Since recently, our student enthusiast team is engaged to develop a stream video surveillance app which to be running on cloud server platform. The app functions include:
- video capture by multiple edge devices (cameras) located in large-scale urban environment (in highlight scenario, our system has to operate over 1000 cameras simultaneouly so they evolve to a widely distributed hardware network);
- pushing video data stream to process on cloud analytics servers (we need a cloud server solution both to manage multiple data streams and to apply various optional algorithms for motion estimation, face recognition etc.);
- providing cloud-based storage to save footage that records relevant events like deviant behavior cases, scenes of motion, road accidents and so on;
- accessing analytically featured records by end user terminals via local browser (so, any user can watch video footage directly in browser window);
- real-time video broadcast in conventional way (i.e. live stream watching with camera change option).
Following our actual cloud module design, the solution will run three servers:
- video data receiver which gets in streams from multiple cameras;
- video analyzer powered by appropriate software;
- footage/live stream provider which connects to user terminal once detecting a client request.
Things we're seeking before kick-off:
- a software toolkit (framework) to make developer experience as enhanced as well (so far we consider Java or Python to code)
- Maybe you know some open source platform, cloud-based service or library which looks topically appropriate to address our specific demand?
- Also any vendor solutions, you are familiar with, may enable faster, advanced development experience (however, no ongoing business offerings like SaaS, PaaS, IaaS would be required to contribute to this project).