TOPOGRAPHIC MEASURING METHOD AND TOPOGRAPHIC MEASURING MACHINE

Description

technical field

[0001] The invention relates to a topographic measurement method and a topographic measurement machine. State of the art

[0002] In many technical fields, protruding elements are used. These protruding elements must have precisely controlled shapes and dimensions. For example, electrically conductive pads serve as electrical contacts for electronic chips or other electronic devices. Accurate dimensional control of these protruding elements ensures reliable electrical contact.

[0003] It is important to be able to measure the shape and dimensions of the multiple protruding elements as accurately and quickly as possible.

[0004] It is known to measure the dimensions of protruding elements using an optical method. Textured light is applied to the surface of a sample containing multiple protruding elements. The textured light defines a plurality of repeating patterns that are applied to the sample surface.

[0005] Several images are acquired of the sample surface illuminated by textured light. By analyzing the different images, it is possible to calculate the dimensions and shape of the protruding elements.

[0006] The quality of the measurement is highly dependent on the quality of the images acquired. Therefore, there is a need to improve the quality of the images used in the calculation of the dimensions and shape of protruding elements.

[0007] US 2021 / 0310791 describes a topographic measuring machine for the optical measurement of a sample containing a plurality of salient patterns. Description of the invention

[0008] One object of the invention is to provide a topographic measurement method that is more efficient than prior art methods.

[0009] This result is achieved through a topographic measurement process which includes: provide a sample comprising a first surface furnished with a plurality of salient patterns; acquire a plurality of images of the first surface of the sample, the plurality of images comprising reference images of the sample, each reference image corresponding to the sample illuminated with structured light, the structured lights of the reference images each defining several repeating patterns and distinguished from each other by the phase and / or frequency of the repeating patterns.

[0010] The process is remarkable in that the plurality of images includes at least a first series of images of the first surface of the sample, the first series of images being acquired when the first surface of the sample is illuminated by a first structured light, the first structured light defining several repeating patterns and the images of the first series of images being distinguished from each other by the angle of application of the first structured light with respect to the first surface; compare the images of the first set of images so as to determine the presence of at least one artifact on at least one image of the first set of images; form at least one working image from the first set of images, the at least one working image being devoid of said at least one artifact or reducing the extent of said at least one artifact; calculate at least one quantity representative of the sample from the at least one working image and the reference images.

[0011] Preferably, the first set of images includes an image of the reference images and at least one working image replaces said image of the reference images.

[0012] Preferably, the image plurality has several first sets of images, each containing one image of the reference images. Each first set of images is used to create a working image that is free of at least one artifact or has a reduced-amplitude artifact. Each working image replaces one of the reference images.

[0013] Preferably, the plurality of images includes at least a second series of images of the first sample surface. This second series of images is acquired when the first sample surface is illuminated by the first structured light. The images in the second series are distinguished from one another by their exposure time, and the angle of application of the first structured light is constant for the second series and corresponds to one of the angles of the first series. The measurement procedure comprises: compare the images of the second set of images so as to determine the presence of at least one artifact on at least one image of the second set of images and form at least one working image from the second set of images, the at least one working image being devoid of said at least one artifact or reducing the extent of the artifact.

[0014] Preferably, the second set of images includes an image of the reference images. At least one working image formed from the second set of images replaces said image of the reference images.

[0015] Preferably, one image from the reference images belongs to both the first and second image sets. The first and second image sets are then used to create a working image that replaces the aforementioned image from the reference images.

[0016] Preferably, the plurality of images has several second sets of images, each containing one image from the reference images. Each second set of images is used to create a working image that is free of at least one artifact or has a reduced-area artifact. Each working image replaces images from the reference images.

[0017] Preferably, the plurality of images includes at least a third series of images of the first sample surface. This third series of images is acquired when the first sample surface is illuminated by a second structured light. The images in the third series are distinguished from one another by the color of the structured light. The angle of application of the second structured light is constant for the third series of images and corresponds to one of the angles of the first series of images. The measurement procedure comprises: compare the images of the third series of images so as to determine the presence of at least one artifact on at least one image of the third series of images and form at least one working image from at least the third series of images, the at least one working image being devoid of said at least one artifact or reducing the extent of the artifact.

[0018] Preferably, comparing the images of the first series of images involves identifying a shift in an area whose brightness is greater than the average brightness of the image of the first series as a function of the angle of application of the first structured light, the artifact being a spurious reflection.

[0019] Preferably, comparing the images of the first series of images involves identifying an evolution in the extent of an area whose brightness is less than the average brightness of the image of the first series as a function of the angle of application of the first structured light, the artifact being a shadow.

