Bonai: Building Footprint Extraction from Off-Nadir Aerial Images

Bonai provides aerial imagery and annotations specifically designed for extracting building footprints from off-nadir perspectives, developed by jwwangchn and updated in June 2024. The data supports research into correcting the geometric displacement between building roofs and their actual ground footprints caused by non-vertical camera angles.

Use Cases

• Training object detection models to identify building footprints in off-nadir imagery
• Developing algorithms to correct geometric displacement between roofs and footprints
• Benchmarking building extraction performance on non-nadir aerial datasets

Strengths

• MIT licensed for open research and commercial use
• Addresses the specific technical challenge of off-nadir perspective correction
• Includes building footprint annotations for object detection and segmentation

Limitations

• Unknown total record count and geographic diversity
• Lack of detailed metadata documentation regarding image resolution or sensor types

Provenance

Source

Learning to Extract Building Footprints from Off-Nadir Aerial Images (Research Paper)

Collection Method

Annotated aerial imagery

Freshness

Last updated June 2024.