节点文献

基于微机械薄膜变形镜的像差校正技术及其应用

Technology and Application of Aberration Correction Based on Micromachined Membrane Deformable Mirror

【作者】 李邦明

【导师】 廖文和;

【作者基本信息】 南京航空航天大学 , 机械工程, 2010, 博士

【摘要】 自适应光学技术是近年来发展迅速的一门新兴综合性技术,将其应用到人眼眼底的显微观察中,可以准确、快速的对人眼像差进行实时测量、控制和校正,提高眼底成像系统的分辨率,为人类眼科疾病及其它全身性疾病的诊断和治疗提供新的手段。论文以自适应光学理论、视光学理论为基础,深入研究基于微机械薄膜变形镜的人眼像差校正理论及其技术,为低成本高精度活体人眼视网膜成像系统的研发提供理论支持和工程应用的新途径。主要内容包括:(1)详细介绍了人眼眼球的结构,定义了人眼波前像差,建立了以Zernike多项式为基函数的人眼像差表达模型,系统研究了基于Hartmann-Shack波前传感器的人眼波前像差重建原理和方法,提出了基于奇异值分解的人眼波前像差重建算法,为人眼像差的校正奠定了基础。(2)详细介绍了微机械薄膜变形镜的结构和原理,理论分析和实验验证了微机械薄膜变形镜镜面形变和驱动电极控制信号之间的关系,建立了以影响函数矩阵为基础的变形镜控制信号和待补偿像差模式系数之间的关系模型,分析了微机械薄膜变形镜对Zernike模式像差的拟合能力和校正范围,为基于微机械薄膜变形镜的像差补偿算法设计提供了理论依据。(3)建立了基于微机械薄膜变形镜空间域像差补偿问题的模型,在开环校正模式下,提出了的基于乘子罚函数方法和基于有效约束集方法的控制信号求解算法,前者通过乘子向量和罚函数将原问题逐步转化为一个无约束的最优化问题,后者通过有效约束集方法将原问题转化为等式约束问题,两种方法具有速度快、精度高等优点,特别适用于系统误差较小、影响函数标定准确的的系统。(4)在闭环校正模式下,提出了基于奇异值分解的控制信号求解算法,构建了微机械薄膜变形镜可校正像差模式空间,详细分析了系统对构造正交基模式像差的校正特性和算法通过模式选择校正以提高变形镜校正能力的原理,通过计算机仿真和模拟眼校正实验,分析了模式像差选择对校正效果的影响,确立基于奇异值分解以实现校正模式滤除的像差补偿算法为最优的人眼像差校正算法。(5)总结了人眼像差的动态特性,建立了自适应光学系统的控制模型,推导了系统的传递函数,以提高控制系统的带宽为优化目标,对比分析了积分控制算法和Smith控制算法下系统的性能,提出了基于Smith控制和奇异值分解的波前校正算法,为基于微机械薄膜变形镜的人眼像差实时校正提供了算法支持。(6)根据论文理论研究成果,设计了国内首套基于微机械薄膜变形镜的视网膜细胞显微镜硬件系统和软件系统,设备具有人眼像差实时测量、校正和视网膜细胞成像功能,通过模拟眼和人眼校正实验,成功拍摄到高清晰度模拟眼眼底图像和活体人眼视网膜细胞图像,目前仪器已顺利通过杭州医疗器械质量监督检验中心的检测。

【Abstract】 As an emerging comprehensive technology which is growing rapidly in recent years, in the application of human fundus microscopic observation, adaptive optics provides a new way of diagnosis and treatment for human ophthalmology diseases and other systemic diseases because of the advantages of exact and rapid real time measurement, control and correction of human aberration, which can improve resolution of imaging system. On the basis of adaptive optics and visual optics theory, this thesis carries out a further research on human aberration correction theory and technology based on micromachined membrane deformable mirror, which can provide theoretical support and engineering practice guidance for the development of low-cost high precision living human retina imaging system. The content of this thesis mainly includes:(1) The human eyeball structure is described in detail. The human wavefront aberration is defined, and its representation model is established on the basis of Zernike polynomial. The principle and method of Hartmann-Shack wavefront sensor based human wavefront aberration reconstruction is systematically studied, and a wavefront aberration reconstruction algorithm of human eyes based on singular value decomposition is put forward, which lays the foundation for human aberration correction.(2) The structure and principle of micromachined membrane deformable mirror is described in detail, and the relationship between micromachined membrane deformable mirror deformation and driving electrode control signal is analyzed theoretically and verified experimentally. On the basis of influence function matrix, the relationship model between deformable mirror control signal and the mode parameter of aberration to be compensated is built, and the micromachined membrane deformable mirror fitting ability and correcting range of Zernike model aberration is analyzed, which provides theoretical basis for the design of aberration compensation algorithm based on micromachined membrane deformable mirror.(3) The spatial domain aberration compensation model based on micromachined membrane deformable mirror is built. In the open-loop correction mode, two control signal solving algorithms based on multiplier penalty function method and effective constraint sets method respectively are put forward. The first method converts the original problem into an unconstrained optimization problem by multiplier vector and penalty function. The second method converts the original problem into an equality constrainted problem. The two methods both have the advantages of high speed and precision, which are very suitable for the system with small system error and a high-accuracy influence function matrix.(4) In the close-loop correction mode, a control signal solving algorithm based on singular value decomposition is proposed, by which the correctable aberration mode space of the micromachine membrane deformable mirror is constructed. The system correction characteristic for the constructed orthogonal basis modes and the principle of improving the deformable mirror correctable ability by selecting aberration correction modes are analyzed in detail. Through computer simulation and model eye correction experiments, the influence of selecting aberration modes is analyzed. At last, the aberration compensation algorithm which can realize aberration modes fitering based on singular value decomposition is determinated as the optimal human eye aberration correction algorithm.(5) The dynamic characteristics of human eye aberration are summarized. The control model of adaptive optics is established and the system transfer function is derived. Taking the improvement of control system bandwidth as the optimization objective, system performance of integral control algorithm and smith control algorithm are comparative analyzed, and a wavefront correction algorithm based on smith control and singular value decomposition is proposed, which offers an appropriate method for human eye aberration correction based on micromachined membrane deformable mirror.(6) According to theoretical research results in this thesis, the hardware system and software system of the first domestic retina cell microscope based on micromachined membrane deformable mirror is designed, which can realize real time measurement and correction of human aberration and acquire retina cell images. By the correction experiments of model eye and human eye, the high definition model eye fundus images and the living human eye retina cell images are obtained successfully. At present, the microscope has successfully passed the test from Supervising and Testing Center of Hangzhou Zhejiang Institute for the control of Medical Device.

  • 【分类号】R11;TP391.7
  • 【被引频次】3
  • 【下载频次】299
  • 攻读期成果
节点文献中: 

本文链接的文献网络图示:

本文的引文网络