This PDF file contains the front matter associated with SPIE Proceedings Volume 11352 including the Title Page, Copyright information, Table of Contents, Introduction, and Conference Committee listing.

Reconfigurable freeform optical systems enable greatly enhanced imaging and focusing performance within nonsymmetric, compact, and ergonomic form factors. In this paper, several improvements are presented for the design, test, and data analysis with these systems. Specific improvements include definition of a modal G and C vector basis set based on Chebyshev polynomials for the design and analysis of non-circular optical systems. This framework is then incorporated into a parametric optimization process and tested with the Tomographic Ionized-carbon Mapping Experiment (TIME), a reconfigurable optical system. Beyond design, a reconfigurable deflectometry system enhances metrology to measure a fast, f/1.26 convex optic as well as an Alvarez lens. Further improvements in an infrared deflectometry system show accuracy around λ/10 of the notoriously difficult low-order power. Working together, the mathematical vector polynomial set, the programmatic optical design approach, and various deflectometry-based optical testing technologies enable more flexible and optimal utilization of freeform optical components and design configurations.

In this contribution we introduce an imaging based measurement setup, that is able to very accurately measure 3D relative positions between moving light sources. The system consists of two highspeed cameras, one equipped with a telecentric, the other with an endocentric lens. To improve accuracy of image based position detection, each lens is upgraded with a computer-generated-hologram (CGH) to replicate a single object point into a predefined pattern of spots on image plane. By averaging the centers of all replications, noise and other error contributions can be reduced. We will show how to apply image processing using two different approaches. The first approach is based on a tracking algorithm running on CPU reaching 330 fps. The second is a FPGA implementation to process whole images with a speed of 390 fps. Furthermore, we will demonstrate how threedimensional calibration can be done using the Nanomeasurement and Nanopositioning Machine NPMM-200. For the calibration, a three-dimensional multivariate polynomial is used. The standard deviations of residual error in object space for a calibration in a volume of 100 mm × 100 mm × 24 mm are σ_x = 0.367 μm, σ_y = 0.373 μm and σ_z = 0.437 μm (polynomial order = 9).

More than 95% of industrial inspection systems still rely on pure 2D information. Different measurement tasks like defect detection, micro-structure characterization, spectral characterization and dimensional metrology are typical application scenarios where standard 2D image processing based measurement systems are used. For dimensional measurements typically edges are used as the primary features for measuring lengths and positions. The accuracy of the corresponding (sub-pixel) edge position measurements is fundamentally limited by photon noise, discretization noise and electronic (camera) noise. For some applications, photon noise and electronic noise can be reduced by temporal averaging. We propose a very simple and cheap modification to improve the accuracy of such edge-based measurement systems. All relevant noise contributions are reduced by using a computer-generated hologram within or in front of the imaging system. The hologram replicates the original image and leads to multiple copies of the image on the image sensor. Therefore, spatial averaging (instead of temporal averaging) can be used to reduce all mentioned statistical measurement uncertainties (including the main limitation, namely discretization), thereby increasing precision. We present the measurement setup and methodology, limiting factors and first results that show the capability to reach accuracies in the range of thousands of a pixel.

Local alterations in UV absorption have been used to investigate subsurface damage in transparent optical materials. As a detection method, a collinear pump-probe arrangement has been utilized and absorption-induced deflections in the nonresonant probe beam has been detected. Depending on the changes in deflection and transmission signals, variations in absorption mappings could be attributed to different origins of material inhomogenities or foreign particles. Also, lightprovoked changes in absorption that can occur with non-linear optical materials have been detected by the proposed method.

Engineered functional surfaces often feature varying slopes on macro- and micro-scales. When surfaces are mirror-like, the highest surface slope that can be measured by a far-field 3D imaging optical surface measuring instrument isthe arcsine of the numerical aperture (NA) of the objective lens, i.e. the acceptance angle of the lens. However, progress in instrument design has allowed for measurement of non-specular surfaces with slopes steeper than this “traditional” NA limit. Nonetheless, there is currently a lack of understanding about the instrument response to surfaces with steep slopes beyond this limit. It is unclear over what surface spatial frequencies we can expect to accurately report fine surface-feature details. Here we present results demonstrating the capability of a commercial coherence scanning interferometer for measuring surface topography of a roughened flat and a blazed grating with tilt angles greater than the NA slope limit. We show that the surface form, i.e. the tilted plane, can be measured correctly. But, while surface texture information that can appear useful is also obtained, tilting significantly influences the measurement accuracy of micro-scale texture, and for asymmetric gratings, can depend on the tilting direction. A simplified surface scattering model suggests that the loss of scattered power captured by the instrument and a low signal-to-noise ratio causes the reduction of measurement accuracy. However, a rigorous three-dimensional instrument model is needed for a full understanding; we will develop this in our future work.

The quality of optical measurements is significantly affected by the reflection properties of the measured component. Therefore, it is important to consider the properties of the reflective surface to obtain accurate measurement results. A common method for the mathematical representation of reflections is the bidirectional reflection distribution function (BRDF). Typically BRDFs are measured via a gonioreflectometer. However, these are often only applicable on flat specimens or objects with previously known geometric properties. This paper presents an approach for the measurement of the BRDF on inhomogeneous freeform surfaces. For this purpose, a robot-assisted multisensor system is used consisting of a fringe projection sensor and an industrial camera, which is modified with six light sources that are evenly distributed around the optical axis and point at the measuring object. The reflection measurement consists of the sequential image acquisition of individual lighting configurations by successively switched on light sources. With the assumption of isotropic surface properties and known position of each individual light source, the relative BRDF can be determined pixel by pixel. This enables the BRDF measurement of freeform surfaces with varying reflection properties. Knowing the transformation between both sensor coordinate systems, the resulting BRDF data can be projected onto the points of the fringe projection measurement for geometrical representation. As an application example, a damage characterization of surfaces, based on the measured BRDF data is presented. For this purpose, a worn turbine blade of an aircraft engine is characterized so that burnt regions on the components’ surface can be detected.

Additive manufactured parts have complex geometries featuring high slope angles and occlusions that can be difficult or even impossible to measure; in this scenario, photogrammetry presents itself as an attractive, low-cost candidate technology to acquire digital form data. In this paper, we propose a pipeline to optimise, automate and accelerate the photogrammetric measurement process. The first step is to detect the optimum camera positions which maximise surface coverage and measurement quality, while minimising the total number of images required. This is achieved through a global optimisation approach using a genetic algorithm. In parallel to the view optimisation, a convolutional neural network (CNN) is trained on rendered images of the CAD data of the part to predict the pose of the object relative to the camera from a single image. Once trained, the CNN can be used to find the initial alignment between object and camera allowing full automation of the optimised measurement procedure. These techniques are verified on a sample part showing good coverage of the object and accurate pose estimation. The procedure presented in this work simplifies the measurement process and represents a step towards a fully automated measurement and inspection pipeline.

Hot forming processes, especially open die forging, are often used for production of high-performance, large-scale objects. The main benefits compared to, e.g. shape cutting methods, include lower material use and higher stress resistance. Inline process control by 3d geometry measurement is an important part of a cost-effective component production. However, there are no automated control systems commercially available for open die forging, which results in a limited precision of the final component geometry. The main challenges for a control system in said conditions are imposed by the temperature influence of the hot object on the measurement systems as well as limited actuator accuracy for the precise handling of hot, heavy objects. Additionally, the tools used in open die forging are kept simple for financial reasons. Comparable tools for, e.g., drop forging, need to be exclusively made for each new object form and therefore cannot be used for a cost-efficient production of low-quantity components. In this paper, we present a production concept in order to control a hot forming method for large scale, low quantity components. The approach combines an adaptable high-resolution 3d geometry measurement system and an incremental open die forging press for cost- and time-efficient production. Forming simulations will need to be conducted prior to the process to gain access to a large database of possible forming steps to reach the desired final geometry. The control system itself compares the measured geometry and temperature to the simulated ones. Occurring deviations are analysed and a sequence of forming steps is calculated from the database. The necessary forging forces and strokes of the actuating system are extracted from the chosen forming sequence and linked back into the system to achieve maximum precision.

When repairing worn components it is crucial to have detailed knowledge of the current object’s state. For this purpose a multi-sensor system was developed to measure objects in different scales and modalities. This work focuses on the 3-D measurement of worn turbine blades using a fringe projection system. The 3-D geometry of turbine blades is crucial for the overall performance and safety of an engine. Therefore it is not sufficient to rely on single fringe projection measurements for a functional evaluation. To obtain a 3-D model the blade has to be measured from multiple directions. Gathered data are combined to form the model. This process is called registration or stitching. To reduce uncertainties during the process markers can be applied on or near the measurement object. However, common methods using markers are insufficient in automatability and feature density and therefore are not applicable in this case. In this work a novel registration strategy based on projected random patterns is developed. Multiple projectors are placed around the object to illuminate its geometry. Keypoints are identified by capturing additional grayscale images and applying state-of-the art feature detection algorithms. Feature matching is performed on consecutive measurements. Matches are preprocessed and a random sample consensus approach is chosen to calculate the rigid body transformation. Multiple measurements of the turbine blade and other geometries have been successfully aligned using the proposed strategy. Beyond that the high density of features allows the alignment of measurements with different scales and resolutions.

The realization of 3D triangulation measurements in inhomogeneous media is challenging, as the sensor light path is not necessarily rectilinear anymore and the triangulation principle is violated. An exemplary measurement scenario under inhomogeneous optical conditions is the geometry characterization through an inspection window. The discrete refractive index variation from air to inspection window and back to air (or even water) can require complicated light path modeling approaches in order to triangulate 3D surface data correctly. As commonly used entocentric lenses “fan out” the projected light rays, the rays’ incidence angles onto the refractive index interface are not constant, and the rays are individually deflected. In consequence, the typically used camera pinhole model does not apply anymore, or can only approximate the actual light path under refraction. In this paper, we present a structured light sensor concept for measurements through inspection windows, which does not require a special adaption of the light path model. The key is the application of a telecentric stereo camera pair and high-quality optical inspection windows. It can be shown theoretically, that no additional parameters are necessary to model light refraction induced by a plane-parallel plate – such as a window –, when used in combination with a telecentric lens. Next to the affine stereo camera pair, the sensor comprises an entocentric projector unit as feature generator. In previous work, the projector was calibrated with refraction model in order to provide a 3D point cloud basis for the affine camera calibration.¹ In the new approach, the projector is merely used as feature generator to solve the correspondence problem between the affine stereo camera pair. Besides the developed sensor hardware concept, we present an overview on the calibration strategy based on an affine self-calibration approach, the solving of the correspondence problem between the cameras, and first calibration and measurement results. In forthcoming work, the sensor is also meant to be used in three-media refraction scenarios.

The design, manufacture and test of a custom-made phase mask for use in a high-speed fringe projection system is presented. The mask produces a controlled anisotropic point spread function (PSF) that blurs binary fringes parallel to the fringe direction, to produce high quality greyscale patterns at up to the maximum projection rate, 22,000 frames s-1, of the DMD (Digital Micromirror Device)-based projector. The paper describes the numerical design method based on a binary scatter plate; a polychromatic Fourier optics model to predict the device’s optical performance; a method to design the fringe patterns; the manufacturing process, based on photolithography and reactive ion etching; and experimental validation of the phase mask performance. Noise in the computed height-encoding phase maps is found to be just 22% higher than for 8-bit greyscale fringe patterns produced by traditional temporal integration through a sequence of bit planes, but the projection rate is increased by over two orders of magnitude.

A 3D measuring endoscope with a small measuring head and parallel arrangement of the fibers can be guided into forming plants and carry out precise measurements of geometries which are unreachable for most three-dimensional measuring systems. The data obtained can be used to quantify the wear of highly stressed structures and thus provide information for maintenance. Due to the compact sensor design and the required accuracy, optics with small working distance and a small measuring volume are used. In addition to in situ single measurements of highly stressed structures, over a hundred individual measurements are conceivable in order to convert large and complex geometries into point clouds. Besides the robust and accurate registration of all measurements, merging is one of the main causes of inaccurate measurement results. Conventional merging algorithms merge all points within a voxel into a single point. Due to the large overlap areas required for registration, points of diverse quality are averaged. In order to perform an improved adaptive merging, it is necessary to define metrics that robustly identify only the good points in the overlapping areas. On the one hand, the 2D camera sensor data can be used to estimate signal-based the quality of each point measured. Furthermore, the 3D features from the camera and projector calibration can evaluate the calibration of a triangulated point. Finally, the uniformity of the point cloud can also be used as a metric. Multiple measurements on features of a calibrated microcontour standard were used to determine which metrics provide the best possible merging.

