Method for calibrating 3D scanners

Abstract

The invention relates to a method for ascertaining a deviation between the location of a first coordinate axis of a 3D line scanner (scanner) and a second coordinate axis of a transport system, said second coordinate axis corresponding to the first coordinate axis, in a 3D scanning assembly which comprises the scanner and the transport system, wherein - in a scanning step, a calibrating body is scanned by the scanner, and in the scanning step - the scanned profiles in a scanning data set are moved along the first coordinate axis and - the calibrating body and the scanner are moved relative to each other along the second coordinate axis. The calibrating body has at least one calibrating object, and the scanning data set generated in the scanning step contains scanning data relating to the calibrating object. Correction values are ascertained from the scanning data, said correction values indicating the deviation between the position of the first coordinate axis and the second coordinate axis. The scanning data set can be modified using the correction values such that the calibrating object in the scanning data set matches the calibrating object of the calibrating body.

Description

[0001] Procedure for calibrating 3D scanners

[0002] Field of the invention

[0003] The invention relates to a method for calibrating a 3D scanning system, consisting of a 3D scanner and a linear transport system, in order to obtain metrically correct and, if necessary, correctly aligned scan data of an object scanned with the 3D scanner.

[0004] Background of the invention

[0005] 3D scanning arrangements with a 3D line scanner and a transport system are well known.

[0006] Figure 1 shows a schematic diagram of a 3D scanning setup in Figure (a). To scan the object O, the 3D line scanner S is moved along a coordinate axis YT. Alternatively, the object to be scanned can be moved along the coordinate axis YT using the transport system.

[0007] The 3D line scanner S generates consecutive scan profiles that have a predetermined offset from one another. The 3D line scanner S assumes that the offset V or feed of the object to be scanned relative to the 3D line scanner S occurs precisely along a transport direction Ys, i.e., that the transport axis of the coordinate system of the 3D line scanner S and the transport axis of the coordinate system of the transport system are identically aligned. Figure (b) of Fig. 1 shows an example of a scan data set SD generated with identically aligned transport axes of both coordinate systems.

[0008] In practice, however, the transport axis of the coordinate system of the 3D line scanner S and the transport axis of the coordinate system of the transport system are only rarely aligned exactly identically, which leads to incorrect or distorted scan data that are intended to represent the object to be scanned.

[0009] Figure (c) in Fig. 1 shows an example of a scan data set SD generated when the transport axes of both coordinate systems are not identically aligned. Compared to the scan data shown in figure (b), the scan data shown in figure (c) are distorted and do not accurately represent the object.

[0010] Object of the invention

[0011] The object of the present invention is therefore to provide a method with which it is possible to generate scan data of an object which correctly represent the object, despite a deviation of the transport axis of the coordinate system of the 3D line scanner S from the transport axis of the coordinate system of the transport system.

[0012] Solution according to the invention

[0013] This object is achieved by a method according to the independent claims. Advantageous embodiments and further developments of the invention are specified in the respective dependent claims. Accordingly, a method is provided for determining a deviation of the position of a first coordinate axis of a 3D line scanner from a second coordinate axis of a transport system corresponding to the first coordinate axis in a 3D scanning arrangement comprising the 3D line scanner and the transport system, wherein

[0014] - in one scanning step, a calibration body is scanned with the 3D line scanner, whereby in the scanning step

[0015] - the scanned profiles of the 3D line scanner in a scan data set along the first coordinate axis and

[0016] - the calibration body and the 3D line scanner are moved relative to each other along the second coordinate axis, wherein the calibration body has at least two calibration objects arranged at a predetermined distance from each other on the calibration body, and wherein the scanning step generates and provides the scan data set containing scan data on the calibration objects, and

[0017] - in an evaluation step, a number of correction values ​​are determined from the scan data contained in the scan data set relating to the calibration objects, which correction values ​​are indicative of the deviation of the position of the first coordinate axis from the second coordinate axis, wherein the scan data set can be changed using the determined correction values ​​such that the distance between the calibration objects in the scan data set corresponds to the predetermined distance between the calibration objects arranged on the calibration body.

