Interactive IO & More

Kinect + 3D Volumetric Display

Taha — Sat, 15 Jan 2011 03:49:18 +0000

So as many of you know i’ve been working on my 3D volumetric display as you can see from my other post, what i’ve recently been able to do is add interactivity to the display by augmenting it using a Xbox360 Kinect Sensor. Enjoy the video.

3D Volumetric Display

Taha — Tue, 04 Jan 2011 05:22:13 +0000

Happy 2011, I hope everyone enjoyed their holidays. On to business; recently i’ve been working on 3d volumetric display that would be capable of rendering a free floating 3d 360° volumetric object/s. Below are a few videos of a quick prototype i knocked up in few days, at the beginning of the holidays. Hope you enjoy it.

The above video shows a simple 3d vector generated and textured primitive, rotating on its X and Y axis.

The above video shows a 3ds file exported from Autodesk (any 3ds file will work even generated from GOOGLE sketchup) rendered in 360° rotating on it Y axis.

The above video demonstrates how the rendering can be enlarged as well as how the rendering medium can be inverted to allow for varying design situations.

Quick Update OptoSensor

Taha — Sat, 11 Dec 2010 08:08:41 +0000

So i just finished a major code overhaul for the Optosensor, i now have pixel mosaic, interpolation, gaussian filtering and ~~tracking~~ (correction blob detection working on tracking at the moment) all done via OpenCV. The image below, shows the raw, scaled output in RGB, the next image is Grey Scale (Not really needed since the the IR sensors only detect a narrow band or IR only), Next is the tracked output, and finally is the reference image used for background subtraction (going clockwise from top left).

Interpolated, Filtered, Tracked Output from OptoSensor

OptoSensor News Update

Taha — Thu, 09 Dec 2010 04:31:40 +0000

So i’ve continued my work on the optosensor, when ever i’ve had time; i’m now at a point where i thought i’d talk about what i’ve been doing during the past few weeks. The past few weeks i’ve mainly been working on is the clean up the code on pc side i.e. updating the tracker for the most part, i’ve cleaned up the code separated it into various classes each representing a specific portion tracker i.e. pixel mosiac, serial communication, tracking, and tuio. There have been various improvements to the back end of code where i’m getting the data, how i’m processing it, etc. Not all that interesting what is interesting is that i’ve managed scale up the raw 16×8 pixel image via standard bicubic interpolation and i’ve further smoothed out the data by applying a Gaussian filter to the interpolated data. Below is what is looks like when i have 5 test points.

To explain it a little better these are the raw pixels that are active pixels (1,1), (1,16), (1,8), (16,8), (4,4), (4,5), (5,4) and (5,5) but anyway below you can see both the interpolated scaled output as well the non interpolated scaled output. Here in this test case I’ve interpolated the data set by a factor of 100 using standard bicubic interpolation, and applied a Gaussian kernel of 10 x 10. However in the final release the interpolation factor will be turned down to 10 and the Gaussian filter will be optional.

Interpolated Scaled Output

No Interpolation Scaled Output

Updated VVVV Tracker

Taha — Sun, 14 Nov 2010 08:43:16 +0000

So recently i was going through some of the old stuff i made and came across a tracker i made in V4 back in 2007. After playing with it for a little while i thought it may prove to be a great resource to others, so i’ve cleaned the v4 patch and made it usable to the average v4 user. This patch is basically an updated version my tracker that i released a while back on the v4 page of my blog.

Features (Added)

Front DI and MTmini Support
Background Subtraction
Dynamic Mesh Calibration
Network Broadcasting (OSC and TUIO)

Note: This tracker was only created as a proof of concept while at university and hasn’t been optimized. I mainly now use it in conjunction with a MTmini when i want to do rapid testing.

This is being released under LGPL for the moment.

Download

Electro-Optical Detection is HERE!!

