数字图像处理课程设计-基于matlab图像增强的比较及其理论分析

课程设计任务书学生姓名:专业班级:指导教师:工作单位:题目:基于matlab图像增强的比较及其理论分析初始条件:1.提供实验室机房及其matlab软件；2.给定一张灰度图片，模拟加入高斯噪声、椒盐噪声等若干噪声，然后使用中值滤波、均值滤波等方法进行图像增强。要求完成的主要任务:(包括课程设计工作量及其技术要求,以及说明书撰写等具体要求):（1）了解对灰度图片处理的基本原理和方法；（2）熟悉掌握matlab软件；（3）对灰度图片模拟加入噪声；（4）利用matlab对加入噪声的该图片进行处理，记录每一种方法的过程，并进行比较，分析哪种噪声的所对应的增强方法效果最好；（5）在进行处理时要对每一步处理进行理论分析；（6）要求阅读相关参考文献不少于5篇；（7）根据课程设计有关规范，按时、独立完成课程设计说明书。时间安排:(1)布置课程设计任务,查阅资料四天；(2)利用软件进行的图片处理，记录其结果一周；(3)完成课程设计报告书三天；指导教师签名:年月日系主任(或责任教师)签名:年月日XX大学《专业课程设计》说明书目录摘要·································································································I1Matlab软件简介··············································································11.1Matlab语言的历史···································································11.2Matlab软件概况······································································12原理分析·······················································································32.1数字图像噪声·········································································32.2空间域图像增强······································································32.3均值滤波···············································································32.4中值滤波···············································································43程序设计·······················································································63.1程序设计思路·········································································63.2要使用的Matlab函数·······························································64程序流程图····················································································84.1添加高斯噪声流程图································································84.2添加椒盐噪声流程图································································84.3均值滤波流程图······································································94.4中值滤波流程图·····································································105源程序·························································································125.1添加高斯噪声再滤波·······························································125.2添加椒盐噪声再滤波·······························································136程序运行结果及分析······································································156.1高斯噪声滤波结果··································································156.2椒盐噪声滤波结果··································································156.3结果分析··············································································167课程设计心得体会··········································································17参考文献·························································································18致谢·······························································································19XX大学《专业课程设计》说明书I摘要图像在实际应用中可能会遇到各种噪声干扰，如高斯噪声、椒盐噪声等。可以采用空间域图像增强的办法对其进行处理，减少噪声的影响。本次课设使用的是空间滤波对图片处理，包括均值滤波和中值滤波。关键词：高斯噪声，椒盐噪声，均值滤波，中值滤波XX大学《专业课程设计》说明书11Matlab软件简介1.1Matlab语言的历史70年代后期,身为美国NewMexico大学计算机系系主任的CleveMoler发现学生用FORTRAN编写接口程序很费时间,于是他开始自己动手,利用业余时间为学生编写EISPACK和LINPACK的接口程序。CleveMoler给这个接口程序取名为Matlab。1984年，为了推广Matlab在数值计算中的应用，CleveMoler、JohonLittle等正式成立了Mathworks公司，从而把Matlab推向市场，并开始了对Matlab工具相等的开发设计。1.2Matlab软件概况Matlab是MatrixLaboratory的缩写，意为矩阵实验室。它具有强大的矩阵处理功能和绘图功能，进还能进行文字处理，绘图，建模仿真等功能。随着版本的不断升级，它在数值计算及符号计算功能上得到了进一步完善。Matlab已经发展成为多学科、多种工作平台的功能强大的大型软件。在欧美等高校，Matlab已经成为线性代数、自动控制理论、概率论及数理统计、数字信号处理、时间序列分析、动态系统仿真等高级课程的基本教学工具。Matlab有以下一些特点：Matlab的帮助功能很强大，自带有详细的帮助手册，基于HTML的完整的帮助功能，也可以用help命令来得到帮助信息。程序语法与C语言类似，设计自由度大，方便我们编程。例如在Matlab里，用户无需对变量预定义就可使用。大量数学函数已经定义好，并且有很强的用户自定义函数的能力。Matlab有高级的程序环境，但程序环境很简单易用，有与其它语言编写的程序结合和输入输出格式化数据的能力；Matlab既具有结构化的控制语句，又有面向对象编程的特性。还有一个原因使Matlab受人们欢迎的，那就是Matlab源程序具有很大的开放性。除了内部函数以外，所有Matlab的核心文件和工具箱文件都是可读可改的源文件，用户可通过对源文件的修改以及加入自己的文件构成新的工具箱。Matlab有强大的的图形绘制功能。在Matlab里，数据可视化的操作非常简XX大学《专业课程设计》说明书2单易用。Matlab还有较强的编辑图形界面的能力。可以用来声成图解和可视化的二维、三维图。Matlab还拥有功能强大的各种工具箱。其工具箱分为两类：功能性工具箱和学科性工具箱。功能性工具箱主要用来扩充其符号计算功能，图示建模仿真功能，文字处理功能以及与硬件实时交互功能。功能性工具箱用于多种学科。而学科性工具箱是专业性比较强的，如（control、signalproceessing、commumnication）toolbox等。这些工具箱都是由该领域内学术水平很高的专家编写的，所以用户无需编写自己学科范围内的基础程序，而直接进行高，精，尖的研究，能极大地促进我们的学习研究工作。虽然Matlab有很多优点，但它也有一些缺点，比如：由于Matlab的程序不用编译等预处理，也不生成可执行文件，程序为解释执行，所以速度较慢。XX大学《专业课程设计》说明书32原理分析2.1数字图像噪声数字图像在图像获取集数字化过程，以及在数字图像传输的过程中，有可能会受到噪声干扰。图像传感器在获取图像中会受到环境和原件质量的影响，产生噪声。例如照相机照相时会受到光照影响。数字图像在传输过程中，由于传输信道回手到噪声干扰，也会产生噪声。噪声有很多种类，例如：高斯噪声，瑞利噪声，椒盐噪声，指数分布噪声等等。本次课设讨论的是高斯噪声和椒盐噪声。高斯噪声符合高斯分布，高斯噪声的产生源于电子电路噪声和由低照明度或高温带来的传感器噪声。其概率密度函数为：222/)(21)(zezp椒盐噪声又称（双极）脉冲噪声，其概率密度函数为：其他,0,,)(bzPbazPazpba,灰度级b在图像中将显示为一个亮点，灰度级a为一个暗点。椒盐噪声最主要表现在成像中的短暂停留中。2.2空间域图像增强对噪声图片处理要使用图像增强的方法。图像增强技术是为了人类视觉系统的生理接受特点而设计一种改善图像的方法，处理图像，使其更适合于特定应用。图像增强的方法有空间域方法和频域方法。本次课设使用的是空间域方法。空间域指图像平面本身，以图像像素直接处理为基础。空间域增强包括灰度变换，直方图处理，用算数、逻辑操作增强，空间滤波等。其中空间滤波包括平滑空间滤波器，锐化空间滤波器。平滑空间滤波器包括平滑线性滤波器和统计排序滤波器。2.3均值滤波平滑线性滤波器，即为均值滤波，其输出是包含在滤波掩模邻域内像素的简单平均值。这种处理减小了图像灰度的尖锐变化，可以减噪，也会有灰度边缘模糊的负面效应。XX大学《专业课程设计》说明书4一幅M×N的图像经过一个m×n（m和n是奇数）的加权平均值滤波器滤波的过程可由下式给出：aasbbtaasbbttswtysxftswyxg),(),(),(),(如下图2.1、2.2显示了两个3×3线性平滑滤波器：111111111图2.1空间均值滤波器121242