In order to rapidly develop computer vision applications, system designers, algorithm and application developers as well as customers and end-users of such systems need architectures that give them the flexibility to build the best possible solution. One very promising approach for supporting algorithm developers is the Connected Vision framework, a dedicated SDK which allows developers to focus on their image processing algorithm rather than worrying about implementation and integration details. It offers a powerful basis for solving complex computer vision tasks, with a modular approach that makes computer vision algorithms and related applications easy to integrate, reusable, secure and combinable.
Key features
- Secure & distributed architecture (modules running on separate physical machines)
- Communication and integration between modules via a RESTful interface using JSON
- Designed to process live (e.g. network camera) as well as archived data (e.g. video file)
- A concept for reusable, secure and combinable approaches
- Fast and efficient data management