Volume 5, Issue 5-6 - Optical 3D Sensing / Guest Editor: Gunther Notni

This paper reviews high-speed optical 3D sensing technologies for obtaining the 3D shape of a target using a camera. The focusing speed is from 100 to 1000 fps, exceeding normal camera frame rates, which are typically 30 fps. In particular, contactless, active, and real-time systems are introduced. Also, three example applications of this type of sensing technology are introduced, including surface reconstruction from time-sequential depth images, high-speed 3D user interaction, and high-speed digital archiving.

Deflectometry is a full-field gradient technique that lends itself very well to testing specular surfaces. It uses the geometry of specular reflection to determine the gradient of the surface under inspection. In consequence, a necessary precondition to apply deflectometry is the presence of at least partially specular reflection. Surfaces with larger roughness have increasingly diffuse reflection characteristics, making them inaccessible to usual deflectometry. However, many industrially relevant surfaces exist that change their reflection characteristic during production and processing. An example is metal sheets that are used as car body parts. Whereas the molded but otherwise raw metal sheets show a mostly diffuse reflection without sufficient specular reflection, the final car body panels have a high specular reflectance due to the lacquering. In consequence, it would be advantageous to apply the same inspection approach both for the raw material and for the final product. To solve this challenge, specular reflection from rough surfaces can be achieved using light with a larger wavelength, as the specular reflectivity of a surface depends on the ratio of the surface roughness and the wavelength of the light applied. Wavelengths in the thermal infrared range create enough specular reflection to apply deflectometry on many visually rough metal surfaces. This contribution presents the principles of thermal deflectometry, its special challenges, and illustrates its use with examples from the inspection of industrially produced surfaces.

Multi-scale optical sensor systems help to overcome the area of conflict between resolution, field size, and inspection time if it comes to the frequent problem of detecting small defects on large areas. The sensors of such systems are chosen according to two main properties: On the one hand, they should measure in opposed scales; on the other hand, their measurement principles should vary as well in order to be suitable for different material and surface properties. However, these systems can only operate at full capacity if it is possible to unify the acquired data from each sensor into one common coordinate system such that an overall analysis is possible, or subsequent sub-measurements can be triggered. In this paper, a general approach for a common sensor referencing is proposed, whose focus lies on microscopic optical sensors for both scattering and reflecting surfaces. The method is able to handle resolutions from the nanometer to millimeter scale in one single system, but is also feasible for a coordinate unification across several single sensor systems.

Structured light projection techniques are well-established optical methods for contactless and nondestructive three-dimensional (3D) measurements. Most systems operate in the visible wavelength range (VIS) due to commercially available projection and detection technology. For example, the 3D reconstruction can be done with a stereo-vision setup by finding corresponding pixels in both cameras followed by triangulation. Problems occur, if the properties of object materials disturb the measurements, which are based on the measurement of diffuse light reflections. For example, there are existing materials in the VIS range that are too transparent, translucent, high absorbent, or reflective and cannot be recorded properly. To overcome these challenges, we present an alternative thermal approach that operates in the infrared (IR) region of the electromagnetic spectrum. For this purpose, we used two cooled mid-wave (MWIR) cameras (3–5 μm) to detect emitted heat patterns, which were introduced by a CO 2 laser. We present a thermal 3D system based on a GOBO (GOes Before Optics) wheel projection unit and first 3D analyses for different system parameters and samples. We also show a second alternative approach based on an incoherent (heat) source, to overcome typical disadvantages of high-power laser-based systems, such as industrial health and safety considerations, as well as high investment costs. Thus, materials like glass or fiber-reinforced composites can be measured contactless and without the need of additional paintings.

A new optical form measurement system for almost rotational symmetric surfaces has been set up. It is based on an interferometric line sensor applying sinusoidal path length modulation in combination with a movement system. With this system, ring-shaped subapertures of the specimens are measured. The system is especially suitable for measuring spheres and aspheres with a broad range of radii (r>50 mm). The individual subapertures are stitched together to yield the full 3D topography. Because the rotation of the specimen by more than 360° has to yield the same results, inherent consistency tests are possible. Example measurements of a sphere are shown and discussed. Reproducibility measurements for one ring scan performed with the system show a standard deviation of 14 nm. The system can be set up at a moderate price as off-the-shelf mechanical and optoelectronic devices can be used. Future improvements of the system are discussed.

