Comparison of phase recovery methods in spiral speckle pattern interferometry correlation fringes

Spiral interferometry can be used as a solution to the problem of sign ambiguity presented in the conventional speckle pattern interferometric technique when the optical phase needs to be reconstructed from a single closed fringe system. Depressions and elevations of the topography corresponding to the object deformation are distinguished by the direction of rotation of the local spiral fringe pattern. In this work, we implement and compare several methods for optical phase reconstruction by analyzing a single image composed of spiral speckle pattern interferometry correlation fringes. The implemented methods are based on contour line demodulation, center line demodulation, Spiral Phase Quadrature Transform and the 2D Riesz transform with multivector structure. Contour line and center line demodulation approaches are exclusively dedicated to images containing a fringe system with spiral structure. The others are based on the 2D Riesz transform, these being well known approaches in conventional interferometry. We examine simulated experiments and analyze some of the emerging drawbacks for solving the phase reconstruction problem by using different mean values of speckle size and background noise levels. We also discuss several numerical procedures that may well improve the efficiency and robustness of the presented numerical implementations. The performance of the implemented demodulation methods is evaluated by using a universal image quality index and therefore a quantitative comparison is also presented.     

扫频光学相干层析视网膜图像配准去噪算法

多帧叠加平均处理是去除扫频光学相干层析系统散斑噪声、获得较为清晰结构信息的有效方法,但眼睛的震颤、漂移、微眼跳等生理特性和系统光路特性会使图像之间存在错位,导致叠加效果不佳、结构稳定性差,为此本文提出一种基于灰度分布信息和目标几何信息相结合的配准算法。该方法根据图像平均灰度分布提取包含目标信息的感兴趣区域,通过相位相关算法和基于分段拟合的灰度投影算法的双重作用校正图像的平移变换;通过拟合视网膜上边界作为特征点迭代确定最佳旋转参数,并再次重新估计平移参数,实现图像的刚性配准;最后通过轴向扫描一对一映射法以能量函数为约束条件实现图像的非刚性配准。对活体兔眼进行实验,结果表明,本文算法配准后的叠加图像边界清晰,结构信息增强,信噪比和对比度平均有效提高一倍多。本算法适用于强噪声视网膜B-Scans图像的配准,能满足多种类型OCT系统的叠加成像需要,具有较高的鲁棒性和图像配准精度。     

天文望远镜像差对斑点成像技术的影响

着重研究了采用斑点成像技术处理天文望远镜图像时,光学系统固定像差对图像恢复结果的影响。在详细研究各种恢复天文图像振幅和相位的理论和方法的基础上,建立了一个包括Labeyrie振幅恢复方法和KnoxThompson相位重构方法的恢复扩展目标的斑点成像处理模型,分析了光学系统固定像差对系统传递函数相位分布和目标相位重构的影响：天文目标通过大气成像,固定像差将会被淹没在大气湍流随机起伏中,像差对相位重构没有显著影响。处理图像结果表明,斑点成像技术能同时消除大气湍流和望远镜系统固定像差的影响,得到高分辨力的扩展目标图像。还提出了一种消除光学系统像差的方法。     

Multiple image encryption using an aperture-modulated optical system

A multiple image cryptosystem based on different apertures in an optical set-up under a holographic arrangement is proposed. The system is a security architecture that uses different pupil aperture mask in the encoding lens to encrypt different images. Based on this approach multiple encryption is achieved by changing the pupil aperture arrangement of the optical system among exposures. In addition to the classical speckle phase mask, the geometrical parameters characterizing the apertures are introduced to increase the system security. Even when an illegal user steals the speckle phase mask, the system cannot be broken into without the correct pupil geometrical parameters. The experimental set-up is based on a volume photorefractive BSO crystal as storing device. Information retrieval is done via a phase conjugation operation. We also have to stress that the multiple storage under this scheme, is only possible with the help of the aperture mask. Simulation and experimental results are further introduced to verify the proposed method.     

Speckle interferometry for measurement of continuous deformations

Improvements of a method for measurement of continuous displacements and deformations with digital phase shifting speckle pattern interferometry are presented. The method is based on an algorithm that, with the knowledge of the initial phase, only needs one image at a time to evaluate continuos phase changes due to object deformations. In the improved method, the initial random phase of the speckle pattern is evaluated using a number of phase-shifted images before the deformation under study. This is used for increasing the accuracy of the initial phase estimation and reducing influences from image noise and other measurement disturbances. The phase-shifted speckle patterns are used as references for comparison with the speckle patterns of the deformed object, thereby increasing the reliability and accuracy of the phase estimations of the deformed patterns. The technique can be used for measuring deformations such as transients and other dynamic events, heat expansion as well as other phenomena where it is difficult to accomplish phase shifting during deformation.     

基于多角度无透镜傅里叶变换数字全息的散斑噪声抑制成像研究

在数字全息成像中,散斑噪声严重影响了再现像的信噪比和成像分辨率,因此为了提高数字全息成像质量,迫切需要抑制再现像的散斑噪声.分析并给出了矩形散射光斑的强度协方差,定量计算了特定实验条件下产生退相关散斑图样的最小角度差.结合透镜的性质设计并搭建了多角度无透镜傅里叶变换数字全息成像系统,利用光纤端面在透镜焦平面的二维机械移动代替传统反射镜的旋转,使照明光束在不改变照明位置和大小的同时,可任意改变光束方向.移动光纤端面使多角度照明满足最小角度差,获取了81幅数字全息图.利用单次快速傅里叶逆变换实现数值再现,对多幅再现像的强度像求平均,实验结果表明散斑对比度降低为单幅再现像的14.6%,使图像信噪比提高了6.9倍.该抑制散斑的多角度数字全息成像方法有效的抑制了散斑噪声,且成像光路结构简单,可操作性强.     

A novel method of high resolution imaging through random perturbation wave front

A novel method is proposed for eliminating the effect of the traditional optical aberration on high resolution imaging by random perturbation wave front and digital image processing. A random phase mask, whose phase spectrum fluctuation is accordant with Kolmogorov distribution, is positioned near the aperture stop of optical system, making the optical aberration image become random perturbation image, that is, the intermediate image. The blind deconvolution algorithm based on maximum-likelihood estimation technique is used to restore the intermediate image acquired by a digital detector. The effects of optical aberrations and the noise on the restoration image are explored. To demonstrate the validity of the method proposed, the computer simulation and laboratory experiments are carried out for the imaging of the optical system with primary aberration. The results have shown that the present method is well suited for effectively improving the imaging quality of the optical system with certain aberration, thus making the optical system resolution close to or reach the diffraction-limit of the optical system.     

Speckle correlation fringes denoising using stationary wavelet transform. Application in the wavelet phase evaluation technique

This paper presents a stationary wavelet transform (SWT) method for speckle noise reduction in digital speckle pattern interferometry fringes. The main advantage of SWT is its translation invariance, which makes it important in statistical image processing applications. This method was used to denoise a simulated speckle fringe patterns, a good fidelity value was obtained. Applied to the wavelet phase evaluation, it has provided a phase distribution with a good accuracy.     

基于多分辨分析的合成孔径声纳相干斑抑制研究

相干斑噪声的存在使合成孔径声纳图像不能正确反映目标的反射特性,严重影响了图像的质量,降低了对图像的分析和理解性能。一个良好的合成孔径声纳图像相干斑抑制算法,必须在有效抑制相干斑噪声的同时,尽量保持图像中的边缘、点目标等细节信息。小波变换具有多分辨特性,可以利用其多分辨特性进行图像噪声消除。由于合成孔径声纳图像相干斑具有乘性噪声的性质,这就需要对其合成孔径声纳图像进行对数变换,把乘性噪声转换成加性噪声。然后对加性噪声进行小波多分辨分析和阈值处理,剔除小于阈值的小波系数后进行逆小波变换和指数变换,从而获得抑制相干斑噪声后的新的合成孔径声纳图像。从相干斑抑制的结果可以看出,基于多分辨分析的相干斑噪声抑制算法能够有效降低合成孔径声纳图像中的相干斑噪声。     

Optimized phase pupil masks for extended depth of field

By properly designing a phase pupil mask to modulate or encode the optical images and then digitally restoring them, one can greatly extend the depth of field and improve image quality. The original works done by Dowski and Cathey introduce the use of a cubic phase pupil mask to extend the depth of field. The theoretical and experimental researches all verified its effectiveness. In this paper, we suggest the use of an exponential phase pupil mask to extend the depth of field. This phase mask has two variable parameters for designing to control the shape of the mask so as to modulate the wave-front more flexible. We employ an optimization procedure based on the Fisher information metric to obtain the optimum values of the parameters for the exponential and the cubic masks, respectively. A series of performance comparisons between these two optimized phase masks in extending the depth of field are then done. The results show that the exponential phase mask provide slight advantage to the cubic one in several aspects.     