In the past, grazing incidence interferometry has been applied to plane, cylindrical, acylindrical and general rod-like surfaces using diffractive beam splitters. Here, we demonstrate that also rough convex steep rotationally symmetric aspherics can be measured along one meridian in a single step using diffractive beam splitters and phase shifting techniques. The measurement of rough surfaces is possible i.e. without the need to polish the surfaces, due to the large effective wavelength (here about 10μm) of the test. The whole surface can be measured by rotating it stepwise around its symmetry axis and measuring successive small meridian regions. These meridian regions have to be stitched together to get the whole surface. Besides the presentation of the measurement principle, simulation results of the procedure for aspherical specimens are given. The corresponding experimental setup with discussion of the measured phase distributions is presented and the misalignment analysis is performed for the special case of spherical objects under test.

The results of experimental and theoretical studies aimed at creating of the device for high-precision measurements of the extended optical fiber length are presented. Now the length of fibers with a length about tens kilometer is measured with an accuracy of several meters. However, in the case of some interferometric measurements it is necessary to know lengths of long fibers with an accuracy of millimeters. The equipment by means of which it can be done is rather expensive. In this work the description of the simple device developed by us which allows measuring a fiber length with extreme accuracy to millimeters is submitted, the results of measurements and the assessment of their accuracy are shown. The principle of work consists of the registration of a phase shift when you change settings of parameters of the laser radiation extended through an optical fiber. The effectiveness of measuring technique was checked on segments of optical fiber of a predetermined length with accuracy about 1 mm.

Dynamic fault identification from fringe patterns is a challenging problem in optical metrology, and is required for applications such as non-invasive condition monitoring and fracture propagation . The paper addresses this problem by proposing a high speed technique for identifying temporally varying faults using graphics processing unit (GPU) accelerated Wigner-Ville distribution method. For this case, a huge stack of fringe patterns need to be processed and the parallel processing ability of GPU provides high computational efficiency and overall execution time improvement. For testing, we simulated a stack of 100 noisy fringe patterns containing time varying defects to mimic the temporal evolution of fracture in a test material. Each fringe pattern has size 4096 by 4096 pixels and signal to noise ratio of 5 dB, and thus the resulting image stack constitutes a large noisy data set. We demonstrate the performance of the proposed method for high speed detection of defects from the fringe patterns, and also show the comparative advantage of the GPU based parallel approach versus the conventional approach of sequential processing. For the given 16 megapixel image size, the sequential implementation using Python’s Numpy scientific library took about 38 minutes for processing a single fringe pattern whereas the same task could be completed within 4.5 minutes using the GPU based parallel implementation. Cumulatively, the reduction in computational cost for processing the complete fringe pattern data set would be substantial. Overall, our results show that the intensive and tedious task of dynamic fault detection can be efficiently processed using the proposed approach with high robustness against noise.

Electromagnetic radiation scattered from an engineering surface carries the information that is related to surface topography by surface measuring instruments such as coherence scanning interferometers, confocal and focus variation microscopes. Although the operating principles of these instruments appear quite disparate, their performance is fundamentally limited by the properties of the illumination and the optics used to measure the scattered field and can be remarkably similar in practice. In recent work we have attempted to characterize the performance of optical instruments using 3D linear systems theory. In this way the measured field is related to the surface form by the 3D point-spread function or equivalently the transfer characteristics expressed in the frequency domain. This paper illustrates and extends this concept by examining traditional contacting metrology and non-contacting optical metrology using the same linear systems framework. Linear systems theory is discussed with reference to the measurement of objects with varying surface gradient and discontinuities and in both cases, similar methods to measure and compensate the transfer characteristics using spherical calibration artefacts can be employed. Finally, we consider the non-linear step of estimating the surface form from raw measurements. We discuss inverse morphological filtering in the case of contacting measurements and inversion using a rigorous vector scattering model with the potential to improve measurements using optical profilometers.

Most 3D metrological microscopes used today require a scanning through the optical axis, which is time consuming. The common techniques are Coherence Scanning Interferometry (CSI), Imaging Confocal Microscopy (ICM), and Focus Variation (FV). If one technique is good for smooth surfaces, it is not for rough ones, while the good for rough is too noisy for smooth ones. Additionally, high local slopes are also dependent on the scattering properties of the surface, making the Numerical Aperture of the objective the most important property of the microscope. Imaging Confocal Microscopy is the best compromise in terms of surface application range (from smooth to rough), high local slopes on shiny surfaces, highest numerical aperture and highest possible magnification. Unfortunately, any kind of Confocal microscope today (laser scan, disc scan or microdisplay scan) requires an in-plane scanning to build up the confocal image in addition to the vertical scan, increasing the total measuring time in comparison to CSI and FV. This is against the needs of quality control in production environments, where scanning speed must be as short as possible. In this paper, we use a Microdisplay Scanning Microscope for obtaining the confocal image only relying on a single image per plane. We use a structured illumination to project a desired pattern onto the surface with a very well-defined frequency and direction. By means of the Hilbert transform, we digitally shift the projected pattern one or many times to recover the bright field and the optical sectioned images. This new method reduces significantly the measurement time, simplifies the overall cost of the system and eliminates the maintenance of scanning devices, while maintaining the optical sectioning properties of each plane. We also studied the performance of the resulting topography in terms of system noise, accuracy, repeatability and fidelity of the surface using different methods to obtain the confocal image. Finally, we also compared the results with true confocal results and with other techniques that require a single image per plane, such as Active illumination Focus Variation (AiFV).

Injection molding is the leading high-volume production method for aspheric lenses used in consumer electronics. The design tolerances of these lenses require careful metrology of pins, resultant lenses and the injection molding plates themselves. Pins determine the form of the lenses, while the alignment of the pins to the mold plate determines apex distance, decenter and tilt. One key parameter of the mold plate is the overall flatness, as it is a critical datum for the pins. We demonstrate flatness metrology on a platform capable of sub-micron lateral resolution that can be used for roughness measurements as well as high resolution measurements of sub-millimeter features such as apex centration. The platform is capable of measuring the full 200 mm by 200 mm mold plate in under 30 minutes. The flatness results are correlated with measurements from a laser Fizeau interferometer and demonstrate better than ±1 μm correlation on samples with 20 microns of departure and better than ±0.2 μm correlation on plates with sub-micron departure.

Metrological stages such as the nano-positioning and nano-measurement machine (NPMM) can position single-digit nanometer accurately on centimeter working volumes. However, their measurement system requires a feedback to the arbitrary shaped specimen by another probe. The differential confocal microscopy (DCM) offers the possibility to have a sensitivity down to that single-digit nanometers but suffers from noise and aberration. Recently the principle of the LockIn filtering could be successfully adapted in DCM and therefore achieved a high SNR. Contrary to the there employed acoustically driven tunable GRIN lens (TAG lens) at the objective, we demonstrate a microelectromechanical system (MEMS), an AFM cantilever, as an ultrafast oscillating pinhole in front of the detector. Its first resonance at 96kHz makes it very competitive regarding acquisition speed, but the low oscillation amplitude lowers contrast. By principle inheriting the possibility to compensate a change in reflectivity, we present another advancement for the evaluation of the resulting differential signal to make it robust against sample induced systematic depth errors, e.g. a tilt-angle. This could be advantageous for DCM with static beam-paths, as well. Potentially, the highest improvement can be achieved in conjunction with the NPMM’s highly accurate measurement interferometers, because the residual error for the depth of a specimen under the influence of varying aberration is kept below 20nm.

