Marcus Große

Coherent two-beam interference fringe projection for highspeed three-dimensional shape measurements

Two-beam interference is a fundamental and well-understood approach to create Fizeaus interference fringes. With a Mach-Zehnder interferometer, we utilize these two-beam interference Fizeau fringes for three-dimensional (3D) shape measurements. By introducing an acousto-optical deflector the phase of the interference fringes can be shifted with a rate of up to 200,000 Hz. When used in conjunction with highspeed cameras, this stereo-photogrammetric approach performs well for highspeed applications in comparison with the commonly used digital light processing projectors for stripe projection. Maximum speed and the achievable accuracy are discussed. Experiments and media substantiate the suitability, accuracy, and speed of this technique for very fast 3D shape measurements.

3D shape measurement of static and moving objects with adaptive spatiotemporal correlation

Common correlation-based photogrammetric 3D shape measurement techniques evaluate either temporal or spatial features. Temporal approaches achieve high accuracies but are limited to the measurement of static objects. Spatial techniques can deal with moving objects but provide relatively inaccurate results. Our goal is to combine these methods in order to measure dynamic scenes that contain static and moving objects. Therefore, we present a spatiotemporal correlation that adapts its temporal and spatial support locally to the motion of the measured objects. In addition, our technique compensates motion by warping the correlated image regions temporally. Our approach is based on structured illumination of random patterns, which are well suited for dynamic scenes due to high possible frame rates. The proposed technique is tested with simulated data and real measurements.

Outdoor three-dimensional shape measurements using laser-based structured illumination

Abstract. Three-dimensional shape measurements using structured illumination can achieve accurate shape representation of arbitrary object scenes. High-speed approaches have been recently developed and are already on the way into industry. Structured illumination for outdoor applications is rarely used as it is difficult to achieve a high-pattern contrast under bright light conditions. Hence, laser scanning or photogrammetry without active illumination is usually preferred; often, measurements are carried out at night. A shape measurement approach is suggested, one that is suitable to quickly and accurately acquire object shapes, even under difficult light circumstances. This approach is based on laser speckle projection and spectral filtering. The results made with this setup are presented, along with media that underlines its capabilities.

Statistical patterns: an approach for high-speed and high-accuracy shape measurements

Abstract. Statistical patterns have been used for structured illumination within a stereo-photogrammetry setup to precisely measure the shape of nearly arbitrary objects in a short time. This contribution gives an overview of recently developed projection setups based on such statistical patterns. Coherent and incoherent approaches as well as the applied reconstruction algorithm are explained. The results show the suitability of the statistical pattern projection approach to replace the commonly used slow digital light processing (DLP) projectors of three-dimensional shape sensors and facilitate measurements in an ultrashort time frame (microsecond range), e.g., to track moving objects.

Camera calibration using time-coded planar patterns

In this paper a novel pattern design for camera calibration using planar patterns with the Zhang calibration method is presented. In contrast to other work related to this calibration technique, which deals with the design and extraction of spatial image features, our focus lies on the precise and fully automated extraction of corresponding points by temporal image features using a sequence of planar patterns, which are displayed by a flat screen. The extraction of correspondences in our approach does not utilize areal properties of the images and therefore is hardly influenced by projective distortion, image distortion, or inhomogeneous illumination. Furthermore, not the whole pattern but only parts of it need to be visible in an individual view. In addition, the planarity as well as the physical precision of the control points of the pattern are ensured by the very nature of the flat screen. The overall calibration time may be reduced, as no human interaction is necessary. We compare the results gained from this novel temporal pattern design with those gained from commonly used checkerboard patterns. The proposed approach resulted in an increased precision for the calibration.

Coherent Pattern Projection for Highspeed 3D Shape Measurements

Since many years laser triangulation is a key technology for metrological measurements. Manifold sensors exist that measure single spot distances or distance profiles via triangulation between the illumination source (e.g. laser diode) and an imaging system incorporating typically a digital detector. The physical limits have been investigated and derived by several researchers [1]. It has been stated that the main limitation for laser triangulation is subjective laser speckle noise that occurs when the laser spot or line is imaged onto the detector.

