<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=utf-8">
<META content="MSHTML 6.00.2900.2963" name=GENERATOR></HEAD>
<BODY style="MARGIN: 4px 4px 1px; FONT: 10pt Tahoma">
<DIV>Don't know if this will meet the standards of the techies on this mailing list, but the website for the State Hermitage Museum in St. Petersburg, Russia has a color search (and a layout search too.)</DIV>
<DIV> </DIV>
<DIV>Here's the link: <A href="http://www.hermitagemuseum.org/fcgi-bin/db2www/qbicSearch.mac/qbic?selLang=English">http://www.hermitagemuseum.org/fcgi-bin/db2www/qbicSearch.mac/qbic?selLang=English</A></DIV>
<DIV> </DIV>
<DIV>Diana<BR><BR>>>> Charlie Hull <charlie@juggler.net> 4/26/2007 9:14 am >>><BR>EVANS, KENNETH (WOODY) wrote:<BR>> I'm a librarian, and I've been thinking about visual searching a lot<BR>> because I get frustrated with the searches I have to settle for... If I<BR>> want a "blue" picture, I have to count on someone using the word "blue"<BR>> as a tag, subject heading in metadata, or in the filename<BR>> (blueboat.png).<BR>> <BR>> I've been working on an idea I was calling "Vizibiz" -- a search engine<BR>> that would let you upload or paste images, and return results that were<BR>> visually similar. The notion was that every pixel would be analyzed,<BR>> and metadata would be created for each pixel -- color, position in pic,<BR>> color of neighboring pixels. Vizibiz would also search in traditional<BR>> metadata and filenames, file sizes, and other attributes (page context).<BR><BR>Unfortunately it's considerably harder than this to work out if images <BR>are 'similar' - I've seen a few projects that have had a go, but I don't <BR>think there is anything that actually works. Think of the problems of <BR>magnification, of different RGB colours that actually look similar to <BR>the human eye, of partial (clipped) images...there's a reason that a <BR>large part of our brain is taken up with processing the data from our <BR>retinas :-)<BR>> <BR>> This idea never got off the ground because 1) I don't write code. I'm<BR>> code ignorant. 2) When I started searching for open software that could<BR>> get me started doing this job, I discovered a couple of companies that<BR>> already seem to be on this track:<BR>> http://www.evisionglobal.com/tech/basic_demo.html<BR>> http://able.mulabs.com/<BR>> <BR>> What if this were web-based? And free for anyone to use? <BR><BR>Then everything would be lovely :-) however I suspect that anyone who <BR>has cracked this problem will be making money selling it to security <BR>agencies and governments, rather than giving it away to the world. Don't <BR>hold your breath.<BR><BR>Charlie<BR><BR>>Is there any traction for this sort of search interface in Search <BR>Wikia? Kind of<BR>> like a "music genome project", but automated and for visual rather than<BR>> audio elements... <BR>> <BR>> Thoughts?<BR>> <BR><BR>_______________________________________________<BR>Search-l mailing list<BR>Search-l@wikia.com<BR>http://lists.wikia.com/mailman/listinfo/search-l<BR>Change options or unsubscribe: http://lists.wikia.com/mailman/options/search-l<BR></DIV></BODY></HTML>