Anonymization

Highlights

Fast even on low-performance hardware
Invariant scale and orientation independent
Street View spherical images supported
Flexible images & video supported

One product, one price, two anonymizations packed in

The anonymization quality is ensured even in the case of low-resolution input via fast image processing. We use detectors based on sequential classifications that allow for extremely fast data evaluation. Our sequential detectors are constructed using an automated learning method based on the WaldBoost algorithm. Based on a wide series of images (millions of positive examples, billions of negative examples) this algorithm finds the optimal decision rule, which maximizes both the detection success and the speed of evaluation. Thanks to this unique process, our detectors are extremely fast and precise.

Car License Plates Anonymization

Car license plate anonymization quickly detects license plates by using our cutting-edge technology, which is also used in our car number plate reading product. The license plate is subsequently blurred.

Human Faces Anonymization

Human faces anonymization quickly detects faces by using our cutting-edge technology, which is also applied in our face recognition products, used for example by police departments. The face is subsequently blurred. 

Feature Parameters
Anonymization
  • License plate blurring: all license types in Europe and North America.
  • Face blurring: multiview face detection.
    • Application examples Street-view like images anonymization
    Privacy protection on images from city webcams
    Anonymization of data for annotation
    Blurring of faces or licence plates on user submitted photos
    Product Type
  • SDK library
  • PC Application
    		•
    Supported Platforms
  • Windows (32 & 64 bit) / Unix (32 & 64 bit)
  • CPU (x_86, ARM) / GPU (only 64bit)
    • *other on request
    Online Demo Yes, available here
    *you can also request offline demo

    Are you interested? Try our demo or...

    Contact us

    Our references