📄 http:^^www.cs.cornell.edu^home^rdz^refinement.html
字号:
MIME-Version: 1.0
Server: CERN/3.0
Date: Sunday, 24-Nov-96 22:03:01 GMT
Content-Type: text/html
Content-Length: 2156
Last-Modified: Thursday, 03-Oct-96 21:45:37 GMT
<html> <head><title>Histogram Refinement for Content-Based Image Retrieval</title></head><center><h1>Histogram Refinement for Content-Based Image Retrieval</h1><h2><!WA0><!WA0><!WA0><!WA0><a href="http://www.cs.cornell.edu/home/gregpass/welcome.html">GregPass</a><br><!WA1><!WA1><!WA1><!WA1><a href="http://www.cs.cornell.edu/home/rdz/rdz.html">Ramin Zabih</a><br></h2></center><hr><br><body> <h2>Abstract</h2><p>Color histograms are widely used for content-based image retrieval.Their advantages are efficiency, and insensitivity to small changes incamera viewpoint. However, a histogram is a coarse characterizationof an image, and so images with very different appearances can havesimilar histograms. We describe a technique for comparing imagescalled histogram refinement, which imposes additional constraints onhistogram based matching. Histogram refinement splits the pixels in agiven bucket into several classes, based upon some local property.Within a given bucket, only pixels in the same class are compared. Wedescribe a split histogram called a color coherence vector (CCV),which partitions each histogram bucket based on spatial coherence.CCV's can be computed at over 5 images per second on a standardworkstation. A database with 15,000 images can be queried using CCV'sin under 2 seconds. We demonstrate that histogram refinement can beused to distinguish images whose color histograms areindistinguishable.<p>A paper entitled <b>Histogram Refinement for Content-Based ImageRetrieval</b> will appear in the Workshop on the Applications ofComputer Vision. This paper is available in PDF format <!WA2><!WA2><!WA2><!WA2><ahref="http://www.cs.cornell.edu/home/rdz/wacv-final.pdf">here</a>.Free PDF readers can be obtained for many platforms from <!WA3><!WA3><!WA3><!WA3><ahref="http://www.adobe.com">Adobe</a>.<p>A slightly older paper entitled <b> Comparing Images Using ColorCoherence Vectors</b> will appear in the 1996 ACM Conference onMultimedia, and is available in PDF format <!WA4><!WA4><!WA4><!WA4><ahref="http://www.cs.cornell.edu/home/rdz/mm96-final.pdf">here</a>.</body><hr><ADDRESS><I>rdz@cs.cornell.edu,<br> gregpass@cs.cornell.edu<BR>Monday, July 23 1996</I></ADDRESS></html>⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -