【作者】 李享；

【导师】 孙亚秀；

【作者基本信息】 哈尔滨工程大学 ， 电子科学与技术， 2019， 硕士

【摘要】 根据谱域中零频分量、实像和共轭像分离情况的不同,数字全息技术分为同轴、离轴和轻微离轴技术,其中离轴数字全息因其可一次曝光采集成像在实时三维定量测量领域受到了广泛青睐。但由于受到零频分量的干扰,传统的傅立叶变换方法恢复出的离轴相位图像分辨率并不高,精度受限;通过相移方法恢复出的相位图像分辨率精度较高,但是需要采集多幅图像,以时间分辨率或成像视场为代价;小波变换等图像处理技术虽然能提高相位恢复质量,但需进行小波基等的选择,缺乏自适应性。而这些问题的存在使得数字全息无法兼顾成像的质量和速度。因此,如何在离轴数字全息技术中实现高分辨率、自适应三维成像测量具有重要的研究价值。本文针对基于离轴结构的白光衍射相位显微系统重建相位过程中不能同时保证恢复相位精度、计算效率和自适应性的问题,引入希尔伯特黄技术,有效地抑制零级及噪声干扰,基于一幅离轴全息图实现相位高分辨率、自适应重建,进而完成组织等折射率测量。具体研究内容如下:首先,根据离轴数字全息技术相位恢复的基本原理,由其频谱特点引入二维经验模态分解,提出基于快速二维经验模态分解的自适应相位重构技术,实现了零频分量的抑制,并通过实验和仿真验证了该方法的优越性,单次曝光即可接近相移方法的重构精度。其次,针对二维经验模态分解过程中存在的耗时性高的问题,以数字全息图像特点为基准,对快速方法进一步做出了改进,结合希尔伯特螺旋变换,提出了基于改进希尔伯特黄的自适应相位重构技术,一方面解决了耗时性高的问题,另一方面也提升了重建像的质量。最后,通过实验和仿真验证了本文提出算法的可行性、快速自适应性和普适性。在定量相位成像的基础上,将改进希尔伯特黄算法获取的相位信息和空间折射率分布结合起来,在未知样本厚度的前提下实现了基于相位的折射率测量,为未来提取医学标本光学特征信息、获取细胞组织病变信息提供了新的思路。更多还原

【Abstract】 According to the difference separation of zero-frequency component,real image and conjugate image in the spectral domain,digital holography can be divided into on-axis,off-axis and slightly off-axis technology.And among them,off-axis digital holography has been widely favored in the field of real-time three-dimensional quantitative measurement because of its single-time exposure imaging.However,due to the interference of the zero-frequency component,the resolution of the off-axis phase image recovered by the traditional Fourier transform method is not high and the accuracy is limited;The phase image recovered by the phase-shifting method has higher resolution accuracy,but needs to acquire multiple images at the cost of time resolution or imaging field of view;Although image processing techniques such as wavelet transform can improve the quality of phase recovery,it is necessary to select wavelet bases and the like,and lacks adaptability.The existence of these problems makes digital holography unable to balance the quality and speed of imaging.Therefore,how to achieve high-resolution,adaptive three-dimensional imaging measurement in off-axis digital holography has important research value.In this paper,Hilbert-Huang technique is introduced to effectively suppress zero-order and noise interference in the phase reconstruction of off-axis phase diffraction phase microscopy system,which can not guarantee the phase recovery accuracy,computational efficiency and adaptability at the same time.Based on an off-axis hologram,phase resolution and adaptive reconstruction are realized,and then refractive index measurement such as organization is completed.The main contents of this article are as follows:Firstly,according to the basic principle of phase recovery of off-axis digital holography technology,two-dimensional empirical mode decomposition is introduced by its spectral characteristics,and an adaptive phase reconstruction technique based on fast two-dimensional empirical mode decomposition is proposed to achieve the suppression of zero-frequency components.The superiority of the method was verified by experiments and simulations,the reconstruction accuracy of the phase-shifting method can be approached only by a single exposure.Secondly,aiming at the time-consuming problem in the process of two-dimensional empirical mode decomposition,based on the characteristics of digital holographic image,the fast method is further improved.Combined with the Hilbert spiral transformation,an adaptive phase reconstruction technique based on improved Hilbert Huang is proposed.On the one hand,it solves the time-consuming problem,and on the other hand,it improves the quality of the reconstructed image.Finally,the feasibility,rapid adaptability and universality of the proposed algorithm are verified by experiments and simulations.On the basis of quantitative phase imaging,the phase information obtained by the improved Hilbert Huang algorithm is combined with the spatial refractive index distribution to realize phase-based refractive index measurement under the premise of unknown sample thickness,and to extract medical specimen optics for the future.Characteristic information and information on obtaining tissue lesions provide new ideas.更多还原