[0018] Using the calibration body and the calibration objects arranged on it, whose distances to one another are known, correction values ​​can be determined with which distorted distances of the calibration objects stored in the scan data set can be corrected, so that the distances of the calibration objects stored in the scan data set correspond to the distance of the calibration objects on the calibration body. The same correction values ​​can then be used for this 3D scanning arrangement to eliminate distortion in the scan data of other scanned objects. It can be advantageous if at least three calibration objects are arranged at predetermined positions on the calibration body. This advantageously also allows the global position of the calibration objects in space to be determined and taken into account when determining the correction values.

[0019] In one embodiment of the invention, the correction values ​​can comprise shear parameters with which the scan data set can be equalized such that the distance between the calibration objects in the scan data set corresponds to the predetermined distance between the calibration objects arranged on the calibration body.

[0020] In one embodiment of the invention, the evaluation step may comprise an extraction step in which the scan data relating to the calibration objects are extracted from the scan data set, with a portion of the extracted scan data being assigned to each calibration object arranged on the calibration body.

[0021] The assignment can be carried out by segmenting the extracted scan data, whereby the segmentation results in a number of point clouds, each point cloud being assigned to exactly one calibration object arranged on the calibration body.

[0022] In the evaluation step, a unique point can be determined for each part of the extracted scan data, each of which represents a calibration object arranged on the calibration body.

[0023] The unique point can represent a center point of the respective calibration object arranged on the calibration body. In one embodiment of the invention, the correction values ​​can be determined iteratively by minimizing a predetermined first error function Fl with respect to the distances of the calibration objects in the scan data set relative to the distances of the calibration objects arranged on the calibration body.

[0024] It has been found to be advantageous if the predetermined first error function Fl = Sij (AK(i,j) - AS(i,j)) 2is, with AK(i,j) equal to the distance of the i-th calibration object from the j-th calibration object of the calibration body and with AS(i,j) equal to the distance of the i-th calibration object from the j-th calibration object in the scan data set.

[0025] In one embodiment of the invention, the scan data set can be changed using the determined correction values ​​so that the shape of the calibration objects in the scan data set corresponds to predetermined shapes of the calibration objects arranged on the calibration body.

[0026] In this case, it may be advantageous if, when determining the correction values, a predetermined second error function F2 is minimized with regard to the deviations of the shape of the calibration objects contained in the scan data set from the respective predetermined shape of the calibration object arranged on the calibration body.

[0027] It has been found to be advantageous if the second error function F2 = Ei (Form(Ki) - Form(Si)) 2is, with Form(Ki) equal to the shape of the i-th calibration object of the calibration body and with Form(Si) equal to the shape of the i-th calibration object in the scan data set.

[0028] In one embodiment of the invention, a total error function can be determined from the first error function F1 and from the second error function F2, preferably a weighted average of the first error function F1 and the second error function F2. The invention further provides a method for determining a deviation of the position of a first coordinate axis of a 3D line scanner from a second coordinate axis of a transport system corresponding to the first coordinate axis in a 3D scanning arrangement comprising the 3D line scanner and the transport system, wherein

[0029] - in one scanning step, a calibration body is scanned with the 3D line scanner, whereby in the scanning step

[0030] - the scanned profiles of the 3D line scanner along the first coordinate axis and

[0031] - the calibration body and the 3D line scanner are moved relative to each other along the second coordinate axis, wherein the calibration body has at least one calibration object arranged on the calibration body, wherein the calibration object has a predetermined shape, and wherein the scanning step generates and provides a scan data set containing scan data on the at least one calibration object, and

[0032] - in an evaluation step, a number of correction values ​​are determined from the scan data contained in the scan data set relating to the calibration object, which correction values ​​are indicative of the deviation of the position of the first coordinate axis from the second coordinate axis, wherein the scan data set can be changed using the determined correction values ​​such that the shape of the calibration object in the scan data set corresponds to the shape of the calibration object arranged on the calibration body.

[0033] The evaluation step may comprise an extraction step in which the scan data relating to the calibration object are extracted from the scan data set, the extracted scan data being assigned to the calibration object arranged on the calibration body.

[0034] The extracted scan data can contain a point cloud, with the point cloud being assigned to the calibration object arranged on the calibration body. It can be advantageous if, when determining the correction values, a predetermined second error function F2 is minimized with respect to the deviation of the shape of the calibration object contained in the scan data set from the predetermined shape of the calibration object arranged on the calibration body.

