Crop Disease Prediction in Agriculture Using Deep Learning: A Comprehensive Review

Abstract

Crop diseases continue to pose a serious danger to agricultural productivity worldwide, resulting in large losses in crop quality, yield, and economic value. For large-scale farming, traditional disease detection techniques, which mostly rely on specialist knowledge and manual examination, are frequently laborious, subjective and ineffective. Deep learning (DL), a branch of artificial intelligence, has become a potent method for automated and precise crop disease prediction because to developments in information technology. With an emphasis on image-based analysis and data-driven modelling, this paper provides a thorough overview of current advancements in deep learning-based methods for crop disease diagnosis and prediction. Convolutional Neural Networks (CNNs), one type of deep learning architecture, have shown exceptional performance in reliably diagnosing different crop illnesses and extracting complicated characteristics from plant photos. The detection accuracy has been further enhanced by advanced versions like ResNet, VGGNet, and EfficientNet, which frequently surpass 95 percentage under controlled circumstances. Precision agriculture techniques have been improved by the real-time monitoring and early disease identification made possible by the integration of deep learning with Internet of Things (IoT) devices, remote sensing technologies, and drone-based imaging systems. Despite these developments, a number of problems still exist, such as the requirement for sizable labelled datasets, high processing demands, overfitting problems, and restricted model generalisation in practical settings. This paper identifies these drawbacks and explores possible remedies, such as explainable deep learning methods, data augmentation, and transfer learning. Future research will focus on integrating intelligent decision-support systems, scalable deployment approaches, and edge computing. All things considered, deep learning-based crop disease prediction systems have enormous potential to revolutionise contemporary agriculture by facilitating early intervention, enhancing crop health management, and encouraging sustainable farming methods.

Plant phenotyping relevance

植物画像から病害状態を推定する深層学習手法を中心に扱うレビューであり、植物フェノタイピング手法レビューに該当する。

Code and data availability

This is a comprehensive review article with no original phenotyping measurements, datasets, images, code, or models of its own. All cited datasets and works belong to prior publications, and no author-deposited assets or availability statements appear in the supplied blocks.

No evidence-backed public reproduction asset is currently recorded.