Taha — Sat, 06 Feb 2010 00:38:26 +0000

Well first of all Happy New Year to all, its been a long time since posting anything. Okay so just over a year ago i started thinking about electro-optical sensing as a form of input detection for Multi-Touch and surface computing applications http://mtaha.wordpress.com/2008/12/25/update/. As such i’ve been working on this primarily by myself and am glad to say that within the next few months will have something to showcase. Till then here is are some titillating items below the break

NUI Group Community Book – Multi-Touch Technologies

Taha — Fri, 05 Jun 2009 23:12:27 +0000

The first edition release of the NUI Group community book Multi-Touch Technologies, is here!

This book comprises state of the art articles from the NUI Group Community Wiki, as well as several chapters written by leading experts in the field specifically for this publication. Detailing basic hardware technologies as well as the process of developing software suited for multi-touch systems, this book provides a technical introduction to the world of multi-touch.

I am proud to have been a contributor to and co-author of this book. Please check it out as it is full of helpful information. Well done to all who took part and made this possible..

Click here to download a copy of the book..

For more info please see the original post here at NUIGroup.

UPDATE:

We have gone ahead with the printing process, through Lulu.com’s online printing services. A this point, we are still waiting for an ISBN number to allow distribution through Amazon.com and other retail channels to achieve maximum readership. Once the ISBN application process is completed, ordering information will be made available. Tentatively, the price for a copy of Multi-Touch Technologies is $7.00 USD, + shipping. We are working to lower this cost: it is not final.

And remember, the online pdf is 100% free! – the version that is going to the presses is available on the NUI code downloads, as described above

Thanks!

Community Earth

Taha — Sun, 31 May 2009 20:27:31 +0000

Well after a long period of development and testing I together with icube and nuigroup are rolling out RC1 of Community Earth for the community to start getting involved in development and give feedback on the project.

We are rolling out RC of Community Earth for the community to start getting involved in development and give feedback on the project.

nuicode.com/projects/earth

Community Earth is our own 3D Earth Viewer built on World Wind SDK the current version is based of the Java SDK.

Nasa World Wind is a complete open source solution that lets you zoom from satellite altitude into any place on Earth. Leveraging Landsat satellite imagery and Shuttle Radar Topography Mission data, World Wind lets you experience Earth terrain in visually rich 3D, just as if you were really there. Virtually visit any place in the world. Look across the Andes, into the Grand Canyon, over the Alps, or along the African Sahara.

This Cross-Platform Version of World Wind – Community Earth is designed to provide users of Windows, Mac OS X and Linux a more naturaluser interface, where users can naturally navigate through the environment using standard gestures. It allows the user to pan, zoom, rotate and tilt the globe, and lets you experience Earth terrain in visually rich 3D.

Best when paired with Community Core Vision as input provider.

Please let us know all your comments, requests and ideas on the software and look forward to working with you on making it better

Download Now!

Before posting any questions, please read the CE – Wiki On nuicode

You can add bug reports, feature requests, or support issues to http://nuicode.com/projects/earth/issues/new

CE Changelog
Compiling Instructions

Use ‘ESC’ key to exit.

Bi-Directional LED Sensing Tips

Taha — Sat, 23 May 2009 03:30:08 +0000

Okay since my last post i’ve been doing a bit of reading and doing some tests on bi-directional led sensing. These are a some Tips i present to people who are interested in taking this further. I personally have given up perusing this route of sensing mainly due to its limitations of use, and sensing capabilities.

Helpful Hints

Helpful Hint #1: Using Inexpensive LEDs as Optical Sensors

Most LEDs can be used as detectors without harming the device in any manner, there are some leds which dont work as well, the ideal typle of led is a red led and with a clear casing.

The trick to get an led to sense is to electrically bias the LED in the proper current-voltage (I-V) quadrant for operation as a detector and to detect an appropriate range of wavelengths.

The LED, as the name (Light Emitting Diode) implies, is electrically a diode and can be used as a detection device similar to a photodiode.

