Folks from the University of North Carolina at Chapel Hill have compiled a bunch of Microsoft Kinect’s into an incredible inexpensive 3D Teleconferencing system.
The setup uses up to four Kinect sensors in a single endpoint, capturing images from various angles before they are processed using GPU-accelerated filters. The video captured by the cameras is processed in a series of steps, filling holes and adjusting colors to create a mesh image. Once the video streams have been processed, they are overlaid with one another to form a complete 3D image.
That description doesn’t do it justice honestly. They’re using GPU’s to realtime process the data from 4 kinects, performing mesh alignment, color correction, and surface smoothing in realtime. The results are then sent to another screen where they can be viewed in 3D (with glasses of course), but another Kinect tracks the head of the viewer to adjust the display to create an even richer experience.
Yet another cool project using the Microsoft Kinect. I am not certain what you can do with it in a snow globe. Perhaps it would be good for viewing the weather around the world all at the same time.
Using a 3D HD projector, two Microsoft Kinect sensors and some additional hardware and software a project at CHI in Vancouver can present a 360 degree view of an object.
Oliver Kreylos has taken a Microsoft Kinect and used it as a 3-D camera to for a pseudo-holographic 3-D video chat. Obviously this is just a proof of concept, and still has a way to go, but it sure does look interesting. Now if they could only keep her face from tearing in two. I find that a bit, distracting.
Real-time “holographic” video chat using two Kinect cameras to capture one participant, and a custom compression algorithm and network protocol to stream the resulting 3D video data across the Internet.
The other side of the conversation was filmed off a consumer 3D TV with a regular video camera; I apologize for the bad video quality.
Note: the Wiimote was not used for head tracking; only to control the program and to move through the virtual space.
The network protocol uses lossless compression using a Hilbert-curve traversal and run-length and delta encoding for the depth stream, and a Theora video codec for the color stream. The resulting bandwidth is about 750 kB/s for one Kinect camera.
The 3D office model was provided by VITAL Environments.
Remember the movie Minority Report where Tom Cruise, playing Chief John Anderton, interfaces with a computer just using his hands? Well, some researchers at MIT are using the Microsoft Kinect to do the same thing. Pretty cool.
This is a graphical interface inspired by the movie “Minority Report”. It uses the Kinect sensor from Microsoft, and the recently released libfreenect driver for interfacing with the Kinect in linux. The graphical interface and the hand detection software were written at MIT to interface with the open source robotics package ‘ROS’, developed by Willow Garage (willowgarage.com). The hand detection software showcases the abilities of the Point Cloud Library (PCL), a part of ROS that MIT has been helping to optimize. The hand detection software is able to distinguish hands and fingers in a cloud of more than 60,000 points at 30 frames per second, allowing natural, real time interaction.
Code available at:
http://www.ros.org/wiki/kinect
http://www.ros.org/wiki/mit-ros-pkg
Work done by CSAIL’s LIS Group (http://lis.csail.mit.edu/) and Robot Locomotion Group (http://groups.csail.mit.edu/locomotion/)
Nick Hardeman has posted a video on Vimeo where he used a Microsoft Kinect to generate a 3-D vector field. This is just an initial test, but it sure is cool.
This is an initial experiment using the kinect to generate a 3D vector field. This would not be possible without the efforts of the OF team and other people openly hacking the kinect.
The Microsoft Kinect is one of those user interfaces that could be great, or it could just as easily flop. Here we have a video showing the Kinect in action at the recent E3 conference.
Comments