Speckle Effects on Statistical Performance of Optical Radar

The statistical performance of optical radar using energy detection techniques has been studied in this paper. The analytical expressions of the speckle number subtended by the receiving aperture are derived. The results show that the speckle number is only determined by the relative size of the speckle cell and the area of the entrance pupil. The receiving aperture will integrate a considerable number of speckle cells when the beam expander is not adopted in the transmitting optical system. Few speckle cells will be intercepted by the aperture when the beam expander is employed and the laser speckle is expected to be the dominant noise source. The detection performance of airborne and space-based optical radar has been analyzed based on two critical parameters--the speckle number and the average photoelectron counts. The probability density function of photoelectrons and the detection probability are derived for a number of occasions.     

Code retrieval via undercover multiplexing

The purpose of this research is to develop an undercover multiplexing technique to give additional protection for optical information encryption. We employ the double random phase mask as our basic optical encryption system. The holographic storage medium of choice is a photorefractive crystal. To achieve the multiplexing we use the aperture size of the pupil in the optical system, as it governs the speckle size. We introduce such variation in order to produce a decorrelation between two consecutively stored speckle patterns. Each stored speckle pattern is associated to an input encrypted image, thus producing a multiplexing of the encrypted information. We implement this operation without altering the setup architecture and the random phase masks. This multiplexing is our undercover operation to encipher a true code behind a fake code. Under this approach, the user can only recover the bulk information stored in the volume hologram. However, he cannot recover the true code without the additional information on the pupil size key, even if accessed in position of the original decoding mask.     

Encryption and decryption using a sandwich phase diffuser made by using two speckle patterns and placed in the Fourier plane: Simulation results

In the present paper, we describe the encryption and decryption of two-dimensional images. The encryption is done by employing a sandwich phase diffuser made by using two elongated speckle patterns, and placed in the Fourier plane of a double random phase encoding system. After encryption, the two constituent phase diffusers of such a sandwich diffuser are separated. During decryption, if the conjugate of either of the two elongated phase speckle patterns is used, the image cannot be retrieved. Correct decryption is also not possible if a sandwich diffuser with any of the phase speckle patterns is shifted in position with respect to the other. For decryption, the encrypted image is Fourier transformed and multiplied by the conjugate of the sandwich diffuser, and then the product is further Fourier transformed. It is possible to generate the image only if both the elongated phase speckle patterns are matched point-to-point and then the proper conjugate is made. The use of elongated speckle patterns in constituting a sandwich phase diffuser makes the system less complicated as compared to the use of a sandwich diffuser made with normal speckle patterns, enabling an easy alignment of the random phase diffuser at the time of rejoining the constituent diffusers for making the right key. The ease of alignment is due to the reduction of the requirement of 360° scanning at the time of rejoining these diffusers with little reduction in the security of the system. Simulation results are presented in support of the proposed idea. For optical implementation, the decrypted image may be obtained by generating a phase conjugate wave by the phase conjugation technique, and passing through the same sandwich phase diffuser. To evaluate the reliability of the technique, mean square error (MSE) between the decrypted and original image has been calculated.     

Influence of the volume speckle on fiber specklegram sensors based on four-wave mixing in photorefractive materials

In this work, the dependence on the speckle size in the performance of a micro displacement sensor based on fiber specklegrams stored in a photorefractive BSO (Bi₁₂SiO₂₀) crystal is experimentally demonstrated. In our experimental setup, a plastic optical fiber (POF) was used to generate a subjective speckle pattern which was recorded in the crystal by using a four-wave mixing arrangement in transmission geometry. The speckle size was controlled by modifying the diameter of a pupil aperture adjacent to a lens producing the image of the speckle. The signal speckle beam was mixed into the crystal with two counter propagating pump beams to generate a fourth beam which is proportional to the conjugate of the original speckle beam. Real time fringe patterns were obtained at the output of the system by producing micro displacements of the fiber output end. Increases of the phase conjugation reflectivity and the visibility of the fringe patterns were appreciated when the speckle length was increased by decreasing the pupil aperture diameter. This behavior allowed recovering the autocorrelation functions of fringe patterns associated to micro displacements that initially led to decorrelation, and therefore, to improve the dynamic range of the metrological system. Until the best of our knowledge this is the first report about the influence of the speckle size on fiber specklegrams sensors recorded on photorefractive materials by four-wave mixing.     