This paper evaluates the robustness of our recently proposed geometric constraint-based phase-unwrapping method to unwrap a low-signal-to-noise ratio (SNR) phase. Instead of capturing additional images for absolute phase unwrapping, the new phase-unwrapping algorithm uses geometric constraints of the digital fringe projection (DFP) system to create a virtual reference phase map to unwrap the phase pixel by pixel. Both simulation and experimental results demonstrate that this new phase-unwrapping method can even successfully unwrap low-SNR phase maps that bring difficulties for conventional multi-frequency phase-unwrapping methods.

Phase-measuring deflectometry (PMD) has become a standard tool to measure the topography of specular surfaces. We implemented PMD for the measurement of the human cornea topography, exploiting an earlier idea of Lingelbach et al. Two problems occur: a large angular dynamical range and a single-shot measurement are required. We solve these problems by an optimized geometry with minimal occlusion and by single sideband demodulation with a pre-distorted fringe pattern with optimal fringe period. An in vivo measurement of an astigmatic cornea displays a deviation from the medical diagnosis of only 0.15 D, which is within the medical quantization step of 0.25 D.

Fringe projection profilometry is an established technique for capturing three-dimensional (3-D)-geometry data with high-point densities in short time. By combining fringe projection with endoscopy techniques, it is possible to perform inline inspection of industrial manufacturing processes. A new fringe projection system is presented, which uses flexible image fiber bundles to achieve versatile positioning of a compact sensor head. When measuring specimens with highly varying reflectivity, such as technical surfaces on tool geometries, measurement errors increase especially due to the crosstalk between individual fibers in the bundle. A detailed analysis of the transmission properties of the utilized fiber bundles is presented. It is shown that aliasing is avoided due to the non-regular grid structure of a bundle. Different techniques are demonstrated to reduce the effect of crosstalk on the phase evaluation. Measurements of highly reflective technical surfaces with different geometrical properties are shown.

The use of optical coherence tomography (OCT) for noncontact three-dimensional aspheric lens profiling and retrieval of aspheric surface parameters is demonstrated. Two commercially available aspheric lenses with different focal length-to-diameter ratio have been imaged using OCT, and the measured optical path length distribution has been least square fitted with the aspheric lens surface retrieving the radius of curvature, aspheric constant, and conic constants. The refractive index of these lenses has also been measured referencing with a standard Zerodur glass flat. The fitted aspheric surface coefficients of the lenses are in close agreement with the manufacturer’s values, thus, envisaging the potential of OCT in rapid screening, testing of aspheric lenses, and other micro-optical components such as those used in illumination optics.

A review of the use of aspherics in the last decades, understood in a broad sense as encompassing single-vision lenses with conicoid surfaces and free-form and progressive addition lenses (PALs) as well, is provided. The appearance of conicoid surfaces to correct aphakia and later to provide thinner and more aesthetically appealing plus lenses and the introduction of PALs and free-form surfaces have shaped the advances in spectacle lenses in the last three decades. This document basically considers the main target optical aberrations, the idiosyncrasy of single lenses for correction of refractive errors and the restrictions and particularities of PAL design and their links to science vision and perception.

Using nonstandard (NS) finite difference time domain (FDTD) scheme to perform 2D electromagnetic (EM) simulations, we investigate how the optical properties of 2D photonic crystals (PCs) are affected by various different kinds of structural noises in the PC lattice. While the transmission spectrum is strongly affected by noises, the position and the depth of the band gap in the transmission spectrum are remarkably robust. It is shown that rather coarse numerical grids can be used to evaluate various PC structures in NS-FDTD EM simulations. The combination of noises affects transmission spectrum in the same way as the most influential individual noise. It is shown that reducing the most influential individual noise is a very efficient method to make PC more accurate.