Case Study - Vehicle Counter & Classifier

The Customer

The customer is a leading intelligent transport system (ITS) provider based in South Africa and serving markets in the African continent. Their offerings cover end to end traffic management solutions.


The existing counting and classifying solution was a manual one with its associated problems.


An automated solution was needed, with the following features:

  • Error-free counting and classification
  • Counting based on image analysis
  • Automatic classification based on previously provided configurations
  • Logs such that user can revisit them and check the same for anomalies


  • Designed a counter classifier that takes video feed from a camera put up at the counting site
  • The video feed was analyzed for vehicles passing through configurable RoIs (Regions of Interest)
  • The video analysis provided length, width, speed and inferred classification of the vehicle based on configurations
  • The results were marked with date, time stamp and were stored in comma separated files for further processing
  • The videos analysis results were stored with various derived parameters so that detailed analysis of the collected data can be done offline and results can be verified offline
  • The video analysis system was trainable to suit various classification schemes and camera angles
  • Standalone as well as server based system can be derived from this solution based on the need of deployment


  • Gained market leadership through first indigenously designed counter-classifier
  • Created a new market for automated counting and classification
  • Kept the product relevant through enhanced technology
  • Complete technical ownership of solution

Technology Areas

  • Image Analysis
  • Video Analysis
  • Vehicle Counter
  • Vehicle Classifier
  • openCV


Vehicle Counter & Classifier


  • 30 fps IP camera with HVGA resolution
  • Intel i5 core (or better) PC with on board proprietary video analysis software


  • Python
  • openCV

Get in touch with us