Method for the optical characterization of a transparent or semi-transparent object and optical coherence tomography system for carrying out the method

Abstract

Method for the optical characterization of a transparent or semi-transparent object (1), in which object dimensions (W, ID, AD), object distances and / or object positions are determined based on optical coherence tomography, wherein a swept-source laser source with a coherence length in the decimeter to meter range is used as the light source (2), and a reference substrate (13) is used to determine absolute distances, dimensions and / or positions in the beam path (12) of the light source (2), which reference substrate comprises a partially reflective optic that is stationary in the beam path (12), characterized in that the object (1) is moving and a scanning path used in scanning the object (1) at least partially follows the movement of the object (1), and in that a spatial jitter is superimposed on the scanning path.

Description

[0001] The underlying invention relates in particular to a method for the optical characterization of transparent or semi-transparent objects and an optical coherence tomography system.

[0002] In the prior art, for example, it is known to examine transparent or translucent objects using optical coherence spectroscopy (OCT). For example, US Pat. No. 10,890,431 B2 discloses a VCSEL (vertical-cavity surface-emitting laser) OCT system for the 3D measurement of transparent objects. In this known method, 3D depth profiles of an object are captured in scans using an optical coherence tomography system. Devices and methods for distance measurement based on optical methods are also known from WO 2016 / 069744 A1, US 2012 / 0 274 943 A1, US 2021 / 0 080 246 A1, and EP2078222 B1.

[0003] There is still room for improvement in known systems and methods for optical coherence spectroscopy and optical characterization of transparent or translucent objects, particularly with regard to the type, size, structure of the objects to be measured and the boundary conditions of the measurement or geometric characterization of the objects.

[0004] Based on this, an improved or new method for optical coherence spectroscopy, in particular for the optical characterization of a transparent, semitransparent, or translucent object, is to be provided. Furthermore, a corresponding optical coherence tomography system is to be provided.

[0005] This problem is solved in particular by the combinations of features of the independent patent claims. Advantageous embodiments emerge in particular from the dependent claims and the following description.

[0006] According to embodiments, a method for the optical characterization of a transparent or semitransparent object is provided. Optically transparent should generally be understood in the sense of light-permeable, and can therefore also include, for example, semitransparency or translucency. In the strict sense, the term "optically transparent" within the scope of the invention can mean transparency with respect to the infrared and / or visible spectrum. In particular, infrared-transparent objects are intended to be included. Preferably, objects made of inorganic materials or non-biological materials are used in the method.

[0007] The proposed optical characterization method determines object dimensions, distances, and / or positions based on optical coherence tomography (OCT). As already explained above, OCT is a well-known method for characterizing objects based on coherent optical radiation.

[0008] Optical coherence spectroscopy is well known. For examples of optical characterization of objects using OCT, see WANG, Zhao [et al.]: Cubic meter volume optical coherence tomography. In: Optica, Vol. 3, 2016, No. 12, pp. 1496-1503. - ISSN 2334-2536 (E). DOI: https: / / doi.org / 10.1364 / OPTICA.3.001496 , and MOREL, Eneas N. [et al.]: Application of a long-range swept source optical coherence tomography-based scheme for dimensional characterization of multilayer transparent objects. In: Optical Engineering, Vol. 56, 2017, No. 8, Article number: 084102 (pp. 1-7). - ISSN 0091-3286 (P); 1560-2303 (E). DOI: https: / / doi.org / 10.1117 / 1.OE.56.8.084102.

[0009] The method proposed herein is characterized by using a swept-source laser source as the light source, in particular a tunable laser source, preferably a wavelength-tunable laser source, which has a coherence length in the decimeter to meter range. Furthermore, the proposed method is characterized by using a reference substrate in the beam path of the light source, which includes a partially reflective optic that is stationary in the beam path. The reference substrate is used to determine absolute distances and / or positions and / or dimensions of the object or to / from the object.

[0010] The underlying invention is based on the finding that the use of a swept-source light source with a reference substrate is suitable for characterizing at least partially transparent or translucent moving objects and provides comparatively reliable and accurate measurement results. At least partially transparent or translucent objects that exhibit movement during the measurement can be measured.

[0011] Furthermore, objects can be measured in particular which emit non-negligible thermal radiation and / or are arranged in an environment which emits non-negligible thermal radiation.

[0012] As mentioned, the proposed combination is suitable for measuring and / or characterizing hot objects, e.g., partially molten glass or objects made of glass. The method is particularly suitable for the optical characterization of glass cylinders or cylindrical glass objects during or during production, whereby the objects may be in an at least partially molten or not yet fully solidified state. For example, the proposed method makes it possible to determine the outer and / or inner diameters and / or wall thicknesses of objects.

[0013] In particular, the proposed method also provides advantageous axial resolution because, unlike known methods, the proposed method operates without Bragg gratings. The advantageous axial resolution results, for example, from the fact that the method proposed here, unlike methods based on Bragg gratings, is essentially unrestricted or limited with regard to the number of reference points or measurement points.

[0014] Furthermore, the proposed method enables the measurement of both absolute and relative distances, thus enabling improved optical characterization.

[0015] The proposed method also enables the characterization of transparent or semi-transparent objects essentially in real time, for example, during the production or manufacturing process. The method also enables the characterization of moving objects.

[0016] According to embodiments, a swept-source laser source is used as the light source, which comprises a MEMS VCSEL-based light source (MEMS: micro-electro-mechanical system; VCSEL: vertical-cavity surface-emitting laser) or an akinetic light source. Such a light source particularly preferably has a coherence length in the range of 0.2 m to 100 m or more, or the coherence length of the emitted radiation can have a corresponding coherence length.

[0017] The proposed method, using a MEMS VCSEL light source and a reference substrate, is advantageously suited for comparatively large coherence length ranges and enables particularly accurate coherence / interference measurements across the entire range. Large coherence lengths can be used, for example, for larger distances between the measuring head or measuring apparatus and the object, for example, when the object is hot, such as (partially) molten glass, and a certain distance from the object must be maintained due to temperature.

[0018] According to embodiments, it can be provided that during signal acquisition and / or signal evaluation of the optical coherence tomography (OCT) signals: a. k-clock resampling is used, b. algorithmically determine minima, maxima or zero crossings of the k-clock of the laser source and use them as sampling clock, or c. a k-clock hardware trigger is used.

[0019] A k-clock enables data to be recorded linearly over time, which can then be transformed into a linear sample in k-space. High-frequency signals, in particular, can be analyzed more easily. Furthermore, it is possible to characterize objects in quasi-real time, which is advantageous, for example, for the geometric characterization of moving transparent or semi-transparent objects during the manufacturing process.

[0020] An exemplary procedure according to b. for the algorithmic determination of minima, maxima or zero crossings of the k-clock can be as follows or include the following steps: b1) Normalization of the k-clock by division by the envelope; b2) performing at least three, preferably immediately consecutive, mathematical operations, such as absolute value formation, addition, absolute value formation; b3) smoothing, e.g. by a median filter or similar; b4) peak finding or determination of maxima; b5) Interpolation of the maxima, minima and zero crossings to the linear k-space.

[0021] Due to the absolute value operation b2), minima, zero crossings and maxima of the k-clock can be determined using the proposed exemplary method steps.

[0022] According to embodiments, the optical coherence spectroscopy method can involve scanning the object with laser light from the light source using a galvanometer scanner, or galvanoscanner for short, using a microlens array and / or a bifocal lens, preferably with a high numerical aperture. Galvanoscanners, microlens arrays, and bifocal lenses, in particular, have proven suitable for the optical characterization of moving objects, for example, during the manufacturing process. Such components are particularly suitable for measuring or characterizing partially or partly melted, moving transparent cylinders, e.g., made of glass.

[0023] With regard to the mode of operation of the galvanometer scanner, it is possible, for example, with moving objects, e.g. rotating and axially moving objects, in particular with comparatively fast rotating objects and comparatively slow axially moving objects, that the galvanometer scanner is or is programmed in such a way that a point on the object or on the surface of the object, in particular a point or area targeted by the galvanometer scanner, is at rest in the reference system of the galvanometer scanner. The object can, for example, be a cylinder, an object with cylindrical geometry, or a cylindrical object. This is possible in particular within the scope of the deflection that can be achieved or is available with the galvanometer scanner. If the galvanometer scanner reaches the end of the deflection orof its deflection range, it is reset in terms of deflection and operated in such a way that it tracks or targets the next point on the surface.

[0024] Synchronization between the galvanometer scanner and data acquisition on the one hand and a rotation of the object on the other hand can be achieved, for example, by an angle encoder. In general, movements of the object, be it translation and / or rotation, can be recorded by a sensor unit or sensor technology, and the recorded movement data can be used for synchronization. Based on such data, for example, a movement path of a point, in particular a targeted point or area, on the object, for example on the surface of the object, can be determined, and the galvanometer scanner and data acquisition on the one hand and the movement of the object on the other hand can be synchronized on the basis of the determined movement path. Preferably, the movement ora movement path is continuously determined, in particular in real time, and used for control and synchronization between movement on the one hand and galvanometer scanner and data acquisition on the other.

[0025] With regard to the mode of operation of the microlens array, it can be provided that the light beam underlying the OCT is aligned along the direction of movement or along a movement path or trajectory of a point on the object, in particular the object surface, and that the scanning takes place simultaneously at or for points or surface points of the trajectory. For example, if the movement or trajectory is known, the data acquisition and the movement can be synchronized, i.e., points on the object or on the object surface can be tracked for data acquisition.

[0026] When using a two-dimensional (2D) microlens array, it is particularly possible to scan the (immediate) vicinity of the trajectory, in particular the locations of interest on the object or an area surrounding the trajectory. This scanning method is particularly suitable for cylindrical or cylindrical objects.

[0027] The scanning by the microlens array is preferably stationary and further preferably at the repetition rate of the laser.

[0028] With regard to a bifocal lens or a sensor head with a bifocal lens, it should be mentioned that this can be used, for example, to optimize the signal strength reflected back from a sample or object. This is particularly true when the sample or object has, for example, (a) periodic fluctuations in the inner diameter (e.g. 50 mm over an axial range of 500 mm), e.g. when there is an axial inclination of the inner surfaces depending on the position of the object, and / or (b) a significant stepwise change in the outer diameter across the axis or when measuring objects with a wide range of outer diameters (e.g. from 200 mm to 400 mm). For condition (a), it is advantageous to use a sensor head with a high numerical aperture, since the light reflected back from an inclined surface propagates at a considerable angle.In addition to the large aperture, it may be useful to separate the excitation of the sample from the detection of the back-reflected light to still allow collimated excitation. For condition (b), a sensor head with a bifocal lens is preferred, which can map the focus of the front and back surfaces to the position of the receptive fiber end, regardless of the diameter / size of the object. For these reasons, a dual-focus design is particularly advantageous for these conditions.

[0029] A bifocal lens or a double-focus asphere can, for example, consist of two semi-convex lenses with focal points fa and fb that are connected to each other, with the connecting surface running parallel to the axis of the object, e.g., a cylindrical object. In another embodiment, a double-focus asphere can be designed similarly to a zone plate with circular sections with alternating focal lengths fa, fb, fa, fb, etc., whereby this embodiment can be used essentially independently of the properties of the object to be examined.

[0030] Another variation of a dual-focal optical setup involves using a single asphere, replicating identical receptive optical elements, such as fibers, (downstream of the asphere in the beam path) in an orthogonal arm, and deflecting the incoming beam into both arms via a beam splitter. The focus of one receiving arm is preferably tuned to the front of the object, while the other is tuned to the back. The signal received by the elements or fibers is captured either by optical multiplexing or with the aid of an optical combiner.

[0031] If the dimensions of the object to be measured are (roughly) known in advance or essentially known, the lens distances of the dual-focus setup can be automatically adjusted to an optimal dual-focus configuration.

[0032] In some embodiments, the probe arm or fiber and the receptive arm or receptive fiber can be formed by different elements or fibers. For example, a collimated beam emerging from the probe fiber can be collinearly coupled into the receptive fiber with a high numerical aperture via a microprism or a glass plate with polished end faces, which has, for example, a bonded, angled, partially reflective surface. This can prevent the light reflected back from the sample from being blocked.

[0033] In cases where an inclined surface of the object to be examined refracts the light beam at a small angle, so that on the opposite surface of the object an area is "excited" which is not exactly opposite with regard to the central axis of the object. This can lead to errors in the determination of the distance, the dimension and / or position. In embodiments it is particularly possible to compensate for the deviation by placing an array camera or a plenoptic camera behind the sample, with which the direction or angle of the transmitted light can be recorded or determined. With the direction or angle the distance, the dimension and / or position can be corrected, for example by ray tracing. According to embodiments a single focus setup can also be used. In this case it can happen that not all four reflections for the surfaces of the object are available for all axial positions of the, for examplecylindrical object can be detected. In such cases, the position of the receptive fiber and / or the lens spacing of the optical setup can be optimized so that at least three out of four reflections of the cylindrical object can be detected, allowing continuous determination of, for example, inner diameter, outer diameter, and wall thickness, e.g., after an initial 180° rotation.

[0034] According to the claim, the object is a moving object. When scanning the moving object, a scanning path used follows at least partially the movement of the object. Advantageously, in this embodiment, the data acquisition of the optical coherence tomography is synchronized with the scanning path.

[0035] In other words, the scanning path can move, at least partially or in part, substantially synchronously with the movement of the object. In particular, it is a finding of the underlying invention that scanning moving objects based on the light source used, coherence length, and reference substrate enables characterization, in particular measurement, of objects or parts thereof, even when the object is moving. As already mentioned above, a galvanometer scanner and / or a microlens array can be used, in particular, for scanning moving objects, wherein synchronization between the galvanometer scanner and / or microlens array and the data acquisition is or will be synchronized with the movement.

[0036] According to the claim, a spatial jitter is superimposed on the scanning path. According to embodiments, a spatial jitter is superimposed on the scanning path in the galvanometer scanner. In particular, such jitter is not only advantageous for irregularly shaped objects, but also supports data acquisition for moving objects. The jitter can be generated, for example, by suitable control of the galvanometer scanner. If the object surface is ideally perpendicular to the light beam of the OCT, a jitter is not absolutely necessary. Since real objects, in particular objects that go through a manufacturing process, fluctuate with regard to the orientation of the surface(s), the accuracy and / or reliability of the determined object data can be improved by using a jitter. In other words, a jitter makes it possible, for example, to detect a back reflection or a reflection at least to a certain extent even with an irregular object surface.Reflected rays from the object can be obtained. Jitter can be adjusted, for example, by superimposing a deflection or movement orthogonal to the trajectory of the point on the object, for example, over the movement of the galvanometer scanner according to the trajectory of the point.

[0037] According to embodiments, the method combines several, in particular a multiplicity of, reflection profiles (or A-scans) obtained along the scanning path, in particular to form a combined reflection profile. Periodogram signals can be determined from the combined reflection profile thus obtained, and at least one object dimension, an object distance, and / or an object position can then be determined from at least one periodogram signal, i.e., from one or more periodogram signals.

[0038] In particular, distances between boundary surfaces of the object and / or their position or location relative to each other relative to a precisely known reference, formed, for example, by the stationary partially reflective optics, can be determined.

[0039] It is therefore possible to determine object dimensions, distances and positions comparatively accurately and reliably, especially when the object is moving and / or when the object structure or geometry varies within certain limits.

[0040] According to embodiments, for example, those A-scans can be filtered out which have a number of back reflections expected from the geometry of the object to be examined, e.g., flat, cylindrical, etc. The expected number of back reflections in OCT can, for example, be determined or derived from the (rough) geometry, target geometry, and / or expected geometry of the object to be examined. The rough, target, and / or expected geometry can be provided to the method, e.g., a corresponding evaluation unit, as a parameter or parameter data set, e.g., as an input parameter or data set, in particular an input parameter data set.

[0041] With regard to the selection of the A-scans to be used or usable for the optical characterization, it is also or additionally possible to use those A-scans, e.g. to filter out those which, for example from a group or set of recorded A-scans, show or have a maximum intensity and / or which have the best S / N (signal-to-noise ratio).

[0042] A periodogram is understood in the usual sense, in particular as a distance signature of the reflections caused by the boundary surfaces of the object, in short distance signature.

[0043] According to embodiments, it can be provided that an infrared filter, preferably a dual-band mirror, is arranged in the beam path of the light source, and thermal radiation emanating from the object and / or the object's surroundings in the range of the far infrared spectrum is at least predominantly filtered out using the infrared filter. The infrared filter is preferably arranged upstream of a measuring head used to detect the light reflected from the object. In particular, it is possible to prevent measured values from being disturbed by radiation or extraneous radiation in the infrared spectral range of the thermal radiation, or the S / N ratio for the light reflected from or at the object can be improved. The use of an infrared filter is advantageous, for example, in the characterization of hot objects orfrom objects that themselves and / or their surroundings emit radiation in the infrared spectral range of thermal radiation, at least to an extent that would or does impair the measured values of the OCT light reflected from the object. With regard to infrared radiation, a spectral range that corresponds to a temperature range of 800 K to 2200 K, assuming a blackbody radiator, is particularly relevant.

[0044] According to embodiments, the optical coherence tomography may include a signal acquisition of the k-clock of the light source with software-based filtering and offset correction, and further, optionally, a subsequent phase extraction, in particular based on a Hilbert transformation, and / or an evaluation of zero crossings and extrema. The filtering may preferably be a low-pass filter.

[0045] Based on the k-clock signal acquisition, a linearization over time is possible, which enables, in particular, real-time characterization of objects. Through subsequent phase interpolation, i.e., a phase interpolation of the phase response of the measurement signals, the data or measurement signals recorded linearly over time due to the k-clock used can be transformed into a linear sample in k-space. This can increase the resolution of the method or of a system executing the method.

[0046] According to embodiments, the phase response can be extracted and the data reinterpolated based on this phase response can be performed for each individual sweep of the swept-source laser. This can, for example, ensure that the system, in particular the signal evaluation and / or processing, is robust with respect to fluctuations in the laser's phase response.

[0047] According to embodiments, filtering, in particular low-pass filtering and / or phase interpolation, can occur after each individual laser sweep. For a 10 kHz laser, for example, at a frequency of 10 kHz. Implementation on an architecture with parallel data processing, for example on graphics processors (GPUs), is particularly advantageous for processing and / or handling the measurement signals at such frequencies. For example, in application cases, the sampling rate of the digitizer, a laser repetition rate of 10 kHz to 100 kHz, and depending on the laser duty cycle, can result in comparatively high data rates in the range of 5 to 7 GB / s (gigabytes per second), which cannot be handled satisfactorily or at all with a conventional CPU. Therefore, the processing and / or handling of the measurement signals is advantageously carried out on a GPU, which can be implemented, for example, in CUDA (Compute Unified Device Architecture).

[0048] Examples of filters are bandpass filters, median filters, etc. Examples of interpolations are polynomial interpolation, Chebychev interpolation, etc.

[0049] According to embodiments, it is provided that periodogram signals are determined from signals of the optical coherence tomography by means of an automatic peak-picking algorithm from periodogram data, and that at least one object dimension, at least one object distance and / or at least one object position are determined from the periodogram signals, in particular taking into account periodogram signals of the reference substrate.

[0050] The term "peak picking" (or peak finding) should be understood in the context of the state of the art as, for example, an algorithm with which local maxima or minima can be determined from signals, in particular measurement signals and / or processed or processed measurement signals. Such an algorithm can, for example, include determining zero crossings of the numerical first derivative of the signal(s), with zero crossings being assigned to individual peaks, in particular local maxima or minima.

[0051] In the peak-picking algorithm, data on the (rough) geometry and / or target geometry and / or expected geometry of the object can be used as input parameters or boundary conditions. In particular, when manufacturing objects, for example glass objects such as glass cylinders, the geometry of the object or a desired geometry of the object is known or specified. If, for example, the target diameter of a cylindrical object to be manufactured, for example the inner and / or outer diameter and / or the wall thickness of the object is known or specified, this data can be used to determine, at least approximately, the location(s) or position(s) of the expected signals, which can be calculated or determined, for example, taking the measuring arrangement into account. If, for example, a cylindrical object is measured with light incident radially, reflection signals, i.e.Peaks are expected for reflections on the outer and inner surfaces of the cylindrical object walls.

[0052] Based on the measurement signals and specified data on the geometry of the object and / or the measurement setup, it is possible to monitor a manufacturing process, i.e., to characterize the object during production, in particular with regard to whether the manufactured object meets the underlying requirements with regard to geometry and / or dimensions. Thus, quality control and / or monitoring of a manufacturing process is possible. Furthermore, for example, a correlation of deviations in the object geometry with operating parameters of the production plant and / or, based on deviations in the object geometry, a fault analysis for the production plant can also be provided.

[0053] The advantage of the proposed method is that it enables real-time characterization, which allows for the optimization of a production process and / or the characterization of the manufactured products during production. Complex object characterization after production can be eliminated.

[0054] In particular, the stationary, partially reflective optics proposed herein make it possible to determine not only relative object distances or object dimensions based on the measurement signals, in particular peaks, but also (absolute) distances, dimensions and / or positions relative to the measuring system or the optical measuring system used.

[0055] By way of example, and based on actual measurements, the proposed method can be used to characterize cylindrical, transparent, or semi-transparent objects, such as cylinders, for example, glass cylinders, with a diameter of 50 mm to 700 mm and a wall thickness of 0.05 mm to 100 mm. It is shown that the proposed method is suitable for a comparatively broad range of applications.

[0056] According to embodiments, an optical coherence tomography system is provided which is designed for the optical characterization of a transparent or semi-transparent object based on optical coherence tomography (OCT). The coherence tomography system comprises at least one swept-source laser source with a coherence length in the range of 0.2 m to 100 m or more, in particular in the meter range, a reference substrate arranged in the beam path of the at least one light source, which reference substrate comprises, for example, a partially reflective optical system stationary in the beam path, and a control unit with a processor, in particular a processor with an architecture for parallel data processing, such as a graphics processor, and a memory assigned to the processor, which memory comprises instructions which, when executed by the processor, effect a method according to one of the method-related embodiments proposed herein.For the purposes of this invention, a processor with associated memory, configured to carry out a method according to one of the embodiments proposed herein, is generally understood to mean an electronic unit which is programmed and / or configured such that, during operation, it carries out a method according to one of the embodiments proposed herein.

[0057] The advantages and beneficial effects of the coherence tomography system and its design arise in particular from the advantages and beneficial effects of the embodiments of the method. Reference is made in this regard and in particular to the above explanations.

[0058] According to one embodiment of the optical coherence tomography system, the swept-source laser source is or comprises a MEMS VCSEL-based light source or an akinetic laser light source, and / or a coherence length of the light source is in the range of 0.2 m to 100 m or more.

[0059] According to one embodiment of the optical coherence tomography system, it further comprises at least one galvanometer scanner configured to scan the object and / or at least one microlens array configured to scan the object. These components, in particular, enable precise scanning of an object to be characterized, especially when the object is moving, for example, when the object moves along a movement path during production.

[0060] Typically, the object's motion path, or a corresponding one, is known or predetermined. The predetermined or known motion path can be used, for example, to control the galvanometer scanner and / or to evaluate the signal data acquired via the microlens array.

[0061] Within the scope of the invention, it is possible or intended to detect or determine the movement of the object, a linear movement and / or a rotational movement, or the movement path of the object, and thus at least indirectly the movement path of the object area or point to be measured. For this purpose, for example, a sensor or measuring system can be provided that is configured to detect the movement of the object.

[0062] According to embodiments of the optical coherence tomography system, it further comprises at least one infrared filter arranged in the beam path of the light source, preferably a dual-band mirror, which is configured to at least predominantly filter out thermal radiation emanating from the object and / or the object's surroundings in the far-infrared spectrum. Such a coherence tomography system is particularly suitable for characterizing hot objects, which emit thermal radiation that is not negligible with regard to OCT or that impairs the OCT result. Such an infrared filter also makes it possible to measure or characterize objects in environments that emit thermal radiation that is not negligible with regard to OCT or that impairs the OCT result.

[0063] According to embodiments of the optical coherence tomography system, the light source has a bandwidth in the range of 20 nm to 100 nm, and / or of at least 40 nm.

[0064] According to embodiments of the optical coherence tomography system, the light source is configured for repetition rates in the range of 4 kHz to 4 MHz, in particular for repetition rates of approximately 10 kHz.

[0065] Compared to known methods for the optical characterization of transparent or semi-transparent objects by means of OCT, the proposed method can, according to embodiments, implement comparatively large axial ranges, i.e. ranges measured parallel to the light radiation, in particular for objects with diameters of up to 100 mm and more, for example with diameters of up to 700 mm and more.

[0066] Furthermore, the proposed method provides for characterizing moving objects, for example, during a manufacturing process in which the object is moved, e.g., performing a rotational and / or translational movement. One example of this is the production of cylindrical bodies or objects made of glass.

[0067] Embodiments of the proposed method particularly and advantageously enable optical characterization by means of OCT of hot objects, i.e. objects that emit comparatively strong blackbody radiation, i.e. thermal radiation. Examples include the production of objects or bodies from molten glass, or more generally based on hot forming. In particular, it is possible to monitor or characterize objects and object geometries during hot forming during production, which on the one hand makes it possible to adapt and monitor process conditions for production in real time and / or to reduce waste. Furthermore, characterization steps that are otherwise usually carried out downstream of production can be omitted, which simplifies the production process and in particular can be shortened.

[0068] The proposed method, in particular in contrast to known methods and devices, enables the measurement or determination of absolute and relative distances. In particular, it is not necessary to provide separate devices for determining relative distances on the one hand and absolute distances on the other. Absolute distances can be relevant, for example, for determining or characterizing the manufacturing process if the object or product is to move along a predetermined path, e.g., in a straight line, during production.

[0069] By determining absolute distances, for example, deviations from the path can be detected and the manufacturing process can be adjusted, or potential sources of error or malfunctions in the manufacturing process or production plant can be identified from deviation data.

[0070] Relative distances can, for example, be used to characterize the geometry of the object itself, such as diameter, wall thickness, cross-sectional shape, etc.

[0071] The proposed method is particularly flexible and suitable for a comparatively broad range of applications. For example, the method is suitable for measurements with a comparatively large axial depth and simultaneously high resolution (e.g., 10 micrometers) and / or with a comparatively small axial depth and simultaneously very high resolution (e.g., 1 micrometer).

[0072] Compared to known methods and devices, the proposed method, in particular the algorithms proposed herein for processing OCT signals, can achieve advantageous resolution. In particular, OCT signals or signals derived therefrom can be achieved with an advantageous half-width.

[0073] Embodiments of the invention are described below with reference to the attached figures. They show: Fig. 1 schematically shows a first device-based OCT implementation for characterizing an object; Fig. 2 schematically shows a second device-based OCT implementation for characterizing an object; Fig. 3 schematically shows a third device-based OCT implementation for characterizing an object; Fig. 4 a procedure for determining periodograms; Fig. 5 exemplary OCT measurement results for the characterization of a glass cylinder based on the third implementation; Fig. 6 exemplary OCT measurement results for the characterization of a glass cylinder based on the first or second implementation; Fig. 7 an exemplary diagram for the characterization of a glass cylinder with a comparatively large axial depth; and Fig. 8 exemplary OCT measurement results for the characterization of a moving, hot glass cylinder.

[0074] Identical or functionally equivalent elements are designated by the same reference numerals in the figures. The figures merely describe exemplary applications or implementations, without limiting the invention thereto or the resulting advantages or advantageous effects.

[0075] Fig. 1 schematically shows a first device-based OCT implementation for characterizing an object 1, which may be, for example, a hollow glass cylinder, also briefly glass cylinder or cylinder.

[0076] The device implementation comprises a swept-source laser 2, also referred to as laser 2 for short, followed by a circulator 3. The three ports of the circulator 3 are connected to the laser 2, a collimator 4, and a beam splitter 5. The beam splitter 5 can have a splitting ratio of 90:10, with a first output 6 being assigned to the splitting ratio 90 and a second output 7 being assigned to the splitting ratio 10.

[0077] A power monitoring unit 8 is connected downstream of the second output 7.

[0078] The first output 6 is followed by a variable optical attenuator 9 (VOA), which is followed by a high-speed detector 10 for generating OCT signals 11.

[0079] Laser light guides, especially so-called TEC fibers (TEC: Thermally Extended Core), can be used to connect the components.

[0080] Downstream of the collimator 4 in the beam path of the laser radiation 12 coming from the laser 2 is a stationary, partially reflective optics 13 or a stationary reference substrate 13. Downstream of the reference substrate 13 in the beam path of the laser radiation 12 coming from the laser 2 is a galvanometer scanner 14, also referred to as scanner 14 for short, for beam deflection with deflection angles φ and θ. The object 1 to be measured or characterized is arranged downstream of the scanner 14 in the beam path.

[0081] During operation for measuring or characterizing object 1, the laser radiation 12 emanating from the collimator 4 first passes through the reference substrate 13 and then strikes the scanner 14, which directs the laser radiation 12 onto the object 1. The laser radiation 12 reflected from the object 1 passes via the scanner 14, the collimator 4, and the circulator 3 to the beam splitter 5, which directs the reflected laser radiation 12 or its signals via the first output 6 through the variable optical attenuator 9 to the detector 10. The detector 10 generates OCT signals from the reflected laser radiation 12 or corresponding signals for the optical characterization of object 1.

[0082] Based on the scanner 14, or by appropriate control of the scanner 14, for example of the angle φ and / or θ, the laser radiation 12 can be directed or irradiated onto the object 1 in such a way that the point of incidence x of the laser radiation 12 on the object 1 follows, for example, a movement of the object 1. In the example shown, the object 1 rotates counterclockwise, which is indicated by a curved arrow. If the object 1 moves linearly at the same time, for example from the figure plane of the Fig. 1, by appropriately controlling the scanner 14, it can also be achieved that the impact point x also follows this linear movement. Thus, moving objects can be optically characterized using OCT.

[0083] Fig. Figure 2 shows a schematic diagram of a second device-based OCT implementation for characterizing an object. In contrast to the implementation according to Fig. 1, a microlens array 15 is arranged downstream of the collimator 4 and the reference substrate, through which the laser radiation 12 is directed onto the object 1. The microlens array 15, for example a 2D microlens array, also makes it possible to track points x on the object 1 or on the object surface, so that the second implementation also makes it possible to characterize objects 1 that exhibit a rotational movement and / or a translational movement (e.g., according to a movement from the figure plane of the Fig. 2 out). Apart from 2D microlens arrays 15, 1D microlens arrays 15 can also be used.

[0084] In further implementation variants, a galvanometer scanner 14 and a microlens array 15 can be used in combination.

[0085] Furthermore, the structure of the implementation corresponds to Fig. 2 the one after Fig. 1.

[0086] The implementations according to Fig. 1 and Fig. 2 correspond to a setup with autocorrelation. Fig. 3 schematically shows a third device-based OCT implementation for characterizing an object 1, wherein this implementation corresponds to a setup with cross-correlation.