[0020] Preferably, comparing the images from the first series of images involves extracting the light intensity changes from the repetitive patterns resulting from the structured light and comparing these light intensity changes.

[0021] The invention also relates to a topographic measuring machine which is more efficient than prior art machines and which allows for more precise calculation of sample quantities.

[0022] This result is achieved using a topographic surveying machine which includes: - - a support having a surface intended to support a sample; projection means configured to emit different structured lights, each structured light defining several repeating patterns and distinguished from each other by the phase and / or frequency of the repeating patterns; image capture means configured to acquire images of the sample illuminated by one of the different structured lights; a control circuit configured to acquire several reference images of the sample illuminated by different structured lights, each structured light of the reference images defining several repeating patterns and distinguished from each other by the phase and / or frequency of the repeating patterns.

[0023] The topographic surveying machine is remarkable in that: The projection means are configured to define several angles of application of the structured light relative to the surface of the support; the control circuit is configured to acquire a first series of images of the sample illuminated by a first structured light, the first structured light defining several repeating patterns and the images of the first series of images being distinguished from each other by the angle of application of the first structured light relative to the surface of the support; the control circuit is configured to compare the images of the first series of images so as to determine the presence of at least one artifact on at least one image of the first series of images; the control circuit is configured to form at least one working image from the first series of images, the at least one working image being free of said at least one artifact or reducing the extent of said at least one artifact;The control circuit is configured to calculate at least one representative quantity of the sample from at least one working image and reference images. Description of the drawings

[0024] Other advantages and features will become clearer from the following description of particular embodiments and implementations of the invention, given by way of non-limiting examples and shown in the accompanying drawings, in which: there figure 1 illustrates, schematically, a topographic surveying machine; the figure 2 is a photograph of a sample with prominent patterns in the form of a hemisphere illuminated by uniform light; the figure 3 illustrates, schematically, a topographic measurement process. Detailed description

[0025] When a sample has one or more protruding features extending from a surface of a substrate, it is useful to characterize the sample and, more specifically, the protruding feature(s). This means calculating representative quantities of the sample and, even more importantly, representative quantities of the protruding feature(s). A quick method is to use optical characterization by taking several images of the sample. Therefore, it is crucial to obtain high-quality images to deduce the sample's size(s) as accurately as possible, particularly those of the protruding features.

[0026] In a first step S1 illustrated at the figure 3 , we provide a sample 1 which includes a first surface furnished with a plurality of salient motifs 2 (see figure 1The sample 1 is placed on a support 3. The protruding motifs 2 can be of any shape. However, it is particularly advantageous for the protruding motifs 2 to be in the form of spheres or hemispheres, cones, rectangular prisms, or other simple shapes, as this facilitates the calculation of the shape as well as the calculation of the transverse dimensions of the protruding motifs 2. The transverse dimensions are dimensions measured along directions perpendicular to the axis of the normal to the support 3, which supports the sample 1 and also represents the normal to the first surface. The normal to the first surface is represented by the axis A shown in Figure 1. figure 1 which represents a topographic measuring machine capable of implementing the measurement process.

[0027] The raised patterns 2 can be made from any material. However, the process appears to be particularly advantageous when the raised patterns 2 are made of metallic material and, more generally, of light-reflecting materials. It has been observed that patterns made of metallic materials reflect light, which generates artifacts. It has also been observed that the emission of textured light from projection devices 4 onto the raised patterns 2 generates a shading phenomenon, resulting in one face of the raised patterns 2 being better illuminated than the others. This phenomenon intensifies as one moves away from the normal to the first surface.

[0028] The topographic measurement method involves illuminating the first surface of sample 1 with structured light. The structured light defines several repeating patterns, for example, several fringes. The illumination of the first surface is achieved by projection means 4. The structured light can be an image in shades of gray or in shades of other colors, and possibly in black and white. The structured light defines patterns that repeat with a predefined frequency and a predefined phase. The projection means 4 advantageously include a projector configured to project the structured light, preferably in the visible range.

[0029] Structured light advantageously forms a fringe pattern, for example, a moiré or shadow moiré pattern. In one embodiment, the projector directly emits a fringe pattern. In another embodiment, the structured light is formed by means of a mask through which the light from the projector passes. The mask has openings and opaque areas, for example, in the form of a Ronchi grid. It is also possible to combine these two techniques.

[0030] The measurement process involves, in a step S2, acquiring a plurality of images of the first surface of sample 1 and in particular a plurality of reference images in a step S2a.

