Deep Learning for Building Segmentation and Change Detection in Urban Monitoring: ‘Study case: Zenata City’

View Full Paper

Ask AI

Bookmark

The proposed hybrid algorithm achieved 98% accuracy and a loss function value of 6%, outperforming U-Net and SegNet.
U-Net and SegNet demonstrated 97% and 95% accuracy, respectively, with their loss function values at 9% and 11%, reflecting their effectiveness.
Evaluations were based on 133 annotated image tiles, using multiple metrics like precision, recall, and mean intersection over union.
Automated detection of buildings from drone imagery may enhance land use planning and infrastructure development in urban areas.

Discussion

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

Comparative analysis evaluates U-Net and SegNet against a new hybrid model, suggesting potential for urban monitoring.

Moulay et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark