Using kinect v2 for OpenCV in C++ on win 10

asked 2017-11-29 08:49:26 -0500

YOUR_HAMSTER gravatar image

I need to read frames from the kinect v2 as a Mat and set the exposure manually to do computer vision for a university semester project. This has proved very difficult for me and even with extensive research I have been unable to find any clear directions. There seems to be a lot of different libraries and ways of doing it but I have not had any luck. Does anyone have any knowledge or know any material that outlines all needed steps in a way that is understandable for a newcomer in computer vision?

Thank you for reading this :)

edit retag flag offensive close merge delete

Comments

I found some Kinect SDK 2.0 code that I wrote before I got the Xbox 360 Kinect with PC adaptor.

I put the code up at: https://github.com/sjhalayka/kinect_o...

The code attempts to grab a main camera frame, depth frame, and infrared frame. It converts the main camera frame from RGBA to BGR, and converts the depth and infrared frames to 32-bit floating point grayscale format.

The code may work right off the bat, but I've never tested it with an actual Kinect, since it doesn't work with the Xbox 360 Kinect.

Your first step is to get the frames from the camera and show them.

If you find that the code works, or needs bug fixes, please let me know.

sjhalayka gravatar imagesjhalayka ( 2017-11-29 11:29:38 -0500 )edit

Another tutorial site: https://homes.cs.washington.edu/~edzh...

Plus there are examples that come with the Kinect SDK 2.0.

Let us know when you've successfully read in the frames and displayed them. I googled for "OpenCV HDR exposure" and got some interesting tutorials regarding exposure.

sjhalayka gravatar imagesjhalayka ( 2017-11-29 12:08:48 -0500 )edit