Enhanced Multiresunet-Based Detection Of Landslides From Remote Sensing Imagery

1K.V.Jhansi Rani,2Geddapu Bhavani,3Chattumala Arjun,4Darsinala Krishnabhavani,5Jujjuvarapu Sai Mani Varma

doi:10.62643/

Authors

1K.V.Jhansi Rani,2Geddapu Bhavani,3Chattumala Arjun,4Darsinala Krishnabhavani,5Jujjuvarapu Sai Mani Varma Author

DOI:

https://doi.org/10.62643/

Keywords:

Landslide Detection, Remote Sensing Imagery, Deep Learning, MultiResUNet, Image Segmentation, Convolutional Neural Networks (CNN), Disaster Monitoring, Satellite Image Analysis, Geospatial Data Processing, Environmental Hazard Assessment, Automated Landslide Mapping, Earth Observation Data.

Abstract

Landslides are among the most destructive natural disasters, causing significant loss of life,
infrastructure damage, and environmental degradation. Early and accurate identification of
landslide-prone and affected areas is essential for disaster management and mitigation. This
paper proposes an automatic landslide disaster identification framework using remote sensing
images based on an improved MultiResUNet architecture. The enhanced model integrates
multi-scale feature extraction, residual connections, and attention mechanisms to improve
segmentation accuracy in complex terrains. By leveraging high-resolution satellite imagery
and advanced deep learning techniques, the proposed system effectively detects landslide
regions with higher precision and reduced false positives. Experimental results demonstrate
improved performance compared to traditional UNet and baseline segmentation models,
making it suitable for real-time disaster monitoring applications.