Conference: Recognition and Pseudonymization of Data Privacy Relevant Areas in Videos for Compliance with GDPR

For the validation of sensors and systems for automated driving, videos are required and recorded. They thereby form an important building block in the development and testing of automated driving. In the last years, due to the General Data Protection Regulation of the EU, the usage of video has seen backlash from company internal compliance departments, as well as from the general public. Often, the video recorded of the external view includes features that concern the privacy of uninvolved third parties. These features include, but are not limited to, faces and license plates. In this paper, we present an approach to process these sensitive areas in the recorded videos. They are detected using state-of-the-art object detection and then made pseudonymous in a way that can be reverted if the video is used in a specially protected viewing or development environment. This way, the development of functions or systems that rely on these personal features can still be done. For all other use cases, the videos can be used without privacy concerns.

IEEE Xplore (DOI)

Recommended citation:

@INPROCEEDINGS{Hiller2019,
  author={Hiller, Johannes and Schuldes, Michael and Eckstein, Lutz},
  booktitle={2019 IEEE Intelligent Transportation Systems Conference (ITSC)}, 
  title={Recognition and Pseudonymization of Data Privacy Relevant Areas in Videos for Compliance with GDPR}, 
  year={2019},
  volume={},
  number={},
  pages={2387-2393},
  doi={10.1109/ITSC.2019.8917267}
}