Method for measuring or inspecting reflective surfaces using deflectometry

Abstract

In a method for measuring or inspecting reflective surfaces using deflectometry, the optical patterns used in deflectometry are automatically calculated in advance based on the target geometry of the surface to be measured and optimized using a cost function for the surface features to be determined. The calculation is preferably performed using a differentiable ray tracing method. The proposed method allows for the high-precision inspection and / or measurement of more complex surface geometries.

Description

Technical Field of ApplicationThe present invention relates to a method for measuring or inspecting specular surfaces by means of deflectometry, in which different optical patterns are displayed one after the other on a screen, which are recorded by means of a spatial resolution optical detector via a reflection on a surface to be measured or inspected and are subsequently evaluated.For the measurement or inspection of specular surfaces, for example painted automobile panels, mirrors or optical components, deflectometry represents an established measurement method. In this case, a plurality of optical patterns are generally displayed one after the other on a screen, which are viewed by a camera via the mirroring in the surface to be measured. By decoding the patterns, a relationship between camera pixels and screen pixels can be established and thus the local inclination profile of the surface can be reconstructed by triangulation and used for surface surveying or for detecting material defects.Prior ArtCurrently, strip-shaped sinusoidal patterns with different spatial frequencies are used as optical patterns in deflectometry, with which good results can be achieved when measuring surfaces with simple geometries, i.e. small curvature and without complicated structures. However, in the case of more complex object geometries, the reconstruction quality and the defect detection accuracy decrease.From S. Keinnel, "Automated optimization of measurement setups for the inspection of specific surfaces", Proc. SPIE 4567, Machine Vision and Three-Dimensional Imaging Systems for Inspection and Metrology II, pages 199 to 206, discloses a method in which the geometric parameters of a deflectometric arrangement, i.e. for example the camera pose, the screen orientation etc., are optimized for the object to be measured in each case via a simulation. The optimization is performed using CAD data of the object to be measured and a model of the imaging function of the optical system using a ray tracing method. Generation of object adjusted optical patterns is not discussed in this publication.The object of the present invention is to specify a method for measuring or inspecting specular surfaces by means of deflectometry, which method enables the measurement or inspection of more complex surface geometries with higher accuracy than is achieved when using strip-shaped sinusoidal patterns.SUMMARY OF THE INVENTIONThe object is achieved by the method according to claim 1. Advantageous embodiments of the method are the subject matter of the dependent patent claims or can be gathered from the following description and from the exemplary embodiment.In the proposed method for measuring or inspecting reflecting surfaces by means of deflectometry, different optical patterns are displayed one after the other on a screen in a known manner, recorded by means of a mirror on a surface to be measured or inspected using a spatially resolving optical detector, in particular a camera, and subsequently evaluated. Depending on the intended application, the evaluation takes place, for example, for the reconstruction of the surface profile or the spatially resolved inclination of the surface and / or for the detection of material defects on the surface. The proposed method is distinguished in that the optical patterns used are automatically calculated beforehand as a function of a setpoint geometry of the surface to be measured or inspected, and are optimized during the calculation with a cost function for features of the surface to be determined. The calculation of the optical patterns is carried out here on the basis of the desired geometry of the surface to be measured or inspected and a deflectometric arrangement predetermined for carrying out the deflectometry, preferably using a differentiable ray tracing method with the technique of gradient descent. Suitable available raytracers, for example the raytracer mitsuba 3 (cf., for example, https: / / mitsuba.readthedocs.io / en / latest / ) can be used for this purpose.The deflectometric arrangement comprises the position and orientation of the screen and camera and of the surface to be measured or inspected or of the object with this surface. A sequence of optical patterns is thus calculated or generated in advance, which are specifically adapted to the specific desired geometry of the surface to be measured or inspected, in order to enable optimum results, in particular in the sense of measurement uncertainty. The method is based on modern techniques of differentiable ray tracing from computer graphics. Ray tracing methods are capable of generating photorealistic images of synthetic scenes. By using a ray tracing method which supports automatic differentiation, a quality measure can be defined on the respectively generated images, for example the measurement uncertainty with respect to which the optical patterns can then be optimized.By specifically adapting the optical patterns used in deflectometry to the surfaces to be measured or inspected and the respective aim of measuring or inspecting these surfaces, test objects or surfaces with a more complex surface geometry can be measured or inspected with higher accuracy than is the case when using the strip-shaped sinusoidal patterns used hitherto.In a possible embodiment of the proposed method, in which the geometry of the surface to be measured or inspected is to be determined, the deviation between a geometry reconstructed when the patterns are used and the desired geometry is minimized during the calculation of the optical patterns. In a further configuration, in which material defects are to be detected in the surface to be measured or inspected, the detection sensitivity or detection accuracy of defects introduced synthetically into the desired geometry and / or a desired reflectance distribution of the surface is maximized during the calculation of the optical patterns.The method can be used advantageously in all applications of the technique of deflectometry, in which more complex surfaces, i.e. surfaces with stronger curvatures or a special structure, are to be measured or inspected.Brief Description of the DrawingsThe proposed method is explained again below with reference to an example in conjunction with the drawings. Shown here are: FIG. 1 shows a schematic representation of a deflectometric arrangement and a framework for calculating the optical patterns used in deflectometry; FIG. 2 is a diagram showing a conventional optical pattern for deflectometry; and FIG. 3 shows an example of an optical pattern for deflectometry used in the proposed method.WAYS OF CARRYING OUT THE INVENTIONIn FIG. 1, a typical deflectometric arrangement is schematically shown on the left-hand side, which comprises a screen 1 and a camera 2, which are arranged such that an optical pattern 3 shown on the screen is captured by the camera 2 via a reflection on the surface 4 to be measured. By comparing the pattern 3 displayed on the screen with the pattern recorded by the camera 2 via the mirroring, the surface shape is reconstructed in a deflectometric reconstruction unit 5. Alternatively or additionally, the pattern modulated by the inspected surface can be used in a defect detection network 6 for detecting defects in the surface 4 to be measured. In principle, any suitable machine learning methods can be used for this purpose, i.e. for example also classic methods which are based on feature extraction and subsequent classification, e.g. support vector machines or the like.In the present example, a complete deflectometric measurement setup modeled by differentiable ray tracing is combined with a likewise differentially formulated reconstruction method in a common framework. For the surface to be measured of a test object geometry given in the present example by a CAD model, an optimization of the screen patterns to be displayed with respect to the measurement uncertainty is then carried out by means of gradient descent methods. There are two possible optimizations for which two optimization functions l 1 and l 2 are used below.Optimization functional I 1 describes the deviation between the reconstructed surface geometry and the predefined desired geometry (given by the 3D model of the test object), also referred to below as ground truth. Optimization functional I 2 describes how well synthetically introduced material defects were detected by a corresponding method (e.g. deep neural network). The optimization is effected by differentiable inverse rendering. The reconstruction is carried out by means of known deflectometric reconstruction methods. Both optimization functions can be derived with respect to the pattern displayed on the screen or the displayed pattern sequence. By gradient descent, the screen patterns can then be updated successively in such a way that the corresponding optimization function is minimized, i.e. the reconstruction accuracy is optimized at l 1 and the defect detection accuracy is optimized at l 2. This is also indicated schematically in FIG. 1.While arbitrary patterns are possible in the optimization function l 2 i.e. all pixels of the displayed pattern are released as optimization parameters, the patterns with respect to the optimization function l 1 must still allow a deflectometric reconstruction, i.e. a determination of the correspondence between camera pixel and screen pixel. For this purpose, for example, a parameterization of the screen patterns can be used, in which deformed wavefronts are permitted in the optical patterns instead of the conventional planar wavefronts. FIG. 2 shows an example of a conventional strip-shaped sinusoidal pattern, while FIG. 3 shows an example of an optical pattern generated according to the proposed method, in which deformed wavefronts are used. In this case, different approaches for modelling the deformation can be selected, for example a polynomials of a certain degree or a Fourier series (composition of cos and sin patterns) with a fixed number of coefficients. As a result, the pattern space is greatly restricted and the optimization is restricted to the determination of the coefficients of the respectively selected model.The proposed method requires a ray tracing method which supports automatic differentiation, also referred to as differentiable ray tracing methods in the present patent application. Such methods are available, wherein, for example, the Raytracer Mitsuba 3 can be used. As input data for the calculation of the optical patterns, the desired geometry of the object to be measured / inspected is provided, for example as a CAD model. Further, the bidirectional reflectance distribution function of the surface of the object to be measured / inspected may be required. The calculation also requires the parameters of the deflectometric arrangement, i.e. the position and orientation of the screen and the optical detector or camera, parameters of the optical detector (e.g. optical parameters such as focal length etc., pixel size etc.), and corresponding parameters of the screen (resolution, dimensions). Optionally, positions / size / properties of synthetic defects can also be predetermined if l 2 is to be optimized.With the proposed method, it is possible to generate and use optical patterns as screen patterns which are optimum for the specific object or the specific surface to be measured or inspected with respect to a cost function which can be selected in a problem-specific manner, for example the measurement imprecision. The desired features or properties can thus be detected with higher accuracy than is the case with the strip-shaped sinusoidal patterns used hitherto.List of reference characters1 Screen 2 Camera 3 Optical pattern 4 Surface 5 to be measured, deflectometric reconstruction unit 6 Defect detection networkReferences included in the specificationThis list of documents cited by the applicant has been produced in an automated manner and is only included for the better information of the reader. The list is not part of the German patent application or utility model application. The DPMA does not take any adhesion for any faults or omissions.Cited Non-Patent LiteratureS. Keinnel, "Automated optimization of measurement setups for the inspection of specific surfaces", Proc. SPIE 4567, Machine Vision and Three-Dimensional Imaging Systems for Inspection and Metrology II, pp. 199-206

