Scholarlink

<p>Landslides severely damage buildings and communities, emphasizing the importance of creating precise maps with accurate predictions to protect against potential hazards. Traditional methods are difficult to apply in landslide-prone areas because of their complexity. This has led to an evaluation of deep learning (DL) techniques, such as the application of recurrent neural networks (RNNs) and convolutional neural networks (CNNs). This review paper explores deep learning in landslide forecasting and enhances mapping to minimize the impact and shows the advancements in landslide susceptibility mapping and forecasting like early warning systems. The study integrates remote sensing technologies such as satellite imagery and Light Detection and Ranging (LiDAR) with DL-trained models to enhance the detection of landslides. The study explores three important objectives (1) the practical implementation of deep learning in real-world landslide scenarios, (2) advancements in remote sensing technologies like Lidar and InSAR, (3) future research directions encompassing uncertainty management, data integration, hybrid models, and model transparency. The review paper grasps the approaches, developments, and uses of deep learning for better landslide&#xa0;mapping and forecasting.</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

Comprehensive review of remote sensing integration with deep learning in landslide forecasting and future directions

  • 期刊论文

  • 发表日期：

    2025年3月10日

查看全文
  • Nilesh Suresh Pawar,
  • Kul Vaibhav Sharma

摘要

Landslides severely damage buildings and communities, emphasizing the importance of creating precise maps with accurate predictions to protect against potential hazards. Traditional methods are difficult to apply in landslide-prone areas because of their complexity. This has led to an evaluation of deep learning (DL) techniques, such as the application of recurrent neural networks (RNNs) and convolutional neural networks (CNNs). This review paper explores deep learning in landslide forecasting and enhances mapping to minimize the impact and shows the advancements in landslide susceptibility mapping and forecasting like early warning systems. The study integrates remote sensing technologies such as satellite imagery and Light Detection and Ranging (LiDAR) with DL-trained models to enhance the detection of landslides. The study explores three important objectives (1) the practical implementation of deep learning in real-world landslide scenarios, (2) advancements in remote sensing technologies like Lidar and InSAR, (3) future research directions encompassing uncertainty management, data integration, hybrid models, and model transparency. The review paper grasps the approaches, developments, and uses of deep learning for better landslide mapping and forecasting.