Registered Image from Kinect v2

edited May 2017 in Kinect

Hey everyone,

I have an issue regarding Open Kinect for Processing. The current project we are building consists of background removal using a kinect. For this purpose we use the RGB information for color substraction as well as the depth information. Due to registration issues between the RGB and depth images provided by the kinect, we need to consider the registered image provided by the kinect (i.e. kinect2.getRegisteredImage().get()). That's where we get stuck...

After 2 days searching around to understand how is the registration information stored in the PImage returned by the kinect, we are still stuck. We were expecting the registered image using RGB channels for RGB information and Alpha channel for the depth values, but from our tests, it is not the case (Alpha channel is always either 255 or 0). We couldn't find any documentation on that matter online...

My question is: do you have any information on how we could extract RGB and depth values from the registered PImage? Currently we can only extract RGB values from it....

Thanks a lot for your feedback!

Laurent

Tagged:

Answers

  • You have to ask for both the registered image and the depth image kinect.getDepthImage(); The depth for any given (x,y) pixel in the registered image is the grayscale value for the (x,y) pixel in the depthImage.

  • So you mean using the registered image as the RGB values and depth image as depth? Mmh... will try it tomorrow, thanks!

Sign In or Register to comment.