Harnessing PDDC-Net for Efficient Plant Disease Detection and Classification through Deep Learning

Abstract

A country's inventive growth is dependent on the agricultural sector. Agriculture, the foundation of all nations, offers food and raw resources. Agriculture is hugely important to humans as a food source. As a result, plant diseases detection has become a major concern. Traditional methods for identifying plant disease are available. However, agriculture professionals or plant pathologists have traditionally employed empty eye inspection to detect leaf disease. This approach of detecting plant leaf disease traditionally can be subjective, time-consuming, as well as expensive, and requires a lot of people and a lot of information about plant diseases. It is also possible to detect plant leaf diseases using an experimentally evaluated software solution. Currently, machine learning and deep learning are using in recent years. This work is focused on implementation of Plant disease detection and classification (PDDC-Net) using deep learning models. The preprocessing operation also performed to remove the different types of noises, which also normalizes the dataset images. Further, the PDDC-Net implements the operation using residual network based convolutional neural network (ResNet-CNN) for feature extraction and classification. Experimental results have shown that proposed PDDC-Net model achieved a good accuracy rate for plant leaf disease detection and classification