DB API¶
DB interactions are done with a specific API. You can either use the default types, or create your own.
Object Recognition API¶
In the Object Recognition Kitchen each algorithm is defined in terms of a pipeline. We have a few interfaces you should follow if you wish to add a new pipeline to reduce the amount of code to write.
Our infrastructure is very flexible thanks to ecto and we have developed several tools to facilitate the traditional
pipelines that contain a training phase and a detection phase. If your pipeline does not require training, just skip
the section.
Once you wrap your code into those base classes (it’s slightly more complex than a base class), you have access to everything
we’ve coded for you. Doing so usually requires two steps: writing an ecto cell in pure C++ (or Python if you know a bit of ecto)
which is the core of your processing, and a simple Python wrapper that enables ORK to exploit your cell fully (by being able to find it,
by getting automatic docs from it ...).