fredrik sydvart

Our thesis investigated how a shared field of view between a set of network cameras can be discovered through motion analysis. A surveillance operator can switch perspective on interested objects if a shared field of view has been found. The method relies on contradicting evidence of motion and will self-learn the shared field of view through time. No initial knowledge of the network is required. A demo was implemented in Python. Title: Finding Field of View Overlap by Motion Analysis Authors: Fredrik Sydvart, Hampus Altvall

The work included solving incoming support tickets from technicians and delegating tickets to second line. The work also included to monitor critical systems. I was also assigned to code review an internal project made in Django with the goal to ensure continued development if needed.

I was responsible for continued development on an internal benchmark system for automated testing of Axis official plugins (ACAPs) for network cameras. I had to find improvements in the benchmark system and workflow, create user stories in a Scrum environment and implement after discussions with the team. The system was built in Python with Jenkins to build and execute automated benchmark tests.