Charles Babbage's (1791-1871, English mathematician, engineer and computer scientist avant la lettre)
"Difference Engine" was at the time when it was invented impossible to build because of the level of technology for making the
detailed mechanical parts, was not existing yet.
The "Difference Engine" was one of the first freely programmable machines like
a computer and was able to solve polynomial (adding, subtracting and multiplication) equations.
Andrew Carol build a fully working Difference Engine out of Lego. Mr Babbage would be amazed that nowadays you can make his at the
time unconstructable machine from children's toys.
Here's Andrew Carol's website with a complete description of his Lego "Difference Engine. The images of this machine are fascinating,
especially when you used to make things out of Lego yourself. >>

10/02/2006
/ OPTIMUS PRIME FOR REAL!
In Japan a few students developed a real working demonstration model of a wheeled vehicle
transforming into a robot and back again. The concept is based on the popular series Transformers
( >>
) from the eighties. This will be spectacular if they put these machines in the robotwar shows!
A movie of this real transformer. >>
08/02/2006
/ GAUSSIAN SMOOTHING
Before applying the edge detection procedure the image has to be free from noise.
Image noise are random speckles on an otherwise smooth surface area. The image looks grainy like a bad
analogue picture taken without good lighting conditions. These grains will withhold a good edge detection function.
By first smoothing out the whole image, the noise problems will be solved.
There are different ways of doing this
but the one which gives most of the times the best result is Gaussian blurring.
Gaussian blur is named to the 18'th
century German mathematician Carl Friedrich Gauss. Gauss developed mathematics that describes the coordinates of bell
shaped curves.
By using these curves mathematics the blur will give very smooth and subtile transitions between every single pixel.
The results are here. >>
02/02/2006
/ SHAPE RECOGNITION PART 1
To identify areas in an image with the same color, saturation or lightness
there's the need for some kind of shape recognition function.
A lot of my graphic work consists out of
research and new ways of screening methods, conversions from bitmap to vector and vice versa.
Till so far
my screenings are based on a direct translation from each pixel object to one or more vector objects.
To overcome this limitation there's to be some search procedure of grouping pixels with the same characteristics.
To do this is to use an edge detection algorithm. Edge detection will find and mark the
borders between sudden changes in the pixel color.
22/01/2006
/ THYPE SCREENING
Building images out of type has got a long history
in the modern Graphic Design.
Beginning with "The Scarecrow" by Schwitters and
Van Doesburg in 1925 to the Ascii Art images made on computers. Typography -
as of course itself already is an image - combined and used as building
blocks of an image will, in most cases achieve aesthetic results.
Projects Mptheunissen
used Ascii Art are Lprint (full color Ascii-screening on Dot-matrix printers)
and Automailer.
A shortcoming of converting an image to Ascii-art, is that the
chosen collection of typographical signs are based on the weight of those signs
and have no textual meaning.
To combine textual meaning with image features
Mptheunissen created a script able to blend images into large amounts of written
text, ready to edit and print in most design programs.
In
this ( >>
) example the icon of the open source internet browser Mozilla Firefox is
blended into a descriptive text from Wikipedia.org.
> http://en.wikipedia.org/wiki/Firefox

10/01/2006
/ APPLE GOES INTEL
Last summer Apple surprised everybody with the announcement
that Intel Cpu's would be used in their new generation of computers.
Yesterday
it finally happened. A brand new line of Powerbooks.... oh ehh no, they called
MacBook Pro's in this new Intel era, and iMac's are introduced by MC Jobs
at the Keynote in San Francisco. I can't wait to test and see them at the
local Apple Centre. This is right on time because I was beginning to wander
that the immensly popular Ipod was the only thing, Apple thought about.
> Apple.com
08/01/2006
/ SCAN GRAPHICS WITH YOUR OPTICAL MOUSE
Sometimes people get bored. If that happens and
you're a student of the technical university of Twente, what else is there
to do then, opening up your mouse and turn it into a "ghetto b&w
handscanner".
Although a good example of some creative hacker techniques,
you can get some serious RSI complaints.
>http://sprite.student.utwente.nl/~jeroen/projects/mouseeye/