Scholarlink

<p>Agri CNN-LSTM Fusion (Agricultural Convolutional Neural Network—Long Short-Term Memory Fusion) is an advanced model that uses real-time data from soil sensors to evaluate a soil's suitability for crop production. The model aims to resolve the issues with conventional farming methods by offering a data-driven substitute for spontaneity in decision-making. A significant research gap is a lack of an integrated method that incorporates temporal and spatial soil features for accurate classification. A data-intensive preprocessing workflow was used to process soil data obtained from the National Institute of Technology, Trichy field, through various sensors. Complex patterns in soil data are identified by this model, which is essential for a precise evaluation of soil health. Hyperparameter optimization further improves the model, which yields precise and consistent predictions that classify soil as “Fit” or “Not Fit” for crop cultivation. The experimental observations showed that AgriCNN-LSTMFusion attained an accuracy of 98.5%, establishing it as a dependable method for real-time soil suitability analysis. The model allows farmers to make informed decisions and optimize resource use by reducing uncertainty in soil assessment. By combining predictive modeling with soil sensor data, agricultural efficiency can be enhanced and sustainable growth encouraged.</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

Smart farming with agri CNN-LSTM fusion: leveraging soil suitability analysis using real-time sensor data

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年10月8日

查看全文
  • G. Srivarshini,
  • M. Sujithra,
  • B. Dhanalakshmi,
  • K. Selvakumar

摘要

Agri CNN-LSTM Fusion (Agricultural Convolutional Neural Network—Long Short-Term Memory Fusion) is an advanced model that uses real-time data from soil sensors to evaluate a soil's suitability for crop production. The model aims to resolve the issues with conventional farming methods by offering a data-driven substitute for spontaneity in decision-making. A significant research gap is a lack of an integrated method that incorporates temporal and spatial soil features for accurate classification. A data-intensive preprocessing workflow was used to process soil data obtained from the National Institute of Technology, Trichy field, through various sensors. Complex patterns in soil data are identified by this model, which is essential for a precise evaluation of soil health. Hyperparameter optimization further improves the model, which yields precise and consistent predictions that classify soil as “Fit” or “Not Fit” for crop cultivation. The experimental observations showed that AgriCNN-LSTMFusion attained an accuracy of 98.5%, establishing it as a dependable method for real-time soil suitability analysis. The model allows farmers to make informed decisions and optimize resource use by reducing uncertainty in soil assessment. By combining predictive modeling with soil sensor data, agricultural efficiency can be enhanced and sustainable growth encouraged.