0

I'm trying to develop a software where can detect face recognition using Kinect.

Until now, I just know I can use OpenCV to do this feature. I just want that the camera can save the face and recognize him later.

I've just use KinectSDK, so I need some help or orientation to implement the feature.

Liam McInroy
  • 4,339
  • 5
  • 32
  • 53
Darf Zon
  • 6,268
  • 20
  • 90
  • 149

2 Answers2

2

I think that you can do that by using depth stream. Each pixel has his own depth (distance from the KInect). Each face has a uniq patter. So you can save a pixel depth pattern and compare it.

Fixus
  • 4,631
  • 10
  • 38
  • 67
  • is it possible that you can show an example Fixus? I'm not so experimented with depth stream. – Darf Zon May 07 '12 at 18:23
  • 1
    i don`t have copy/paste code. you need to get the join (in this case it will be head). That get the pixel array and analyze it using the depth. Read more about the depth stream. IMO it`s the only way. It`s not so easy but it can be done – Fixus May 07 '12 at 19:44
0

Yes it is possible by the help of calling NU-GET id is(Sacknet.KinectV2FacialRecognition.x64) packages in to your c# program. I'm also trying to do the facial recognition with kinect using these libraries but I'm little bit confused about it. how sequentially put my code into program with kinect SDK-2 libraries with Sacknet libraries. or sacknet have the features of complete SDK-2 so that we can tackle the frame and manipulate according to our requirement. If any one have idea about this please let me know.here is the link for Sacknet pakages

raja abdulmoeez
  • 21
  • 5