Influence of the structured illumination frequency content on the correspondence assignment precision instereophotogrammetry

Stereophotogrammetric 3D shape measurement using structured illumination is an established class of methods for industrial inspection. One essential step in the measurement process for all stereophotogrammetric techniques is the assignment of corresponding points between the stereo views. As the purpose of the used structured illumination is to ease and improve the correspondence assignment, the choice of said sequence is of utmost importance. The precision of the correspondence assignment directly affects the noise of the final point cloud and therefore this assignment should be conducted with the highest precision possible. Depending on the chosen structured illumination sequence, different degrees of freedom for the pattern design exist and may affect the precision of the correspondence assignment and thus the noise of the 3d point cloud. In our contribution we want to discuss the influence of the frequency content of the structured illumination for a scheme employing bandlimited statistical patterns, which have been fruitfully used for highspeed applications in the past years. To evaluate the limits of the correspondence assignment accuracy we created a simple numerical signal-detector model. Using this model the correspondence assignment in dependence of the chosen structured illumination can be compared to ground truth-data. Furthermore, the noise of point clouds in real measurements is investigated to validate the results of the used simulation. Therefore, illumination sequences using different spatial frequency bands are created and projected onto a reference object. Afterwards, the noise of the resulting pointcloud is evaluated. The results indicate that it is advisable to optimize the pattern design depending on the used sensor and object properties.

Streifenprojektionsgenauigkeit mit hoher Messrate – 3D-Sensorik für die schnelle, dichte und genaue Formvermessung

Zusammenfassung Dieser Beitrag stellt ein Verfahren zur optischen 3D‐Vermessung von Oberflächenformen auf Basis der Stereophotogrammetrie und Projektion statistischer Muster vor. Ein Vorteil im Vergleich zu üblichen Streifenprojektionssystemen besteht in der hohen Rate, bei welcher Muster auf das Messobjekt projiziert werden können. Dadurch kann, wie gezeigt wird, eine sehr kurze Messzeit realisiert werden. Des Weiteren wird die algorithmische Auswertung der Messdaten vorgestellt, die neben der kurzen Messzeit auch eine kurze Rekonstruktionszeit ermöglicht. Experimentelle Vergleichsdaten mit einem etablierten Streifenlichtsensor sowie Anwendungsbeispiele belegen, dass die hohe Aufnahmegeschwindigkeit nicht zulasten der Messpräzision geht und eine äquivalente Messpräzision wie unter Verwendung der phasenschiebenden Streifenprojektion erreicht werden kann.

Stereophotogrammetric Image Field Holography

In conventional holography, it is difficult to gain access on the three-dimensional surfaces of objects under test [1]. In contrast, there are well established techniques like the stereophotogrammetry that reconstruct surfaces of objects in dense three-dimensional point-clouds by using at least two cameras and structured illumination [2]. Recent works demonstrated, that reconstructed digital holograms can be used as images in the stereophotogrammetrical determination of object surfaces resulting in a three-dimensional point-cloud of the object [3]. Additionally, it is possible to combine the point-cloud with holographic techniques, such as holographic interferometry or phase-shifting techniques [4]. Due to such a combination, full three-dimensional deformation vectors can be calculated, providing the user with a method of high precision deformation measurements [5]. To accomplish these calculations, four cameras are used to connect three point-clouds with each other, so that the deformations measured, depending on the sensitivity-vector of each camera, can be combined to a full three-dimensional vector.

Accurate outdoor 3D shape measurements with structured illumination using narrow spectral filtering

3D shape measurements using structured illumination are rarely used for outdoor measurements, although they offer a big potential for accurate and dense shape acquisitions. Usually such setups are used indoor, avoiding the influence of extraneous light sources, e.g. sun. The crucial difficulty is to preserve the image pattern contrast of the illumination while covering larger areas to achieve appropriate measurement volumes. We have developed and will present an outdoor 3D shape measurement setup using stereo-photogrammetry that is capable of acquiring dense and accurate shape representations by narrow spectral filtering. Media that substantiates the suitability of this approach will be shown.