[0031] The reference images show the deformation of the repeating patterns of structured light on the sample surface. Knowing the angle of application of the structured light and the repetition interval of the patterns, it is possible to deduce the deformation of the repeating patterns and thus derive information about the sample's properties. By using several repeating patterns that differ from each other in their repetition interval and phase (and possibly their size), it is then possible to calculate quantities related to the sample.

[0032] In a step S4, the reference images are used to calculate at least one quantity of the sample and preferably a quantity representative of the plurality of salient patterns 2. Preferably, the reference images are used to calculate at least one quantity which is chosen from the coplanarity of the salient features, the shape of the salient features in a cross-section plane, i.e. perpendicular to the axis A, the repetition step of the salient features in one or more directions perpendicular to the axis 1, a cross-sectional dimension of at least one of the salient features, i.e. along a direction perpendicular to the axis A and the height of one or more salient features i.e. the dimension along the axis A.

[0033] To acquire images, it is advantageous to use image capture means 5 configured to acquire images of the first surface of the sample 1. The image capture means 5 advantageously include a camera arranged to capture structured light applied to the surface of the sample 1. The camera is adapted to provide surface topography measurements of the system from the captured structured light. In the illustrated embodiment, the optical axis of the image capture means 5 coincides with the normal A to the surface of the support 3.

[0034] To characterize the sample, it is advantageous to successively apply different structured lights. Structured lights define repeating patterns, preferably fringes. To characterize the sample, a plurality of reference images of the first surface of sample 1 will be acquired in step S2a. Each reference image is acquired with a different structured light. The structured lights are distinguished from one another by the phase and / or the frequency of the repeating patterns. By varying the frequency, the repetition interval of the patterns is varied, and it is also advantageous to vary the width of the patterns, particularly the fringes. By varying the phase, the position of the fringes on the sample is varied; a white area becomes a black area or a shade of gray.By varying the phase and frequency of the patterns, a person skilled in the art can characterize sample 1. The evolution of the intensity representing the passage from a white area to a black area can be sinusoidal, triangular, stepped or any other suitable shape.

[0035] Acquiring several reference images that differ from each other by the phase and frequency of repeating patterns, preferably fringes, allows the calculation of representative quantities of the sample.

[0036] The various reference images are analyzed together to determine the topography of the sample using a known technique. The deformation of the fringes on the surface of the sample is observed to deduce the characteristics of the sample and preferentially any salient patterns.

[0037] It is possible to calculate the spacing between two salient patterns 2, the height of each salient pattern 2, the lateral dimensions of each salient pattern 2, the shape of each salient pattern 2, and the flatness of the surface receiving the salient patterns 2. It is also possible to calculate the mathematical area passing through the highest point of the salient pattern (along axis A) for each salient element 2. Furthermore, it is possible to calculate the coplanarity of the vertices of the salient patterns 2, that is, to determine whether the vertices of the salient patterns 2 lie in the same plane or not, and possibly the extent (along axis A) of the area passing through all the vertices of the salient patterns 2. Finally, it is possible to calculate characteristics of the surface passing through the vertices of the salient patterns 2 and / or characteristics of the surface of the substrate supporting the salient patterns 2.The calculation of the characteristics can be a calculation of the type of deflection (BOW) and torsion (TWIST).

[0038] However, these images contain artifacts that prevent a more precise characterization of sample 1, so it is important to acquire additional images in order to deliver a finer characterization of the representative quantities of the sample and preferentially of the salient patterns.

[0039] Structured light is emitted at a first angle α relative to the first surface. The projection means 4 project the structured light at a first angle α relative to the first surface of the sample 1, also called the application angle α. The first angle α is measured relative to the normal to the surface of the support 3 that supports the sample and which corresponds approximately to the normal to the first surface of the substrate that receives the raised patterns. Other configurations are possible.

[0040] To improve the quality of characterization, it is advantageous to acquire several images under different conditions to determine the presence of artifacts and to be able to remove them or at least reduce their significance. The additional images acquired can be used to replace a reference image with another calculated reference image in order to reduce the extent of at least one artifact, or even remove at least one artifact, or to add a reference image to the existing reference images to reduce the impact of the artifacts. This allows for a more precise calculation of the sample's characteristic parameters.

[0041] In step S2b, it is advantageous to acquire at least a first series of images of the first surface of sample 1. The first series of images is acquired when the first surface of sample 1 is illuminated by a first structured light. The first structured light defines several predefined repeating patterns, and the images in the first series are distinguished from one another by the angle of application α of the first structured light relative to the first surface.

[0042] The projection means 4 are configured to apply the same first structured light according to different values ​​of the first angle α. The first angle α is preferably measured between the optical axis of the projection means 4 and the normal to the plane of the support 3 that receives the sample. Depending on the configuration, the projection means 4 include a movable projector that moves to define the multiple angles α. It is also possible to have several projectors, each projector associated with a predefined angle. It is also possible to have several movable projectors.

[0043] The first angle α is advantageously between 5° and 175°, preferably between 10° and 170°, more preferably between 25° and 155°, and even more preferably between 25° and 55° and between 115° and 155°. Angles between 0° and 90° correspond to projection means 4 arranged on one side of the sample, that is, on one side of the support and therefore on one side of the image capture means 5, for example on the right side of the figure 1 while angles between 90° and 180° correspond to projection means 4 arranged on the other side of the sample, i.e. on the other side of the support and therefore on the other side of the image capture means 5, for example on the left side of the figure 1 By using projection means 4, for example two projectors, capable of successively illuminating the two opposite faces of the salient patterns 2, it is easier to better define the contours of the salient patterns 2.

[0044] From the plurality of images, at least one initial series of images is acquired of the first surface of sample 1. This first series of images is acquired when the first surface of sample 1 is illuminated by a first structured light. The first structured light defines several repeating patterns.

[0045] It has been observed that when light reflection phenomena occur between the salient patterns 2, some of the emitted structured light is reflected back to the capture means 5 as specular reflection, resulting in a very bright area in at least part of the images. This very bright area is defined as an area whose brightness exceeds the average brightness of the image. This area of ​​high brightness is characterized by low contrast, making it more difficult or impossible to observe the characteristics of the structured light patterns, particularly the evolution of light intensity. Stray reflection, and especially specular reflection, generates poor local contrast in the projected repetitive patterns.

[0046] Poor contrast introduces artifacts and reduces the accuracy of the topographic reconstruction calculation. It then becomes more difficult to characterize the sample, for example, the characteristics of salient patterns, particularly the edge of salient pattern 2, the shape of salient pattern 2, or an edge. This very bright area creates an artifact that makes it more difficult to measure the representative dimensions of salient patterns 2. By modifying the application angle α, it is possible to shift the reflection points of the emitted structured light and thus move the position of the very bright area between the images of the first series.By comparing the position of very bright areas on the different images of the first series as a function of the value of the angle of application, it is possible to identify a displacement of a very bright area as a function of the angle of application α which makes it possible to determine the presence of an artifact, in this case a spurious reflection or a refraction.

[0047] Modulating the application angle α also allows for modification of shading phenomena, which makes it possible to better define the edges of the salient patterns 2 while maintaining structured light emitted from the same side of the support and therefore of the sample. Shading results in the appearance of a dimly lit area, i.e., a low-contrast area. Due to this low contrast, it is difficult to observe the repetitive patterns of the structured light and therefore to calculate relevant information about the sample, particularly the salient patterns 2.

[0048] Modulating the application angle α also helps reduce the influence of multiple reflections between salient patterns. Multiple reflections will reflect back to the capture means a projected intensity that should originate from another location on the sample. This locally distorts the reconstruction calculation. To further reduce the influence of multiple reflections, it is also advantageous to modify the frequency and / or the pitch of the textured light.

[0049] For example, the figure 2Figure 1 illustrates a sample with a substrate surmounted by protruding patterns roughly in the form of hemispheres. The sample is illuminated by uniform white light applied at an angle of 45°. It appears that the protruding patterns do not all reflect light in the same way and that each pattern exhibits a specular reflection zone and a penumbra zone. It also appears that the central protruding pattern has several reflection points representing the reflections of adjacent protruding patterns.

[0050] In general, modulating the application angle α allows us to determine the presence of at least one artifact, which appears as a high-brightness or low-brightness area and is a consequence of reflection, refraction, etc.

[0051] Modulating the application angle α allows the acquisition of a plurality of images which make it possible to better determine the representative quantities of the sample by facilitating the determination of the characteristics of the repeating patterns, by reducing the extent of artifact and / or by moving artifacts which makes it possible to obtain sufficient contrast to identify the presence of at least part of a repeating pattern of the textured light.

[0052] The images in the first series are distinguished from one another by the value of the application angle α of the first structured light relative to the first surface. Modulating the application angle α allows for modification of the interaction between the sample and the structured light. Each image corresponds to a known value of the application angle α. Modulating the application angle α of the first light allows for modification of the extent of the shadowed areas as well as the projection angle of the structured light onto the lateral walls of the salient patterns 2, and therefore the reflections and refractions.