It is well known that neural networks including deep learning have been widely employed to solve the problems in recognition and classification. It was not until recently that people started to use them to solve imaging problems. In this talk, we focus on how to use deep learning to solve phase retrieval problems.

Model based approaches in dimensional metrology have great potential in terms of better accuracy. In some cases they may even help to overcome classical resolution criteria. A famous example is optical scatterometry for measuring critical dimensions on semiconductor chips in the tenth-nanometer range. Basically, these techniques rely on the solution of the inverse problem, i.e. retrieve the measured profile from a signal, e.g. a spectrum in optical scatterometry. Here, the appropriateness and accuracy of the model is of great importance to achieve the goals of quantitative metrology. On the other hand, the implemented models shall enable fast turnaround cycles for practical applications. Thus, long computation times and huge memory consumption can hardly be accepted. We investigate and compare different scenarios of model complexity and rigorousness related to application in two likewise well-established optical profiling techniques, laser focus scanning (LFS) and coherence scanning interferometry (CSI). Especially, two electromagnetic diffraction methods are considered, Finite Element Method (FEM) and Modal Methods (MM). In general, rigorous methods are rather expensive and time consuming compared to methods based on analytical approximations. Particularly, full 3D models require huge efforts and computing resources. Thus, some alternatives shall be discussed in this paper such as reduction of dimensionality and various other methods for acceleration such as symmetry usage and for MM. Moreover, the different approaches are compared and conclusions are drawn with respect to their practical applicability in both, CSI as well as LFS.

Light scattering methods are promising for in-process surface measurement. Many researchers have investigated light scattering methods for evaluating surface texture. Researchers working on scatterometry have developed methods to derive surface texture parameters by solving the inverse scattering problem. However, most of the research has been focused only on texture measurement or determination of critical dimensions where feature sizes are less than the wavelength of the light source. In this paper, we propose a new light scattering method to reconstruct the surface topography of grating patterns, using a cascaded machine learning model. The experimental scattering signal can be fed into the machine learning model as the input and the surface topography can be determined as the output. The training dataset, i.e. scattering signals of different surfaces, are generated through a validated rigorous surface scattering model based on a boundary element method (BEM). In this way, the machine learning model can be trained using a big data approach including tens of thousands of datasets, which represent most of the scenarios in real cases. The cascaded machine learning model is designed as a combined top-down, two-layer model implemented using neural networks. The first layer consists of a classification model designed to determine which type of structured surface is being measured, amongst a set of predefined design variants. The second layer contains a regression model, designed to determine the values of the design parameters defining the specific type of structured surface which has been identified, for example the nominal pitch and height of its periodic features. We have developed a prototype system and conducted experiments to verify the proposed method. Structured surfaces containing grating patterns were considered, and different types of gratings were analysed. The results were validated by comparison with measurements performed with atomic force microscopy.

This paper presents a deep-learning-based algorithm dedicated to the processing of speckle noise in phase measurement from holography. The deep learning architecture consists in a pre-trained residual convolution neural network, initially devoted to de-noising of natural images. In order to adapt the network to de-noise phase maps, a database is constituted by a set of noise-free and noisy-phase maps corresponding to realistic noise conditions (non-Gaussian, non-stationary, controlled speckle size). The algorithm is qualified according to quantitative metrics such as the phase error, the error method, the Qindex and the computation time. This paper demonstrates that the proposed de-noising algorithm yields stateof-the-art results in terms of phase error and error method. In addition, the processing efficiency in term of computation time appeared to be better. So, de-noising of phase maps using such deep-learning-based approach is expected to yield very promising results in optical metrology. Application of the method to the characterization of vibrations over surface about 400cm² is presented when dealing with vibration amplitude of 20nm at 17kHz recorded at 100kHz by a high speed in-line Fresnel configuration.

None-tactile metrology systems for inner radius measurements of cylindrical objects with large diameters are often based on the triangulation principle, using a laser source as illumination unit and a camera as detection unit. Different approaches have been presented in the past in order to generate a complete profile section of the measurement object’s inner radius. A standard light-section sensor cannot provide a 360° view of the radius without sensor rotation around the cylinder axis. The additional rotation axis needs to be calibrated and the captured point clouds registered in the same coordinate frame. To spare the necessity of a rotational axis, we developed a prototype sensor based on the hardware approach suggested by Yoshizawa et al.,¹ using a cylinder cone mirror and a laser illumination unit in order to generate a line circle projected onto the inner radius. In combination with a wide-angle camera, the laser line can be captured in one shot. Unlike the approach by Yoshizawa et al. , we present a model-based calibration routine for the triangulation sensor by mathematically describing the laser light path. The cone mirror expands the laser light into a disc (plane) or into a cone – depending on the incidence angle between laser and mirror. In our model, the light cone is parametrized by the right circular cone equation to reduce the number of unknowns in regression calculus. The necessary 3D support points to approximate the model parameter are gained by recording planar calibration pattern poses with and without laser line. The intersection calculation between the camera’s line-of-sight and the projected laser light geometry is derived, and the mathematical ambiguity in the line-cone intersection successfully solved. We present first experimental calibration and measurement data of a cylinder. By intentionally misaligning sensor and cylinder axes with arbitrarily chosen angles, the robustness of the suggested procedure is demonstrated.

Microsphere-assisted microscopy is a new two-dimensional super-resolution imaging technique, which allows the diffraction limit to be overcome by introducing a transparent microsphere in a classical optical microscope. This super-resolution technique makes it possible to reach a lateral resolution of up to one hundred nanometres (~λ/6) in air. Furthermore, microsphere-assisted microscopy distinguishes itself from others by being able to perform label-free and full-field acquisitions and requires only slight modifications of a classical white light microscope. This presentation gives an overview of the results of simulations and experiments we have performed over the last four years. The influence of the photonic jet on the image nature and the unconventional behaviour of the magnification factor is presented. In addition, the phenomenon behind the super resolution, which is still not fully understood, is discussed. Moreover, a new microsphere-based imaging device which enables the metrology of transparent nano-structures is shown. Finally, interferometry through microspheres is demonstrated for the 3D reconstruction of nano-elements. In order to provide a better understanding of the phase behaviour through microspheres, a numerical model of microsphere-assisted interference microscopy is being implemented.