[0035] It has been found to be advantageous if, when several calibration objects are arranged on the calibration body, the second error function

[0036] F2 = Egg (Form(Ki) - Form(Si)) 2 where Form(Ki) is equal to the shape of the i-th calibration object of the calibration body, and Form(Si) is equal to the shape of the i-th calibration object in the scan data set. It can be advantageous to determine correction values ​​iteratively.

[0037] In one embodiment of the invention, the calibration body can have at least two calibration objects arranged at a predetermined distance from one another, wherein the scan data set can be changed with the determined correction values ​​such that the distance from one another of the calibration objects in the scan data set corresponds to the predetermined distance from one another of the calibration objects arranged on the calibration body.

[0038] Optionally, three calibration objects can be arranged at predetermined positions on the calibration body. This advantageously allows the global position of the calibration objects in space to be determined and taken into account when determining the correction values.

[0039] The correction values ​​can include shear parameters that can be used to equalize the scan data set so that the distance between the calibration objects in the scan data set matches the predetermined distance between the calibration objects arranged on the calibration body. In the extraction step, a portion of the extracted scan data can be assigned to each calibration object arranged on the calibration body.

[0040] It may be advantageous if the assignment is carried out by means of a segmentation of the extracted scan data, whereby a number of point clouds are obtained by the segmentation, whereby each point cloud is assigned to exactly one calibration object arranged on the calibration body.

[0041] In the evaluation step, a unique point can be determined for each part of the extracted scan data, each of which represents a calibration object arranged on the calibration body.

[0042] It is advantageous if the unique point represents a center point of the respective calibration object arranged on the calibration body.

[0043] In one embodiment of the invention, the correction values ​​can be determined iteratively by minimizing a predetermined first error function Fl with respect to the distances of the calibration objects in the scan data set relative to the distances of the calibration objects arranged on the calibration body.

[0044] It may be advantageous if the predetermined first error function Fl = Sij (AK(i,j) - AS(i,j)) 2 is, with AK(i,j) equal to the distance of the i-th calibration object from the j-th calibration object of the calibration body and with AS(i,j) equal to the distance of the i-th calibration object from the j-th calibration object in the scan data set.

[0045] It may be advantageous to determine a total error function from the first error function F1 and the second error function F2, preferably a weighted mean of the first error function F1 and the second error function F2. In one embodiment of the invention, a transformation step can be carried out after the evaluation step, wherein a transformation matrix is ​​determined in the transformation step, wherein the scan data set modified with the correction values ​​is aligned with the transformation matrix into a coordinate system defined by the calibration body.

[0046] The invention further provides a method for scanning an object with a 3D scanning arrangement comprising a 3D line scanner and a transport system, wherein during the scanning process the object is moved along the second coordinate axis of the transport system relative to the first coordinate axis of the 3D line scanner, wherein the scan data of the object obtained during the scanning process are adjusted with correction values, and wherein the correction values ​​are determined using one of the above-mentioned methods according to the invention.

[0047] Short description of the characters

[0048] Details and features of the invention, as well as specific embodiments of the invention, will become apparent from the following description taken in conjunction with the drawing. It shows:

[0049] Fig. 1 a schematic diagram of a 3D scanning arrangement in Figure (a), an example of a correct scan data set in Figure (b) and an example of a scan data set with distorted scan data in Figure (c)

[0050] Fig. 2 a schematic diagram of a calibration body with a calibration object;

[0051] Fig. 3 a schematic diagram of a calibration body with three calibration objects;

[0052] Fig. 4 actual positions and shapes of three calibration objects on the calibration body in Figure (a), the positions and shapes of the three calibration objects from Figure (a) in the scan data before the correction of the scan data in Figure (b), and the positions and shapes of the three calibration objects from Figure (a) in the scan data after the correction of the scan data in Figure (c);

[0053] Fig. 5 is a flowchart showing the essential steps of the method according to the invention in a first embodiment;

[0054] Fig. 6 is a flowchart showing the essential steps of the method according to the invention according to a second embodiment; and

[0055] Fig. 7 is a flowchart showing the essential steps of the method according to the invention in a third embodiment;

[0056] Detailed description of the invention

[0057] The method according to the invention is based on a calibration body on which a calibration object is arranged or, according to a further embodiment of the method, on which at least two calibration objects are arranged at a distance from one another. The exact geometry of the calibration body and the calibration objects arranged on it is known. It is also assumed that the individual scan profiles of a scanning process are metrically correct.

