Smart City
AI + Computer Vision for Sidewalks
This project applies AI and computer vision to sidewalk feature detection and assessment, with a strong focus on accessibility analysis and reducing the heavy manual effort required by traditional inspection workflows.
Back to Research
Project Summary
Data-fusion ML using aerial imagery, LiDAR, IR depth, and RGB cameras for feature assessment.
Sidewalk assessment is subjective, labor-intensive, and requires considerable manual processing, which makes it slow to scale and difficult to standardize. Build a data-fusion machine learning system for sidewalk feature detection and assessment that can support accessibility and compliance analysis more efficiently. The sensing pipeline combines aerial imagery, LiDAR, depth (IR) sensors, and RGB cameras rather than relying on a single modality. The project targets sidewalk feature detection and ADA-PROWAG style assessment tasks relevant to accessible pedestrian infrastructure. A key outcome is replacing slow, subjective manual inspection with scalable automated analysis.