Video Anonymization

Anonymization is a process of blurring or pixelating personal data in videos and photos to comply with GDPR (RODO) requirements. This service is essential when publishing or sharing traffic measurement videos that contain identifiable information.

Why is anonymization required?

Many governmental and regulatory bodies require video anonymization before public release:

  • GDDKiA (Generalna Dyrekcja Dróg Krajowych i Autostrad) in Poland - requires anonymization for General Traffic Measurement (GPR) publications

  • Department of Transport and road authorities in EU countries

  • Municipal authorities when publishing traffic studies

  • Private companies sharing video evidence with third parties

Blurring individuals, advertisements, or license plates is a recurring requirement in EU General Traffic Measurementarrow-up-right and a must-have for road photo registration.

Anonymization methods

AITracker offers two anonymization methods:

Pixelation

The default anonymization effect. Provides the fastest processing time and is suitable for most use cases. The pixelated area completely obscures the original content while maintaining the overall scene context.

Blurring (Gaussian blur)

Get a nicer visual effect by using Gaussian blur. This method provides a smoother appearance but requires slightly more processing time.

What can be anonymized?

You can configure exactly what categories of objects should be anonymized:

License Plates

Complete blurring/pixelation of license plate text. Useful when you need to preserve vehicle visibility but hide identification data.

Pedestrians

Full profiles of pedestrians and cyclists are anonymized. This includes faces, body shapes, and any identifying features.

Vehicles

All vehicle categories can be completely anonymized. Use this when you want to focus on road infrastructure rather than traffic participants.

Configuration in AITracker

In the scene configuration, you can set anonymization options:

Select any combination of categories based on your requirements.

Additional effects

AITracker also offers additional visual effects that can be combined with anonymization:

Tilt-shift effect

Add a tilt-shift effect to view your video as a miniature/thumbnail model. Creates a unique aesthetic impression that makes the scene appear as if it were a small-scale model.

Snow effect

A festive falling snow effect that can add atmosphere to your video. Originally added for Christmas promotions but available year-round.

Horizontal flip

Mirror the video horizontally. Useful for creating variations or adjusting scene orientation.

Downsize (Compression)

All videos are compressed by default to reduce file size and transfer speed. This is automatically applied to optimize output files.

Output format and location

Anonymized videos are saved with the following specifications:

  • Codec: H.265 (HEVC) for optimal compression and quality

  • Maximum resolution: 1280px (longer edge) - videos are automatically downscaled if larger

  • Output location: ConfigurationX/ folder in the results directory, where X is the configuration number (e.g., Configuration1/, Configuration2/)

Combining counting with anonymization

If you need both traffic counting results and anonymized video output, you can use the multiple configurations feature:

  1. Configuration 1: Set up traffic counting service (e.g., Traffic measurement, ALPR)

  2. Configuration 2: Set up anonymization service with desired settings

Both configurations will process the same input video, and you will receive:

  • Traffic measurement reports in Configuration1/ folder

  • Anonymized video in Configuration2/ folder

This approach allows you to get complete analytics while also producing GDPR-compliant video for publication or sharing.

Accuracy

The anonymization service achieves up to 97% accuracy in detecting and anonymizing personal data. Each frame of the video is processed individually to ensure consistent protection throughout the entire recording.

Processing workflow

  1. Upload your video or photo to AITracker

  2. Select what categories to anonymize (license plates, pedestrians, vehicles)

  3. Choose anonymization method (pixelate or blur)

  4. Apply optional effects (tilt-shift, snow, flip)

  5. Process - receive a new anonymized version of your video

The original video remains unchanged. You receive a new version with all selected personal data anonymized frame by frame.

Last updated

Was this helpful?