[0058] Fig. 2 shows a schematic diagram of a calibration body KK with a calibration object Ko mounted on it. The calibration body KK consists of a support on which a calibration object Ko, designed as a sphere, is mounted. The size of the sphere is known or predetermined. Instead of a sphere, any other geometric shape can be chosen as the calibration object Ko, such as a cube or a tetrahedron.

[0059] Fig. 3 shows a schematic diagram of a calibration body KK with several (here three) calibration objects Ko arranged on it. The calibration body KK consists of a carrier. The three calibration objects Ko are each made of spheres and arranged on the carrier. The size of the spheres is known or predetermined, although it is also possible for each sphere to have a different diameter. Providing different diameters for the spheres can be advantageous in one embodiment of the method according to the invention in order to simplify correct assignment of the sphere data contained in the scan data set to the individual calibration objects Ko. Instead of spheres as calibration objects Ko, any other geometric shape can be selected here, such as cubes or tetrahedrons, although different geometric shapes can also be provided, such as a sphere, a cube, and a tetrahedron.The shape of the calibration objects Ko is preferably selected so that a unique point (e.g. center point) can be determined for each calibration object Ko from the data on the calibration objects Ko contained in the scan data.

[0060] In the case of a calibration body KK with several calibration objects Ko arranged on it, the distances d between the calibration objects Ko are also known, whereby the distance d can be the distance between the centers of two calibration objects Ko.

[0061] Both the calibration body KK shown in Fig. 2 and the calibration body KK shown in Fig. 3 can be designed as a plate on which the calibration objects Ko are arranged. Instead of a plate, any other support structure can be used that is suitable for arranging the calibration objects Ko on the support structure in such a way that the position of the individual calibration objects Ko relative to each other does not change during intended use.

[0062] In the embodiment shown in Fig. 2 with only one calibration object Ko, the calibration object Ko can simultaneously also form the calibration body KK. A calibration object Ko configured in this way can, for example, be a flattened sphere or a cube. According to the method according to the invention, the calibration body KK with the calibration objects Ko arranged thereon is used to determine correction values ​​from the scan data generated by scanning the calibration body KK, with which these scan data can be modified such that the modified data of the calibration objects in the scan data correspond to the calibration objects Ko of the calibration body KK. "Correspond" in this context means that

[0063] (a) the calibration objects in the scan data set each have the same size and shape as the corresponding calibration objects Ko of the calibration body KK, or

[0064] (b) the calibration objects in the scan data set are each at the same distance from each other as the corresponding calibration objects Ko of the calibration body KK, or

[0065] (c) the calibration objects in the scan data set satisfy conditions (a) and (b).

[0066] The correction values ​​thus determined can then be used in a scanning process in which a real object O is scanned to modify the associated scan data set. According to the invention, the correction values ​​are used to generate modified, i.e., corrected, scan data from the scan data, in which the distortion resulting from the deviation of the transport axis of the coordinate system of the 3D line scanner S from the transport axis of the coordinate system of the transport system T is eliminated.

[0067] First, a first embodiment of the method according to the invention is described. Subsequently, a second embodiment of the method according to the invention is described. Both embodiments of the method according to the invention can be combined with one another.

[0068] In the following, the coordinate axis Ys of the 3D line scanner S is referred to as the first coordinate axis and the coordinate axis YT of the transport system T is referred to as the second coordinate axis.

[0069] As explained above, these two coordinate axes are each the axis along which the 3D line scanner offsets the scanned profiles, or the object to be scanned or the calibration body KK to be scanned is moved. For the method according to the invention, it is irrelevant whether only the 3D line scanner is moved relative to the object / calibration body, only the object / calibration body relative to the 3D line scanner, or whether both the 3D line scanner and the object / calibration body are moved. If both the 3D line scanner and the object / calibration body are moved, the relative movement to each other results in a transport movement that can be treated as the transport direction. As stated above, the offset V between two scan profiles is known, regardless of how the 3D line scanner and the object / calibration body are moved relative to each other along their respective transport axes.

[0070] A first embodiment of the method according to the invention is described in more detail below with reference to Fig. 5.

