数字图像处理技术的学习总结知识大纲 4.61MB

kilroy_7需要积分:8(1积分=1元)

资源文件列表:

数字图像处理技术.zip 大约有6个文件

数字图像处理技术/MATLAB常用操作.docx 8KB
数字图像处理技术/MATLAB常用操作.html 23.58KB
数字图像处理技术/数字图像处理.I.docx 1.42MB
数字图像处理技术/数字图像处理.I.html 88.23KB
数字图像处理技术/数字图像处理.II.docx 3.15MB
数字图像处理技术/数字图像处理.II.html 80.08KB

资源介绍:

内容残缺，我也不知道归哪一类了

<link href="/image.php?url=https://csdnimg.cn/release/download_crawler_static/css/base.min.css" rel="stylesheet"/><link href="/image.php?url=https://csdnimg.cn/release/download_crawler_static/css/fancy.min.css" rel="stylesheet"/><link href="/image.php?url=https://csdnimg.cn/release/download_crawler_static/89538390/raw.css" rel="stylesheet"/><div id="sidebar" style="display: none"><div id="outline"></div></div><div class="pf w0 h0" data-page-no="1" id="pf1"><div class="pc pc1 w0 h0"><img alt="" class="bi x0 y0 w1 h1" src="/image.php?url=https://csdnimg.cn/release/download_crawler_static/89538390/bg1.jpg"/><div class="t m0 x1 h2 y1 ff1 fs0 fc0 sc0 ls0 ws0">数字图像处理<span class="ff2">.II</span></div><div class="t m0 x2 h3 y2 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">VI<span class="ff1">图像编码</span></span></div><div class="t m0 x3 h3 y3 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">6.1<span class="ff1">概述</span></span></div><div class="t m0 x4 h3 y4 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">动机<span class="ff2">/</span>原因：</span></div><div class="t m0 x5 h3 y5 ff1 fs1 fc1 sc0 ls0 ws0">表达数字图像所需数据量通常很大</div><div class="t m0 x4 h3 y6 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">图像编码：</span></div><div class="t m0 x5 h3 y7 ff1 fs1 fc1 sc0 ls0 ws0">采用对图像的新的表达方法以减小所需的数据量</div><div class="t m0 x4 h3 y8 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">数据和信息：</span></div><div class="t m0 x5 h3 y9 ff1 fs1 fc1 sc0 ls0 ws0">数据是信息的载体。<span class="ff2"> </span>对给定量的信息可用不同的数据量来表示；<span class="ff2"> </span></div><div class="t m0 x5 h3 ya ff1 fs1 fc1 sc0 ls0 ws0">对给定量的信息，设法减少表达这些信息的<span class="ff2"> </span>数据量称为<span class="ff4">数据压缩</span></div><div class="t m0 x4 h3 yb ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">图像压缩方法的分类<span class="ff2"> </span>：</span></div><div class="t m0 x6 h3 yc ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">信息保存型：</span></div><div class="t m0 x7 h3 yd ff1 fs1 fc1 sc0 ls0 ws0">在压缩和解压缩过程中没有信息损失<span class="ff2"> </span>压缩率一般在<span class="ff2">2 ~ 10</span>之间</div><div class="t m0 x6 h3 ye ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">信息损失型：</span></div><div class="t m0 x7 h3 yf ff1 fs1 fc1 sc0 ls0 ws0">常能取得较高的压缩率（几十<span class="ff2">~</span>几百）<span class="ff2"> </span>压缩后并不能经解压缩恢复原状</div><div class="t m0 x3 h3 y10 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">6.2<span class="ff1">无损压缩编码</span></span></div><div class="t m0 x4 h3 y11 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">6.2.1<span class="ff1">数据冗余</span></span></div><div class="t m0 x6 h3 y12 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">（<span class="ff2">1</span>）编码冗余</span></div><div class="t m0 x6 h3 y13 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">概念</span></div><div class="t m0 x7 h3 y14 ff1 fs1 fc1 sc0 ls0 ws0">数据是信息的载体<span