Machine learning and object detection

asked 2016-10-31 05:53:26 -0500

Apologies if a similar question has been asked before but the material in the documentation seems to be really advanced - in fact I may have to pay someone to do this job! The requirements are essentially quite simple but I suspect that the implementation may not be.

What I want to do is to detect the position of 'known' objects within a fairly constant context - e.g., items within a room, where the camera angle doesn't change. That presumably requires prior training over the objects in question, so that they can be subsequently detected within a composite image and their location derived. I'm sorry if that seems a bit vague but I don't want to reveal details in public.

Finally, I wish to use a Python API but I believe that is already available. I just don't know how complete or usable it is.

Thanks ...

Richard Prosser

edit retag flag offensive close merge delete

Comments

  • what are your objects ?
  • can you simply put markers on them ?
  • are there "moving things" ?
  • would human interaction be possible (i.e initially marking position & tracking then) ?

yes, you can use python. i'd say, it's 96% complete.

berak gravatar imageberak ( 2016-10-31 06:10:27 -0500 )edit