Privacy Preservation of Plant Disease Detection using Hashing based Convolution Neural Network

Agriculture is the main source of the Andhra Pradesh state economy. The agriculture industry heavily depends on crop production. But, many of the plants suffer from several diseases. Contaminated crops are mostly to blame for food loss, which consequently slows down the development of the agricultural industry. Recently in 2022, Both AP & Telangana face 40-80% of damage to chili crops due to pest attacks also informed in Parliament. It is very challenging for farmers to identify diseases in plants at the beginning stages. The assembly of the product and the market's economic value will be severely harmed by erroneous identification. There are many existing solutions to this problem associated with a mobile phones. But most of them are limited to the detection of plant diseases with a smaller number of categories of plants. The goal of this system is to detect plant diseases from images of the plant's leaves by developing a mobile application for the Andhra Pradesh state region, which will assist farmers in using appropriate pesticides for that disease. This involves the detection of plant disease based on the classification of leaf images using the Convolutional Neural Network (CNN) model. So, farmers can easily detect plant diseases in the early stages and use pesticides to treat those plants, which helps gain the economy. The dataset is the PlantVillage dataset under Kaggle which contains leaf images of various plants. The model is enhanced by hashing using the SHA-256 algorithm. The input data to the hash function is a leaf image which is in bytes format. The proposed model achieves integrity by using SHA-256. This model allows the identification of plant diseases that are tamper-proof.