Kinect Tangible Table Prototype
baxpace writes "The first open source prototype of a tangible table using the Microsoft Kinect sensor. The hack is essentially a proof of concept that can serve a multitude of purposes including a real-time analysis on urban models. The program uses the Kinect point cloud which is mapped onto a flat surface. The upper layer of the point cloud will apply a colour to anything that is placed on the table and is recognized by the Kinect depth sensor. Every object that is placed on the table is detected automatically and in turn becomes trackable."
Seriously? Who cares? I've yet to see a kinect device in front of me, none of my friends have one nor I intend to buy one. Is it only me or this whole kinect stuff is going out of proportion?
This has been possible with the Reactable for some time. No proprietary hardware required, either.
http://www.reactable.com/
opencv has done object and base plane recognition and tracking for a long time and doesn't need a depth sensor, just a webcam
As far as I can tell, he's just projecting the depth (as a few color bands) on top of the table. About five lines of Python with libfreenect and OpenCV. He isn't even tracking anything, just projecting the raw depth quantized to a few layers and roughly calibrated onto the table. Seriously, there are probably hundreds of Kinect hacks more impressive than this one.
The only odd part about this Kinect hack is that he's using the ugly proprietary CodeLabs NUI drivers instead of OpenKinect/libfreenect or OpenNI/SensorKinect.
What's a tangible table? Neither TFA nor TFS say. Also, what makes a tangible table more so than the wooden one in my living room?
If God forks the Universe every time you roll a die, he'd better have a damned good memory.
There were enough people yesterday complaining about Microsoft stories suspiciously appearing in close succession, and it seems pretty clear now that Slashdot has become a shill for the company.
The device in itself is not that big a deal, and many of these stories are really simple hacks, tbh. I don't see why we get 2-3 stories a day about the kinect, and Slashdot should be able to see by now that stories that only get 30 comments half of which say "this story shouldn't have been posted" should not be posted.
This could be cool if you had a clay model of a landscape and were to simulated floods, tsunamis, etc. You could quickly mold new landscape modifications to try out. Or with a detailed enough sensor you might be able to simulate a wind tunnel (ie this pic of the Tesla Model S - http://bit.ly/Tesla_Model_S) - of course the model could only be reduced in one direction, with a single sensor (the typical drawback of a topographical map).
Sure the code may be open but the framework, the hardware and SDK behind the whole Kinect is NOT open source in either spirit or word. There is no way to implement the Kinect without blessing or buying from Microsoft or adhering to their patents or whatever limitations they want to apply to the system and software.
The whole SDK expects you to have Visual Studio in order to develop for it. Let me know if there is a truly open source and cross platform implementation of both drivers, hardware and software that I am free to implement in my own package without getting sued.
Custom electronics and digital signage for your business: www.evcircuits.com
This could be used during surgery to make sure that all tools are accounted for. It could recognize all tools and keep an inventory of all tools in use.
This looks similar to Microsoft Surface that Microsoft was taking about a few years ago
The future is here... and it's a big-ass table.