Phase-shifting interferometry is a high-precision and commonly used phase retrieval method. In practical applications, phase shift errors are usually introduced due to factors such as environmental disturbances and phase shifter error. In this paper, we propose a deep learning method for estimating phase shift error from phase-shifting interferograms. This method manages to process the three interferograms with phase shift π/2, and uses neural network to extract phase shift errors from three interferograms. The analysis shows that the proposed method can effectively estimate the phase shift error under the noisy interferograms. This method can be used to correct phase-shift errors for phase retrieval (e.g., the least squares phase retrieval method) and calibrate phase shifters.
Wavelength-tuning phase-shifting interferometry is a modern optical measurement method based on the propagation characteristics of light, which can determine sample topography by wavelength-level non-contact method with high-precision. The accuracy of this method can be up to nanometer and sub-nanometer precision. The measurement of transparent parallel plates is important and significant in optical measurement. Based on the design of window and sampling functions, the introduced 36 steps phase-shifting algorithm can effectively extract the target information. The influences of constant and first-order phase-shifting errors are analyzed. The least square separation method based on 19 steps iteration can determine the surface phase information accurately and separate surfaces including front, rear, thickness information, even the parasitic term. The influence of different iteration steps for the PV and RMS values is considered, which provides the basis and reference for the design of the separation algorithm.
Fringe pattern denoising is an important process for fringe pattern analysis. In this paper, fringe pattern denoising using the convolutional neural network (CNN) is introduced. We use Gaussian functions to generate the various phase distributions, and then the required training samples are simulated according to theoretical formulas. The noisy fringe pattern can directly obtain the clean fringe pattern using the trained model. The denoising performance has been verified, which can recover high-quality fringe pattern.
Spatial particle distribution can be recorded by holography technology and can be constructed from multi-layer hologram. Due to the influence of holographic recording and reconstruction process, each tomography of multi-layer reconstruction from holography also contains noise in addition to containing spatial particle distribution information. How to denoise each tomography is a key problem. The existing methods either have a long operation time or the noise reduction effect is not obvious. In order to solve the above problems, we proposed a denoising method based on deep learning in this paper. A deep neural network is built to train and test with simulated spatial particle tomography on multi-layer holography reconstruction. According to the simulation results, the method proposed in this paper is effective in denoising the reconstruction results of spatial particles. The proposed method has the advantages of rapidity and high efficiency.
Deep learning is an extension of machine learning,deep learning uses multilayer neural networks to analyze data[1,2].Convolutional Neural Networks (CNN) is a commonly used neural network structure in deep learning. It is widely used in various fields, especially in the field of machine vision. In the field of optics,Monochromatic aberrations include spherical aberration, coma, astigmatism, defocus and so on, the common way to interpret interferograms is the Zernike polynomials, it generally used to describe the wavefront characteristics.In this paper, the convolutional neural network algorithm is used to identify astigmatism and defocus of the typical monochromatic interferograms, Zernike polynomial is multiplied by the aberration coefficient to represent the wavefront, the wavefront into the light intensity formula to obtain the aberration interferogram, the use of the above method to obtain the defocus interferogram, the result shows that the recognition accuracy is very high.The method of deep learning algorithm used to identify monochrome interferograms is simple and fast, and the training samples do not need manual calibration.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.