Scholarlink

In response to the growing challenges of climate change and pest infestations, we have successfully developed and implemented a novel system that integrates advanced pest detection techniques with real-time climate data to enhance food security, specifically within rice crops. Our approach employs the ResNet50 model, a deep learning architecture known for its robust image classification capabilities, fine-tuned to accurately detect pests within complex crop images. This model’s high precision in identifying and classifying various pests is complemented by the incorporation of environmental data, specifically temperature and precipitation records from the AgMERRA dataset. By integrating these climate variables, our system not only detects pests but also contextualizes their potential impact based on prevailing and forecasted weather conditions. This allows for more informed and climate-aware decision-making regarding pest management practices. For instance, the system can determine the optimal timing for pesticide application, avoiding periods where high temperatures might reduce efficiency or when impending rainfall could wash away treatments. This strategic approach minimizes unnecessary pesticide use, reduces environmental impact, and significantly enhances crop preservation. Our methodology has been rigorously tested and demonstrated a 95% accuracy rate in pest detection, leading to the preservation of 50–55% of crop yield that would otherwise be lost to pest damage. This performance not only surpasses traditional pest management techniques but also outperforms other prominent pre-trained models, such as VGG-16, VGG-19, and ResNet101, in the specific context of agricultural disease detection. The success of our system underscores its potential as a critical tool in the global effort to secure food supplies against the dual threats of pest infestations and climate variability.

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

A Novel Predictive Model for Integrating Pest Detection and Environmental Factors for Food Security Assessment

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Muhammad Qasim,
  • Danish Mehmood,
  • Asifa Bibi,
  • Ghufran Ahmed,
  • Adnan Akhunzada

摘要

In response to the growing challenges of climate change and pest infestations, we have successfully developed and implemented a novel system that integrates advanced pest detection techniques with real-time climate data to enhance food security, specifically within rice crops. Our approach employs the ResNet50 model, a deep learning architecture known for its robust image classification capabilities, fine-tuned to accurately detect pests within complex crop images. This model’s high precision in identifying and classifying various pests is complemented by the incorporation of environmental data, specifically temperature and precipitation records from the AgMERRA dataset. By integrating these climate variables, our system not only detects pests but also contextualizes their potential impact based on prevailing and forecasted weather conditions. This allows for more informed and climate-aware decision-making regarding pest management practices. For instance, the system can determine the optimal timing for pesticide application, avoiding periods where high temperatures might reduce efficiency or when impending rainfall could wash away treatments. This strategic approach minimizes unnecessary pesticide use, reduces environmental impact, and significantly enhances crop preservation. Our methodology has been rigorously tested and demonstrated a 95% accuracy rate in pest detection, leading to the preservation of 50–55% of crop yield that would otherwise be lost to pest damage. This performance not only surpasses traditional pest management techniques but also outperforms other prominent pre-trained models, such as VGG-16, VGG-19, and ResNet101, in the specific context of agricultural disease detection. The success of our system underscores its potential as a critical tool in the global effort to secure food supplies against the dual threats of pest infestations and climate variability.