黄巢菊花
超声检测在临床医学中具有广泛的应用。医学超声检测利用超声波的反射和衍射特性，通过观察显示在超声检测仪上的反射波的位置、强度变化，来判定被检器官内部和表面是否存在异常。超声检测具有无创，实时，双模等优点。
超声回波反射技术是超声检测中常用的技术之一。然而，在使用回波反射技术时，由于回波的干涉效应以及散射回的超声波束之间存在相互干扰，当相关反射源反射出的两束回波重叠时，在超声图像上呈现为颗粒状，即论文说论述的超声图像斑点噪声。斑点噪声影响诊断信息的提取，给诊断带来困难。因此，使用超声检测时，消除斑点噪声对于获取准确的诊断信息具有重要的意义。
计算机技术的发展，使得计算机可以应用到更广的领域。数字图像处理就是其中一种应用。使用计算机处理超声图像，研究去斑算法，降低噪声，能提高获取诊断信息的准确性。利用计算机模型仿真斑点噪声，为研究、验证去斑算法提供了客观对象。而对于超声图像的研究应该建立在二维图像的基础上，因此，研究计算机仿真的二维超声图像斑点模型就更有现实意义。
目前针对斑点噪声模型的研究主要集中于斑点的一维分布特性。主要有两种分布模型。根据Burckhardt的研究发现，斑点噪声的幅值遵循瑞利（Rayleigh）分布，而斑点的位置分布会随着分辨单元的增多而发生变化。Burckhardt的理论详细描述了斑点的分布特性，许多研究均建立在此理论基础之上。但Burckhardt只给出了斑点噪声的一维分布模型。Cramblitt在对Burckhardt模型的研究基础上，用一个泊
松点过程描述一维斑点噪声的位置分布特性。而对噪声幅值则没有采用Burckhardt模型中给出的瑞利（Rayleigh）分布，转而以一个平稳高斯过程表示噪声的幅值统计特性。Cramblitt模型对Burckhardt模型进行了改进，但同样只给出了一维的统计分布和一维斑点噪声模型的实现。因此，探索一种易于实现的二维斑点噪声模型计算机仿真，对研究二维超声图像的去斑算法很重要。
直播行业Hilbert于1891年发现了Peano曲线的一种变形，把单位正方形分成四个小正方形。然后，类似地再把四个小正方形的每一个分成四个更小的正方形，以此类推。最终得到Hilbert曲线，这种曲线具有空间填充的特性。Hilbert曲线是一种重要的数字图像处理工具，在图像处理中被广泛的使用。
本文在研究Burckhardt模型和Cramblitt模型的基础上，给出了一种易于实现的利用Hilbert空间填充曲线生成二维斑点噪声图像的计算机模型。生成过程主要分三个步骤：仿真一维斑点噪声；生成Hilbert空间填充矩阵和Hilbert空间填充曲线；利用Hilbert空间填充矩阵生成二维斑点噪声图像。在一维斑点噪声的仿真中，对斑点噪声位置分布的描述保留了Cramblitt模型中给出的泊松点过程，本文还保留了Cramblitt给出的在点过程中，不同散射密度对斑点位置分布的影响结果。而幅值特性则采用Burckhardt模型中给出的瑞利（Rayleigh）分布来描述。生成Hilbert
空间填充矩阵和Hilbert空间填充曲线是通过采用矩阵计算递归算法编程实现。然后利用生成的空间填充矩阵将一维噪声转换为二维斑点噪声图像。最后，通过对不同散射密度下生成的二维斑点噪声图像
做图像分析对比，通过对不同散射密度下生成的二维斑点噪声图像做直方图分析，通过对不同散射密度下生成的二维斑点噪声图像做功率谱密度分析，通过与临床超声图像中的斑点分布作比较分析以及通过采用Burckhardt给出的斑点噪声模型复杂度计算公式，对Burckhardt模型，Cramblitt模型和本文给出的模型做模型复杂度柱状图比较，验证本文给出的二维斑点噪声图像模型的可行性和易实现性。实验结果表明，该模型生成的二维斑点噪声，能保持斑点噪声的统计特性，与真实噪声接近，且生成斑点噪声图像的算法相对容易。
关键词：超声斑点；瑞利分布；伽马分布；Hilbert空间填充；斑点图像
Abstract
Ultrasound detection has been widely ud in clinical applications. The medical ultrasound detection us the echo and diffraction characteristics of ultrasound wave to determine whether the abnormity of the internal organs surface occurs by obrving the changes of echo location and strength which are displayed on the ultrasonic detector. Ultrasonic detection has many advantages such as non-invasive, real-time and dual-mode.
