Abstract :

Landslides Are One Of The Most Destructive Natural Disasters, Causing Significant Damage To Human Lives, Infrastructure, And The Environment. Early And Accurate Identification Of Landslide-prone Areas Is Essential For Disaster Management, Risk Assessment, And Mitigation Planning. Remote Sensing Imagery Provides An Effective Means For Monitoring Large Geographical Regions And Detecting Landslide Occurrences. However, The Complex Terrain Characteristics, Variations In Illumination, Vegetation Cover, And Image Noise Make Automatic Landslide Detection A Challenging Task. To Address These Issues, This Study Proposes An Automatic Identification Model For Landslide Disaster Detection Using Remote Sensing Images Based On An Improved MultiResUNet Architecture. The Proposed Model Enhances The Traditional MultiResUNet By Incorporating Advanced Feature Extraction Mechanisms, Multi-scale Convolutional Blocks, And Improved Residual Connections To Effectively Capture Both Local And Global Contextual Information From Remote Sensing Images. The Model Performs Semantic Segmentation To Accurately Distinguish Landslide Regions From Nonlandslide Areas, Thereby Improving Detection Accuracy And Boundary Delineation. Experimental Results Demonstrate That The Proposed Approach Achieves Superior Performance In Terms Of Accuracy, Precision, Recall, F1-score, And Intersection Over Union (IoU) When Compared With Conventional Deep Learning Models Such As CNN, U-Net, And Standard MultiResUNet. The Enhanced Segmentation Capability Of The Improved MultiResUNet Enables Reliable Identification Of Landslide-affected Regions, Making It A Valuable Tool For Disaster Monitoring, Environmental Management, And Decision-making Processes. The Proposed System Contributes To The Development Of Intelligent And Automated Landslide Detection Frameworks That Support Timely Disaster Response And Risk Reduction Strategies.

Published:

06-7-2026

Issue:

Vol. 26 No. 7 (2026)

Page Nos:

1561-1566

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite