Scholarlink

<p>In the realm of agricultural production, plant diseases precipitated by bacterial, fungal, and viral pathogens represent a significant impediment, culminating in substantial losses in crop yields. A myriad of research endeavors has sought to mitigate the deleterious impacts of such plant diseases, focusing on augmenting plant resilience through the application of cutting-edge technologies in Artificial Intelligence (AI) and Deep Learning (DL). These technologies, notably AI and DL, have emerged as pivotal tools in the early detection and classification of crop diseases, thereby minimizing potential damage. This paper delves into an array of models within the domain of deep learning, with a particular emphasis on Convolutional Neural Networks (CNNs), which are employed to classify and detect plant diseases. This investigation is further enriched by the inclusion of a case study focusing on sunflower crop diseases, employing a dataset specifically curated for this purpose. Through a comparative analysis, this study delineates four distinct approaches to classifying sunflower crop diseases. Among these, the fourth approach—characterized by the utilization of image augmentation, Transfer Learning (TL), and regularization techniques in conjunction with various models—stands out for its superior efficacy. This approach not only yields the most favorable outcomes in comparison to its counterparts but also addresses the challenge of overfitting, thereby enhancing model generalizability. The empirical results of this study are noteworthy, with the accuracy of disease classification in the fourth approach varying between 85.11% and 98.95% across different models. The Xception model, in particular, demonstrates exceptional performance, achieving the highest recorded accuracy of 98.95%. Such findings underscore the potential of integrating sophisticated AI and DL methodologies in the battle against plant diseases, offering a promising direction for safeguarding crop yields. In conclusion, this paper contributes to the burgeoning body of knowledge on the application of AI and DL techniques in agriculture, providing a foundation for future research in this area. It not only highlights the effectiveness of these technologies in enhancing disease detection and classification but also opens directions for the development of more resilient agricultural systems. Further research is encouraged to explore the scalability of these models across different crops and disease types, and to refine these technologies for real-world application in agricultural practices.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Revolutionizing plant disease management: integrating AI and deep learning for enhanced detection and classification in agriculture applications

  • 期刊论文

  • 发表日期：

    2025年12月15日

查看全文
  • Rania A. Ahmed,
  • Walid El‑Shafai,
  • Ezz El-Din Hemdan,
  • Zeinab A. Ahmed,
  • El-Sayed M. El-Rabaie,
  • Fathi E. Abd El-Samie

摘要

In the realm of agricultural production, plant diseases precipitated by bacterial, fungal, and viral pathogens represent a significant impediment, culminating in substantial losses in crop yields. A myriad of research endeavors has sought to mitigate the deleterious impacts of such plant diseases, focusing on augmenting plant resilience through the application of cutting-edge technologies in Artificial Intelligence (AI) and Deep Learning (DL). These technologies, notably AI and DL, have emerged as pivotal tools in the early detection and classification of crop diseases, thereby minimizing potential damage. This paper delves into an array of models within the domain of deep learning, with a particular emphasis on Convolutional Neural Networks (CNNs), which are employed to classify and detect plant diseases. This investigation is further enriched by the inclusion of a case study focusing on sunflower crop diseases, employing a dataset specifically curated for this purpose. Through a comparative analysis, this study delineates four distinct approaches to classifying sunflower crop diseases. Among these, the fourth approach—characterized by the utilization of image augmentation, Transfer Learning (TL), and regularization techniques in conjunction with various models—stands out for its superior efficacy. This approach not only yields the most favorable outcomes in comparison to its counterparts but also addresses the challenge of overfitting, thereby enhancing model generalizability. The empirical results of this study are noteworthy, with the accuracy of disease classification in the fourth approach varying between 85.11% and 98.95% across different models. The Xception model, in particular, demonstrates exceptional performance, achieving the highest recorded accuracy of 98.95%. Such findings underscore the potential of integrating sophisticated AI and DL methodologies in the battle against plant diseases, offering a promising direction for safeguarding crop yields. In conclusion, this paper contributes to the burgeoning body of knowledge on the application of AI and DL techniques in agriculture, providing a foundation for future research in this area. It not only highlights the effectiveness of these technologies in enhancing disease detection and classification but also opens directions for the development of more resilient agricultural systems. Further research is encouraged to explore the scalability of these models across different crops and disease types, and to refine these technologies for real-world application in agricultural practices.