[0053] Acquiring several images to form the first series of images allows for better observation of the repetitive patterns of structured light, which allows for better identification of the dimensional characteristics of the salient patterns 2 and preferentially when the first series of images includes images where the first structured light is emitted in opposite directions, i.e. with angles whose values ​​are positive and negative, i.e. with projection means 4 which are arranged on one side or the other of the sample 1.

[0054] It is particularly advantageous to have images from the first series that correspond to different values ​​of the application angle α and that are emitted from the same side of the sample and / or from the same side of the capture means 5. For example, at least two or at least three images are acquired when the first light is applied to the sample 1, the images being distinguished from each other by the value of the application angle α, the first light being emitted from the same side of the support 3. For example, the three images are acquired for three different values ​​of angle α that differ from each other by at least 10°. For example, the three images are acquired for angles α between 80° and 10°, preferably between 70° and 20°, even more preferably between 60° and 30°, for example at 55°, 45° and 35°.

[0055] Using several images associated with angle values ​​α preferably whose difference is less than 30°, preferably less than 20° or even 15° or 10° makes it easier to track the displacement of artifacts as a function of the application angle a. The difference in angle α between the images is preferably greater than 1° and more preferably greater than 5°.

[0056] It is also advantageous that the first series of images has images at different angles α with structured light emitted from one side of the sample as well as several images at different angles α with structured light emitted from the other side of sample 1. Both sides of sample 1 are illuminated under several different conditions to improve the quality of the calculation of the characteristic quantities of sample 1.

[0057] In step S3a, the images from the first series of images are used to form a working image which will form an additional reference image or replace one of the reference images. In step S3a, it is advantageous to compare the images from the first series of images in order to identify the presence of one or more artifacts which correspond to areas with low brightness or areas with high brightness, i.e. areas with low contrast.

[0058] Once the artifact(s) have been identified, the images from the first series are used to form a working image that represents the sample and is devoid of at least one of the identified artifacts or whose extent of the artifact is reduced.

[0059] The images from the first series can be used in various ways. It is possible to identify areas representing artifacts in the images and search in other images for the same area without the artifact in order to create a working image that corresponds to a stacking of the multiple initial images that form the first series, added together, for example in the form of layers with areas of varying weights.

[0060] It is particularly advantageous to process the images from the first series to identify the patterns of the first structured light and, preferably, to extract the intensity variations of the structured light that define the repeating patterns. For example, if the structured light defines grayscale fringes represented by a sinusoidal intensity variation, it is beneficial to search for this sinusoidal variation in the images. Since the fringes are distorted by the sample, the variations are no longer necessarily sinusoidal, but the variations are transposable between images because the frequency and the application angle α are known. It is then possible to compare the images by comparing the intensity variations of the structured light in order to estimate the distortions of the patterns on the sample surface, given the known application angle α.This configuration is particularly interesting because it allows us to disregard variations in light intensity on the salient patterns 2 due to variations in the application angle α. The working image is represented by the light intensity changes on the sample. This working image is used in conjunction with reference images, which also contain the light intensity changes on the sample, to provide a higher-quality characterization.

[0061] Advantageously, one image from the first set of images is one of the reference images. The working image is then created to replace the reference image from the first set of images. The working image is of higher quality than the reference image it replaces because it has eliminated or reduced the artifact.

[0062] Even more advantageously, several initial image series are acquired, and each initial series includes one of the reference images. Each initial series allows for the creation of a working image intended to replace a reference image, thereby improving the calculation of the characteristic quantities of sample 1.

[0063] Preferably, the first several series of images use the same values ​​of application angles a.

[0064] A particularly advantageous feature is that the images in the first series are acquired by varying only the application angle α, thus facilitating comparison between bright and dark areas. For example, the images in the first series are acquired without varying the exposure time, the textured light patterns, the textured light color, or the angle of the acquisition means. It is also advantageous for the sample to be at the same temperature for all images in the first series. This simplifies the comparison of the images in the first series.

[0065] To further improve the characterization of the sample, it is advantageous to acquire at least a second series of images in an S2c step. Among the plurality of images, at least a second series of images of the first surface of sample 1 is acquired. The second series of images is acquired when the first surface of sample 1 is illuminated by the first structured light.

[0066] The images in the second series are distinguished from one another by different exposure times of the image capture devices. The brightly lit areas correspond to overexposed areas of the image. Conversely, the dimly lit areas correspond to underexposed areas. By varying the exposure time between the images in the second series, it is possible to obtain images that exhibit artifacts whose surface area changes as a function of the exposure time.

