Abstract

Plant diseases pose a significant menace to agricultural output and global food security. Detecting these diseases promptly and precisely is pivotal for effective disease control and optimization of crop yield. This survey paper presents a comprehensive analysis of recent advancements in plant disease detection leveraging deep learning and wavelet neural network approaches. Thirty notable papers are reviewed, emphasizing the diverse methodologies and datasets employed. The analyzed papers collectively showcase the efficacy of convolutional neural networks, wavelet transformations, and deep learning models in accurately identifying and classifying plant diseases. While conventional Convolutional Neural Networks (CNNs) have proven effective in image-based disease detection, they might encounter challenges in capturing frequency-domain nuances within plant images. Several papers emphasize specialized datasets such as Field Plant, Plant Village, and others, catering to real-world scenarios. Moreover, research indicates a continuous trend towards enhancing model performance, optimizing accuracy, and enabling real-time disease detection, promising a significant impact on agricultural practices. Precision, Recall, and F1-Score are essential performance metrics that can be used to conduct a thorough evaluation of the presented models. Additionally, the Confusion Matrix and Area Under the Precision-Recall Curve (AUC-PR) can be used to measure key parametric performance and real-world feasibility