class="ff2"> </span>同量的数据可表达不同量的信息<span class="ff2"> </span>同量的信息可用不同量的数据表达</div><div class="t m0 x8 h3 y15 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">冗余</span></div><div class="t m0 x9 h3 y16 ff1 fs1 fc1 sc0 ls0 ws0">数据表达了无用的信息<span class="ff2"> </span>数据表达了已表达的信息</div><div class="t m0 x6 h3 y17 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">图像压缩原理</span></div><div class="t m0 x7 h3 y18 ff1 fs1 fc1 sc0 ls0 ws0">由于一幅图像存在数据冗余和主观视觉冗<span class="ff2"> </span>余，所以压缩方式就可以从这两方面着手<span class="ff2"> </span></div><div class="t m0 x7 h3 y19 ff1 fs1 fc1 sc0 ls0 ws0">开展。</div><div class="t m0 x8 h3 y1a ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff5">◼<span class="ff1">改变图像信息的描述方式，以压缩掉图像中的数据冗余。</span></span></div><div class="t m0 x8 h3 y1b ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff5">◼<span class="ff1">忽略一些视觉不太明显的微小差异，以压缩掉图像中的视觉冗余</span></span></div><div class="t m0 x6 h3 y1c ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">数据冗余类别</span></div><div class="t m0 x7 h3 y1d ff1 fs1 fc1 sc0 ls0 ws0">减少<span class="ff2">/</span>消除其中的一种<span class="ff2">/</span>多种冗余，就能取得数据压缩的效果。</div><div class="t m0 x8 h3 y1e ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">(1) <span class="ff1">编码冗余：与灰度分布的概率特性有关；</span></span></div><div class="t m0 x8 h3 y1f ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">(2) <span class="ff1">像素相关冗余：空间冗余，几何冗余；</span></span></div></div><div class="pi" data-data='{"ctm":[1.611830,0.000000,0.000000,1.611830,0.000000,0.000000]}'></div></div><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="/image.php?url=https://csdnimg.cn/release/download_crawler_static/89538390/bg2.jpg"><div class="t m0 xa h3 y20 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">原始图像越有规则，各像素之间的相关性越强，它可能压缩的数据就越多。（规则</span></div><div class="t m0 xb h3 y21 ff1 fs1 fc0 sc0 ls0 ws0">冗余大，不规则冗余小）</div><div class="t m0 xb h3 y22 ff1 fs1 fc1 sc0 ls0 ws0">为减少像素间冗余，需要将常用的<span class="ff2">2D</span>像素矩阵<span class="ff2"> </span></div><div class="t m0 xb h3 y23 ff1 fs1 fc1 sc0 ls0 ws0">表达形式转换为某种更有效（但可能不直观）<span class="ff2"> </span></div><div class="t m0 xb h3 y24 ff1 fs1 fc1 sc0 ls0 ws0">的表达形式，这种减少像素间冗余的转换常称<span class="ff2"> </span>为映射。<span class="ff2"> • </span></div><div class="t m0 xb h3 y25 ff1 fs1 fc1 sc0 ls0 ws0">如果原始图像能从转换后的数据集合完全恢复<span class="ff2"> </span></div><div class="t m0 xb h3 y26 ff1 fs1 fc1 sc0 ls0 ws0">出来，则这种映射是可反转的（压缩是无损的<span class="ff2"> </span>），否则就是不可反转的。</div><div class="t m0 x8 h3 y27 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">(3) <span class="ff1">心理视觉冗余：与主观感觉有关。</span></span></div><div class="t m0 x9 h3 y28 ff1 fs1 fc1 sc0 ls0 ws0">眼睛对某些视觉信息更敏感<span class="ff2"> </span>人对某些视觉信息更关心<span class="ff2"> </span>（损失不可逆转）</div><div class="t m0 xa h3 y29 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">1.<span class="ff1">图像在色彩，亮度，空间频率等方面超出了人眼的接收范围和可区分程度。</span></span></div><div class="t m0 xa h3 y2a ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">2.<span class="ff1">图像中某些目标或区域的信息不太容易引起人眼的注意。