[0067] By reducing the exposure time, the overexposed area shrinks, thus limiting the extent of the artifact and therefore the amount of unusable information. By increasing the exposure time, the underexposed area shrinks, thus limiting the extent of the artifact and therefore the amount of unusable information. By varying the exposure time between the multiple images in the second series, it is easier to track the changing intensity of the repetitive patterns in the first textured light.

[0068] Processing the images from the second series allows for a more precise calculation of the light intensity variations representative of the repeating patterns of structured light. It is then possible to create a working image in which the repeating patterns of structured light are better defined. Advantageously, the light intensity variations of structured light on the sample were extracted from the images in the second series. These light intensity variations are used to define a new image that is free of at least one artifact. Preferably, at least four images with four different exposure times are acquired and used to create a working image.

[0069] In step S3b, the images from the second image set are used to create at least one working image, which is intended to form an additional reference image or replace a reference image. Preferably, the second image set includes at least one of the reference images; preferably, the reference image from the second image set is also common to the first image set, if applicable. This allows the first and second image sets to be used to create a higher-quality working image by reducing the number and / or extent of artifacts.

[0070] It is particularly advantageous to acquire several second sets of images, each with a reference image, and even more preferably, a reference image that is also common to one of the first sets, if applicable. In this way, it is possible to significantly improve the quality of several reference images, which in turn improves the accuracy of the calculation of representative sample values.

[0071] Preferably, the exposure time is shortened compared to the duration used for the reference image in order to facilitate the measurement of the repetitive patterns of structured light on the surface of the substrate which receives the salient patterns 2. This image makes it possible to improve the measurement relative to the flatness of the surface which receives the salient patterns 2.

[0072] Preferably, the exposure time ratio between the shortest and longest durations is at least 1.3, preferably at least 1.6, more preferably at least 2, or even at least 5. The higher the exposure time ratio value, the easier it is to discriminate information between very bright areas and low-light areas, which generally correspond to information about the substrate supporting the salient patterns 2 and information about the reflective areas of the salient patterns 2.

[0073] When acquiring the second series of images, it is particularly advantageous that the other parameters are preferably constant, and in particular the parameters related to structured light, the angle of application and the angle of observation.

[0074] In yet another advantageous embodiment, at least a third series of images is acquired. The process includes a step S2d, which consists of acquiring at least one image of the sample illuminated by a second structured light. The second structured light differs from the structured light used to acquire one of the reference images, and preferably the first series, by the color of the emitted light. The repeating patterns are identical to those of the structured light in at least one reference image, i.e., in phase and frequency. The color modification can be used to improve contrast depending on the materials used to form the salient patterns 2 and / or the substrate that receives the salient patterns 2.

[0075] In step S3c, the images from the third image set are used to create an image that is free of at least one of the artifacts. It is advantageous for the third image set to include one of the reference images. The images from the third image set are then used to create a working image, which is intended to form an additional reference image or to replace one of the reference images.

[0076] It can be advantageous to create several third sets of images, each containing one of the reference images. These multiple third sets of images are then used to replace the reference images.

[0077] The projection devices 4 can be configured to change the color of the patterned light. It is advantageous for the projection devices 4 to be configured to deliver white light with black or grayscale patterns. It is also advantageous for the projection devices 4 to be configured to deliver light of another color, for example, red, blue, green, yellow, orange, or purple with black or grayscale patterns, or of another color.

[0078] In yet another embodiment, at least a fourth series of images is acquired. The process includes a step S2e which consists of acquiring at least one image of the sample illuminated by the first structured light. The images of the fourth series of images are distinguished from the reference images by the use of an acquisition angle that is different from that used by the image capture means 5.

[0079] The image capture means 5 are arranged to acquire at least one image at a second angle β, also called the capture angle, which differs from the first angle. The second angle β is the angle between the optical axis of the image capture means 5 and the axis normal to the support 3, which is intended to support the sample. The capture means are then mounted movably and / or multiple capture means are used. Changing the capture angle allows the position of overexposed and underexposed areas to be modified.

