Distributed category/instance/face/body matching [closed]

asked 2017-08-17 05:38:20 -0500

Luccio gravatar image

Hi,

I was wondering if there is any module allowing to distribute the video input, from iOS or Android devices, to a recursive tree of recognition services. My question is pretty similar to this blog post, focusing on meta based search, - https://daniel-j-h.github.io/post/dis..., but applied to computer vision.

Refs:

Goals:

  • Optimize memory usage (shared memory)
  • Dynamic topology calls
  • Remote / local recognition services. (pre-processing, focused instance matching per classes)
  • Async aggregation of results (timeout)

Luc

edit retag flag offensive reopen merge delete

Closed for the following reason question is off-topic or not relevant by berak
close date 2017-08-18 08:02:06.120052

Comments

There is not a ready made module for this, since OpenCV as a framework does not focus on parallel computing. That being said it does not seem impossible, but it will require you to build the distribution around the OpenCV supplied functionality yourself using a third party distribution framework.

StevenPuttemans gravatar imageStevenPuttemans ( 2017-08-17 06:20:21 -0500 )edit
1

not at all an opencv problem, right ?

berak gravatar imageberak ( 2017-08-18 08:01:51 -0500 )edit

ah well, I guess if he plans to distribute OpenCV routines, it slightly is :D

StevenPuttemans gravatar imageStevenPuttemans ( 2017-08-18 08:13:12 -0500 )edit