•   over 7 years ago

Access the image location or bounding box of found objects

Is there a way to access the image location or bounding box of objects recognized by ReKognition API?

  • 2 comments

  •   •   over 7 years ago

    Assuming you're recognizing faces - The API returns an array of faces - each object is a struct / rect of points for each corner, with the context being the original image's rect. The object is going to be "tight" on the face - so if you will want to expand the area -- in object C you can use CGRectInset (faceRect, -10,10) - to expand the rect outwards by 10px - for example.

  •   •   over 7 years ago

    Hi yosun:
    We don't have a bounding box for objects right now (we're working on it), but we do have bounding box for faces. Filmfest made a detailed explanation about the face bounding box so I'll not repeat it. :)
    Thanks Filmfest!

Comments are closed.