-
I want to thank all the members that have upgraded your accounts. I truly appreciate your support of the site monetarily. Supporting the site keeps this site up and running as a lot of work daily goes on behind the scenes. Click to Support Signs101 ...
Anyone know this font?
- Thread starter New to Printing
- Start date
The Hobbyist
New Member
I wish some talented programmer would create a font identification database, and register ALL fonts into it. Maybe this could be a project for some large college of graphic arts or something.
I am not sure exactly how that would work. Perhaps you would import certain letters, and plot the nodes that make up the chosen letters. Each node would have a specific X,Y coordinate. Then a computer could search for letters that have a node in the immediate vicinity of each node of YOUR letter, and show you the choices. For instance, an upper case R pretty much has the same node positions. Maybe the program could plot the nodes, and then create a "community" for each node, based on an X,Y coordinate from 0,0. Any font that also have one node in each individual node community would be presented as a possible match?
The size of the communities (the diameter of variance for the position of each node) could be expanded as needed. For example, the letter 'M' has many forms, but basically, the bottom outside nodes are either directly beneath the top outside nodes, or spaces equally further apart, while the nodes making the center bottom of the letter could range from at the base of the letter to nearly the top. A letter that has many fancy curls at the end would also qualify as a potential match, if the nodes for that letter were still in the community of the nodes of your letter. The greater distance +/- X and +/- Y for each node, and perhaps a general overall score for that letter's node "fingerprint" would score the match accordingly.
When you think about it, the FBI already has software and a database of TRILLIONS of individual fingerprints dating back 100 years or more, and it can still match up YOUR fingerprint or even a PARTIAL fingerprint found on a gun to one of YOUR fingerprints on file, if your prints are in the database, as mine are. I have a CDW, and I used to have a security guard clearance at a government facility. So there are about 40 sets of my fingerprints in file cabinets somewhere! I wonder if the software that Ancestry.com uses to match your DNA "ribbon" to that of Alexander the Great could be modified to search for font matches. I cannot imagine that matching two letter 'E's would require more computer power than matching my DNA to Genghis Khan!
So the coding to search for a match based on POINTS of a random image or a plotted series of nodes is already written. We just need to use the same or similar software to search for the "fingerprint" of specific letters.
Then all we would need to do is submit samples of every known font and any newly created fonts to a team of people who would create the files for the search engine. How hard could that be?
I can visualize the code, but I am not talented enough to code it myself.
I am not sure exactly how that would work. Perhaps you would import certain letters, and plot the nodes that make up the chosen letters. Each node would have a specific X,Y coordinate. Then a computer could search for letters that have a node in the immediate vicinity of each node of YOUR letter, and show you the choices. For instance, an upper case R pretty much has the same node positions. Maybe the program could plot the nodes, and then create a "community" for each node, based on an X,Y coordinate from 0,0. Any font that also have one node in each individual node community would be presented as a possible match?
The size of the communities (the diameter of variance for the position of each node) could be expanded as needed. For example, the letter 'M' has many forms, but basically, the bottom outside nodes are either directly beneath the top outside nodes, or spaces equally further apart, while the nodes making the center bottom of the letter could range from at the base of the letter to nearly the top. A letter that has many fancy curls at the end would also qualify as a potential match, if the nodes for that letter were still in the community of the nodes of your letter. The greater distance +/- X and +/- Y for each node, and perhaps a general overall score for that letter's node "fingerprint" would score the match accordingly.
When you think about it, the FBI already has software and a database of TRILLIONS of individual fingerprints dating back 100 years or more, and it can still match up YOUR fingerprint or even a PARTIAL fingerprint found on a gun to one of YOUR fingerprints on file, if your prints are in the database, as mine are. I have a CDW, and I used to have a security guard clearance at a government facility. So there are about 40 sets of my fingerprints in file cabinets somewhere! I wonder if the software that Ancestry.com uses to match your DNA "ribbon" to that of Alexander the Great could be modified to search for font matches. I cannot imagine that matching two letter 'E's would require more computer power than matching my DNA to Genghis Khan!
So the coding to search for a match based on POINTS of a random image or a plotted series of nodes is already written. We just need to use the same or similar software to search for the "fingerprint" of specific letters.
Then all we would need to do is submit samples of every known font and any newly created fonts to a team of people who would create the files for the search engine. How hard could that be?
I can visualize the code, but I am not talented enough to code it myself.
bannertime
Active Member
Geneva Olson
Expert Storyteller
Try whatthefont.com You drop your image into it and it finds close matches. If the matches aren't close enough, you can ask in the forum section. I usually get an answer fairly quickly.