[0080] In an S3d step, the images from the fourth image set are used to create an image that is free of at least one of the artifacts. It is advantageous for the fourth image set to include one of the reference images. The images from the fourth image set are then used to create a working image intended to form an additional reference image or to replace one of the reference images. Defining a pair of capture and application angles allows for the definition of a measurement sensitivity along the A-direction. By modifying either or both of the capture and application angles, it is possible to determine the pair of angles that allows for a measurement sensitivity along the A-direction that is greater than a threshold value, or even the optimal threshold value.However, a drawback of this method is that superimposing the images is more difficult because it requires additional calculations to distort the acquired images to make them more similar to the previous images. In the fourth series of images, it is advantageous to acquire another first series of images, or even a second or third series of images, in order to calculate a different surface area for the salient features. It is then possible to use the two calculated surfaces for the same salient feature under different conditions to improve the accuracy of the calculation.

[0081] In general, the measurement process includes a step S3 which consists of forming at least one working image using the first series of images and advantageously at least one of the second, third and fourth series of images.

[0082] It is advantageous to take a second set of images for the different conditions present in the first set of images, that is, for the different values ​​of the application angle a. The multiple second sets of images improve the quality of the images in the first set, making it possible to create a working image of better quality with only the first set of images or with only the first set of images and a single second set of images.

[0083] Combining the first set of images with one or more second sets of images that share images allows for better elimination of artifacts and / or a greater reduction in the surface area of ​​the artifacts.

[0084] Reference images from a first acquisition, from a replacement by a working image and / or corresponding to an added working image are used as is known to the person skilled in the art to identify the distortions of the repetitive patterns of textured lights in order to calculate at least one quantity representative of the sample.

[0085] Preferably, the images in the first series are compared to identify the brightest and darkest pixels in order to locate reflections and / or refractions. The position of the brightest pixels is advantageous for detecting specular reflections.

[0086] The measuring machine includes a structured light emitter 4 configured to illuminate at least a first surface of a sample 1 with repeating patterns and an image acquisition device 5 configured to capture a plurality of images of the first surface of the sample 1.

[0087] The measuring machine also includes a control circuit 6 configured to define the conditions for acquiring the plurality of images by the image acquisition device 5 and to acquire at least a first image of the first surface of the sample 1. The measuring machine may also include a memory 7 configured to store the plurality of images.

[0088] It is also preferable that the measuring machine includes a first displacement device 8 configured to change the angle of inclination α of the light emitter 4 with respect to the first surface and therefore to the surface of the support 3. Preferably, the measuring machine includes a second displacement device 9 configured to change the angle of inclination of the image acquisition device 5 with respect to the first surface and therefore to the surface of the support 3.

[0089] The control circuit 6 is configured to acquire several reference images of sample 1 illuminated by different structured lights. Each structured light in the reference images defines several repeating patterns and is distinguished from one another by the phase and / or frequency of these repeating patterns. The control circuit is configured to acquire a first series of images of the sample illuminated by a first structured light. The first structured light defines several repeating patterns, and the images in the first series are distinguished from one another by the angle of application α of the first structured light relative to the surface of the support 3.

[0090] The control circuit 6 is configured to compare the images of the first series of images in order to determine the presence of at least one artifact on at least one image of the first series of images.

[0091] A processing circuit 10 is configured to form at least one working image from the first set of images, this working image being free of said at least one artifact or reducing the extent of said at least one artifact. The processing circuit 10 is configured to calculate at least one representative quantity of the sample from the at least one working image and the reference images.

[0092] In a preferred embodiment, sample 1 is placed in an enclosure located within a topographic surveying machine. The enclosure has a first glass surface, and the projector is positioned outside the enclosure. Structured light passes through the transparent part of the enclosure to reach the first surface of sample 1. Image capture means 5 may include a camera, which is arranged outside the enclosure if necessary. The camera is positioned to capture the structured light applied to the surface of sample 1 and propagating through the transparent part of the enclosure, specifically through the second glass surface of the enclosure.

Claims

1. Method for performing topographic measurement that comprises: - providing a sample (1) comprising a first surface provided with a plurality of salient patterns (2); - acquiring a plurality of images of the first surface of the sample (1), the plurality of images comprising reference images of the sample (1), each reference image corresponding to the sample (1) illuminated with a structured light, the structured lights of the reference images each defining several repetitive patterns and differing from one another by the phase and / or the frequency of the repetitive patterns; characterised in that the plurality of images comprises at least one first series of images of the first surface of the sample (1), the first series of images being acquired when the first surface of the sample (1) is illuminated by a first structured light, the first structured light defining several repetitive patterns and the images of the first series of images differing from one another by the application angle (α) of the first structured light with respect to the first surface; - comparing the images of the first series of images so as to determine the presence of at least one artefact on at least one image of the first series of images; - forming at least one working image at least from the first series of images, the at least one working image being devoid of said at least one artefact or reducing the extent of said at least one artefact; - calculating at least one quantity representative of the sample from the at least one working image and the reference images.

