Unique Colors
+ Personal
+ Movies
My Favorite Movies
Every one I've ever seen
SciFi movies that occur in a specific year
My Favorite SciFi movies
My NetFlix Queues
+ TV
What I'm currently watching
Shows that I've seen every episode
Favorite episodes
Favorite Star Trek episodes
+ Music
My Favorite Bands
What I'm currently listening to
All Beatles songs/albums rated
+ Adventure
Places I want to travel
Driving Across America
Mt. Saint Helens: snowboarding
Mt. Saint Helens: hiking
Dave and Scott Run the USA
+ My Adventures in Asia
My Wedding Photo Album
Kelly's Wedding Photo Album
My Flickr photos
My wife's Flickr photos
Wedding Newspaper article 1
Wedding Newspaper article 2
Wedding Newspaper article 3
Wedding Newspaper article 4
Photo Gallery 1
Photo Gallery 2
Photo Gallery 3
Photo Gallery 4
Photo Gallery 5
Photo Gallery 6
Photo Gallery 7
Photo Gallery Malaysia
Photo Gallery Thailand
Photo Gallery Singapore
Photo Gallery Indonesia
Asian Adventure Blog 1
Asian Adventure Blog 2
Bad English Pirated DVD
Congklak (oldest game in the world) defeated
+ Random
Recent Updates
About Me
Contact Me
Where I have lived
Languages I've tried to learn
Everyone I've ever Known (almost)
YouTube: Favorite Channels
Pivotal People and Moments in my Life
My Photo Album
My Digital Devices
My Links Page
Google Map Favorites
Rating Ben & Jerry's Flavors
Stephen King
+ Games
Portal 2
SimCity 2000
SimCity 2000 XTreme
SimTower
The SIMs
Oldest game in the world defeated
World's Hardest Word Search Puzzles
World's Hardest Word Guess Game
+ Misc
Sitemap: Every Page on this site
Domain Names
- my search for the perfect domain name
Cool Web Stuff
- cool stuff I've seen or done on web pages
Funny and Interesting Stuff
Tivo vs ReplayTV
Tool: View Photos
on another web site
Tool: View Images
on this web site
Tool: Date Math
Tool: Name That Color
Unique Color Images
This gallery is my attempt to convert images so that no two colors are the same and yet have the image look exactly as the original.
There are about 16 million colors in the 24 bit RGB color scheme. Our eyes can only see between 35,000 and 55,000 unique colors. This means that you will not be able see small changes in color. If there are a large number of pixels of a single color, then it will require a lot of change from the original color and become noticeable.
Galleries
[
Best
] [
Worst
] [
Average
]
I searched Google extensively for a program that already does this but I failed to find anything. I was curious to see if (A) it could be done and (B) how well it could be done.
I learned that it is possible to convert an image so that every color is unique and yet have it look exactly like the original but only for images that had a high percentage of unique colors to start. If the image had a lot of one color for example black or white, then the deviation would be so great as to introduce noticeable noise. Blurred areas also proved to be a problem since when you blue an image, you reduce the number of unique colors.
Examples of Improving Algorithm
[
Grinder Girl
] [
Iron Man
] [
Dominoes
]
I learned that if I purposely add noise to the image before converting it to all unique colors, that it could eliminate the jpeg artifacts and smooth out the blotchy patches so the resulting picture would look like a grainy version of the original. These are some examples of how my algorithm is improving as I learn from my experiments.
The algorithm is described on my
developer blog
.
Date Created November 4, 2008
Last Updated November 4, 2008
Contact: patcoston@gmail.com