微血管显微图像采集及处理系统的应用研究

发布时间：2018-01-22 14:01

  本文关键词： 微血管 图像 自动聚焦 稀疏去噪　出处：《东北石油大学》2017年硕士论文　论文类型：学位论文

【摘要】：微血管参与有机体内很多疾病的发生、发展过程,对于疾病的预警有很大的参考价值,并且提示了疾病临床治疗的新研究方向。要探讨微血管在有机体内的作用,首先要获得清晰的微血管显微图像。但研究对象为活体状态,由于其生理活动,会导致图像采集困难,并且图像在采集过程中会因受到噪声污染而使得成像质量变差,为了达到采集清晰图像的目的,设立了本课题,并进行了以下研究:首先,设计并实现了微血管显微图像采集系统。活体小动物自身的生理活动会产生图像待采集区域的平面位移或改变焦距,为了能采集到相对清晰的图像,本研究应用Visual Studio 2005在MFC的基础上应用C++程序设计语言编程实现了PC主机与二维运动平台和工业摄像机的接口设计,并应用Visual Studio 2005在DirectX套件的基础上应用C++程序设计语言编程实现了PC主机与平台运动控制器之间的接口设计,保证平台运动控制器对于二维运动平台和工业摄像机的控制,实现了系统的图像采集功能。然后,设计并实现了自动调焦系统。在图像采集系统的基础上引进一个步进电机,通过比较常用调焦函数特性,确定应用Variance函数作为自动调焦系统的运算函数,并基于逐次逼近法获取了采集图像的正焦位置,并在此基础上设计了自动调焦程序流程,应用C++程序设计语言编程实现了图像采集系统的自动调焦功能。最后,对于获取的微血管显微图像进行稀疏去噪处理。图像的稀疏去噪处理要解决稀疏去噪方法的选择和字典的设计两个关键性问题,通过对于常用算法收敛性和在构造字典的过程中所用到的原子数的比较,本研究中选择K-SVD算法训练出的字典对微血管显微图像进行稀疏去噪处理。并基于这种理论进行了仿真实验,K-SVD训练出的学习字典的成像质量好,本文提出的基于K-SVD训练出的学习字典具有良好的自适应性。
[Abstract]:Microvasculature is involved in the occurrence and development of many diseases in organic body, which has great reference value for disease warning. In order to explore the role of microvasculature in the organism, we must obtain a clear micrograph of the microvessel, but the object of study is living state, because of its physiological activities. It will lead to the difficulty of image acquisition, and the image quality will become worse because of the noise pollution in the process of image acquisition. In order to achieve the purpose of collecting clear images, this subject has been set up. The following studies are carried out: firstly, the microvascular microimage acquisition system is designed and implemented. The physiological activities of the living small animals will result in the plane displacement of the image to be collected or the change of focal length. In order to capture relatively clear images. In this study, Visual Studio 2005 is applied to realize the interface design of PC host computer with two-dimensional motion platform and industrial camera based on MFC. And the application of Visual Studio 2005 in the DirectX suite based on the application of C. The interface design between PC host and platform motion controller is realized by programming language. Ensure that the platform motion controller for two-dimensional motion platform and industrial camera control, achieve the system image acquisition function. Then. An automatic focusing system is designed and implemented. Based on the image acquisition system, a step motor is introduced, and the characteristics of the common focusing function are compared. The Variance function is used as the operation function of the auto-focusing system, and the positive focus position of the collected image is obtained based on the successive approximation method, and the program flow of auto-focusing is designed on this basis. C programming language is used to realize the automatic focusing function of the image acquisition system. Finally. The sparse denoising of the microvascular microimages obtained is done. The two key problems of sparse denoising are the selection of sparse denoising methods and the design of dictionaries. The convergence of common algorithms and the number of atoms used in the construction of dictionaries are compared. In this study, the dictionary trained by K-SVD algorithm is selected to deal with sparse denoising of microvascular image, and the simulation experiment is carried out based on this theory. The learning dictionary trained by K-SVD has good imaging quality, and the learning dictionary based on K-SVD has good adaptability.
【学位授予单位】：东北石油大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R318;TP391.41
，

本文编号：1454865