Speckle noise reduction in digital holography with spatial light modulator and nonlocal means algorithm

An integrated method based on optical and digital image processing is presented to suppress speckle in digital holography. A spatial light modulator is adopted to introduce random phases to the illuminating beam. Multiple holograms are reconstructed and superimposed, and the intensity is averaged to smooth the noise. The adaptive algorithm based on the nonlocal means is designed to further suppress the speckle. The presented method is compared with other methods reduction is improved, and the proposed method is effective The experimental results show that speckle and feasible.     

散斑干涉相位条纹图的频域滤波处理

为了解决在数字散斑干涉技术测量时,散斑干涉相位条纹图像中大量噪声对相位解包裹结果和精度产生严重影响的问题,介绍了一种条纹正余弦分解和频域低通滤波结合的方法,实现了散斑干涉相位条纹图的高精度滤波。该方法的基本思路是在对相位图像进行滤波处理前,先将相位图通过正余弦函数进行映射转换成两幅图,分别经过频域滤波,然后再合成为相位图。这种分解频域滤波方法可以在滤波的同时,有效保留相位跳变信息。实验结果表明：与传统的图像降噪方法相比,该方法能够在保留图像“尖峰”信息的基础上,较好地滤除图像中的散斑噪声,方法简单有效,有效解决了传统滤波方法应用在相位条纹图中,相图灰度信息丢失10%~40%的问题。     

微型激光光学引擎中光学噪音的抑制

分析了微型激光投影引擎的两种主要光学噪音|干涉网格和激光散斑产生的基理|提出采用振动的随机位相板同时实现干涉网格和激光散斑抑制的方法并进行了理论分析|自行设计了微型激光投影引擎,利用微型电控马达带动随机位相板振动,用CCD相机在不同的F#及曝光时间下对投影图像进行了采集,并对图像的散斑对比度行了测试.结果表明,通过静止的随机位相板可以很好地抑制干涉网格噪音|通过使随机位相板以大于视觉暂留效应形成的融合频率进行振动,可以同时抑制干涉网格和激光散斑,使得散斑对比度小于5%,达到投影观察的要求.     

基于超分辨率重建的图像增强算法研究

 为了提高图像分辨率,从软件的角度出发,对低分辨率图像序列重建高分辨率图像的原理和算法实现开展了研究,提出从多个低分辨率图像序列中获取更高分辨率图像的方法。通过采用基于光流场的金字塔分层结构实现由粗到精的图像配准,获取了亚像素的运动估计。在采用多帧低分辨率图像进行亚像素级配准后,提出采用动态自适应确定正则化参数的方法,构造了简单的正则化代价函数,建立了低分辨率图像与高分辨率图像之间关系的重建模型,仿真实验结果表明该超分辨率重建算法水平和垂直方向上分辨率各增加一倍,与其他算法相比清晰度更高,计算时间不到传统POCS算法的一半。     

Reduction of speckle noise in digital holography by using digital image processing

A fundamental problem in optical and digital holography is the presence of speckle noise in the reconstruction process. Many approaches have been carried out in order to overcome such a problem ranging from modifying the spatial coherence of the illumination (optical techniques) to image processing techniques (digital techniques). This work shows the merged use of digital image processing techniques in order to reduce the speckle noise in digital reconstruction of optically recorded Fresnel's holograms. The proposed filtering techniques are illustrated with experimental results.     

部分相干X射线衍射相位成像与相位复原的模拟研究

借助于光学传递函数(OTF)概念，分析了微分相位成像过程和直接相位成像过程，讨论了通过图像处理进行相位复原的方法，得到了具有较高成像对比度和分辨率的实验参量区间，并利用数字方法模拟研究了相位成像的过程，对多种相位复原方法进行了比较。