GEOSPATIAL AI FOR LANDSLIDE SUSCEPTIBILITY ANALYSIS USING REMOTE SENSING DATA
Keywords:
Landslide classification , satellite image classification , supportvector machine , fuzzy based classification, landslide prediction , land cover classificationAbstract
The paper's landslide prediction capabilities are founded on high-resolution satellite imagery and deep learning algorithms. It utilizes the Fine-Tuned ResNet101 deep learning architecture to classify images as either landslides or non-landslides and employs Image Processing methods to enhance the quality of satellite photos. The system can be accessed online by both administrators and users through an application. The experimental results indicate that the proposed approach for Disaster Risk Management & Early Warning Systems is effective, as it accurately predicts landslide disasters with a 96.88% success rate using Deep Learning technology.