We present a method to determine the two-dimensional spatial distribution of alignment as well as retardation of a nematic liquid crystal layer. Recently-developed novel alignment techniques for liquid crystals enable the definition of arbitrary spatial alignment patterns, which can then be used to shape optical beams for applications such as q-plates or diffractive optical elements. In order to quantify the alignment quality, a measurement method to determine the alignment of the major axis of the liquid crystals (between 0° and 180° ) with high spatial resolution over a large two-dimensional area is essential. Our approach is based on measuring the change in polarization of light that occurs on passing through a liquid crystal film. We then show a method to deduce the alignment direction and the retardation of the film from the change in polarization. To demonstrate the capabilities of this method, we measure specific alignment patterns that are difficult to quantify unambiguously with other methods.

PISTIL (Piston and Tilt) is a recent interferometric system that computes the absolute piston and tip/tilt map of a segmented wavefront. Its high precision makes it usable as a metrology tool for wavefront sensing of coherently-combined laser arrays for example. This interferometer needs to correctly address high dynamic piston sensing, while dealing with fringes wrapping that leads to ambiguous phase estimations. We derived a mathematical combination for two measurements at different wavelengths and did a technical demonstration of it, using a IRIS-AO PTT111 Deformable Mirror as a segmented wavefront generator. We have verified that the loss of accuracy is slightly increased for a larger piston compared to a previous study, and we got a standard error of λ/160 with a Peak-to-valley of λ/50. This technique could be extended to a broader spectrum.

Multiphoton microscopy (MPM) is an approach now well established in biomedical sciences, especially thanks to its excitation spectrum in the near infrared range (NIR). The simultaneous imaging of numerous of these substances imposes the use of a wideband excitation spectrum, indispensable in the case of in vivo and in live imaging or for detecting phenomena at video rates. A unique spectral bandwidth, covering the range between 750 and 1000 nm has been recently demonstrated and has made emerging a simplification in MPM: the excitation system is now no longer an lock for generating multiphoton images of numerous fluorophores. But such a solution might be highly sensitive to chromatic distortions and diffraction limit which might result in detrimental effects on image quality and especially on resolution performance. This question is at the core of the current presentation. A point-spread function (PSF) estimation is realized with a standard computational tool. Our experimental strategy has shown two interesting points. First, the resolution is preserved in the lateral plan (xy) regardless of the excitation procedure chosen. Second, a significant deterioration of the resolution is observed in the axial direction (z), with a factor 4 between the best resolution obtained with a standard imaging procedure and the worst one obtained with the wider spectral bandwidth. Starting with this result, the role of a computational solution of image reconstruction is highlighted for reducing the gap observed in axial resolution between standard and wideband excitation solution of MPM. The illustration of the interest of a large spectral bandwidth of excitation is then shown on a mouse muscle sample presenting 3 fluorophores having a spectral bandwidth of excitation spread along 300 nm. This set of experiments illustrates the impact of chromatic distortions and diffraction limit on the deterioration of resolution. As a conclusion, a basic protocol for image reconstruction is used in order to highlight the interesting level of improvement of the visual image quality generated by a standard computational image restoration.

With the persistent progress in nanotechnology the importance to accurately characterize nanoscale structures steadily increases. Optical measurement techniques have proven to be well suited for this purpose. We investigate different approaches to gain more information about nanoscale features by combining polarimetric setups with specially designed nanostructures. Our experimental setup combines a dual-rotating compensator Mueller matrix ellipsometer with an optical microscope. One arm of the ellipsometer is rigid and consists of a light source, polarization state generator, and optics to focus the light onto a sample. Samples under investigation are mounted on top of a combination of rotation and translation stages to precisely adjust them into the microscope’s focus. The other arm forms the microscope part with a long working distance objective, a polarization state analyser and a CCD camera. A large aperture rotation stage allows to rotate this arm around the sample stage. Thus, measurements in reflection and transmission under different angles of incidence can be performed. The setup measures Mueller matrices for each pixel in the obtained image. Therefore, it allows to examine the polarizing properties of the sample spatially and helps to gain further topological information. This information can be additionally enhanced by supporting nanostructures placed close to the sample to extract information from the near field. Therefore, we designed plasmonic lenses for different measurement configurations. The investigations are complemented by numerical finite element simulations. These are performed to validate the design of the nanostructures and to compare them with measured values. Up to now, we characterized the experimental setup and designed and validated the supporting nanostructures and reference structures. Future steps include extending the setup with a monochromator to ensure flexible illumination, measurement of the reference structures, and the combination of setup and plasmonic lenses to further enhance the sensitivity to subwavelength sized features.

Hyperspectral imaging opens a wide field of applications. It is a well established technique in agriculture, medicine, mineralogy and many other fields. Most commercial hyperspectral sensors are able to record spectral information along one spatial dimension in a single acquisition. For the second spatial dimension a scan is required. Beside those systems there is a novel technique allowing to sense a two dimensional scene and its spectral information within one shot. This increases the speed of hyperspectral imaging, which is interesting for metrology tasks under rough environmental conditions. In this article we present a detailed characterization of such a snapshot sensor for later use in a snapshot full field chromatic confocal system. The sensor (Ximea MQ022HG-IM-SM5X5-NIR) is based on the so called snapshot mosaic technique, which offers 25 bands mapped to one so called macro pixel. The different bands are realized by a spatially repeating pattern of Fabry-P´erot filters. Those filters are monolithically fabricated on the camera chip.

We provide an overview of optical coherence tomography (OCT) applied in the field of non-destructive testing (NDT). The applications include examples for structural polymer material testing at mesoscopic size scale, visualizing and validating internal sub-surface micro-defects, or characterizing the coating and bonding quality in multilayer samples. Furthermore, dynamical processes observed by OCT in their temporal progress can be demonstrated in the context of NDT as well as multimodal settings for gaining structural and specific insights on the materials tested. We will also regard the challenges for performing dimensional measurements and optical metrology by OCT imaging. In addition, novel developments and trends at illumination and detection sites will be discussed as essential components and requirements for the progress in OCT technology. This includes newly accessible spectral ranges under the view point of sensing as well as adaptable learning tools under the view point of data and image processing, which will round off the topic.