Ultrasonic-echo detecting technique is one of ultrasonic detection methods commonly ud. However, when using this method, the interference effect and the echo back scattering between the
番薯做法
ultrasonic beams interfere with each other. When the two related echo overlap, the speckle is prent on the ultrasound images. It is that the ultrasound images speckle noi discusd in this paper. Speckle noi impacts on the extraction of diagnostic information and make diagno difficult. Therefore, speckle reducing is important to obtain the exact diagnostic information when using ultrasound detection.
With the developing of computer technology, the computers are applied to a broader area. Digital image processing is one of the applications. Using computer to process the ultrasound image, to study the speckle reducing algorithm for reducing the noi could increa the accuracy of obtaining the diagnostic information. Using computer to simulate the speckle noi can provide an effective model to rearch and validate the speckle reducing algorithm. And the ultrasound image is a two-dimensional image. Therefore, it has more practical significance to study the 2-D ultrasound image speckle noi simulation using computer.
Current rearches for speckle model are focus on the one-dimensional distribution of scatters. There are two main models. According to the Burckhardt, the amplitude of speckle noi follows the Rayleigh probability density distribution, and the location of the scatters will change with the increas e of the scatting cell. The Burckhardt’s theory has described the distribution of the speckle in detail.
Many studies have established on the ba of this theory. But Burckhardt only gave the one-dimensional speckle noi distribution model. Cramblitt ud the Poisson point process to describe the location distribution of the scatters and ud a stationary Gaussian process to describe the amplitude of speckle noi instead of the Rayleigh distribution that the Burckhardt gave. Cramblitt improved on the Burckhardt’s model. However, Cramblitt just gave the
one-dimensional speckle noi distribution model and the realization of the 1-D model. Therefore, it is important to find an easier way to simulate the 2-D speckle noi for speckle reducing algorithm study.ronaldo
In 1891, Hilbert found a Peano curve’s deformation. He puts the unit square into four smaller squares and each small square into four smaller squares as the same way. Finally, the Hilbert curve was obtained. The Hilbert curve has the characteristic of space-filling, which is one of the important tools using in digital image processing.
In prent study, an easier way to simulate the 2-D speckle noi image using Hilbert space-filling curve is given. The method prent in this paper is bad on the Burckhardt’s and Cramblitt’s stud ies. The process of the simulation contains three steps. Firstly, the 1-D speckle noi is simulated. S
econdly, the Hilbert curve and Hilbert space-filling matrix are obtained. At last, the 2-D speckle noi image is obtained by using the Hilbert curve obtained in the cond step. In the 1-D speckle noi simulation, the location distribution of the scatters is the Poisson point process which was given by Cramblitt and the amplitude of speckle noi follows the Rayleigh distribution given by Burckhardt. We program to obtain the Hilbert curve and Hilbert matrix by using the matrix calculating algorithm. Then we u the obtained Hilbert curve to take the 1-D speckle noi into the 2-D speckle noi image. We compare the 2-D images obtained in the ca of different scattering density and then we analyze the image histogram and the image power spectral density of the obtained images. We u the speckle noi complexity formula given by Burckhardt to compare the complexity of the model prent in this paper with the Burckhardt’s model and Cramblitt’s model. We compare the obtained speckle image with the clinical ultrasound image at last. Experimental results show that the 2-D speckle noi model prent in this paper could prerve the statistical characteristics of speckle noi and the speckle image obtained is similar to the clinical speckle noi.
Key words:ultrasound speckle; Rayleigh distribution; Gamma distribution; Hilbert space-filling; speckle image
目录
无线ap模式
摘要 (Ⅰ)
ABSTRACT (Ⅲ)
目录 (i)
第一章引言 (1)
1.1 研究目的和意义 (1)
炒黑米饭1.2 超声检测技术 (2)
1.2.1 超声波的概念及产生与接收 (2)
1.2.2 超声波的传播特性 (2)
1.2.3 医学超声检测 (3)
1.2.4 超声成像原理 (4)
1.2.5 超声图像斑点的产生原理 (5)
1.2.6 计算机在超声检测中的应用 (7)
1.3 超声图像斑点模型在国内外的研究现状 (7)
1.4 论文的主要工作 (10)
第二章数学理论及算法 (12)
2.1泊松分布 (12)
2.2瑞利分布 (13)
2.3伽马分布 (14)运动会通讯稿100字
2.4泊松点过程与散弹噪声 (15)
2.4.1泊松点过程 (15)
2.4.2散弹噪声 (17)
2.5希尔伯特空间填充 (18)
2.5.1希尔伯特空间填充的历史 (18)
手机号查姓名2.5.2 Hilbert曲线的生成算法 (19)
2.5.3 Hilbert空间填充矩阵的生成 (19)
第三章二维斑点噪声图像的仿真 (22)
3.1一维斑点噪声信号的仿真 (22)
3.1.1一维斑点噪声的散射点位置分布特性 (23)
3.1.2一维斑点噪声的幅值分布特性 (27)
