pasteraudio.blogg.se

Matlab r2013a image analysis toolbox
Matlab r2013a image analysis toolbox









  1. #Matlab r2013a image analysis toolbox how to
  2. #Matlab r2013a image analysis toolbox update

With my search for some solution (at the time for my Master Thesis) I stumbled upon open-source undocumented library for Java, published on October 2013 (unbelievable, right?). For some reason there exists libraries working withy C++, Java and Python but so far there is (officialy) nothing for MatLab. Unfortunately it seems that it is not so straightforward at it sounds, and we are not able to use Kinect v2 in MatLab in with what we have for Kinect v1. This means that the core library works on the same principle, and we are able to get data in similar fashion. You might know that Kinect v2 is based on Kinect v1 develpoed by PrimeSense. I wanted to post more about the skeletal tracking first, but this is more interesting at this time. As I promised in one of my previous posts, I will go through my own way to connect Kinect v2 (and also Kinect v1, but that is not so important).

#Matlab r2013a image analysis toolbox how to

Kinect v2 for Windows With real time viewĪs you may already found, there are many ways how to get data from kinect, save them as binary or mat files and then load them to MatLab.

matlab r2013a image analysis toolbox

I will test everything for Kinect v2 for Xbox too, as I fear a bit that official Kinect v2 support will be only for Kinect for Windows version as it was with Kinect v1. Still a lot of coding to do to have whole thing functional, but I am quite satisfied for now.

#Matlab r2013a image analysis toolbox update

I did get to update and optimize few things in Toolbox, and got very good results for Depth Maps. I still need to find out how fast it will be with saving, but at this point I is enough for me to be happy about it. I did get over my 30 FPS at geting depth data to MatLab.

matlab r2013a image analysis toolbox

So you can imagine I was happy with where my optimalization did get me today. But after some testing, I found that it is true, at least at point where I dont gather the data. I didn’t believe it because at that time 30 FPS was what I wished for. In some of oficial report was writen that Kinect v2 was capable of 50 FPS.

matlab r2013a image analysis toolbox

As Kinect v2 is more complex then Kinect v1 i worked only with the new one, and I will later do the same test with the older one. In previous post I wrote that the FPS was about 15 FPS for Kinect v2 and 20 FPS for Kinect v1. My aim was to speed up the FPS of depth maps in MatLab. After few weeks and two conferences, I had again a bit of time to work on Kinect Universal Toolbox.











Matlab r2013a image analysis toolbox