August 7, 2020

Parking lot detection

Since the parking lot in Leusden could be very busy, we start an experiment to see if it is possible to use image recognition to find empty spaces. The model exists of two levels, one detecting the parking lots and the second model to detect if the spot is free.

The model is trained based on images found on the internet of several parking lots. The angle of the camera is important to recognize free parking lots. We discovered that the angle of the camera for the training pictures were very different from the angle of the camera in Leusden. Therefore we were not able to a sufficient parking lot detection model.

  • Stephan Schrijver

    DataOps Engineer
  • Kasper Meijjer

    Senior Data Scientist