【作者】 高峰；

【导师】 汪平河；

【作者基本信息】 电子科技大学 ， 光学工程（专业学位）， 2016， 硕士

【摘要】 光学相干层析技术(Optical Coherence Tomography,OCT)是一种新型的光学成像技术,将光的低相干技术和共焦技术结合并利用外差测量的方法获取生物样品的断层图像,具有非侵入,非接触式,高分辨率,无损等特点。扫频源OCT(Swept Source OCT,SSOCT)是傅里叶相干层析系统(Fourier domian OCT,FD-OCT)的一种,与传统的时域相干层析系统(Time Domain OCT,TD-OCT)相比,由于采用扫频光源和点探测的方式,减少参考臂的移动,提高了成像速度和光能的利用率,改善图像质量。由于OCT的优良特性,已广泛应用于临床眼科诊断和测量以及材料的高分辨检测。特别是眼科应用,OCT技术能够实现对全眼球(包括角膜,巩膜,视网膜)的清晰成像,通过测量角膜的曲率、厚度,房角大小,视网膜多层结构等为医生检测眼科疾病提供便利。为了实现对人眼成像,本文主要内容是针对中心波长为1060nm的SS-OCT成像技术进行研究,包括许多图像过程中的关键技术,主要有:1、系统的扫描成像技术,利用函数发生卡等硬件控制SS-OCT系统对样品进行多种模式的扫描并进行数据的采集。2、色散校准技术,由于参考臂和样品臂中的光路不匹配等原因将引入系统色散,影响图像的信噪比,通过对系统的二阶和三阶色散进行补偿,从而提高系统的轴向分辨率。3、GPU图像处理加速,基于CUDA开发语言,利用图像处理器(GPU)的高度并行性,对系统采集数据进行加速处理,加速比达到20倍以上,基本实现系统的实时显示。4、基于相位相关的方法进行采样图像的配准。通过计算相邻帧图像的平移量,并进行多帧对准后叠加,消弱OCT图像的散斑噪声,提高图像信噪比和对比度。最后,利用搭建的SS-OCT系统对人眼的眼前节、巩膜和视网膜进行扫描成像,获得良好的图像效果,并对研究工作内容进行总结和分析。更多还原

【Abstract】 Optical Coherence Tomography technology(OCT) is a new kind of Optical imaging technology, the light of low Coherence technique is combined with confocal technique and the method of using heterodyne measurement for biological sample cross-sectional imagin, with the advantages of non-invasive, non-contact, high resolution, etc. Sweep Source OCT(SS-OCT) belong to Fourier domian OCT(FD-OCT), compared with the traditional Time Domain OCT(TD- OCT), due to adopted frequency sweep light sources and the method of point detection, reducing the movement of the reference arm, improves the imaging speed and light energy utilization, improve image quality. Because of the excellent characteristics, OCT has been widely used in clinical diagnosis and measurement and high-resolution test of material.In order to image the human eyes, in this paper, on the basic of the center wavelength of 1060 nm SS-OCT system, some key technology of imaging method are studied, mainly includes:1. Scanning imaging technology, use of function cards and capture card etc hardware to control SS-OCT system to collect the samples.2. Dispersion calibration technology, the unmatch in the reference arm and sample arm will introduce dispersion, which affect the signal-to-noise ratio of image, through compensating the system’s second order and third order dispersion would improve the axial resolution of the system.3. GPU acceleration, based on CUDA development language, combine with the high parallelism of graphics processor(GPU) to accelerate the system process speed, speed ratio reaches more than 20 times, reach the goal of real-time display.4. Based on phase correlation to sampling image registration. Through the calculation amount of translational adjacent frames, then alignment frames, weaken the OCT image speckle noise and enhance image SNR and contrast.Finally, our SS-OCT system used to capture the human eye imaging, and then summarize my research work.更多还原