AI crack detection
From a photographto the length and widthof every crack.
AI reads hundreds of facade photographs to one standard, finds the fine concrete cracks, and calculates the length and width of each. It reduces both what gets missed and how much the judgement drifts.
Contact D-TECT ↗Everything read to one standard
Human review shifts with time and fatigue. AI reads hundreds of images against the same criteria, so less is missed, and every detection can be checked against the original photograph.
- Fine concrete cracks detected by automated image analysis
- Detections reviewed side by side with the original
- Incorrect areas corrected by a person and fed back in
Not just found — measured
Marking a crack and measuring it are different problems. Capture metadata and capture spatial information are combined to convert a crack in the image into its real-world size.
- Real crack length, following the path rather than the straight line
- 0.3mm-class width measurement
- Concrete spalling area and repair quantities compiled automatically
Frequently asked questions
How can crack length and width be measured from photos?
High-resolution imagery is combined with capture metadata and capture spatial information to convert image measurements into real-world dimensions. Detailed calculations are handled inside the protected D-TECT engine.
What happens when a detection is wrong?
Detections are reviewed alongside the original image. Incorrect areas are corrected, and any crack the analysis missed can be marked by hand and carried into the results.