http://www.360doc.com/content/13/0425/15/12119205_280853575.shtml
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type" />
<meta content="360doc" name="classification" />
<link rel="alternate" media="only screen and (max-width: 640px)" href=">
<meta content="图像,检索,特征" name=keywords>
<meta content="关于图像检索的相关汇报" name=description>
<meta name="mobile-agent" content="format=html5;url=http://www.360doc.cn/article/12119205_280853575.html"/>
<meta content="
www.360doc.com" name="author" /> <title>
关于图像检索的相关汇报
</title>
<link href="http://bdimg.share.baidu.com/static/css/bdsstyle.css?cdnversion=20130704" rel="stylesheet" type="text/css" />
<script type="text/javascript" src="http://libs.baidu.com/jquery/1.4.4/jquery.min.js"></script>
<script type="text/javascript" src="http://www.360doc.com/js/jQuery.md5.js?t=2013121101"></script>
<script type="text/javascript" language="javascript">
window.onerror = ignoreError;
function ignoreError() {
return true;
}
</script>
<link href="http://css.360doc.com/wzhead-cleaned.css?t=2014081202" rel="stylesheet" type="text/css" />
<link href="http://css.360doc.com/wz20130912-cleaned.css?t=2014081204" rel="stylesheet" type="text/css" />
<script type="text/javascript" src="/js/indexhead.js?t=2013081206"></script>
<script type="text/javascript" src="/js/showarticlefollow20130912.js?t=2014072302"></script>
</head>
<body style="font-family: Arial, Helvetica, sans-serif;" οnmοuseup="setDragEnd()"
οnmοusemοve="DragDiv();" οnclick="delAllDiv(event)">
<span id="LayerLogin"></span>
<div style="position:fixed; width:100%; z-index:1000;">
<div class="mainnav">
<div class="mainnav1">
<div class="lf360" style="padding-left: 5px;">
<img src="http://pubimage.360doc.com/newsite/logoo.gif" /></div>
<div style="width: 625px; padding-top: 0px; padding-left: 4px; position: absolute;">
<table height="34" border="0" cellspacing="0" cellpadding="0">
<tbody>
<tr id="daohangtr">
<td style="width: 185px; _width: 0px;">
</td>
<td class="navwz1" width="95">
<a href="http://www.360doc.com/index.html">首 页</a>
</td>
<td width="100" class=" navwz1">
<a href="http://www.360doc.com/readroom.html">阅览室</a>
</td>
<td width="88" class=" navwz1">
<a href="http://www.360doc.com/weekstar.html">馆友</a>
</td>
<td width="132" class=" navwz1">
<a href="http://www.360doc.com/my360doc.aspx">我的图书馆</a>
</td>
</tr>
</tbody>
</table>
</div>
<div class="rt360" style="padding-top: 6px; width: 171px;" id="outerdiv">
</div>
<div class=" rt360" style="padding-top: 6px; padding-right: 0px;">
<div class="lf360 serch">
<input name="input2" class="serchpt" id="txtSearchword" οnkeydοwn="ongosearch()"
οnkeyup="showIntelliSense()" οnfοcus="if(this.value=='搜文章 找馆友') this.value='';this.style.color='#272727'"
οnblur="if(!this.value) {this.value='搜文章 找馆友';this.style.color='#b2b2b2';}" οnpaste="showIntelliSense1()"
type="text" value="搜文章 找馆友"></div>
<div class="lf360 ssuo">
<a id="searchlinkid" href="javascript:gosearch();"></a>
</div>
<div id="divsearch" style="margin-top: 25px; margin-left: 1px; display: block; position: absolute;
z-index: 100;">
</div>
</div>
</div>
</div>
<div class="navxt">
</div>
</div>
<div style=" height:44px;_height:0px;width:100%;"></div>
<div class="both">
</div>
<div class="mainzt">
<table class="index_main" border="0" cellpadding="0" cellspacing="0">
<tbody>
<tr>
<td align="left" valign="top" width="727px">
<div id='ADAboveArtContent'>
</div>
<div style="height: 7px;">
</div>
<table border="0" cellpadding="0" cellspacing="0" class="mainbj" id="docArtClass">
<tbody>
<tr>
<td>
<table border="0" cellpadding="0" cellspacing="0" style="width: 100%; text-align: center;">
<tbody>
<tr>
<td>
<table style="width: 100%; text-align: left;" border="0" cellpadding="0" cellspacing="0">
<tbody>
<tr>
<td class="biaoti">
<div class="biaoti2 lf360">
关于图像检索的相关汇报
</div>
</td>
</tr>
<tr>
<td style="height: 32px;">
<div style="width: 670px; height: 28px; padding-left: 26px; color: #777b7a;">
<table>
<tr>
<td>收藏人：<span class="name" οnclick="wzhitnew(1);"><A href="http://www.360doc.com/userhome/12119205" target="_blank">nuomunie</A></span>
</td>
<td>
<div class="btzctsg" id="spanfollowstatus">
</div>
<script>isfollow(12119205);</script>
</td>
<td>
<div class="ytdh1"><a οnclick="doctalk();wzhitnew(106);" href="javascript:void(0)"></a></div>
</td>
</tr>
</table>
</div>
<div class="wztopxx">
<div style="float: left; height: 28px; padding-top: 7px;">
<table cellpadding="0" cellspacing="0" style="position: relative; z-index: 4">
<tbody>
<tr>
<td>
2013-04-25 | 阅：<span id="360doc_Readnum"></span> <span
style="position: relative;">转：<span οnclick="wzhitnew(3)" id="360doc_saverNum"></span><span
id="360doc_saverUser"></span></span>
</td>
<td>
<div style="position: relative;">
| <span id="docsource"></span>
</div>
<div class="sekuaibox" style="display: none; margin-left: 48px;" id="zihaoboxID">
<div style="padding-top: 10px; padding-left: 5px;">
<div class="lf360 danxunbt">
<input id="artdocfont3" οnclick="ChangeArtSize(3, this); wzhitnew(15); " name="radio"
type="radio" value="1" />
</div>