[0087] In the Fig. 3 shown structure is analogous to Fig. 1 and Fig. 2, a beam splitter 5 is present, which is connected on the one hand to a circulator 3 and on the other hand to a high-speed detector 10 or a power monitoring unit 8. In contrast to Fig. 1 and Fig. 2, the port of the circulator 3 closest to the port of the laser 2 in the circulation direction of the circulator 3 is connected to a first port 17.1 of a further beam splitter 16. A second port 17.2 of the further beam splitter 16 is connected to a further variable optical attenuator 18 and a retroreflector 19. A third port 17.3 is connected to a collimator 4, which, as in Fig. 1 and Fig. 2 is configured to collimate laser radiation 12 onto the object 1 to be measured or characterized, or onto laser radiation 12 reflected from the object 1. A fourth connection 17.4 is connected to the detector 10 via a variable optical attenuator 9. In the implementation according to Fig. 3 with cross-correlation, the detector 10 is coupled to both beam splitters 5, 16 for receiving laser signals. Reference substrate 13 and microlens array 15 are in Fig. 3 not shown, but may be present individually or in combination.

[0088] Fig. Figure 4 shows an exemplary process flow for determining periodograms. The algorithm uses k-clock 20-based resampling 21 for the OCT signals 11. For performance reasons, the algorithm is executed on a processor architecture with parallel data processing, such as a GPU.

[0089] For resampling 21, the start times for the k-clock and OCT signals are adjusted or aligned 22. Before resampling, the k-clock is filtered 23 with a zero-phase low-pass filter, followed by a calculation 24 of the phase evolution and an extraction 25 of the phase profile (so-called "phase unwrapping"). After the k-clock-based resampling 21 of the OCT signals, a window function is applied 26 to the resampling result, followed by a fast Fourier transform (FFT) 27 to determine OCT periodograms 28. From the OCT periodograms 28, geometric parameters or quantities characterizing the object 1, such as diameter, wall thickness, etc., can be determined.

[0090] To measure or characterize moving objects 1, which, for example, perform a comparatively fast rotational movement superimposed on a comparatively slow translational movement, the mirrors of the galvanometer scanner 14 are synchronized with the data acquisition in such a way that a point on the surface of the moving object 1, for example, a cylinder, is tracked in space. If necessary, a certain spatial jitter can be added to the mirrors of the galvanometer scanner 14, in particular to compensate for any non-perpendicular orientation of the surface of the object 1 relative to the incident laser beam.

[0091] In particular, tracking the point in space and spatial jitter contribute to reducing the probability of signal loss, or to essentially avoiding signal loss. Signal losses would, for example, lead to an incomplete characterization of the geometry of the moving object 1.

[0092] All A-scan signals obtained from tracking the point in space are combined and used to calculate periodogram signals. Combining the A-scan signals can achieve robustness against signal losses.

[0093] For example, an automatic peak-picking algorithm can be used to identify maxima of the OCT periodograms 28.

[0094] Fig. Figure 5 shows exemplary OCT measurement results for the characterization of a glass cylinder based on the third implementation (“cross-correlation”). In the diagram of the Fig. 5, the abscissa (x-axis) denotes time t, and the ordinate (y-axis) denotes wall thickness W, inner diameter ID, and outer diameter AD, each in arbitrary units, e.g., mm. The OCT measurement uses a rotating glass cylinder as object 1 to be measured. The diagram shows the curve or the determined values for wall thickness W, inner diameter, and outer diameter AD over time from the OCT measurement, which in turn is linked to the movement of the glass cylinder.

[0095] Fig. 6 shows exemplary OCT measurement results for the characterization of the glass cylinder according to Fig. 5 based on the first or second implementation (“autocorrelation”). In the diagram of the Fig. 6, analogous to Fig. 5, the abscissa (x-axis) represents time t and the ordinate (y-axis) also represents wall thickness W, inner diameter ID and outer diameter AD, each in any unit, e.g. in mm.

[0096] Fig. Figure 7 shows an example diagram for characterizing a glass cylinder as the object 1 used at a comparatively large axial depth, in particular at a comparatively large distance between object 1 and the measuring device. In the example shown, this is approximately 400 mm. In the diagram of the Fig. In Figure 7, the abscissa (x-axis) denotes the distance D of object 1 from the measuring device in millimeters (mm), and the ordinate (y-axis) denotes the OCT signal strength OS (in arbitrary units). As can be seen from the diagram, signals with a low signal-to-noise ratio can be obtained even at comparatively large distances. Furthermore, it can be seen that, for example, by using the reference substrate 13, absolute distances can also be determined.

[0097] Fig. Figure 8 shows exemplary OCT measurement results for the characterization of a moving, hot glass cylinder. An infrared filter was used for the measurement, which is designed to filter out thermal radiation emanating from the object and / or its surroundings in the far infrared spectrum. In the diagram of the Fig. 8 the abscissa (x-axis) denotes time t and on the ordinate (y-axis) wall thickness W, inner diameter ID and outer diameter AD are each given in arbitrary units, e.g. mm. Fig. 8 shows in particular that the proposed method is also suitable for measuring or characterizing hot, transparent or semi-transparent objects.