To realize continuous blood pressure (BP) measurement, we think the vibrational signal of pulsation waveform has good potential for BP model improvement. In this paper, we used a structured light projection method and a fringe analysis method to develop a non-contact measuring tool which used vibrational waveforms of arterial pulsation signals. Our method, based on a triangular configuration, used a digital light processing (DLP) projector and a camera with frame rate of 46 fps. The fringe pattern with pre-defined spatial carrier frequency was projected on the subject’s wrist. In our configuration design, instantaneous pulsation-induced skin vibrations of subtle amplitude can be observed and recorded within each frame of the fringe pattern. Using a two-dimensional Fourier transform, we chose a frequency region of interest (ROI) filter to collect the spectrum magnitude embedded with the deformation data to deliver the phase retrieval. The phase map was unwrapped using a non-iterative unweighted algorithm based on Fast Fourier Transform (FFT), we obtained the unwrapped phase map. After using a phase-to-height conversion, the results of the full-field dynamic vibrational field were analyzed. Several indicators such as heart rhythm (HR), heart rate variable (HRV), and root mean squared errors (RMSE) were adopted to compare the pulsation signal with the ECG and PPG signals. Our results demonstrated that the peak-to-peak arterial of a pulsation waveform amplitude was about 50-70μm which confirmed the suitability of structured light projection method for continuous pulsation signal monitoring.

The acquisition of surface information such as height, roughness and waviness is crucial in production accompanying metrology. This work aims to demonstrate an alternative approach to gather surface roughness information of profiles with millimeter lengths while having sub-nm resolution based on a low-coherence interferometer. The surface height information is encoded by spectral dispersion with a well-defined phase minimum of the interference data. By applying an imaging approach, the captured information on the surface profiles allows an assessment without any scanning along one lateral dimension. The axial resolution is dependent but not limited by the dispersive element. The combination of the determination of the phase minimum and fitting of spectral interference data allows for sub-nm resolution while the axial measurement range is several ten micrometers. This results in a significantly higher aspect ratio than comparable approaches. During this work, initial experiments were performed on a calibrated surface roughness standard from the German national metrology institute PTB. It could be shown that a roughness of Ra = (21.15 ± 0.8) nm and Rq = (26.58 ± 1.0) nm was measurable on a lateral measurement range of 1.5 mm. Due to the application of advanced analysis methods, such as auto-convolution function analysis it was proven that these values correspond well to the values measured for calibration on a tactile profilometer. Additionally, investigations on a polished glass substrate with an aluminum mirror coating are presented. With these measurements roughness differences of Ra = 0.1 nm could be determined within one measurement between different parts of the sample. All data acquisition was carried out in a one-shot fashion.

Optical scatterometry has matured to become a routine technique in semiconductor and submicron metrology. Due to its rather simple instrumentation, the spectral approach is the preferred solution in submicrometrology. Several years ago, Fourier (backplane) scatterometry was proposed taking advantage of a strongly focused light beam in order to simultaneously illuminate the sample from various directions. The tiny focal spot enables to address a much smaller area of interest as compared to state-of-the-art OCD that operates with metrology boxes of 20 to 50 microns side length. We propose a scanning Coherent Fourier Scatterometry approach where the wave front diffracted by a non-periodic sample is recorded by means of a Shack-Hartmann sensor. It can be observed that measuring the wave front distortion rather than the intensity distribution is more sensitive to slightest profile (or scan position) variations. Several measurements on periodic as well as non-periodic sub-resolution patterns are presented. It is demonstrated that our approach is highly sensitive even when scanning a red laser spot across non-periodic profile features as small as 100 nm and below. In similarity to standard OCD, the sample profile has to be reconstructed by solving the inverse problem. To this end, we developed a rigorous model based on modal diffraction methods such as RCWA and C-method in combination with a physical optics ray tracing model. Our simulations show good agreement to the measurements. We believe that our approach may have the potential to help abandon the usage of reference patterns and instead pave the way to direct in-die measurement. Particularly, the measurement of non-periodic features will become possible by scanning the spot across the sample profile. The application of Zernike coefficients is suggested to reduce the complexity of the approach with regard to the solution of the inverse problem. First experiments show that only a few (mostly low order) coefficients are very sensitive whereas other show only small change during scanning.

The accurate determination of the azimuth of a given direction, e.g., the true (geographic) North, is of fundamental importance in many fields. Just as few examples, it guides buildings construction in civil engineering, supports environmental and cartographic surveys, allows the correct positioning and stability control of concentrating solar power plants, as well as of airport installations, provides the geographic North reference for geomagnetic measurements, contributes to the interpretation of the orientation choices of ancient constructions in archeoastronomy, can be the primary benchmark to calibrate other compasses or gyroscopes. When aiming at reaching azimuth measurements with accuracies well below 1°, magnetic compasses are unreliable: firstly, they indicate the magnetic North rather than the geographic one; secondly, they are heavily influenced by possible surrounding ferromagnetic items.

Subject of the present paper is the application of a multichannel optical spectrometer performing spectral measurements in specified areas of the optical range using a set of narrow-band interference optical filters that are tuned to certain wavelengths. A special construction principle of the developed spectral-selective device allows to the transmission of the analyzed radiation through not only one optical fiber, but a bundle of fibers, which increases the sensitivity of the device without compromising its resolution. Tasks of monitoring and controlling the combustion processes and the prospects of using non-contact optical spectroscopy for their solution are considered. Results of the analysis of informative features in the radiation spectrum of various combustion processes are presented. Recommendations are given both on the choice of materials for infrared and visible filters, and on the parameters of the primary geometric structure that is used for the synthetization. A technique for the evaluation of the manufacturing quality or of the results of the filter structure design is proposed. A visible range filter based on a combination of zirconium oxide and silicon oxide layers has been developed. The results of the influence analysis of the layers number on obtaining the required spectral characteristics are presented.

The technique proposed in this paper provides a quality control components surface flatness by non-destructive and contactless way, with high resolution and increased sensitivity. The control is done in real time and instantaneously on all inspected surface. The accuracy of components geometry is the one of parameters which influences precision of the function. Moiré inteferometry is full-field optical technique in which the shape of object surfaces is measured by means optical interference generated by optical device. The technique has found various applications in diverse fields, from biomedical to industrial and scientific applications. In many industrial micro metrology applications, non-destructive and non-contact shape measurement is a desirable tool for, quality control and contour mapping. This method of optical scanning presented in this paper is used for precision measurement deformation in shape or absolute forms in comparison with a reference component form, of optical or mechanical components, on surfaces that are of the order of few mm2 and more. The principle of the method is to project the image of the source grating to palpate optically surface to be inspected, after reflection; the image of the source grating is printed by the object topography and is then projected onto the plane of reference grating for generate moiré fringe for defects detection. The optical device allows a significant defect magnification of up to 1000 times the inspected micro defect on micro surfaces, which allows easy processing and achieves an exceptional nanometric imprecision of measurements. According to the measurement principle, the sensitivity for displacement measurement using moiré technique depends on the frequency grating, for increase the detection resolution.