<div class="lf360 danxunwz" οnclick="ChangeArtSize(3,this);wzhitnew(15); " style="cursor: pointer;">
大
</div>
<div class="lf360 danxunbt">
<input id="artdocfont2" οnclick="ChangeArtSize(2, this); wzhitnew(16); " name="radio"
type="radio" value="2" />
</div>
<div class="lf360 danxunwz" οnclick="ChangeArtSize(2,this);wzhitnew(16); " style="cursor: pointer;">
中
</div>
<div class="lf360 danxunbt">
<input id="artdocfont1" οnclick="ChangeArtSize(1, this); wzhitnew(17); " name="radio"
type="radio" value="3" />
</div>
<div class="lf360 danxunwz" οnclick="ChangeArtSize(1,this);wzhitnew(17); " style="cursor: pointer;">
小
</div>
</div>
</div>
<div class="sekuaibox" style="margin-left: 76px; display: none;" id="sekuaiboxID">
<div style="padding-top: 10px; padding-left: 5px;">
<img src="http://pubimage.360doc.com/wz/cr1.gif" id="docArtC1" οnclick="ChangeArtClass(1,this);wzhitnew(18);"
style="cursor: pointer;" /><img src="http://pubimage.360doc.com/wz/cr2.gif" id="docArtC2"
οnclick="ChangeArtClass(2,this);wzhitnew(19);" style="cursor: pointer;" /><img src="http://pubimage.360doc.com/wz/cr3.gif"
id="docArtC3" οnclick="ChangeArtClass(3,this);wzhitnew(20);" style="cursor: pointer;" /><img
src="http://pubimage.360doc.com/wz/cr4.gif" id="docArtC4" οnclick="ChangeArtClass(4,this);wzhitnew(21);"
style="cursor: pointer;" /><img src="http://pubimage.360doc.com/wz/cr5.gif" id="docArtC5"
οnclick="ChangeArtClass(5,this);wzhitnew(22);" style="cursor: pointer;" /><img src="http://pubimage.360doc.com/wz/cr6.gif"
id="docArtC6" οnclick="ChangeArtClass(6,this);wzhitnew(23);" style="cursor: pointer;" />
</div>
</div>
</td>
<td>
<div style="padding-left: 10px;">
<a href="javascript:void(0)">
<img οnmοuseοver="changedivdisplay(1);" src="http://pubimage.360doc.com/wz/wz1.gif" /></a>
</div>
</td>
<td>
<div style="padding-left: 10px;" οnmοuseοver="changedivdisplay(2);">
<a href="javascript:void(0)">
<img src="http://pubimage.360doc.com/wz/wz2.gif" /></a>
</div>
</td>
<td> | <a href="#" οnclick=" setTimeout(showShareLayer, 10);wzhitnew(4);">分享</a> </td>
<td>
<a href="#" οnclick=" setTimeout(showShareLayer, 10);wzhitnew(5);">
<img style="cursor: pointer;" src="http://pubimage.360doc.com/wz/tb10.gif"></a><div
class="fenxiang" id="fenxiangLayer" style="display: none; border: none; width: 110px; margin-left: -28px; _margin-left: -40px;">
</div>
</td>
</tr>
</tbody>
</table>
</div>
<div id="resavelayer1" class="rt360">
</div>
<div id="shareweixinabove" class="rt360">
</div>
<table cellpadding="0" cellspacing="0" style="text-align: center; margin: 0px auto;">
<tbody>
<tr>
<td>
<div id="AlertArt1" style="display: none; position: absolute;">
</div>
<div id="SendToFriends1" style="display: none; position: absolute; z-index: 100;">
</div>
<div class="lf360 zcbutton" style="margin-left: 100px;">
</div>
<div class="lf360">
</div>
</td>
</tr>
</tbody>
</table>
</div>
</td>
</tr>
<tr>
<td></td>
</tr>
</tbody>
</table>
</td>
</tr>
<tr>
<td>
<table class="wenzhangcss" border="0" cellpadding="0" cellspacing="0" width="100%">
<tbody>
<tr>
<td>
<div class="xiantao">
</div>
</td>
</tr>
<tr>
<td align="left" valign="top" width="670px" style="height: 200px;">
<span id="articlecontent" οnmοusedοwn="newhighlight = true;" οnmοuseup="NewHighlight(event)">
<table>
<tbody>
<tr>
<td width="670px" style="position: relative;">
<p class="MsoNormal" style="text-indent:28.1pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><b><span style="font-size:14.0pt;font-family:宋体;mso-ascii-font-family:
Calibri;mso-hansi-font-family:Calibri">关于图像检索的相关汇报</span></b></p>
<p class="MsoNormal" style="text-indent:21.1pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><b><span style="font-family:宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:
Calibri">一、图像相似度计算</span><span lang="EN-US"><o:p></o:p></span></b></p>
<p class="MsoNormal" style="text-indent:21.0pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><span style="font-family:
宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">基于内容的图像检索技术是根据内容见的相似度进行检索，也就是采用相似度匹配，实质是计算对应特征之间的相似程度。通常图像特征用矢量表示，可看作是多维空间中的点，很自然地，就可以用特征空间中点与点之间的距离度量其相似程度。距离值越小，图像越相似。如果用相关计算计算特征矢量之间的相关度，通过相关值的大小可说明图像的相似度，这主要用于颜色特征的相似性度量；由于图像中有些特征是二值型的，如状态型特征，对此应采用关联系数的方法，常用</span><span lang="EN-US">Gower</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">关联系数法；实际上，图像特征是多维复杂体，多特征图像的相似度可用单特征相似度的线性加权和表示。当图像同时具有二值型特征时，还要用关联系数法（设有</span><span lang="EN-US"><span style="position:relative;top:2.0pt;mso-text-raise:-2.0pt"><!--[if gte vml 1]><v:shapetype
id="_x0000_t75" coordsize="21600,21600" o:spt="75" o:preferrelative="t"
path="m@4@5l@4@11@9@11@9@5xe" filled="f" stroked="f">
<v:stroke joinstyle="miter"/>
<v:formulas>
<v:f eqn="if lineDrawn pixelLineWidth 0"/>
<v:f eqn="sum @0 1 0"/>
<v:f eqn="sum 0 0 @1"/>
<v:f eqn="prod @2 1 2"/>
<v:f eqn="prod @3 21600 pixelWidth"/>
<v:f eqn="prod @3 21600 pixelHeight"/>
<v:f eqn="sum @0 0 1"/>
<v:f eqn="prod @6 1 2"/>
<v:f eqn="prod @7 21600 pixelWidth"/>
<v:f eqn="sum @8 21600 0"/>
<v:f eqn="prod @7 21600 pixelHeight"/>
<v:f eqn="sum @10 21600 0"/>
</v:formulas>
<v:path o:extrusionok="f" gradientshapeok="t" o:connecttype="rect"/>
<o:lock v:ext="edit" aspectratio="t"/>
</v:shapetype><v:shape id="_x0000_i1025" type="#_x0000_t75" style='width:12.75pt;
height:12.75pt' o:ole="">
<v:imagedata src="file:///C:\Users\ADMINI~1\AppData\Local\Temp\msohtmlclip1\01\clip_image001.wmz"
o:title=""/>
</v:shape><![endif]--><!--[if !vml]--><img width="17" height="17" src="file:///C:/Users/ADMINI~1/AppData/Local/Temp/msohtmlclip1/01/clip_image002.gif" v:shapes="_x0000_i1025"><!--[endif]--></span><!--[if gte mso 9]><xml>
<o:OLEObject Type="Embed" ProgID="Equation.DSMT4" ShapeID="_x0000_i1025"
DrawAspect="Content" ObjectID="_1428409002">
</o:OLEObject>
</xml><![endif]--></span><span style="font-family:宋体;mso-ascii-font-family:
Calibri;mso-hansi-font-family:Calibri">个特征，第</span><span lang="EN-US"><span style="position:relative;top:2.0pt;mso-text-raise:-2.0pt"><!--[if gte vml 1]><v:shape
id="_x0000_i1026" type="#_x0000_t75" style='width:6.75pt;height:12.75pt'
o:ole="">
<v:imagedata src="file:///C:\Users\ADMINI~1\AppData\Local\Temp\msohtmlclip1\01\clip_image003.wmz"
o:title=""/>
</v:shape><![endif]--><!--[if !vml]--><img width="9" height="17" src="file:///C:/Users/ADMINI~1/AppData/Local/Temp/msohtmlclip1/01/clip_image004.gif" v:shapes="_x0000_i1026"><!--[endif]--></span><!--[if gte mso 9]><xml>
<o:OLEObject Type="Embed" ProgID="Equation.DSMT4" ShapeID="_x0000_i1026"
DrawAspect="Content" ObjectID="_1428409003">
</o:OLEObject>
</xml><![endif]--></span><span style="font-family:宋体;mso-ascii-font-family:
Calibri;mso-hansi-font-family:Calibri">个特征的的相似度为</span><span lang="EN-US"><span style="position:relative;top:6.0pt;mso-text-raise:-6.0pt"><!--[if gte vml 1]><v:shape
id="_x0000_i1027" type="#_x0000_t75" style='width:12pt;height:18pt' o:ole="">
<v:imagedata src="file:///C:\Users\ADMINI~1\AppData\Local\Temp\msohtmlclip1\01\clip_image005.wmz"
o:title=""/>
</v:shape><![endif]--><!--[if !vml]--><img width="16" height="24" src="file:///C:/Users/ADMINI~1/AppData/Local/Temp/msohtmlclip1/01/clip_image006.gif" v:shapes="_x0000_i1027"><!--[endif]--></span><!--[if gte mso 9]><xml>
<o:OLEObject Type="Embed" ProgID="Equation.DSMT4" ShapeID="_x0000_i1027"
DrawAspect="Content" ObjectID="_1428409004">
</o:OLEObject>
</xml><![endif]--></span><span style="font-family:宋体;mso-ascii-font-family:
Calibri;mso-hansi-font-family:Calibri">，有</span><span lang="EN-US"><span style="position:relative;top:5.0pt;mso-text-raise:-5.0pt"><!--[if gte vml 1]><v:shape
id="_x0000_i1028" type="#_x0000_t75" style='width:8.25pt;height:15.75pt'
o:ole="">
<v:imagedata src="file:///C:\Users\ADMINI~1\AppData\Local\Temp\msohtmlclip1\01\clip_image007.wmz"
o:title=""/>
</v:shape><![endif]--><!--[if !vml]--><img width="11" height="21" src="file:///C:/Users/ADMINI~1/AppData/Local/Temp/msohtmlclip1/01/clip_image008.gif" v:shapes="_x0000_i1028"><!--[endif]--></span><!--[if gte mso 9]><xml>
<o:OLEObject Type="Embed" ProgID="Equation.DSMT4" ShapeID="_x0000_i1028"
DrawAspect="Content" ObjectID="_1428409005">
</o:OLEObject>
</xml><![endif]--></span><span style="font-family:宋体;mso-ascii-font-family:
Calibri;mso-hansi-font-family:Calibri">个二值特征，相似度为</span><span lang="EN-US"><span style="position:relative;top:5.0pt;mso-text-raise:-5.0pt"><!--[if gte vml 1]><v:shape
id="_x0000_i1029" type="#_x0000_t75" style='width:12pt;height:12.75pt' o:ole="">
<v:imagedata src="file:///C:\Users\ADMINI~1\AppData\Local\Temp\msohtmlclip1\01\clip_image009.wmz"
o:title=""/>
</v:shape><![endif]--><!--[if !vml]--><img width="16" height="17" src="file:///C:/Users/ADMINI~1/AppData/Local/Temp/msohtmlclip1/01/clip_image010.gif" v:shapes="_x0000_i1029"><!--[endif]--></span><!--[if gte mso 9]><xml>
<o:OLEObject Type="Embed" ProgID="Equation.DSMT4" ShapeID="_x0000_i1029"
DrawAspect="Content" ObjectID="_1428409006">
</o:OLEObject>
</xml><![endif]--></span><span style="font-family:宋体;mso-ascii-font-family:
Calibri;mso-hansi-font-family:Calibri">，则整个相似度为</span><span lang="EN-US"><span style="position:relative;top:11.0pt;mso-text-raise:-11.0pt"><!--[if gte vml 1]><v:shape
id="_x0000_i1030" type="#_x0000_t75" style='width:129pt;height:23.25pt' o:ole="">
<v:imagedata src="file:///C:\Users\ADMINI~1\AppData\Local\Temp\msohtmlclip1\01\clip_image011.wmz"
o:title=""/>
</v:shape><![endif]--><!--[if !vml]--><img width="172" height="31" src="file:///C:/Users/ADMINI~1/AppData/Local/Temp/msohtmlclip1/01/clip_image012.gif" v:shapes="_x0000_i1030"><!--[endif]--></span><!--[if gte mso 9]><xml>
<o:OLEObject Type="Embed" ProgID="Equation.DSMT4" ShapeID="_x0000_i1030"
DrawAspect="Content" ObjectID="_1428409007">
</o:OLEObject>
</xml><![endif]--></span><span style="font-family:宋体;mso-ascii-font-family:
Calibri;mso-hansi-font-family:Calibri">）。图像检索算法大都要求具有相同的成像光照条件，而在实际中很难保证光照一致性，则具有强光照变化鲁捧性的图像检索算法研究就是一个仍待研究的课题。</span></p>
<p class="MsoNormal" style="text-indent:21.1pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><b><span style="font-family:宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:
Calibri">二、对图像内容基于语义特征的研究</span><span lang="EN-US"><o:p></o:p></span></b></p>
<p class="MsoNormal" style="text-indent:21.0pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><span style="font-family:
宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">上面所述的是对图像低层视觉特征的研究。对图像内容基于其语义特征的研究需要人机交互和相关反馈。相关反馈是从用户处进一步挖掘检索需求信息，通过多次信息回馈，来获取用户的确切需求。最近的研究将其称为密度估计、学习或分类问题。与相关反馈有关的短期和长期学习方法，分类模型通过建立低层视觉特征和图像类别语义的映射，对克服语义鸿沟有一定作用。</span></p>
<p class="MsoNormal" style="text-indent:21.1pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><b><span style="font-family:宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:
Calibri">三、基于视觉注意模型的感兴趣区检测</span><span lang="EN-US"><o:p></o:p></span></b></p>
<p class="MsoNormal" style="text-indent:21.0pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><span style="font-family:
宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">基于视觉注意模型的感兴趣区检测，根据感兴趣区域提取方法的不同可分为</span><span lang="EN-US">4</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">类：</span><span lang="EN-US">1</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">，基于人机交互的感兴趣区检测是通过人机交互方式由用户选择图像中的若干区域作为感兴趣区，对于医学图像等比较单一的应用有较好的检索效果；</span><span lang="EN-US">2</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">，基于低层视觉特征的感兴趣区检测过于客观，与人们的主观评价具有一定差距；</span><span lang="EN-US">3</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">，基于注意力的感兴趣区检测可以使用视觉注意模型得到图像中显著度较高的区域作为检测结果，但没有考虑到图像对象的信息；</span><span lang="EN-US">4</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">，基于对象的感兴趣区检测算法，检测效果较好。</span></p>
<p class="MsoNormal" style="text-indent:21.0pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><span style="font-family:
宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">感兴趣区检测的本质是图像区域分割问题，考虑将上面的</span><span lang="EN-US">3</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">和</span><span lang="EN-US">4</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">结合，提出自动提取图像感兴趣区的检测方法。为此，实际应用中以</span><span lang="EN-US">Itti</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">模型为基础，对其改进后可实现。基于改进</span><span lang="EN-US">Itti</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">模型与进化规划的感兴趣区检测方法包括视觉显著度量、注意焦点的选择和转移以及感兴趣区生成三部分，属于自底向上（</span><span lang="EN-US">Bottom-Up</span><span style="font-family:宋体;mso-ascii-font-family:
Calibri;mso-hansi-font-family:Calibri">）的视觉注意模型，具体不再介绍。与此相反，</span><span lang="EN-US">Top-Down</span><span style="font-family:宋体;mso-ascii-font-family:
Calibri;mso-hansi-font-family:Calibri">模型是针对观察任务建立一个对象模型，找到与对象模型匹配的期望对象，再进一步处理期望对象。基于相似距离的</span><span lang="EN-US">Top-Down</span><span style="font-family:宋体;mso-ascii-font-family:
Calibri;mso-hansi-font-family:Calibri">模型包括给定观察任务、相似区域的判断和视觉期望</span><span lang="EN-US">3</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">个部分，具体也不再介绍。以上所介绍的都是无需用户参与的非监督检测方法。</span></p>
<p class="MsoNormal" style="text-indent:21.1pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><b><span style="font-family:宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:
Calibri">四、视点跟踪技术</span><span lang="EN-US"><o:p></o:p></span></b></p>
<p class="MsoNormal" style="text-indent:21.0pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><span style="font-family:
宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">视点跟踪技术是利用视点跟踪设备，对人眼进行检测和定位，实时跟踪视线运动，并得到注视点空间位置的信息。现有的视点跟踪技术主要分为两类，分别以硬件和软件为基础。光电传感技术逐渐发展为以硬件为基础的视点跟踪技术的主流技术。基于软件的视点跟踪大多使用神经网络的方法。基于</span><span lang="EN-US">Gabor</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">变换的人眼定位方法是在对图像进行归一化处理以减少光照影响的基础上，利用</span><span lang="EN-US">Gabor</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">变换后的图像在人眼处幅值较大的特点，通过投影得到人眼的坐标，这种方法对一定的表情、姿态、关照变化有较好的鲁捧性，但不能很好处理光照变化和戴眼镜的情况。基于双阈值的低复杂度视点跟踪方法在高分辨率的人眼图像下，不需要经过人脸检测，用双阈值处理直接定位人眼，提高了视点跟踪的速度，但在训练与运行期间，当用户头部位置相差较大时，误差将明显增大，精度下降。基于</span><span lang="EN-US">Web Camera</span><span style="font-family:宋体;mso-ascii-font-family:
Calibri;mso-hansi-font-family:Calibri">的自适应模板视点跟踪方法采用快速的类</span><span lang="EN-US">Haar</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">特征人脸检测和自适应人眼模板来提取人眼特征，然后围绕瞳孔或人眼中心全搜索定位反射点，最后用两眼位置、瞳孔中心和反射点矢量作为神经网络输入参数，预测视点位置，准确度介于传统模板和双阈值法之间。视点跟踪技术为相关反馈提供了新的交互途径，也就是在获取人眼运动数据的基础上进行相关反馈。</span></p>
<p class="MsoNormal" style="text-indent:21.1pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><b><span style="font-family:宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:
Calibri">五、感兴趣区检测</span><span lang="EN-US"><o:p></o:p></span></b></p>
<p class="MsoNormal" style="text-indent:21.0pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><span style="font-family:
宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">在图像检索的实际应用中，感兴趣区检测主要有两种情况：用户未携带观察任务、用户携带或产生了观察任务。基于此，提出的基于视觉感知的感兴趣区检测主要由视觉注意模型和视点跟踪系统等组成。视觉注意模型对输入图像进行非均匀采样，提取图像的显著特征，构成特征关注图，经合并的显著区作为候选感兴趣区。视点跟踪系统首先通过视点跟踪设备捕获人脸图像，然后计算人眼特征参数，再经神经网络训练预测视点，构成注视点分布，最后将</span><span lang="EN-US">aROI</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">和注视点分布图融合，得到客观化的感兴趣区</span><span lang="EN-US">ROI</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">。基于视觉感知的图像检索是一个较新的研究方向，具有重要的研究意义和实用价值。</span></p>
<p class="MsoNormal" style="text-indent:21.1pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><b><span style="font-family:宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:
Calibri">六、个性化图像检索</span><span lang="EN-US"><o:p></o:p></span></b></p>
<p class="MsoNormal" style="text-indent:21.0pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><span style="font-family:
宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">“个性化图像检索”是图像检索的发展趋势，考虑了用户的不同，根据用户兴趣定制个性化检索结果，以提高检索结果的准确度，实现个性化图像检索的关键是获取用户兴趣，有效表示用户兴趣，并能根据用户兴趣反馈个性化的检索结果。个性化图像检索由人机交互、用户兴趣模型及个性化推荐</span><span lang="EN-US">3</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">部分组成。人机交互是通过系统和用户的交互来获取用户兴趣，利用相关反馈技术逐步求精，优化检索结果。用户兴趣模型的作用是进行用户个性化分析以产生个性化文件（</span><span lang="EN-US">Agent</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">推理机通过学习用户日志信息，自定义感兴趣信息以及反馈信息，推理出用户兴趣，从而构建用户兴趣模型，产生个性化文件）。个性化推荐是一种个性化过滤机制，将信息空间与个性化文件进行匹配，实现个性化信息过滤，向用户推荐个性化检索结果，其技术主要分为基于规则的系统和信息过滤系统两种。信息过滤系统又可分为基于内容的过滤和协作过滤，若将基于内容的过滤系统和协作过滤系统融合，将能克服两者的缺陷。个性化图像检索的发展趋势是考虑如何构建反映用户高层语义的用户兴趣模型，提供更加友好的相关户型去的反馈模式以进一步克服语义鸿沟。</span></p>
<p class="MsoNormal" style="text-indent:21.1pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><b><span style="font-family:宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:
Calibri">七、用户兴趣的获取</span><span lang="EN-US"><o:p></o:p></span></b></p>
<p class="MsoNormal" style="text-indent:21.0pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><span style="font-family:
宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">在图像检索中引入用户兴趣模型，有利于实现自适应的检索。用户兴趣模型通过记录和学习用户的检索行为，获取检索结果与用户需求之间的相关性，以提高用户的满意度。用户兴趣模型的构建是一个机器学习的过程，包括收集兴趣和反馈兴趣两部分。检索系统用相关反馈算法根据用户在当前反馈中给出的相关评价信息调整相似性度量，得到短期兴趣中的视觉特征，并在每次相关反馈的过程中进行更新（短期兴趣中的语义特征表示为</span><span lang="EN-US"><span style="position:relative;top:7.0pt;mso-text-raise:-7.0pt"><!--[if gte vml 1]><v:shape
id="_x0000_i1031" type="#_x0000_t75" style='width:177.75pt;height:20.25pt'
o:ole="">
<v:imagedata src="file:///C:\Users\ADMINI~1\AppData\Local\Temp\msohtmlclip1\01\clip_image013.wmz"
o:title=""/>
</v:shape><![endif]--><!--[if !vml]--><img width="237" height="27" src="file:///C:/Users/ADMINI~1/AppData/Local/Temp/msohtmlclip1/01/clip_image014.gif" v:shapes="_x0000_i1031"><!--[endif]--></span><!--[if gte mso 9]><xml>
<o:OLEObject Type="Embed" ProgID="Equation.DSMT4" ShapeID="_x0000_i1031"
DrawAspect="Content" ObjectID="_1428409008">
</o:OLEObject>
</xml><![endif]--></span><span style="font-family:宋体;mso-ascii-font-family:
Calibri;mso-hansi-font-family:Calibri">）。对用户短期兴趣中的视觉特征提取之后需建立从图像低层视觉特征到图像高层语义特征之间的映射关系以得到图像的语义信息。具体可采用机器学习手段，实现图像的分类，于是便引入了支撑向量机（</span><span lang="EN-US">SVM</span><span style="font-family:宋体;mso-ascii-font-family:Calibri;
mso-hansi-font-family:Calibri">），它是建立在统计学习理论基础上的分类器，不需要特定问题的先验知识，在有限的训练样本情况下，可以很好地控制学习机器的推广能力，其基本思想是通过内积函数定义的非线性变换将输入矢量变换到一个高维特征空间，而后在高维特征空间中构造最优分类超平面。用户的长期兴趣是一个兴趣积累的过程，可用兴趣度作为其定量化指标。同样，用户的长期兴趣也表示为用视觉特征和语义特征共同构造的长期兴趣矢量，其获取需对积累的用户兴趣进行分析推理，并借助专家系统为理论框架，通过推理机推理，具体就是先通过逐步非线性遗忘的兴趣推理算法获取用户长期兴趣中的视觉特征，然后由聚类算法得到用户长期兴趣中的语义特征。</span></p>
<p class="MsoNormal" style="text-indent:21.0pt;mso-char-indent-count:2.0;
layout-grid-mode:char;mso-layout-grid-align:none"><span style="font-family:
宋体;mso-ascii-font-family:Calibri;mso-hansi-font-family:Calibri">个性化图像检索应以数据挖掘与知识发现为主要手段，加上图像理解和自然语言理解技术，对检索结果进行进一步分析，滤掉与用户需求不相关或弱相关的信息，从而提高系统性能和检索的精度与效果。</span></p>
</td>
</tr>
</tbody>
</table>
<div id="viewerPlaceHolder" style="width: 717px; height: 700px; display: none;">
</div>
</span>
<table>
<tbody>
<tr>
<td></td>
</tr>
</tbody>
</table>
</td>
</tr>
<tr>
<td align="left">
<a name="anchordoc1" id="anchordoc1"></a>
<div style="height: 28px">
<div style="font-size: 14px; color: #797b7a; float: left;">
来自：<span class="name" οnclick="wzhitnew(59)"><A href="http://www.360doc.com/userhome/12119205" target="_blank">nuomunie</A></span> > <span
id="cname" οnclick="wzhitnew(60)"></span>
</div>
<div style="float: left;">
<div class="btzctsg" id="spanfollowstatus1">
</div>
<script>isfollow(12119205);</script>
</div>
<div style="float: left;">
<div class="ytdh1"><a οnclick="doctalk();wzhitnew(107);" href="javascript:void(0)"></a></div>
</div>
</div>
<div>
<div class="modlist lf360 list" style="padding-top: 0px;">
<ul>
<div id='lastart'>
</div>
<div id='nextart'>
</div>
</ul>
</div>
</div>
</td>
</tr>
<tr>
<td align="left">
<div style="padding-top: 12px;">
<script type="text/javascript">
/*文章页正文下方标签云，2013-8-21*/
var cpro_id = "u1350774";
</script>
<script src="http://cpro.baidustatic.com/cpro/ui/c.js" type="text/javascript"></script>
</div>
</td>
</tr>
<tr>
<td style="text-align: left;">
<div style="padding-top: 15px; width: 670px;">
<table style="width: 670px;" cellpadding="0" cellspacing="0">
<tbody>
<tr>
<td style="width: 530px;">
<div class=" lf360" style="cursor: pointer" οnmοuseοver="ChangeBtnStyle(1);" οnmοuseοut="ChangeBtnStyle(2);">
<div class="zscbt">
<a href="javascript:void(0);" οnclick="SaveArt();wzhitnew(26);">
<div id="zscbtword" class="lf360 xianhua" style="padding-left: 38px;">
转藏到我的图书馆
</div>
</a>
</div>
</div>
<div style="margin-left: 10px; cursor: pointer;" class="lf360 xhmainbox" id="sendflowerdiv"
οnclick='Showflowerlayer("sendedLayer1");wzhitnew(27)' οnmοuseοver="ChangeFlowerBtnStyle(1);"
οnmοuseοut="ChangeFlowerBtnStyle(2);">
<div id="flowimg2" class="lf360">
<div style="padding-left: 10px; padding-top: 10px;">
<img id="flowimg3" src="http://pubimage.360doc.com/wz/huahuanew.gif">
</div>
</div>
<div id="flowimg5" class=" lf360 xianhua">
献花(<span id="articleflowernum">0</span>)<div id='flowernumadd' class="addtionone">
+1
</div>
</div>
</div>
<div class="lf360" style="cursor: pointer; padding-left: 10px;" οnmοuseοver="ChangeBtnStyle(3);" οnmοuseοut="ChangeBtnStyle(4);">
<div class="fxwxbt">
<a href="javascript:void(0);" οnclick="shareWeixin();wzhitnew(64);">
<div id="fxwxword" class="lf360 xianhua" style="padding-left: 38px;">
分享到微信
</div>
</a>
</div>
</div>
<div class="lf360" style="cursor: pointer; padding-left: 10px;" οnmοuseοver="ChangeBtnStyle(5);" οnmοuseοut="ChangeBtnStyle(6);">
<div class="ywzwbt">
<a href=" http://www.baidu.com/s?wd=%e5%9b%be%e5%83%8f+%e6%a3%80%e7%b4%a2+%e7%89%b9%e5%be%81&tn=SE_hldp01010_6r0gix1f" οnclick="wzhitnew(100);" target="_blank">
<div id="ywzwword" class="lf360 xianhua" style="padding-left: 44px;">
以文找文
</div>
</a>
</div>
</div>
</td>
<td style="vertical-align: bottom; font-size: 12px;" width="140">
<div class="rt360" id="sharediv2">
</div>
</td>
</tr>
</tbody>
</table>
</div>
</td>
</tr>
<tr>
<td height="10"></td>
</tr>
</tbody>
</table>
</td>
</tr>
</tbody>
</table>
</td>
</tr>
</tbody>
</table>
<div class="lswzbox" style="padding-top: 10px;">
<div>
<table cellpadding="0" cellspacing="0">
<tbody>
<tr>
<td width="330" valign="top">
<div class="lswz" style="height: 300px;">
<div class="lstwtop">
<div class="lswztit">
<div class=" lf360">
<div class="titwx1">
类似文章
</div>
</div>
<div class="rt360" style="padding-top: 5px; padding-right: 9px;">
<a href=http://www.360doc.com/relevant/280853575_more.shtml target='_blank' οnclick='wzhitnew(38)'>更多</a>
</div>
</div>
<div class="modlist2 list">
<ul>
<li style="height: 30px; line-height: 30px;"><span>
<a href=http://www.360doc.com/content/08/0708/11/10521_1409518.shtml target=_blank οnclick='wzhitnew(37)'>一些有用的网站 - 标定、立体视觉、图像...</a></span></li>
<li style="height: 30px; line-height: 30px;"><span>
<a href=http://www.360doc.com/content/10/1206/13/4905888_75475039.shtml target=_blank οnclick='wzhitnew(37)'>基于内容的图像检索技术2</a></span></li>
<li style="height: 30px; line-height: 30px;"><span>
<a href=http://www.360doc.com/content/07/0407/09/6394_434619.shtml target=_blank οnclick='wzhitnew(37)'>多媒体信息检索技术简介</a></span></li>
<li style="height: 30px; line-height: 30px;"><span>
<a href=http://www.360doc.com/content/10/0225/20/617416_16817561.shtml target=_blank οnclick='wzhitnew(37)'>信息熵蚁群算法在特征提取和图像识别中的....</a></span></li>
<li style="height: 30px; line-height: 30px;"><span>
<a href=http://www.360doc.com/content/07/0407/09/6394_434613.shtml target=_blank οnclick='wzhitnew(37)'>著名的图像检索系统介绍</a></span></li>
<li style="height: 30px; line-height: 30px;"><span>
<a href=http://www.360doc.com/content/07/0407/09/6394_434635.shtml target=_blank οnclick='wzhitnew(37)'>视频流处理和检索技术概述</a></span></li>
<li style="height: 30px; line-height: 30px;"><span>
<a href=http://www.360doc.com/content/05/0922/12/126_13986.shtml target=_blank οnclick='wzhitnew(37)'>搜索引擎发展的随想--Software Test 软件....</a></span></li>
<li style="height: 30px; line-height: 30px;"><span>
<a href=http://www.360doc.com/content/07/0329/14/3323_419579.shtml target=_blank οnclick='wzhitnew(37)'>基于因特网的个性化信息服务研究</a></span></li>
</ul>
</div>
</div>
</div>
</td>
<td width="330" valign="top">
<div class="lswz" style="margin-left: 9px; height: 300px;">
<div class="lstwtop">
<div class="lswztit">
<div class=" lf360">
<div class="titwx1">
热门推广
</div>
</div>
<div class="rt360" style="padding-top: 5px; padding-right: 9px;">
</div>
</div>
<div class="modlist2 list" style="width: 300px; overflow: hidden; margin-top: 5px;">
<script type="text/javascript">
var sogou_ad_id = 268217;
var sogou_ad_height = 242;
var sogou_ad_width = 298;
</script>
<script type='text/javascript' src='http://images.sohu.com/cs/jsfile/js/c.js'></script>
</div>
</div>
</div>
</td>
</tr>
</tbody>
</table>
<div style="background-color: #fafafa; width: 670px; height: 317px; margin-top: 10px;">
<div style=' padding-top:7px; padding-left:12px; height:182px;'> <div style=' height:30px;'><div class='titwx3 lf360'>猜你喜欢</div><div class=' rt360' style=' padding-right:18px;'><img οnclick='changefiveart("19999","19999");wzhitnew(62);' src='http://pubimage.360doc.com/wz/hyz.gif' οnmοuseοver='this.src="http://pubimage.360doc.com/wz/hyza.gif"' οnmοuseοut='this.src="http://pubimage.360doc.com/wz/hyz.gif"' width='76' height='24' style='cursor:pointer;' /></div></div><table cellpadding='0' cellspacing='0'><tr id='fivearttdid'><td valign='top' width='131'><div><a οnclick='wzhitnew(39);' href='http://www.360doc.com/content/13/0306/00/1117434_269562103.shtml' target='_blank'><img src='http://artpic.360doc.com/13/1023/12/269562103_201310231223280842.jpg' width='114' height='86'/></a></div><div class='lan twentit'><a οnclick='wzhitnew(40);' href='http://www.360doc.com/content/13/0306/00/1117434_269562103.shtml' target='_blank'>不补习，也能第一名的方法</a></div></td><td valign='top' width='131'><div><a οnclick='wzhitnew(39);' href='http://www.360doc.com/content/13/0504/00/7593597_282804630.shtml' target='_blank'><img src='http://artpic.360doc.com/13/0830/15/7593597_201308301543300593_a282804630.jpg' width='114' height='86'/></a></div><div class='lan twentit'><a οnclick='wzhitnew(40);' href='http://www.360doc.com/content/13/0504/00/7593597_282804630.shtml' target='_blank'>一篇值得读N次的好文章</a></div></td><td valign='top' width='131'><div><a οnclick='wzhitnew(39);' href='http://www.360doc.com/content/13/0510/06/5138922_284286608.shtml' target='_blank'><img src='http://artpic.360doc.com/13/0830/15/5138922_201308301544590328_a284286608.jpg' width='114' height='86'/></a></div><div class='lan twentit'><a οnclick='wzhitnew(40);' href='http://www.360doc.com/content/13/0510/06/5138922_284286608.shtml' target='_blank'>颂扬母亲的古诗词十五首</a></div></td><td valign='top' width='131'><div><a οnclick='wzhitnew(39);' href='http://www.360doc.com/content/13/1129/11/10403862_333127078.shtml' target='_blank'><img src='http://artpic.360doc.com/13/1206/00/10403862_201312060010050766.jpg' width='114' height='86'/></a></div><div class='lan twentit'><a οnclick='wzhitnew(40);' href='http://www.360doc.com/content/13/1129/11/10403862_333127078.shtml' target='_blank'>男人的眉毛吉凶</a></div></td><td valign='top' width='131'><div><a οnclick='wzhitnew(39);' href='http://www.360doc.com/content/13/0308/14/7625451_270178037.shtml' target='_blank'><img src='http://artpic.360doc.com/13/1027/13/270178037_201310271347120498.jpg' width='114' height='86'/></a></div><div class='lan twentit'><a οnclick='wzhitnew(40);' href='http://www.360doc.com/content/13/0308/14/7625451_270178037.shtml' target='_blank'>女星四十清醒纪</a></div></td></tr></table></div>
<div style="padding-left: 12px;">
<iframe marginwidth="0" marginheight="0" frameborder="0" scrolling="no" width="650" height="100" src="http://www.360doc.com/ad/artadali4.html?t=2014012301"></iframe>
</div>
<div id="adabtest">
</div>
</div>
<div id="Reflction" style="padding-top: 15px; padding-left: 15px;">
<div id="360docRefTN">
</div>
<div id="360docRefCT">
</div>
<div id="360docRefPB" align="center">
</div>
</div>
<a name="sf"></a>
<div id="ReflectionPart">
<div style="height: 206px; margin: 0px 0px;">
<div class="plbox">
<div class="plmain">
<div class="titwx" style="height: 23px;">
发表评论：
</div>
<textarea name="SendRefTB" id="SendRefTB" style="width: 100%; height: 75px; border: 1px solid #e7e7e7; background-color: #fff; overflow: auto; font-size: 12px;"
οnfοcus="testContent(1);" οnblur="testContent(2)"></textarea>
<div style="text-align: right; padding-right: 3px; padding-top: 6px;">
<div class="huifubt">
<a id="ImgSendPL" href="javascript:void(0);"></a>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</td>
<td align="center" valign="top" width="18px"></td>
<td align="left" valign="top" width="252px">
<div style='height: 98px' id="artrightad1">
</div>
<div class="remenwz" style="height:219px;"><div class="rementop"><div class="titwx" style=" padding-top:10px; padding-left:24px;">最新文章</div></div><div class="modlist3 list" style=" padding-left:15px;"><ul><li><span><a href=http://www.360doc.com/content/14/0220/12/276037_354161911.shtml target=_blank οnclick='wzhitnew(51)'>戏曲电影大全</a></span></li><li><span><a href=http://www.360doc.com/content/14/0224/21/6017453_355400349.shtml target=_blank οnclick='wzhitnew(51)'>不看此文 遗憾终生</a></span></li><li><span><a href=http://www.360doc.com/content/13/1029/06/10101824_324945910.shtml target=_blank οnclick='wzhitnew(51)'>生活中打圆场的方法</a></span></li><li><span><a href=http://www.360doc.com/content/13/1023/01/7858166_323415362.shtml target=_blank οnclick='wzhitnew(51)'>知道了这些，你就是中医！</a></span></li><li><span><a href=http://www.360doc.com/content/14/0705/16/2550590_392202327.shtml target=_blank οnclick='wzhitnew(51)'>六年级数学毕业总复习应用题大...</a></span></li><li><span><a href=http://www.360doc.com/content/13/1205/22/13437033_334810611.shtml target=_blank οnclick='wzhitnew(51)'>比尔 · 盖茨：当我出门在外 我...</a></span></li></ul> <div class=" rt360" style=" padding-right:23px;"><a href="http://www.360doc.com/userhome/12119205" target="_blank" οnclick="wzhitnew(52);">更多</a></div></div></div>
<div class="both" style="padding-top: 10px; height: 260px;">
<iframe id="ifartad1" height="250px" width="250px" frameborder="0" scrolling="no"></iframe>
<div style="height: 10px;">
</div>
</div>
<div class="remenwz">
<div class="rementop">
<div class="titwx" style="padding-top: 10px; padding-left: 24px;">
热门文章</div>
</div>
<div class="modlist3 list" style="padding-left: 15px;">
<ul>
<li><span>
<a href=http://www.360doc.com/content/13/1028/16/9986257_324828943.shtml target=_blank οnclick='wzhitnew(55)'>【人际关系】细品有深度的话，...</a></span></li>
<li><span>
<a href=http://www.360doc.com/content/13/1114/15/14555072_329167551.shtml target=_blank οnclick='wzhitnew(55)'>高级学习方法另有一套</a></span></li>
<li><span>
<a href=http://www.360doc.com/content/14/0124/10/9272557_347507166.shtml target=_blank οnclick='wzhitnew(55)'>用老干妈做出来的14道菜，立马...</a></span></li>