</span></span></div><div class="t m0 x6 h3 y2b ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">相对数据冗余</span></div><div class="t m0 x8 h4 y2c ff3 fs1 fc0 sc0 ls0 ws0">●</div><div class="t m0 x4 h3 y2d ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">6.2.2 <span class="ff1">图象编码中的保真度准则</span> </span></div><div class="t m0 x5 h3 y2e ff4 fs1 fc1 sc0 ls0 ws0">图像保真度<span class="ff6">;<span class="ff1">信息保存型<span class="ff2">/</span>信息损失型<span class="ff2"> </span>描述解码图象相对于原始图象的偏离程度<span class="ff2"> </span></span></span></div><div class="t m0 x5 h3 y2f ff1 fs1 fc1 sc0 ls0 ws0">对信息损失的测度<span class="ff2"> </span></div><div class="t m0 x6 h3 y30 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">1. <span class="ff1">客观保真度准则</span></span></div><div class="t m0 x7 h3 y31 ff1 fs1 fc1 sc0 ls0 ws0">主观测量图象的质量，因人而异，应用不方便</div><div class="t m0 x6 h3 y32 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">2. <span class="ff1">主观保真度准则</span></span></div><div class="t m0 x7 h3 y33 ff1 fs1 fc1 sc0 ls0 ws0">用编码输入图与解码输出图的某个确定函数表<span class="ff2"> </span>示损失的信息量，<span class="ff2"> </span>便于计算或测量<span class="ff2"> </span></div><div class="t m0 x4 h3 y34 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">6.2.3 <span class="ff1">图象编码模型</span> </span></div><div class="t m0 x6 h4 y35 ff3 fs1 fc0 sc0 ls0 ws0">●</div></div><div class="pi" data-data='{"ctm":[1.611830,0.000000,0.000000,1.611830,0.000000,0.000000]}'></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="/image.php?url=https://csdnimg.cn/release/download_crawler_static/89538390/bg3.jpg"><div class="t m0 x6 h4 y36 ff3 fs1 fc0 sc0 ls0 ws0">●</div><div class="t m0 x6 h3 y37 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">信源编码器和解码器的模型</span></div><div class="t m0 x8 h3 y38 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">•<span class="ff1">映射器</span> <span class="ff1">：减少像素间冗余，或进行图像变换。</span></span></div><div class="t m0 x8 h3 y39 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">•<span class="ff1">量化器</span> <span class="ff1">：减少视觉心理冗余，仅用于有损压缩。</span></span></div><div class="t m0 x8 h3 y3a ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">•<span class="ff1">符号编码器：减少编码冗余使用哈夫曼编码</span></span></div><div class="t m0 x4 h3 y3b ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">6.2.4 <span class="ff1">信息论的基础理论</span></span></div><div class="t m0 x3 h3 y3c ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">6.3<span class="ff1">无损压缩编码</span></span></div><div class="t m0 x4 h3 y3d ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">6.3.1 <span class="ff1">基本编码定理</span></span></div><div class="t m0 x6 h3 y3e ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">概念</span></div><div class="t m0 x7 h3 y3f ff4 fs1 fc1 sc0 ls0 ws0">编码：<span class="ff1">需建立码本来表达数据<span class="ff2"> </span></span>码本<span class="ff1">（字典）：用来表达一定量的信息或一组事件所<span class="ff2"> </span></span></div><div class="t m0 x7 h3 y40 ff1 fs1 fc1 sc0 ls0 ws0">需的一系列符号（如字母、数字等）<span class="ff2"> <span class="ff4">码字</span></span>（单词）：对每个信息或事件所赋的码符号序列<span class="ff2"> </span></div><div class="t m0 x7 h3 y41 ff1 fs1 fc1 sc0 ls0 ws0">码字的长度（字长）：<span class="ff2"> </span>字母个数<span class="ff2">/</span>每个码字里的符号个数<span class="ff2"> </span></div><div class="t m0 x7 h3 y42 ff1 fs1 fc1 sc0 ls0 ws0">如果一个图像的灰度级编码，使用了多于实际需要的编码符号，就称该图像包含了编码冗</div><div class="t m0 x7 h3 y43 ff1 fs1 fc1 sc0 ls0 ws0">余。