[0098] From the exemplary embodiments, it becomes particularly clear that the method proposed herein and the optical coherence tomography system enable reliable measurement and geometric characterization of objects and are suitable for determining not only relative object geometries, but also absolute distances with respect to the measuring system and the object. Furthermore, the method is suitable for characterizing hot and / or moving objects. In particular, the method is suitable for characterizing objects, e.g., during a manufacturing process, wherein the characterization can be carried out essentially in real time. REFERENCE SYMBOL 1 object 2 swept-source lasers or lasers 3 Circulator 4 Collimator 5 beam splitters 6 first exit 7 second exit 8 Performance monitoring unit 9 variable optical attenuator 10 High-speed detector 11 OCT signals 12 Laser radiation 13 stationary, partially reflective optics, stationary reference substrate 14 galvanometer scanners or scanners 15 microlens array 16 additional beam splitters 17 connections for additional beam splitters 18 additional variable optical attenuator 19 Retroreflector 20 K-Clock 21 Resampling 22 Synchronization of start times 23 zero-phase low-pass filtering 24 Calculation of phase development 25 Extraction of the phase curve (“phase unwrapping”) 26 Applying a window function 27 Fast Fourier Transformation (FFT) 28 OCT periodogram t time W wall thickness ID inner diameter OD outer diameter D Distance OS OCT signal strength X point of impact, point

Claims

[1] Method for the optical characterization of a transparent or semi-transparent object (1), in which object dimensions (W, ID, AD), object distances and / or object positions are determined based on optical coherence tomography, wherein a swept-source laser source with a coherence length in the decimeter to meter range is used as the light source (2), and a reference substrate (13) is used to determine absolute distances, dimensions and / or positions in the beam path (12) of the light source (2), which reference substrate comprises a partially reflecting optic that is stationary in the beam path (12), characterized by that the object (1) is moving and a scanning path used in scanning the object (1) at least partially follows the movement of the object (1) and that a spatial jitter is superimposed on the scanning path. [2] Method according to claim 1, characterized bythat the swept-source laser source (2) comprises a MEMS VCSEL based light source or an akinetic laser light source and / or that the coherence length is in the range of 0.2 m to 100 m or more. [3] Method according to claim 1 or 2, characterized by that during signal acquisition and / or signal evaluation of coherence tomography a k-clock resampling (21) is used, b. algorithmically determine minima, maxima or zero crossings of the k-clock of the laser source (2) and use them as sampling clock, or c. a k-clock hardware trigger is used. [4] Method according to one of the preceding claims, characterized bythat in optical coherence spectroscopy, the object (1) is scanned with laser light (12) from the light source (2) using a galvanometer scanner (14), using a microlens array (15) and / or a bifocal objective, preferably with a high numerical aperture. [5] Method according to claim 4, characterized by that the optical coherence tomography data acquisition is synchronized with the scanning path. [6] Method according to one of the preceding claims, characterized by that several, in particular a plurality of reflection profiles obtained along the scanning path are combined, periodogram signals are determined from the combined reflection profile thus obtained, and at least one object dimension (W, ID, AD), at least one object distance and / or at least one object position are determined from at least one periodogram signal. [7] Method according to one of the preceding claims, characterized bythat an infrared filter, preferably a dual-band mirror, is arranged in the beam path (12) of the light source (2) and thermal radiation emanating from the object (1) and / or the object's surroundings in the range of the far infrared spectrum is at least predominantly filtered out using the infrared filter. [8] Method according to one of the preceding claims, characterized by that the optical coherence tomography comprises a signal recording of the k-clock (20) with a software-based filtering (23) and an offset correction (24), and further, optionally, a subsequent phase extraction (25), in particular based on a Hilbert transformation and / or an evaluation of zero crossings and extrema, wherein, further optionally, the filtering (23) comprises a low-pass filtering. [9] Method according to one of the preceding claims, characterized bythat periodogram signals are determined from signals of the optical coherence tomography by means of an automatic peak-picking algorithm from periodogram data (28), and at least one object dimension (W, ID, AD), at least one object distance and / or at least one object position are / is determined from the periodogram signals, in particular taking into account periodogram signals of the reference substrate (13). [10] Optical coherence tomography system designed for the optical characterization of a transparent or semi-transparent moving object (1) based on optical coherence tomography, comprising at least one swept-source laser source (2) with a coherence length in the range from 0.2 m to 100 m or, in particular in the meter range, a reference substrate (13) arranged in the beam path of the at least one light source (2), which reference substrate (13) comprises a partially reflective optical system that is stationary in the beam path (12), and a control unit with a processor, in particular a processor with an architecture for parallel data processing, and a memory assigned to the processor, which memory comprises instructions that, when executed by the processor, bring about a method according to one of claims 1 to 9. [11] Optical coherence tomography system according to claim 10, wherein the swept-source laser source (2) comprises a MEMS VCSEL based light source or an akinetic laser light source, and / or a coherence length of the light source (2) is in the range of 0.2 m to 100 m or more. [12] Optical coherence tomography system according to claim 10 or 11, further comprising at least one galvanometer scanner (14) configured to scan the object and / or at least one microlens array (15) configured to scan the object (1). [13] Optical coherence tomography system according to one of claims 10 to 12, further comprising at least one infrared filter arranged in the beam path (12) of the light source (2), preferably a dual-band mirror, which is designed to at least predominantly filter out thermal radiation emanating from the object (1) and / or the object environment in the range of the far infrared spectrum. [14] Optical coherence tomography system according to one of claims 11 to 13, wherein the light source (2) has a bandwidth in the range of 20 nm to 100 nm, and / or of at least 40 nm, and / or is configured for repetition rates in the range of 4 kHz to 4 MHz, in particular of approximately 10 kHz.

Images