<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0"
    xmlns:dc="http://purl.org/dc/elements/1.1/"
     xmlns:admin="http://webns.net/mvcb/"
     xmlns:content="http://purl.org/rss/1.0/modules/content/"
     xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">
	<channel> 

	<title>Comments on: Image Matching</title>
	<link>http://ask.metafilter.com/4783/Image-Matching/</link>
	<description>Comments on Ask MetaFilter post Image Matching</description>
	<pubDate>Tue, 20 Jan 2004 18:27:50 -0800</pubDate>
	<lastBuildDate>Tue, 20 Jan 2004 18:27:50 -0800</lastBuildDate>
	<language>en-us</language>
	<docs>http://blogs.law.harvard.edu/tech/rss</docs>
	<ttl>60</ttl>

	<item>
		<title>Question: Image Matching</title>
		<link>http://ask.metafilter.com/4783/Image-Matching</link>	
		<description>Is there an image search engine that allows me to take a given image and see if there are other copies online (the image name may be different and perhaps even resized?) (more inside) &lt;br /&gt;&lt;br /&gt; I found an uncredited image online which I&apos;d like to track down a better copy of. The image name is nothing unique (like 2.jpg). Is there a way (based on colors, resolution maybe?) that i can track down other copies of this image floating around the internet?</description>
		<guid isPermaLink="false">post:ask.metafilter.com,2004:site.4783</guid>
		<pubDate>Tue, 20 Jan 2004 17:28:00 -0800</pubDate>
		<dc:creator>vacapinta</dc:creator>
		
			<category>images</category>
		
			<category>searchengines</category>
		
	</item> <item>
		<title>By: substrate</title>
		<link>http://ask.metafilter.com/4783/Image-Matching#107488</link>	
		<description>Not really, maybe if you linked it here somebody would recognize it and be able to point you towards a better copy. A computer is a big dumb box and so can only do what a person can tell it how to do. I&apos;ll bet that unless it&apos;s a relatively well known picture (the Mona Lisa, Starry Night, goatse.cx) that you couldn&apos;t even describe it well enough for us to find and most of us are smarter than the best big dumb box you can find. &lt;br&gt;
&lt;br&gt;
Even if you did figure out how to describe it you&apos;d have to describe it in a manner that the computer already searched for. Examining every image on the internet to see if it matches your criteria would take astounding amounts of bandwidth and computational resources.&lt;br&gt;
&lt;br&gt;
So, post it here, maybe it&apos;ll ring a bell.</description>
		<guid isPermaLink="false">comment:ask.metafilter.com,2004:site.4783-107488</guid>
		<pubDate>Tue, 20 Jan 2004 18:27:50 -0800</pubDate>
		<dc:creator>substrate</dc:creator>
	</item><item>
		<title>By: whatnotever</title>
		<link>http://ask.metafilter.com/4783/Image-Matching#107489</link>	
		<description>People are working on it.&lt;br&gt;
&lt;br&gt;
Search google for &quot;query by image&quot; &quot;search by example&quot; &quot;search by sketch&quot;, etc.  But I highly doubt there is any public search engine implementing it.  I&apos;d guess that the best you could hope for right now is an academic (*maybe* commercial) application with a royalty-free image database.&lt;br&gt;
&lt;br&gt;
If you can describe it, do so in as many ways as possible on Google&apos;s image search.  That&apos;s my suggestion.</description>
		<guid isPermaLink="false">comment:ask.metafilter.com,2004:site.4783-107489</guid>
		<pubDate>Tue, 20 Jan 2004 18:27:54 -0800</pubDate>
		<dc:creator>whatnotever</dc:creator>
	</item><item>
		<title>By: vacapinta</title>
		<link>http://ask.metafilter.com/4783/Image-Matching#107494</link>	
		<description>Yeah,&lt;a href=&quot;http://forensic.to/databases/&quot;&gt; i can see that this is available&lt;/a&gt; for private image databases but nothing for the web as far as I can see.</description>
		<guid isPermaLink="false">comment:ask.metafilter.com,2004:site.4783-107494</guid>
		<pubDate>Tue, 20 Jan 2004 18:44:17 -0800</pubDate>
		<dc:creator>vacapinta</dc:creator>
	</item><item>
		<title>By: vacapinta</title>
		<link>http://ask.metafilter.com/4783/Image-Matching#107518</link>	
		<description>I guess i should re-ask. is there a way to find other copies online of the &lt;i&gt;exact same image file&lt;/i&gt;, that is same size, same exif information etc. Only the name may have changed.&lt;br&gt;
&lt;br&gt;
It seems this would be useful in finding other people that have taken images from your site and are hosting them somewhere else.</description>
		<guid isPermaLink="false">comment:ask.metafilter.com,2004:site.4783-107518</guid>
		<pubDate>Tue, 20 Jan 2004 20:23:41 -0800</pubDate>
		<dc:creator>vacapinta</dc:creator>
	</item><item>
		<title>By: amberglow</title>
		<link>http://ask.metafilter.com/4783/Image-Matching#107522</link>	
		<description>Maybe you could search for the hex info or whatever that stuff is?&lt;br&gt;
&lt;br&gt;
There was a visual stock photography search thing online, but I can&apos;t find it (and that was only for images close to what you had that was in their database, and commercial)</description>
		<guid isPermaLink="false">comment:ask.metafilter.com,2004:site.4783-107522</guid>
		<pubDate>Tue, 20 Jan 2004 20:30:54 -0800</pubDate>
		<dc:creator>amberglow</dc:creator>
	</item><item>
		<title>By: oissubke</title>
		<link>http://ask.metafilter.com/4783/Image-Matching#107527</link>	
		<description>I have a vague recollection of &quot;Yahoo Images&quot; from the old days.  It would let you search an image, and then find images with similar colors, shapes, etc.&lt;br&gt;
&lt;br&gt;
Not sure whetever happened to that.</description>
		<guid isPermaLink="false">comment:ask.metafilter.com,2004:site.4783-107527</guid>
		<pubDate>Tue, 20 Jan 2004 20:48:05 -0800</pubDate>
		<dc:creator>oissubke</dc:creator>
	</item><item>
		<title>By: yerfatma</title>
		<link>http://ask.metafilter.com/4783/Image-Matching#107582</link>	
		<description>&lt;i&gt;is there a way to find other copies online of the exact same image file, that is same size, same exif information etc.&lt;/i&gt;&lt;br&gt;
&lt;br&gt;
I suppose you could try a P2P app that checks hashes.</description>
		<guid isPermaLink="false">comment:ask.metafilter.com,2004:site.4783-107582</guid>
		<pubDate>Wed, 21 Jan 2004 05:13:34 -0800</pubDate>
		<dc:creator>yerfatma</dc:creator>
	</item><item>
		<title>By: substrate</title>
		<link>http://ask.metafilter.com/4783/Image-Matching#107765</link>	
		<description>google might work on querying exif data. I don&apos;t know the exif format but I&apos;d try something like:&lt;br&gt;
&lt;br&gt;
ext:jpg &quot;exif-token: value&quot;</description>
		<guid isPermaLink="false">comment:ask.metafilter.com,2004:site.4783-107765</guid>
		<pubDate>Wed, 21 Jan 2004 13:39:53 -0800</pubDate>
		<dc:creator>substrate</dc:creator>
	</item>
	</channel>
</rss>
