none
Kinect Fusion Head Scanning RRS feed

  • Question

  • Hello,

    Do you have any information on whether Kinect Fusion Head Scanning sample will be released. It's pretty useful feature in the Kinect version1 for our application. Thanks for the help.

    Thanks,

    Santosh

    Tuesday, September 30, 2014 7:14 PM

Answers

  • not at this time, but I gave the general steps above. The compiler will guide you in the right direction to where the code needs to change since the v1 Kinect SDK is very different from v2. As a general guide to writing v2 Kinect apps you can have a look at the Jumpstart videos:

    http://aka.ms/k4wv2jumpstart


    Carmine Sirignano - MSFT


    Tuesday, October 7, 2014 6:25 PM

All replies

  • Are you asking about the Multi-Camera Fusion sample from v1? If so, there is currently a limitation with v2 where the run-time only supports 1 sensor connected to a machine.  THere are several other threads that discuss the technical reasons for this.

    Since you already have the code for v1, porting that to v2 wouldn't be that hard, but enabling a second sensor for data integration is going to be the hard part. There are no plans in the near term to enable that. We have seen at different hackathons, where people have done multi-camera scanning using several machines networked together and integrating the data at a main system controller.


    Carmine Sirignano - MSFT


    Monday, October 6, 2014 5:23 PM
  • I meant the Fusion sample where it detects the face of the person and scans just the face.
    Monday, October 6, 2014 7:17 PM
  • What is the sample name? Launch the toolkit and provide the name of the same. There isn't a face detection fusion sample I am aware of. The only face detection sample is a face tracking library, but that doesn't use Fusioin.

    If you want face tracking and 3D mesh of the face(HD Face) those are already part of the v2 SDK. Face is our 2D facial feature tracking library and HD Face is the one where it morphs a face model to match that of the user.


    Carmine Sirignano - MSFT

    Monday, October 6, 2014 7:25 PM
  • The name of the sample is Kinect Fusion Head Scanning-WPF which is available in toolkit v1.8.
    Tuesday, October 7, 2014 12:55 PM
  • There are currently no plans to port this sample to v2 at this time. If this is something you need, the samples provide full source code, therefore the porting effort should be straight forward. As long as there are no dependencies on multi-sensor support in the v1 code, it would be a matter of removing all the v1 dependencies and replacing them with the v2 libraries. From there, the code changes For initializing the readers and copying data is similar then you would find in the Fusion Basics/Explorer v2 samples.

    Carmine Sirignano - MSFT

    Tuesday, October 7, 2014 5:14 PM
  • Is there a tutorial or more information for porting the code from versioni 1 to version 2?
    Tuesday, October 7, 2014 5:23 PM
  • not at this time, but I gave the general steps above. The compiler will guide you in the right direction to where the code needs to change since the v1 Kinect SDK is very different from v2. As a general guide to writing v2 Kinect apps you can have a look at the Jumpstart videos:

    http://aka.ms/k4wv2jumpstart


    Carmine Sirignano - MSFT


    Tuesday, October 7, 2014 6:25 PM
  • Thanks for the information. I will start digging into to code for porting the sample.
    Tuesday, October 7, 2014 6:32 PM