Kinect for Video Wall - Photo Booth app RRS feed

  • Question

  • Hi Guys,
    Do you recommend Kinect be used at the bottom of video wall or on top ? We have used on the top but that results in chopping off of the feet when trying to extract the full body RGB in photo booth application.

    If on the bottom what is the minimum height it should be at ? Anything from your experience would help us satisfy our customer. Ofcourse it would have its own challenge. One problem could be that user would seem to look in the sky when we take photo so are there any techniques you could recommend to adjust the warp angle ? 

    So far we have spent lot of time and effort in extracting the feet with BG subtraction techniques and what not but we are not able to reach to satisfactory solution and have unhappy customer and have started doubting the capability of Kinect solution in such a use case.

    We had to spend significant amount of time in feathering the edge which was part of Kinect V1 but not part of Kinect V2 SDK.
    Any guidance on this front will be helpful to make our customer happy.
    Also I wanted to know if Kinect provides way providing the composite view of multiple Kinects and combine different views into unified view ?

    Wednesday, August 12, 2015 3:46 PM

All replies

  • this is more like a photography opinion rather than technical,

    usually the best portrait is made by placing the camera at the same height of your subject face,

    if you want to include the feets then i would place the users at least at 2 meter of distance from the kinect

    but  i will guess you already tried that, maybe what you want is to get your subject close to get better detail of their face and then i would think to use some lens correction against the RGB pixels of the "user mask".

    Thursday, August 13, 2015 2:57 AM