Two-Phase Architecture For Detection Of Landslides Via Satellite Images

View Full Paper

Ask AI

Bookmark

The proposed method enhances landslide detection accuracy compared to traditional techniques, making it vital for risk assessment.
Using Faster-RCNN, the model identifies landslide locations within large satellite images, demonstrating a significant accuracy improvement.
The technique segments loess landslides into precise boundaries, addressing the challenges posed by unclear boundary information.
Integration of dilated inception blocks in the U-Net framework boosts the effectiveness of segmentation, yielding superior results against benchmarks.

Discussion

No takes yet. Share an insight, caveat, or question.

Novel approach shows improved landslide detection and segmentation in satellite imagery, indicating better risk assessment methods.

Swapnalaxmi et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark