shuzituxiangchuli.zip_阈值分割程序_图像增强_matlab例程

<html xmlns="http://www.w3.org/1999/xhtml"> <head> <meta charset="utf-8"> <meta name="generator" content="pdf2htmlEX"> <meta http-equiv="X-UA-Compatible" content="IE=edge,chrome=1"> <link rel="stylesheet" href="https://static.pudn.com/base/css/base.min.css"> <link rel="stylesheet" href="https://static.pudn.com/base/css/fancy.min.css"> <link rel="stylesheet" href="https://static.pudn.com/prod/directory_preview_static/62688ba04f8811599e141040/raw.css"> <script src="https://static.pudn.com/base/js/compatibility.min.js"></script> <script src="https://static.pudn.com/base/js/pdf2htmlEX.min.js"></script> <script> try{ pdf2htmlEX.defaultViewer = new pdf2htmlEX.Viewer({}); }catch(e){} </script> <title></title> </head> <body> <div id="sidebar" style="display: none"> <div id="outline"> </div> </div> <div id="pf1" class="pf w0 h0" data-page-no="1"><div class="pc pc1 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/62688ba04f8811599e141040/bg1.jpg"><div class="c x0 y1 w2 h2"><div class="t m0 x1 h3 y2 ff1 fs0 fc0 sc0 ls0 ws0">一、设计开发要求</div><div class="t m0 x2 h4 y3 ff1 fs1 fc0 sc1 ls0 ws0">用<span class="_ _0"> </span><span class="ff2">matlab<span class="_ _0"> </span></span>编程<span class="_ _1"></span>工<span class="_ _1"></span>具，<span class="_ _1"></span>编写<span class="_ _1"></span>具有<span class="_ _1"></span>操作<span class="_ _1"></span>界<span class="_ _1"></span>面的<span class="_ _1"></span>应用<span class="_ _1"></span>程序<span class="_ _1"></span>，<span class="_ _2"></span>通<span class="_ _2"></span>过</div><div class="t m0 x1 h4 y4 ff1 fs1 fc0 sc1 ls0 ws0">菜<span class="_ _2"></span>单<span class="_ _2"></span>、<span class="_ _1"></span>对<span class="_ _2"></span>话<span class="_ _2"></span>框<span class="_ _2"></span>，<span class="_ _1"></span>选<span class="_ _2"></span>项<span class="_ _1"></span>框<span class="_ _2"></span>等<span class="_ _2"></span>界<span class="_ _2"></span>面<span class="_ _1"></span>控<span class="_ _2"></span>制<span class="_ _1"></span>，<span class="_ _2"></span>对<span class="_ _2"></span>算<span class="_ _2"></span>法<span class="_ _1"></span>进<span class="_ _2"></span>行<span class="_ _1"></span>演<span class="_ _2"></span>示<span class="_ _2"></span>。<span class="_ _2"></span>需<span class="_ _1"></span>要<span class="_ _2"></span>在</div><div class="t m0 x1 h4 y5 ff1 fs1 fc0 sc1 ls0 ws0">应用系统中实现的算法及功能包括：</div><div class="t m0 x1 h4 y6 ff2 fs1 fc0 sc1 ls0 ws0">1<span class="ff1">、图像增强算法</span></div><div class="t m0 x3 h4 y7 ff1 fs1 fc0 sc1 ls0 ws0">（<span class="ff2">1</span>）<span class="_ _2"></span>灰度线<span class="_ _2"></span>形变换<span class="_ _2"></span>：亮<span class="_ _2"></span>度及对<span class="_ _2"></span>比度可<span class="_ _2"></span>以调节<span class="_ _2"></span>（通过<span class="_ _2"></span>控制参<span class="_ _2"></span>数</div><div class="t m0 x1 h4 y8 ff1 fs1 fc0 sc1 ls0 ws0">的改变，能够实时预览变化结果）；</div><div class="t m0 x3 h4 y9 ff1 fs1 fc0 sc1 ls0 ws0">（<span class="ff2">2</span>）直方图均衡：可在界面上对比均衡前后的效果。</div><div class="t m0 x1 h4 ya ff2 fs1 fc0 sc1 ls0 ws0">2<span class="ff1">、图像变换算法</span></div><div class="t m0 x3 h4 yb ff1 fs1 fc0 sc1 ls0 ws0">（<span class="ff2">3</span>）对一<span class="_ _2"></span>幅彩色图像进<span class="_ _2"></span>行<span class="_ _0"> </span><span class="ff2">DC<span class="_"> </span>T<span class="_ _0"> </span></span>变换和反变换，<span class="_ _2"></span>对比结果和原</div><div class="t m0 x1 h4 yc ff1 fs1 fc0 sc1 ls0 ws0">图；</div><div class="t m0 x3 h4 yd ff1 fs1 fc0 sc1 ls0 ws0">（<span class="ff2">4</span>）<span class="_ _2"></span>仅保留<span class="_ _2"></span>左上角<span class="_ _3"> </span><span class="ff2">16X16<span class="_ _0"> </span></span>的<span class="_ _4"> </span><span class="ff2">DCT<span class="_ _4"> </span></span>系<span class="_ _2"></span>数，进<span class="_ _2"></span>行反变<span class="_ _2"></span>换，<span class="_ _2"></span>计算结</div><div class="t m0 x1 h4 ye ff1 fs1 fc0 sc1 ls0 ws0">果图的信噪比<span class="_ _4"> </span><span class="ff2">SNR</span>。</div><div class="t m0 x1 h4 yf ff2 fs1 fc0 sc1 ls0 ws0">3<span class="ff1">、图像分割算法</span></div><div class="t m0 x3 h4 y10 ff1 fs1 fc0 sc1 ls0 ws0">（<span class="ff2">5</span>）采用最优阈值算法对灰度图像进行分割；</div><div class="t m0 x3 h4 y11 ff1 fs1 fc0 sc1 ls0 ws0">（<span class="ff2">6</span>）用<span class="_ _4"> </span><span class="ff2">Canny<span class="_ _4"> </span></span>算法对灰度图像进行边缘检测；</div><div class="t m0 x3 h4 y12 ff1 fs1 fc0 sc1 ls0 ws0">（<span class="ff2">7</span>）<span class="_ _2"></span>用<span class="_ _0"> </span><span class="ff2">hough<span class="_ _4"> </span></span>变换<span class="_ _2"></span>，检测<span class="_ _2"></span>边缘图<span class="_ _2"></span>像中<span class="_ _2"></span>的直线<span class="_ _2"></span>，并用<span class="_ _2"></span>不同颜<span class="_ _2"></span>色</div></div></div><div class="pi" data-data='{"ctm":[1.611850,0.000000,0.000000,1.611850,0.000000,0.000000]}'></div></div> </body> </html>

<div id="pf2" class="pf w0 h0" data-page-no="2"><div class="pc pc2 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/62688ba04f8811599e141040/bg2.jpg"><div class="c x0 y1 w2 h2"><div class="t m0 x1 h4 y13 ff1 fs1 fc0 sc1 ls0 ws0">将检测出的直线叠加显示到原图像上。</div><div class="t m0 x1 h3 y14 ff1 fs0 fc0 sc0 ls0 ws0">二、设计原理及分析</div><div class="t m0 x1 h4 y15 ff3 fs1 fc0 sc1 ls0 ws0">1<span class="ff1 sc0">、灰度线性变换的原理</span></div><div class="t m0 x1 h4 y16 ff1 fs1 fc0 sc1 ls0 ws0">（<span class="ff4">1</span>）全域线性变换</div><div class="t m0 x4 h4 y17 ff1 fs1 fc0 sc1 ls0 ws0">若大<span class="_ _2"></span>部分<span class="_ _2"></span>像素<span class="_ _2"></span>的灰<span class="_ _2"></span>阶分<span class="_ _2"></span>布<span class="_ _2"></span>在<span class="_ _5"></span><span class="ff5">[a, b]<span class="_ _2"></span></span>，小<span class="_ _2"></span>部分<span class="_ _2"></span>灰<span class="_ _2"></span>度级<span class="_ _2"></span>超出<span class="_ _2"></span>了</div><div class="t m0 x1 h4 y18 ff1 fs1 fc0 sc1 ls0 ws0">此区域，为了改善增强效果，可以用如下所示的变换关系：</div><div class="t m0 x1 h4 y19 ff1 fs1 fc0 sc1 ls0 ws0">（<span class="ff5">2</span>）分段线性变换</div></div></div><div class="pi" data-data='{"ctm":[1.611850,0.000000,0.000000,1.611850,0.000000,0.000000]}'></div></div>

<div id="pf3" class="pf w0 h0" data-page-no="3"><div class="pc pc3 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/62688ba04f8811599e141040/bg3.jpg"><div class="c x0 y1 w2 h2"><div class="t m0 x1 h4 y1a ff3 fs1 fc0 sc1 ls0 ws0">2<span class="ff1 sc0">、直方图均衡的原理</span></div><div class="t m0 x2 h4 y1b ff1 fs1 fc0 sc1 ls0 ws0">直<span class="_ _2"></span>方<span class="_ _1"></span>图<span class="_ _2"></span>是<span class="_ _1"></span>图<span class="_ _2"></span>像<span class="_ _1"></span>的<span class="_ _2"></span>灰<span class="_ _1"></span>度<span class="_ _5"></span>—<span class="_ _2"></span>—<span class="_ _1"></span>像<span class="_ _2"></span>素<span class="_ _1"></span>数<span class="_ _1"></span>统<span class="_ _2"></span>计<span class="_ _1"></span>图<span class="_ _2"></span>，<span class="_ _1"></span>即<span class="_ _2"></span>对<span class="_ _1"></span>于<span class="_ _2"></span>每<span class="_ _1"></span>个<span class="_ _2"></span>灰<span class="_ _1"></span>度</div><div class="t m0 x1 h4 y1c ff1 fs1 fc0 sc1 ls0 ws0">值<span class="_ _2"></span>，<span class="_ _2"></span>统<span class="_ _1"></span>计<span class="_ _2"></span>在<span class="_ _2"></span>图<span class="_ _2"></span>像<span class="_ _1"></span>中<span class="_ _2"></span>具<span class="_ _1"></span>有<span class="_ _2"></span>该<span class="_ _2"></span>灰<span class="_ _2"></span>度<span class="_ _1"></span>值<span class="_ _2"></span>的<span class="_ _1"></span>像<span class="_ _2"></span>素<span class="_ _2"></span>个<span class="_ _2"></span>数<span class="_ _1"></span>，<span class="_ _2"></span>并<span class="_ _1"></span>绘<span class="_ _2"></span>制<span class="_ _2"></span>成<span class="_ _2"></span>图<span class="_ _1"></span>形<span class="_ _2"></span>，</div><div class="t m0 x1 h4 y1d ff1 fs1 fc0 sc1 ls0 ws0">称为灰度直方图（简称直方图）。</div><div class="t m0 x5 h4 y1e ff1 fs1 fc0 sc1 ls0 ws0">设<span class="_ _2"></span>一<span class="_ _1"></span>幅<span class="_ _2"></span>给<span class="_ _1"></span>定<span class="_ _2"></span>图<span class="_ _1"></span>像<span class="_ _3"> </span><span class="ff6">f<span class="_ _3"> </span></span>的<span class="_ _2"></span>灰<span class="_ _2"></span>度<span class="_ _1"></span>级<span class="_ _1"></span>分<span class="_ _2"></span>布<span class="_ _1"></span>在<span class="_ _3"> </span><span class="ff4">0≤<span class="ff7">r</span>≤1<span class="_ _0"> </span></span>范<span class="_ _1"></span>围<span class="_ _2"></span>内<span class="_ _1"></span>。<span class="_ _1"></span>可<span class="_ _2"></span>以<span class="_ _1"></span>对</div><div class="t m0 x1 h4 y1f ff1 fs1 fc0 sc1 ls0 ws0">［<span class="ff4">0, 1</span>］区间内的任一个<span class="_ _4"> </span><span class="ff7">r<span class="_ _4"> </span></span>值进行如下变换： <span class="ff7">s<span class="ff4">=T(</span>r<span class="ff4">) </span></span></div><div class="t m0 x5 h4 y20 ff1 fs1 fc0 sc1 ls0 ws0">也就<span class="_ _2"></span>是<span class="_ _2"></span>说<span class="_ _2"></span>，<span class="_ _2"></span>通<span class="_ _2"></span>过<span class="_ _2"></span>上<span class="_ _2"></span>述<span class="_ _2"></span>变<span class="_ _2"></span>换，<span class="_ _2"></span>每<span class="_ _2"></span>个<span class="_ _2"></span>原<span class="_ _2"></span>始<span class="_ _2"></span>图<span class="_ _2"></span>像<span class="_ _2"></span>的<span class="_ _2"></span>像<span class="_ _2"></span>素灰<span class="_ _1"></span>度值<span class="_ _6"> </span><span class="ff7">r<span class="_ _0"> </span></span>都</div><div class="t m0 x1 h4 y21 ff1 fs1 fc0 sc1 ls0 ws0">对应产生一个<span class="_ _4"> </span><span class="ff7">s<span class="_ _4"> </span></span>值。变换函数<span class="_ _4"> </span><span class="ff4">T(<span class="ff7">r</span>)</span>应满足下列条件：</div><div class="t m0 x5 h4 y22 ff1 fs1 fc0 sc1 ls0 ws0">（<span class="ff4">1</span>） 在<span class="_ _4"> </span><span class="ff4">0≤<span class="ff7">r</span>≤1<span class="_ _4"> </span></span>区间内，<span class="ff4">T(<span class="ff7">r</span>)</span>值单调增加；</div><div class="t m0 x5 h4 y23 ff1 fs1 fc0 sc1 ls0 ws0">（<span class="ff4">2</span>） 对于<span class="_ _4"> </span><span class="ff4">0≤<span class="ff7">r</span>≤1</span>， 有 <span class="ff4">0≤T(<span class="ff7">r</span>)≤1</span>。</div><div class="t m0 x5 h4 y24 ff1 fs1 fc0 sc1 ls0 ws0">从<span class="_ _4"> </span><span class="ff7">s<span class="_ _4"> </span></span>到<span class="_ _4"> </span><span class="ff7">r<span class="_ _4"> </span></span>的反变换可用下式表示： <span class="ff7">r<span class="ff4">=</span>T</span></div><div class="t m0 x6 h5 y25 ff4 fs2 fc0 sc1 ls0 ws0">-1</div><div class="t m0 x7 h6 y24 ff4 fs1 fc0 sc1 ls0 ws0">(<span class="ff7">s</span>)</div><div class="t m0 x5 h4 y26 ff1 fs1 fc0 sc1 ls0 ws0">由概<span class="_ _2"></span>率论<span class="_ _2"></span>理论<span class="_ _2"></span>可知<span class="_ _2"></span>，如<span class="_ _2"></span>果已<span class="_ _2"></span>知随<span class="_ _2"></span>机变<span class="_ _2"></span>量<span class="_ _3"> </span><span class="ff4">ξ<span class="_ _0"> </span></span>的概<span class="_ _2"></span>率密<span class="_ _2"></span>度函<span class="_ _2"></span>数为</div></div></div><div class="pi" data-data='{"ctm":[1.611850,0.000000,0.000000,1.611850,0.000000,0.000000]}'></div></div>

<div id="pf4" class="pf w0 h0" data-page-no="4"><div class="pc pc4 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/62688ba04f8811599e141040/bg4.jpg"><div class="c x0 y1 w2 h2"><div class="t m0 x1 h6 y13 ff7 fs1 fc0 sc1 ls0 ws0">p</div><div class="t m0 x8 h5 y27 ff7 fs2 fc0 sc1 ls0 ws0">r</div><div class="t m0 x9 h4 y13 ff4 fs1 fc0 sc1 ls0 ws0">(<span class="ff7">r</span>)<span class="ff1">，而随<span class="_ _2"></span>机变量<span class="_ _0"> </span><span class="ff7">η<span class="_ _0"> </span></span>是<span class="_ _4"> </span><span class="ff7">ξ</span></span> <span class="_ _2"></span><span class="ff1">的函数<span class="_ _2"></span>，即<span class="_ _0"> </span><span class="ff7">η</span></span>=<span class="ff7">f</span>(ξ)<span class="ff1">，<span class="_ _2"></span><span class="ff7">η</span></span> <span class="_ _2"></span><span class="ff1">的概率<span class="_ _2"></span>密度<span class="_ _2"></span>为</span></div><div class="t m0 x1 h6 y28 ff7 fs1 fc0 sc1 ls0 ws0">p</div><div class="t m0 x8 h5 y29 ff7 fs2 fc0 sc1 ls0 ws0">s</div><div class="t m0 xa h4 y28 ff7 fs1 fc0 sc1 ls0 ws0"> <span class="ff4">(</span>s<span class="ff4">)<span class="ff1">，可由<span class="_ _4"> </span></span></span>p</div><div class="t m0 xb h5 y29 ff7 fs2 fc0 sc1 ls0 ws0">r</div><div class="t m0 xc h4 y28 ff4 fs1 fc0 sc1 ls0 ws0">(<span class="ff7">r</span>)<span class="ff1">求出<span class="_ _4"> </span><span class="ff7">p</span></span></div><div class="t m0 xd h5 y29 ff7 fs2 fc0 sc1 ls0 ws0">s</div><div class="t m0 xe h4 y28 ff4 fs1 fc0 sc1 ls0 ws0"> (<span class="ff7">s</span>)<span class="ff1">。 </span></div><div class="t m0 x5 h4 y2a ff1 fs1 fc0 sc1 ls0 ws0">因<span class="_ _1"></span>为<span class="_ _6"> </span><span class="ff7">s<span class="ff4">=</span>f<span class="ff4">(</span>r<span class="ff4">)<span class="_ _5"></span></span></span>是<span class="_ _1"></span>单<span class="_ _5"></span>调<span class="_ _5"></span>增<span class="_ _5"></span>加<span class="_ _5"></span>的<span class="_ _5"></span>，<span class="_ _5"></span>由<span class="_ _5"></span>数<span class="_ _1"></span>学<span class="_ _5"></span>分<span class="_ _5"></span>析<span class="_ _5"></span>可<span class="_ _5"></span>知<span class="_ _5"></span>，<span class="_ _1"></span>它<span class="_ _5"></span>的<span class="_ _5"></span>反<span class="_ _5"></span>函<span class="_ _5"></span>数</div><div class="t m0 x1 h6 y2b ff7 fs1 fc0 sc1 ls0 ws0">r<span class="ff4">=</span>f</div><div class="t m0 xf h5 y2c ff4 fs2 fc0 sc1 ls0 ws0">-1</div><div class="t m0 x10 h4 y2b ff4 fs1 fc0 sc1 ls0 ws0">(<span class="ff7">s</span>)<span class="_ _1"></span><span class="ff1">也<span class="_ _5"></span>是<span class="_ _1"></span>单<span class="_ _5"></span>调<span class="_ _1"></span>函<span class="_ _5"></span>数<span class="_ _1"></span>。<span class="_ _5"></span>在<span class="_ _1"></span>这<span class="_ _1"></span>种<span class="_ _5"></span>情<span class="_ _1"></span>况<span class="_ _5"></span>下<span class="_ _1"></span>，<span class="_ _5"></span>如<span class="_ _1"></span>图<span class="_"> </span></span>4-8<span class="_ _3"> </span><span class="ff1">所<span class="_ _1"></span>示<span class="_ _1"></span>，<span class="_ _5"></span><span class="ff7">η<span class="_ _5"></span></span>＜<span class="_ _1"></span></span>s</div><div class="t m0 x1 h4 y2d ff1 fs1 fc0 sc1 ls0 ws0">且仅当<span class="_ _4"> </span><span class="ff4">ξ</span>＜<span class="ff7">r<span class="_ _4"> </span></span>时发生，所以可以求得随机变量<span class="_ _4"> </span><span class="ff7">η<span class="_ _4"> </span></span>的分布函数为<span class="ff4">:</span></div><div class="t m0 x11 h4 y2e ff1 fs1 fc0 sc1 ls0 ws0">对上式两边求导，即可得到随机变量<span class="ff7">η</span>的分布密度函数<span class="ff7">p</span></div><div class="t m0 x12 h5 y2f ff4 fs2 fc0 sc1 ls0 ws0">s</div><div class="t m0 x13 h6 y2e ff4 fs1 fc0 sc1 ls0 ws0"> </div><div class="t m0 x14 h4 y30 ff4 fs1 fc0 sc1 ls0 ws0">(<span class="ff7">s</span>)<span class="ff1">为</span></div><div class="t m0 x15 h4 y31 ff1 fs1 fc0 sc1 ls0 ws0">通过变换函数<span class="ff7">f<span class="ff4">(</span>r<span class="ff4">)</span></span>可以控制图像灰度级的概率密度函数，</div><div class="t m0 x14 h4 y32 ff1 fs1 fc0 sc1 ls0 ws0">从而改变图像的灰度层次。这就是直方图修改技术的理论基</div><div class="t m0 x14 h4 y33 ff1 fs1 fc0 sc1 ls0 ws0">础。</div><div class="t m0 x1 h4 y34 ff3 fs1 fc0 sc1 ls0 ws0">3<span class="ff1 sc0">、</span>DCT<span class="_ _4"> </span><span class="ff1 sc0">变换及其反变换的原理</span></div></div></div><div class="pi" data-data='{"ctm":[1.611850,0.000000,0.000000,1.611850,0.000000,0.000000]}'></div></div>