Hello everyone!
I'am working on a project where I need to match the videos which are recorded from two different Kinect v2 sensors from different view angles. Which algorithm is best for this approach? Basically I'am programming this in Visual Studio using c/c++. Can we convert the .xef files to any other format? so that I can use those videos in my code.
Any help will be highly appreciated.
Thanks in advance
Are you referring to correspondence matching to reconstruct a full 3D model out of two videos with depth? Or you just want to know whether the two videos are from the same room/environment?
You may try using Kinect Studio API to read .xef files. I never tried it and probably there may no any straightforward method. But there must be some workaround like hooking kinect service while playing the file in KStudio.
I have to reconstruct a 3D model with the two videos.
I have tried it using kinect studio. But am unable to use the recorded videos in my code. Is there any other alternative to extact xef files?
Correspondence matching with keypoint descriptors can be used to match two videos. If you consider the depth information, you can do this with extreme accuracy. SIFT, SURF, ORB etc. are possible keypoint descriptors. Additionally RANSAC method can be used to eliminate the mismatched points.
Unfortunately there is no other known method to extract video from xef file directly. It is mandatory to use their API AFAIK.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages