Skip to content
Let's Talk

Bruges Road Surface Damage & Texture Classification (Case Study)

Location: Bruges, Belgium

Project Overview

Nazru’s AI platform was deployed in Bruges, Belgium, to automatically detect and classify road surface damages and texture types using satellite and panchromatic (pan) imagery. This project involved identifying cracks, patches, and seams, as well as classifying surface material into three categories: asphalt, pavers (cobblestone), and concrete.

The Challenge

Bruges has a historic city centre with a mix of modern asphalt roads and traditional cobblestone pavements. The client needed a comprehensive inventory of surface distresses (cracks, patches, seams) differentiated by pavement type. Manual inspection was slow, subjective, and disruptive to traffic. No automated solution existed to simultaneously map damages and recognise surface material from satellite and panchromatic imagery.

The Nazru Solution

Nazru’s platform integrated high‑resolution satellite imagery and panchromatic (pan) imagery with deep learning models. Our algorithms automatically:

  • Detected and classified three damage types:

    • Cracks – fatigue, transverse, edge, or random cracking

    • Patches – repaired or overlaid areas

    • Seams – longitudinal construction joints or working cracks

  • Classified road surface texture into:

    • Asphalt – smooth, blacktop surface

    • Pavers (Cobblestone) – traditional stone blocks, often in historic zones

    • Concrete – rigid, jointed cementitious pavement

  • Fused panchromatic (high‑resolution structural detail) with multispectral satellite data for improved accuracy

  • Geolocated each damage instance and surface segment

Key Results & Benefits

  • City‑wide digital map of road damages and surface types for Bruges

  • Clear distinction between asphalt, cobblestone, and concrete areas

  • Damage prioritisation tailored to each surface type

  • Automated workflow reducing manual survey time and improving safety

Key Technologies Used

  • AI‑based semantic segmentation on fused satellite + panchromatic imagery

  • Material recognition using spectral and textural features

  • Linear and polygonal feature extraction

 

Sample image for Bruges, Belgium – AI‑based detection of road damage and surface texture using satellite and panchromatic (pan) imagery