Home Page |
IFR White Paper |
Applying IFR |
Proof of Concept |
About Colorcom |
|
Home Page |
IFR White Paper |
Applying IFR |
Proof of Concept |
About Colorcom |
 |
 |
| IFR Proof of Concept |
|
Main Topic Menu |
 |
|
 |
|
|
The test patterns
that follow demonstrate IFR's ability
to sequence across an entire picture,
while converting raster to vector
without any loss. Other raster to
vector converters can only sequence
across a few pixels, usually about
three.
Our tests include running a vector rendering program on our vector data to verify its quality. A significant rendering test is the process of changing the resolution of the picture. Subsequently, it is possible to trace along the border to see if the border is continuous and fluid. As shown by the images below, IFR converts all of the test patterns with no discontinuities. In comparison, the closest competitor has about 100 discontinuities. We are further able to test our vector quality by running the vector data through a resolution enhancement program that produces the resulting test patterns illustrated in the right column below. |
|
|
|
|
Original size (1:1):
 |
|
 |
 |
|
|
|
|
| |
Original size (1:1):
 |
|
 |
 |
|
| |
Original size (1:1):
 |
|
 |
 |
|
|
|
|
IFR takes no
shortcuts. All combinations of
sequences are taken into account.
This offers a completely different
paradigm to approach the raster to
vector conversion problem. Flight
provides a good analogy. Without an
airplane, a person's time in the air
is limited by the length of a jump;
but with powered flight, the time in
the air is instead limited by the
amount of fuel.
Other technologies are limited to the ability to sequence over a few dots because they do not prepare the data while considering every possibility. IFR's raster to vector conversion is able to accomplish its global sequencing because the developers of IFR have spent many years searching, defining, and cataloging all possibilities. This research has resulted in the new paradigm technology called IFR. Test pattern 9 demonstrates that IFR exceeds human ability in some areas. In other words, it is becoming possible for a machine to see things in a picture that humans cannot see.
How many lines do you
see in the Test Pattern 9 picture?
Most people see five diagonal lines
but there are actually eight. Three
of the five apparent lines are
actually two lines in one, making a
total of eight lines.
|
|
|
Original size (1:1):
 |
 |
 |
|
|
|
|
IFR determines that
there are eight lines in Test Pattern
9. It locates the slopes of long
lines by placing two points along the
lines. The slope that IFR determines
does not violate any coarse pixels.
IFR also determines the beginning and
the ending of each line.
IFR is fast. It determines the lines, slopes, and extremities of Test Pattern 9 in a single pass and without any iteration! Remember that IFR accounts for all possibilities before it starts to process the data.
Although humans
perform raster to vector conversion
effortlessly and on the fly, the
human eye can not detect the slight
grade differences that make a
significant difference in the quality
of resolution enhancement. IFR
technology reads the vector data and
correctly defines the image more
accurately, thereby allowing for
lossless conversion and fantastic
resolution enhancement.
|
|
 |
|
|
|
|
| |
Original size (1:1):
|
|
|
|
|
|
|
|
| |
Original size (1:1):
|
|
|
|
|
| |
Original size (1:1):
|
|
|
|
| Test Pattern 10 | |
Original size (1:1):
|
|
|
|
| Test Pattern 11 | |
Original size (1:1):
|
|
|
|
| © 2003 - 2010 Colorcom, Ltd, All rights reserved. |
|