[0004] https: / / mitsuba.readthedocs.io / en / latest /

[0007]

Claims

Method for measuring or inspecting reflecting surfaces by means of deflectometry, in which different optical patterns (3) are displayed one after the other on a screen (1), which are recorded by means of a spatial resolution optical detector (2) via a reflection on a surface (4) to be measured or inspected and are subsequently evaluated, characterized in that the optical patterns (3) are calculated in advance automatically as a function of a desired geometry of the surface (4) to be measured or inspected and are optimized during the calculation with a cost function to features of the surface (4) to be determined.Method according to Claim 1, characterized in that the calculation of the optical patterns (3) takes place on the basis of the desired geometry of the surface (4) and of a deflectometric arrangement which is predetermined for carrying out the deflectometry, using a differentiable ray tracing method using the gradient descent technique.Method according to Claim 1 or 2, characterized in that, in order to determine a geometry of the surface (4) as a feature to be determined, a deviation between a geometry reconstructed when the patterns are used and the desired geometry is minimized during the calculation of the optical patterns (3).Method according to one of Claims 1 to 3, characterized in that, in order to determine material defects in the surface (4) as a feature to be determined, a detection sensitivity or detection accuracy of defects introduced synthetically into the desired geometry and / or a desired reflectance distribution is maximized during the calculation of the optical patterns (3).Method according to one of Claims 1 to 4, characterized in that the desired geometry of the surface (4) is provided as a CAD model.

Images