<li><span>
<a href=http://www.360doc.com/content/13/1209/00/5473201_335587094.shtml target=_blank οnclick='wzhitnew(55)'>每一个故事都那么悲伤（下）</a></span></li>
<li><span>
<a href=http://www.360doc.com/content/14/0723/17/1542087_396559237.shtml target=_blank οnclick='wzhitnew(55)'>高中数学必修5 全集</a></span></li>
<li><span>
<a href=http://www.360doc.com/content/13/1114/13/10058454_329143453.shtml target=_blank οnclick='wzhitnew(55)'>妻贤夫病少 好妻胜良药</a></span></li>
<li><span>
<a href=http://www.360doc.com/content/14/0625/20/6877033_389717886.shtml target=_blank οnclick='wzhitnew(55)'>偶可是萌死人不偿命的哦</a></span></li>
<li><span>
<a href=http://www.360doc.com/content/14/0717/19/11400841_395105509.shtml target=_blank οnclick='wzhitnew(55)'>迷人鲍天琦</a></span></li>
<li><span>
<a href=http://www.360doc.com/content/13/1112/20/12156292_328712468.shtml target=_blank οnclick='wzhitnew(55)'>王羲之《兰亭序》的笔法训练</a></span></li>
<li><span>
<a href=http://www.360doc.com/content/14/0108/14/305549_343583999.shtml target=_blank οnclick='wzhitnew(55)'>全素年夜饭【甲午马年.第一桌】</a></span></li>
<li><span>
<a href=http://www.360doc.com/content/14/0224/19/15275972_355355617.shtml target=_blank οnclick='wzhitnew(55)'>《夏天能让“蚊子”蟑螂绝子绝...</a></span></li>
<li><span>
<a href=http://www.360doc.com/content/14/0510/10/269089_376317169.shtml target=_blank οnclick='wzhitnew(55)'>【微情书319】风在摇它的叶子，...</a></span></li>
</ul>
<div class=" rt360" style="padding-right: 23px;">
<a href="http://www.360doc.com/readroom.html" target="_blank" οnclick='wzhitnew(56)'>更多>></a></div>
</div>
</div>
<div class="both" style="padding-top: 10px;">
<iframe id="ifartad2" height="250px" width="250px" frameborder="0" scrolling="no"></iframe>
</div>
<br />
<script type="text/javascript">
var sogou_ad_id = 261623;
var sogou_ad_height = 328;
var sogou_ad_width = 250;
</script>
<script language='JavaScript' type='text/javascript' src='http://images.sohu.com/cs/jsfile/js/c.js'></script>
<script>
function change(pro_name, i) {
var pro = document.getElementById(pro_name);
var pro_arr = pro.getElementsByTagName("span");
for (var m = 0; m < pro_arr.length; m++) {
pro_arr[m].className = "unclecked link_black";
}
pro_arr[i].className = "checked link_bai";
if (i == 1) {
document.getElementById("qh_ul_1").style.display = "";
document.getElementById("qh_ul_0").style.display = "none";
}
else if (i == 0) {
document.getElementById("qh_ul_1").style.display = "none";
document.getElementById("qh_ul_0").style.display = "";
}
}
</script>
<div class="both" id="divad5" style="padding-top: 11px; zoom: 1;">
<div id="divad4" style="display: none;">
<div>
<iframe id="ifartad" height="250px" width="250px" frameborder="0" scrolling="no"></iframe>
</div>
<div class="ggwz">
<a href="javascript:void(0);" οnclick="closead(2);">关闭</a>
</div>
<div style="height: 13px;">
</div>
</div>
<div style="clear: both; display: none;" id="divad6">
<div>
<script type="text/javascript">
/*文章页右下滚屏200*200，创建于2013-7-22*/
var cpro_id = "u1324965";
</script>
<script src="http://cpro.baidustatic.com/cpro/ui/c.js" type="text/javascript"></script>
</div>
<div class="ggwz">
<a href="javascript:void(0);" οnclick="closead(3);">关闭</a>
</div>
<div style="height: 13px;">
</div>
</div>
</div>
</td>
</tr>
</tbody>
</table>
</div>
<input type="hidden" id="docarttitle" value="%e5%85%b3%e4%ba%8e%e5%9b%be%e5%83%8f%e6%a3%80%e7%b4%a2%e7%9a%84%e7%9b%b8%e5%85%b3%e6%b1%87%e6%8a%a5" />
<input type="hidden" id="artreadroomid" value="0" />
<script>var asw='%e5%9b%be%e5%83%8f+%e6%a3%80%e7%b4%a2+%e7%89%b9%e5%be%81'</script>
<div id="docdiviframe"></div>
<script language="javascript" type="text/javascript">
var appName = navigator.appName.toLowerCase();
if (appName.indexOf("microsoft internet explorer") > -1) {
document.write("<scr" + "ipt type='text/javascript' src='http://www.360doc.com/js/StickySystemIE20130911.js?t=2014081101'></sc" + "ript>");
}
else {
document.write("<scr" + "ipt type='text/javascript' src='http://www.360doc.com/js/StickySystemOther20130911.js?t=2014081101'></sc" + "ript>");
}
</script>
<script>GerLookingUserInfo(1,12119205,1,1,'-1',-1,10,0);OutputSource('http%3a%2f%2fwww.360doc.com%2fuserhome%2f12119205');OutputCategory(12119205,3,'%e3%80%8a%e5%8e%9f%e5%88%9b%e3%80%8b');</script>
<script type="text/javascript">
var _bdhmProtocol = (("https:" == document.location.protocol) ? " https://" : " http://");
document.write(unescape("%3Cscript src='" + _bdhmProtocol + "hm.baidu.com/h.js%3Fd86954201130d615136257dde062a503' type='text/javascript'%3E%3C/script%3E"));
</script>
<script type="text/javascript" id="bdshare_js" data="type=tools&uid=6857724" src="http://bdimg.share.baidu.com/static/js/bds_s_v2.js?cdnversion=384393"></script>
<script type="text/javascript">
document.getElementById("bdshell_js").src = "http://bdimg.share.baidu.com/static/js/shell_v2.js?cdnversion=" + Math.ceil(new Date() / 3600000);
</script>
<!-- Baidu Button END -->
<script src="http://bdimg.share.baidu.com/static/js/logger.js?cdnversion=384393"></script>
</body>
</html>
3173
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
