Scholarlink

Plant disease prediction is essential for enhancing agricultural productivity, sustainability, and food security. In this study, we present an integrated multimodal deep learning approach to predict diseases in rice and potato crops. The model utilizes a combination of leaf imaging, environmental parameters (temperature, light intensity), and soil metrics (humidity, NPK values, pH) to improve prediction accuracy. Convolutional Neural Networks (CNNs) are used for image processing, and a Long Short-Term Memory (LSTM) model is applied for environmental and soil time-series data. The features extracted are then fused via a Multi-Layer Perceptron (MLP) for final disease classification. Experimental results show significant performance improvement over state-of-the-art methods, demonstrating the potential of multimodal data integration in agritech solutions.

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

Multimodal Deep Learning Approach for Disease Prediction in Rice and Potato Crops Using Leaf Imaging, Environmental, and Soil Data

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Hiranmoy Roy,
  • Soumyadip Dhar,
  • Arpan Deyasi,
  • Poly Saha

摘要

Plant disease prediction is essential for enhancing agricultural productivity, sustainability, and food security. In this study, we present an integrated multimodal deep learning approach to predict diseases in rice and potato crops. The model utilizes a combination of leaf imaging, environmental parameters (temperature, light intensity), and soil metrics (humidity, NPK values, pH) to improve prediction accuracy. Convolutional Neural Networks (CNNs) are used for image processing, and a Long Short-Term Memory (LSTM) model is applied for environmental and soil time-series data. The features extracted are then fused via a Multi-Layer Perceptron (MLP) for final disease classification. Experimental results show significant performance improvement over state-of-the-art methods, demonstrating the potential of multimodal data integration in agritech solutions.