</div><div class="t m0 x8 h4 y44 ff3 fs1 fc0 sc0 ls0 ws0">●</div></div><div class="pi" data-data='{"ctm":[1.611830,0.000000,0.000000,1.611830,0.000000,0.000000]}'></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="/image.php?url=https://csdnimg.cn/release/download_crawler_static/89538390/bg4.jpg"><div class="t m0 x8 h4 y45 ff3 fs1 fc0 sc0 ls0 ws0">●</div><div class="t m0 x8 h3 y46 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">冗余度：<span class="ff2">1-</span>编码效率</span></div><div class="t m0 x8 h4 y47 ff3 fs1 fc0 sc0 ls0 ws0">●</div></div><div class="pi" data-data='{"ctm":[1.611830,0.000000,0.000000,1.611830,0.000000,0.000000]}'></div></div><div id="pf5" class="pf w0 h0" data-page-no="5"><div class="pc pc5 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="/image.php?url=https://csdnimg.cn/release/download_crawler_static/89538390/bg5.jpg"><div class="t m0 x6 h3 y20 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">最佳编码原理<span class="ff2">·<span class="ff4">熵</span></span></span></div><div class="t m0 x8 h4 y48 ff3 fs1 fc0 sc0 ls0 ws0">●</div><div class="t m0 x8 h4 y49 ff3 fs1 fc0 sc0 ls0 ws0">●</div><div class="t m0 x4 h3 y4a ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff2">6.3.2 <span class="ff1">霍夫曼编码</span></span></div><div class="t m0 x5 h3 y4b ff1 fs1 fc1 sc0 ls0 ws0">可变长最佳编码：<span class="ff2"> </span>给出现概率高的符号分配较短的码字，<span class="ff2"> </span></div><div class="t m0 x5 h3 y4c ff1 fs1 fc1 sc0 ls0 ws0">给出现概率低的符号分配较长的码字。<span class="ff2"> </span></div><div class="t m0 x6 h3 y4d ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">算法步骤</span></div><div class="t m0 x8 h3 y4e ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">（<span class="ff2">1</span>）将信源符号出现的概率按从小到大的顺序排列。</span></div><div class="t m0 x8 h3 y4f ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">（<span class="ff2">2</span>）把两个最小的概率进行组合相加，形成一个新的概率。</span></div><div class="t m0 x8 h3 y50 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">（<span class="ff2">3</span>）重复第（<span class="ff2">1</span>）（<span class="ff2">2</span>）步，直到概率和达到<span class="ff2">1</span>为止。</span></div><div class="t m0 x8 h3 y51 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">（<span class="ff2">4</span>）合并符号时，将被合并的符号赋以<span class="ff2">1</span>和<span class="ff2">0</span>（大概率赋<span class="ff2">1</span>，小</span></div><div class="t m0 x8 h3 y52 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">概率赋<span class="ff2">0</span>，或者相反）。</span></div><div class="t m0 x8 h3 y53 ff3 fs1 fc0 sc0 ls0 ws0">●<span class="_ _0"> </span><span class="ff1">（<span class="ff2">5</span>）寻找从每一个信源符号到概率为<span class="ff2">1</span>处的路径，记录下路径</span></div></div><div class="pi" data-data='{"ctm":[1.611830,0.000000,0.000000,1.611830,0.000000,0.000000]}'></div></div>

100+评论昵称:

captcha

类型	标题	大小	时间
	pdf电子发票提取主要内容转excel表格	47.31MB	9月前

	巴比达内网穿透、免费内网穿透	12.01MB	9月前
	巴比达9.0.8版免费内网穿透	12.01MB	9月前
	西南交通大学模拟电子技术课程资料	18.26MB	9月前
	rocketmq-dashboard-1.0.0.zip	50.22MB	9月前

	rocketmq-all-4.4.0-bin-release.zip	11.73MB	9月前
	CADe_SIMU 3.zip	4.52MB	9月前
	PDF_Converter.zip	71.79MB	9月前