Rule of thumb #1: LEDs will not respond as rapidly to optical rise and fall times and PN or PIN photodiodes.
	A properly biased, good quality photodiode will have sub-nanosecond time response and be able to detect the rise and fall time of light pulses shorter than 1 nanosecond in duration, an LED used as a photodiode may only be able to detect 100 microsecond optical rise times and may not do well in detecting pulses shorter than a millisecond. However, if you are trying to detect if the room lights were turned on or if an outside door was opened during the day, this might be the ticket to an inexpensive, readily available optical detector.
Rule of thumb #2: LEDs will only detect light of wavelength shorter than the wavelength of light that the LED would emit if it was put in a circuit that forward biased the LED.
	For example, a red LED will detect light emitted by a yellow LED and a yellow LED will detect light emitted by a green LED but a green LED will not detect light emitted by a red or yellow LED. All three LEDs will detect “white” light or light from a blue LED. White light contains a blue light component which can be detected by the green LED. Recall that visible light wavelengths can be listed from longest wavelength to shortest wavelength as Red, Orange, Yellow, Green, Blue, Indigo, Violet (remember the mnemonic “Roy G. Biv”). Violet is the shortest wavelength light with the most energetic photons and red has the longest wavelength light with the least energetic photons of all of the visible colors of light.
Rule of thumb #3:To use the LED as an optical detector, do not forward bias the LED into quadrant # 1 of the current-voltage (I-V). (Quadrant 1 is when the operating voltage and current are both positive.) Allow the LED to operate in the solar cell mode, quadrant #4 (operating voltage is positive, current is negative), or in the photodiode mode quadrant #3 (operating voltage is negative, current is negative).
	Photodiodes are typically operated in quadrant 3 of their I-V characteristics. In this mode, a small reverse bias voltage is applied to the device and the incident light linearly increases the leakage current proportional to the intensity (actually the number of above bandgap photons) of light falling on the photodiode die. Although LEDs are not intended to experience large reverse bias voltages, most can be reverse biased by a few volts (3v to 7v) and operate in the photodiode mode. Make sure you limit the magnitude of the reverse current so that you do not damage the LED. The photodiode mode will give the best time response and the most linear sensitivity to light. You can expect sensitivities of a few tenths of a microamp of current for each microwatt of optical power directed at the LED (or photodiode die). In the solar cell mode, no applied bias voltage is used. The solar cell (or LED in this case) generates its own current and voltage. To make the LED act as a current generating device proportional to the magnitude of light incident on the LED die, use the “short circuit mode”. External circuit conditions and the amount of light incident on the LED die determines the current-voltage operating point of the device. If you keep the generated voltage across the LED low by using circuit techniques to effectively put a short circuit across the LED, the generated current will be approximately linearly related to the amount of light incident on the LED. This becomes similar to the photodiode mode. The simplest approach to this circuit is to put a relatively small (know value) resistor across the two terminals of the LED and measure the small voltage generated by the current flowing through your known resistor. Experiment with the size of the resistor; the best value will depend on the amount of light you are detecting and the optical arrangement to capture this light onto your LED. Try 100 ohms, 1,000 ohms and 10,000 ohms. If you are using the NEEM-112 to detect this signal, use one of the +/- 2.5 volt channels for 16 bit sensitivity (approximately 75 microvolts resolution). To make the LED act as a voltage generating device, operate in the open-circuit mode by connecting either no load resistor or as large a resistive load to the LED terminals as possible. In the no load resistor mode, the voltage measuring instrument will serve as the load. If the input impedance of the measuring instrument is high enough, the circuit conditions attached to the LED will appear as an open circuit condition. If you are using the NEEM-112 to detect this signal, use any of the voltage sensing channels including either of the 0 to 5 volt channel, either of the +/- 2.5 volt channels, or the +/- 10 volt input channel. In this voltage generating, open circuit mode, the LED will generate an open circuit voltage between 0.0 volts and approximately 1.6 volts – depending on the amount of light incident on the LED die and the color of the LED selected for this measurement. Remember, in the open circuit mode, the magnitude of the voltage is NOT linearly related to the amount of light incident on the LED die.

LED Sensing

Taha — Sun, 03 May 2009 19:38:30 +0000

Alrite to update from my previous post heres an update video of my 3×2 matrix using led sensing