2. Method for performing topographic measurement according to claim 1 wherein the first series of images comprises an image of the reference images and wherein the at least one working image replaces said image of the reference images.

3. Method for performing topographic measurement according to claim 2 wherein the plurality of images has several first series of images that each comprise an image of the reference images, wherein each first series of images is used to form a working image devoid of at least one artefact or having an artefact of reduced extent and wherein each working image replaces one of the reference images.

4. Method for performing topographic measurement according to one of claims 1 to 3 wherein the plurality of images comprises at least a second series of images of the first surface of the sample (1), the second series of images being acquired when the first surface of the sample (1) is illuminated by the first structured light, the images of the second series of images differing from one another by their exposure time, the application angle (α) of the first structured light being constant for the second series of images and corresponding to one of the angles of the first series of images, wherein the method for performing measurement comprises: - comparing the images of the second series of images so as to determine the presence of at least one artefact on at least one image of the second series of images and - forming at least one working image at least from the second series of images, the at least one working image being devoid of said at least one artefact or reducing the extent of the artefact.

5. Method for performing topographic measurement according to claim 4 wherein the second series of images comprises an image of the reference images and wherein the at least one working image formed from the second series of images replaces said image of the reference images.

6. Method for performing topographic measurement according to the combination of claim 2 and 5 wherein an image of the reference images belongs to the first series of images and to the second series of images and wherein the first series of images and the second series of images are used to form a working image that replaces said image of the reference images.

7. Method for performing topographic measurement according to one of claims 4 to 6 wherein the plurality of images has several second series of images that each comprise an image of the reference images, wherein each second series of images is used to form a working image devoid of at least one artefact or having an artefact of reduced extent and wherein each working image replaces images of the reference images.

8. Method for performing topographic measurement according to any one of claims 1 to 7 wherein the plurality of images comprises at least a third series of images of the first surface of the sample (1), the third series of images being acquired when the first surface of the sample (1) is illuminated by a second structured light, the images of the third series of images differing from one another by the colour of the structured light, the application angle (α) of the second structured light being constant for the third series of images and corresponding to one of the angles of the first series of images, wherein the method for performing measurement comprises: - comparing the images of the third series of images so as to determine the presence of at least one artefact on at least one image of the third series of images and - forming at least one working image at least from the third series of images, the at least one working image being devoid of said at least one artefact or reducing the extent of the artefact.

9. Method for performing topographic measurement according to any one of claims 1 to 8, wherein comparing the images of the first series of images comprises identifying a displacement of an area presenting a greater luminosity than the mean luminosity of the image of the first series according to the application angle (α) of the first structured light, the artefact being a spurious reflection.

10. Method for performing topographic measurement according to any one of claims 1 to 9, wherein comparing the images of the first series of images comprises identifying a modification of the extent of an area presenting a luminosity that is lower than the mean luminosity of the image of the first series according to the application angle (α) of the first structured light, the artefact being a shadow.

11. Method for performing topographic measurement according to any one of claims 1 to 10, wherein comparing the images of the first series of images comprises extracting the light intensity modifications of the repetitive patterns originating from the structured light and comparing said light intensity modifications.

12. Topographic measuring machine comprising: - a support (3) having a surface designed to support a sample (1); - projection means (4) configured to emit different structured lights, the structured lights each defining several repetitive patterns and differing from one another by the phase and / or the frequency of the repetitive patterns; - the image capture means (5) configured to acquire images of the sample (1) illuminated by one of the different structured lights; - a control circuit (6) configured to acquire several reference images of the sample (1) illuminated by different structured lights, the structured lights of the reference images each defining several repetitive patterns and differing from one another by the phase and / or the frequency of the repetitive patterns. topographic measuring machine characterised in that: - the projection means (4) are configured to define several angles of application (α) of the structured light with respect to the surface of the support (3); - the control circuit (6) is configured to acquire a first series of images of the sample illuminated by the first structured light, the first structured light defining several repetitive patterns and the images of the first series of images differing from one another by the application angle (α) of the first structured light with respect to the surface of the support (3); - the control circuit (6) is configured to compare the images of the first series of images so as to determine the presence of at least one artefact on at least one image of the first series of images; - a processing circuit (10) is configured to form at least one working image at least from the first series of images, the at least one working image being devoid of said at least one artefact or reducing the extent of said at least one artefact; - the processing circuit (10) is configured to calculate at least one quantity representative of the sample from the at least one working image and the reference images.