On the road with self learning system machine learning techniques

Automatic pavement distress detection

  • robustly detect cracks, potholes and patches
  • identify intact infrastructure
  • reduce the human amount of work
  • speed up the inspection process
  • reduces costs






Output options:

  • Confidence maps: image with the same size as the input image, where every single pixel stores a probability value of belonging to a distress class or a to intact road surface.
  • Axis-aligned bounding boxes: the results are saved as XML files in PASCAL VOC format (the format used by ImageNet).
  • Distress-grid: the results are saved as XML files as defined by the Road and Transportation Research Association (FGSV) and the German Federal Highway Research Institute (BASt). This outpot option requires input images that comply with the FGSV and BASt regulations.

Self learning system / machine learning techniques „Of course, other distress features can be detected as well. It only requieres a training step in the forefront.“

How to Get Pavement Distress Detection Ready for Deep Learning?

A Systematic Approach with Pavement Profile Scanner