits in real time to receive instant classification reports related to plant or fruits disease.