In this work we report the design of a conical corneal null-screen compact topographer, which uses a mobile device to capture null-screen reflection produced by the posterior corneal surface. The instrument features a head holder like those of virtual reality headset with the aim of align the topographer. For corneal topography the device is calibrated by testing a reference surface where the geometrical parameters such as the radius of curvature and the conic constant, are obtained. We present examples of surface topography measurements on some human corneas.

Interference microscopy is a non-destructive full-field imaging method, mainly used to measure the surface topography of different samples. In this work, two designs for improving the signal quality are described. The first consists of an original vertically orientated breadboard interferometer, in a Linnik configuration. The mechanical design of the arms allows the independent control and alignment of the coherence and the focal plane positions for optimizing fringe contrast. A low noise 16-bit camera is used to improve the sensitivity. The second interferometer is based on a Thorlabs tube system, with a Nikon Mirau Objective and a white LED, all controlled with IGOR Pro software or Labview, with the aim of being more compact, flexible and mobile. For both systems, an evaluation of the interferometric signal quality is performed, whereas the difference in lateral resolution by considering the 3D nature of the interferometric system, or only its 2D imaging abilities, is explored.

In this work we derive the system of equations that describe the passage of light through the conical corneal null-screen topographer from the Fermat principle; specifically, it computes the surface coordinates in which the reflection of the null-screen is produced, so that its normal field may be computed; since most of the surface information is obtained from it, this is an essential step on the evaluation process; the formulation here presented is an upgrade from previous works.

Amplitude-modulated continuous-wave laser scanner with focusing optics can realize extremely high-precision 3D measurement. Since amplitude-modulated continuous-wave scheme employs periodical modulation, the longitudinal resolution and the maximum unambiguous range are in a trade-off. Our system utilizes dual-frequency modulation compromise such trade-off. However, such an attractive laser scanner suffers from ranging ambiguity due to aliasing, which is the systematic error inherent in amplitude-modulated continuous-wave scheme. We have removed the ranging ambiguity by aliasing synthesis. Secondly, the acquired 3D point clouds contain phase jumping at the maximum unambiguous range. With leveraging the relationship between the intensity and spatial information, the phase jumping was unwrapped to recover the spatial continuity. Thirdly, the 3D point clouds in the defocused region of the amplitude-modulated continuous-wave laser scanner distort since the depth-of-focus of focusing optics is generally cm order. The 3D point clouds in the defocused area are contaminated by aliasing which can also be regarded as a ranging ambiguity problem. We have experimentally restored the 3D point clouds by aliasing synthesis with the assistance of intensity information. The ranging area can be elongated by at least ten times of the depth-of-focus with such data processing. With the above-all mentioned configuration and data processing, we have compromised the ranging ambiguity inherent in the amplitude-modulated continuous-wave laser scanner comprehensively. We expect that our results contribute to high-precision industrial inspection for Industry 4.0.

The experimental studies of 3D images such as rough surface profile have been described. These studies have been performed using acousto-optic tunable filters and optical components with strong chromatic aberration. The basic characteristic of the experiment mock-up is longitudinal resolving power (by z-distance) which has to be defined according to the certain criterion. The proposed criterion is connected with the admissible probability of missing of information unit which relates to the z-distance characterization of the device. The experimental circuit providing the 3D image studies by means of z-distance measurements is described, and the experimental results are listed and discussed. The most interesting result is that defocusing is distinguished with 90% probability at electric frequency deviation of 200 kHz. The perspectives of the further improvement of information transmission by this device are discussed. It has been found that this improvement can be attained by means of the noise level decreasing to the level taking place for electric frequency of 94 MHz. The device possible applications in medical diagnostics are discussed.

Predicting cropland latent heat flux (LHF) from commonly measured low-cost meteorological parameters (MPs) like net solar radiation, soil and air temperature, vapor pressure deficit, wind speed, and canopy temperature of the crops is essential for modeling crop production and managing water resources economically. In this treatise, we explore the deep reinforcement learning framework for short-term LHF trend estimation from the above MPs. The problem is reformulated as a classification problem, where each MP is acquired for a cost, and the objective is to optimize the trade-off between the predicted trend error and the relative MP acquisition cost. A sequential trend forecasting problem is evaluated via Q-learning with a linear guesstimate and a deep Q-learning scheme via neural network, where the distinct actions are the individual request for the MP values, and each episode is terminated by anticipating a trend. The proposed methodology is validated on the acquired farm-data, collected from the field experiments conducted on the cropland monitoring sites at Bidhan Chandra Krishi Viswavidyalaya State Agricultural University, Kalyani, West Bengal, India. The three non-rice crops, namely the yellow Sarson (mustard), potato, and green-gram, are studied owing to their similar energy balance partitioning patterns.

In this work we describe how to perform virtual experiments by deforming our ideal device with translations and rotations of each component, then we determine which are the minimal deformations that can be detected (sensibility) and how much does they affect the results of the measurement (sensitivity), a necessary endeavor since systematic errors due to misalignment of the components may lead to poor performance of optical systems, especially those used to measure optical components. The simulation of the passage of light is computed using a system of equations obtained from the vector reflection law.

Iterative calculation is a necessary step in the calibration of stereo deflectometry. Inaccurate input can result in the iterative process converging in a wrong position or unconvergence. Image distortion is an important factors affecting the accuracy of the input. In order to reduce the influence of image distortion and increase the robustness of the calibration, a method based on a search algorithm is investigated for stereo deflectometry. Because there is few distortion at an image center, a search window with a certain border length is positioned at the image center to obtain a group of data for the iterative process. The size of the window is determined based on an algorithm proposed in this paper. Due to the fact that the centers of distortion and image are not coincident, the window is consecutively relocated within the image. A function is proposed to evaluate the input accuracy. Along with the window moving, the calculated data which makes the proposed function reach the minimum is selected to compute the following iterative process. Experimental results affirm the presented method can significantly enhance the robustness of the calibration accuracy of a stereo deflectometry system. By applying the proposed method, the RMS (root mean square) of calibration error can be increased from 0.31 pixels to 0.05 pixels.

