CODEWITHSUNDEEP
3dkinectpoint-clouds
Matt
Matt

Reputation: 2853

Using Kinects Depth Camera to see in the Dark

I'm trying to read the depth camera data from the kinect sensor with the MS Kinect SDK and create a depth map so that you can make out features in the dark, kind of a night vision feedback. Ive been mucking around with the standard monochrome look provided with the samples and it is very hard to make out any features really is good just to see outlines of objects, you cant really make out rises and falls in objects. Looking at ending up something like the 3d image done here 3d Kinect Depth Map.

Just wondering if this is possible? And if so how to go about it with the new MS Kinect SDK?

Upvotes: 2

Views: 752

Answers (3)

Shlomo Zippel
Shlomo Zippel

Reputation: 681

It sounds like you want to perform histogram equalization to the depth data. The basic idea is assigning a color range (yellow->black) to a depth range (min distance from sensor->max distance from sensor).

You can find plenty of reference code:

  1. Zigfu OpenNI bindings for Unity3d (C#/Unity3D).
  2. OpenNI NiSimpleViewer sample (C++/OpenGL).

Upvotes: 0

Liam McInroy
Liam McInroy

Reputation: 4356

The light the Kinect sensor sends out and retrieves to make the depth map is infa red, so normal light does NOT effet how the sensor should work. So you wouldn't have do anything to the code since kinect is already set up to do that

Upvotes: 1

mankoff
mankoff

Reputation: 2301

I think you want to highlight "surface normals" which makes everything much much clearer in 3D. See "cocoaKinect" app on OS X or check the "normals" checkbox on RGBDemo on Windows or OS X (although that implementation isn't quite as smooth as the cocoaKinect app).

Upvotes: 0

Related Questions