Identification of Plant Species and Their Associated Diseases from Leaf Images using Machine Learning Approaches

Abstract

The automatic identification of plant diseases from leaf images remains a significant challenge for researchers. Plant diseases adversely affect growth, leading to reduced agricultural productivity and economic losses. Early and accurate disease detection is crucial for implementing timely preventive measures. Traditional image processing techniques have been widely used, but recent advancements in deep learning, particularly Convolutional Neural Networks (CNNs), have revolutionized image analysis. Deep learning architectures consist of multiple processing layers that learn hierarchical data representations, making them highly effective compared to conventional methods. This paper presents a methodology for identifying plant species and detecting diseases from leaf images using deep CNNs. Specifically, we adopt the GoogLeNet architecture, a powerful deep learning model, for disease classification. Transfer learning is utilized to fine-tune a pre-trained model, enhancing its performance. The proposed system achieves an accuracy of 85.04% in identifying four disease classes in plant leaves. Additionally, a comparative analysis with other models is conducted to demonstrate the effectiveness of our approach in improving accuracy and efficiency in plant disease detection.