The economical optical sensor for measurements of drone coordinates in three-dimensional space is presented. Now drones use digital cameras to perform accurate taking off, landing or dumping of goods. Stabilized digital cameras with electronics for image compression and transmission via high speed wireless channel make drones expensive, reduce their weight payload and battery charge. The proposed optical sensor guarantees the coordinate measurements with ten centimeter accuracy in a volume with dimensions of several meters. The illumination part of this sensor is installed round a landing pad or a goods delivery pad. It forms a set of low-energy optical beams of definite shapes. Each beam transmits a digital code that characterizes its location relatively the pad. The receiving part of this sensor is a set of miniature photodetector units that are fixed under a drone. The proposed technique of the beam code comparison helps to calculate the drone coordinates relatively the pad. As a result, this sensor closes the gap between the accuracy of the Global Positioning System and the centimeter accuracy necessary for accurate drone taking off or landing without usage of a digital camera. The paper describes the sensor design and the experimental testing of this optical sensor. The advantages and possible applications of this sensor are also discussed.

A double-diffraction grating interferometer for precision displacement and rotation angle measurement is proposed. The grating interferometer uses electro-optical modulation technology to generate heterodyne light sources, then the "double-diffraction" optical path is established by the cooperation of the mirror and the grating, so that the diffracted light passes through the grating again to generate a diffraction, thereby introducing a doubled phase change to improve the system resolution of the measurement effectively. Then we can obtain the measurement result of the in-plane displacement and the in-plane rotation angle by the measurement result of the light detection. Under this optical path system, the interferometer system can continue to provide correct in-plane displacement and rotation angle measurement information as the grating moves out of plane. The experimental results show that the double-diffraction grating interferometer measurement technology can provide measurement information of double-degree-of-freedom displacement and rotation angle without changing the optical structure. The resolution of the displacement and rotation angle can reach 2 nm and 300 nrad respectively, the repeatability is better than 2 nm and 100 nrad respectively, and the measurement speed limit can reach 160 μm/s, which has excellent measurement performance.

Digital holography is one of the most powerful methods used in metrological applications for non-destructive testing of various components as it provides higher precision up to several nanometres at high speed. As there are many industrial applications such as gear metrology, surface tracing of planar components and so on, which involve dynamic objects, and holographic measurements on such objects is a challenging task. The interference pattern is no longer stable, resulting in low contrast and resolution of the recorded hologram thus degrading the recorded information. In this paper, lensless Fourier transform digital holography is used for analysing the interference contrast as a function of velocity for planar moving objects. Numerical simulations have been carried out to study how the size of reference source and the exposure time of camera affects the contrast of the interference pattern of a moving object. Experimentally, lensless Fourier transform holographic geometry is realised via Sagnac interferometer which provides robustness and immunity against the external vibrations during the recording. The maximum extent of velocity is estimated by analysing the variations in contrast such that there is minimal loss of information from the recorded hologram.

The regeneration of aircraft engines provides a key approach in order to reduce operating costs of these complex capital goods. The efficient and resource-saving maintenance of aircraft engines is particularly challenging, as they are made from a variety of different parts with locally adapted properties and different functionalities, which therefore also differ in the requirements in surface quality. The engine parts are manufactured and repaired on different machines, such as milling or turning machines. The subsequent quality check is realized with the help of highly specialized metrology systems. In case of so-called aircraft blisks, tactile coordinate measuring machines are used to measure the blisks surface data in regions with limited accessibility. The state-of-the-art regeneration procedure is therefore time-consuming and costly, as the parts need to be moved and mounted several times. In this paper, we present a blisk regeneration approach, which is meant to allow the inline quality monitoring of the regeneration process within a milling machine by means of a rigid endoscopic fringe projection system. The presented approach is meant to be time- and resource-saving as the fringe projection system is directly used in the blisk repair environment. To this end, we present the developed measurement system, outline the planned system integration into the milling machine and define challenges that arise due to the calibration of optical measurement systems within the production environment. The borescope with a chip-on-the-tip camera at the measuring head offers the possibility to drive into limited space and to perform high-precision 3-D measurements.

Optical retrieval of the structure of transparent objects at the nano-scale requires adapted methods capable of probing their interaction with suitable light. Scattering events, for instance, depend on the content and the arrangement of the medium encountered by the light along its propagation path. Physical models of scattering were used in the past to understand image formation and phase retrieval in transparent objects. Here, we considered one of them, which is based on the acquisition of defocused images obtained with partially coherent illumination, and explored its phase retrieval capability over a wide range of illumination angles. We used a basic transmission bright-field microscope with a bandpass filter and an adjustable illumination aperture. The bandpass filter was used to select a specific wavelength range to avoid light dispersion in the sample and maximize illumination capabilities. The adjustable illumination aperture was used to probe and assess the calibration over a wide range of illumination angles, which give access to different parts of the spatial frequency spectrum of the sample. We subsequently employed a computational algorithm to retrieve the local 3-dimensional phase-shift induced on the light field by the scattering through the sample. We imaged several types of samples to explore the calibration and the results in different experimental configurations. We employed: (1) commercial dielectric nanospheres to assess the phase calibration when measured along the optical axis, (2) custom-made nanosteps micropatterned in a glass substrate to assess the phase calibration when measured along the transversal axis. We first verified the model prediction in the spatial frequency domain for the scattering induced by our samples, and subsequently obtained a consistent and linear phase-calibration for illumination numerical apertures ranging from 0.1 to 0.5. This enabled us to calibrate a very simple optical system in a wide range of illumination angles to obtain quantitative 3D metrology of transparent samples at the nanoscale.

The application of precision interferometers is generally restricted to expensive and smooth high-quality surfaces. Here, we offer a route to ultimate miniaturization of interferometer by integrating beam splitter, reference mirror and light collector into a single optical element, an interference lens (iLens), which produces stable high-contrast fringes from in situ surface of paper, wood, plastic, rubber, unpolished metal, human skin, etc. The iLens splits the incoming beam to generate two beams, viz. reference beam and the object beam, at the front surface of the iLens. The iLens then collects back-scattered light from the sample and project towards the screen where it combines with the reference beam to produce high-contrast fringes. A self-referencing real-time precision of sub-20 picometer (∼λ/30000) is demonstrated with simple intensity detection under ambient conditions. The principle of iLens interferometry has been exploited to build a variety of compact devices, such as a paper-based optical pico-balance, with a weighing accuracy of sub-100 pg, having 1000 times higher sensitivity and speed when compared with a high-end seven-digit electronic balance. Furthermore, we used cloth, paper, polymer-films to readily construct broadband acoustic sensors possessing matched or higher sensitivity when compared with piezo and electromagnetic sensors. Our work opens path for a new class of ultra-affordable yet ultra-precise frugal photonic devices for diverse applications, such as material processing, high-resolution imaging, etc., in science and education.