Scholarlink

The present work uses unmanned aerial vehicle (UAV) images for flood detection in urban areas. Two Machine Learning (ML) techniques, namely k-nearest neighbours (KNN) and random forest, are initially employed to classify flood-affected areas based on extracted features from the UAV images. Subsequently, two Deep Learning (DL) methods, UNet and DeepLabv3+, are implemented for semantic segmentation of the images to obtain precise flood maps. The evaluation metrics include accuracy, precision, recall, and F1-score. This work demonstrates the effectiveness of different ML and DL techniques for flood detection using UAV images in urban areas. The comparison reveals that UNet offers superior accuracy compared to KNN, random forest, and DeepLabv3+. The higher accuracy of UNet is attributed to its ability to capture both local and global contextual information, enabling precise delineation of flood extents in urban areas. The findings also emphasize the significance of deep learning approaches in solving similar problems.

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

Flood Detection in UAV Images of Urban Area Using Machine Learning and Deep Learning Techniques

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Tirthankar Ghosh,
  • Arati Paul,
  • Nabendu Chaki

摘要

The present work uses unmanned aerial vehicle (UAV) images for flood detection in urban areas. Two Machine Learning (ML) techniques, namely k-nearest neighbours (KNN) and random forest, are initially employed to classify flood-affected areas based on extracted features from the UAV images. Subsequently, two Deep Learning (DL) methods, UNet and DeepLabv3+, are implemented for semantic segmentation of the images to obtain precise flood maps. The evaluation metrics include accuracy, precision, recall, and F1-score. This work demonstrates the effectiveness of different ML and DL techniques for flood detection using UAV images in urban areas. The comparison reveals that UNet offers superior accuracy compared to KNN, random forest, and DeepLabv3+. The higher accuracy of UNet is attributed to its ability to capture both local and global contextual information, enabling precise delineation of flood extents in urban areas. The findings also emphasize the significance of deep learning approaches in solving similar problems.