[0071] In a first step Sl.l (scanning step) of the process, a calibration body KK is scanned using the 3D line scanner S. The 3D line scanner S and the calibration body KK are moved relative to each other along the coordinate axis YT.

[0072] According to the first embodiment of the method according to the invention, the calibration body KK has at least two calibration objects Ko arranged at a predetermined distance from one another and at predetermined positions on the calibration body KK, as described with reference to Fig. 3. In step S 1.1, a scan data set is generated and made available for further processing in the method according to the invention. The scan data set contains scan data for the calibration objects Ko. If the alignment of the first coordinate axis Ys deviates from the second coordinate axis YT, the scan data contained in the scan data set are distorted, as explained with reference to figure (b) of Fig. 4. When the scan data are visualized, the visualized calibration body would then be displayed distorted compared to the physical calibration body KK.

[0073] In the subsequent step S1.2 (evaluation step) of the method, correction values ​​are determined from the scan data contained in the scan data set for the calibration objects Ko. The determined correction values ​​are indicative of a deviation of the orientation of the first coordinate axis Ys from the orientation of the second coordinate axis YT. Using these correction values, the scan data set can be modified such that the distance between the calibration objects Ko in the scan data set corresponds to the predetermined distance between the calibration objects Ko arranged on the physical calibration body KK, as shown with reference to illustration (c) of Fig. 4. The determined correction values ​​can therefore be used to rectify the scan data contained in the scan data set in order to eliminate the distortion of the scan data caused by the different orientation of the two coordinate axes.

[0074] If the alignment of the two coordinate axes is not changed for a further scanning process, for example for scanning another object, the scan data for this object will also show a distortion, which can be corrected using the correction values ​​determined for the calibration body KK.

[0075] As a result, the correction values ​​are used to perform a translational correction of the positions of the points in the scan data set, which ultimately depends only on the coordinate of the transport axis Ys. Each scanned point is thus shifted in such a way that the resulting distortion is eliminated.

[0076] Mathematically, this translation corresponds to a matrix multiplication with a shear matrix M (shear parameter), which offsets the X, Y, and Z coordinates of the points in the scan data set depending on Y:

[0077] 1 cx 0'

[0078] M = 0 1 + cy 0

[0079] 0 cz 1. with cx, cy, cz = correction values ​​of the respective coordinate.

[0080] The evaluation step S1.2 can include an extraction step in which the scan data for the calibration objects Ko are extracted from the scan data set. A portion of the extracted scan data is assigned to each calibration object Ko arranged on the calibration body KK. From the scan points contained in the scan data, those points that can be assigned to a calibration object Ko are extracted and each assigned to a calibration object Ko. This has the advantage that the entire scan data set is no longer used to determine the correction values, but rather only those scan data that can be assigned to a calibration object Ko. This enables faster determination of the correction values ​​and leads to more precise correction values.

[0081] The assignment itself can be performed by segmenting the extracted scan data. This segmentation results in a number of point clouds, with each point cloud being assigned to exactly one calibration object Ko arranged on the calibration body KK. Each point cloud therefore represents a calibration object Ko of the calibration body KK, albeit in a distorted form (if the alignment of the two transport axes differs). The point clouds obtained by segmentation can be optimized as needed, for example, by removing outliers to smooth the surface of the point clouds.

[0082] In the evaluation step S1.2, a unique point can be determined for each part of the extracted scan data or for each point cloud, each of which represents a calibration object Ko arranged on the calibration body KK. A calibration object Ko is thus no longer represented by a number of scan points, but by a single or designated point. The unique point can, for example, be the center of a point cloud, which can be determined from the points in the point cloud. The prior removal of outliers can lead to a more precise unique point, as these are ignored when determining the unique point. Other types of unique points can also be used.

[0083] The distance between the individual calibration objects Ko in the scan data set can now be determined based on the respective unique points. As explained above, the correction values ​​are then determined based on the distance values ​​thus determined. Applying the determined correction values ​​to the scan data of the scan data set now results in the distances between the unique points in the scan data set corresponding to the distances between the corresponding unique points of the calibration objects Ko in the calibration body KK.

[0084] In one embodiment, the correction values ​​can be determined iteratively by minimizing a predetermined first error function Fl with respect to the distances of the calibration objects Ko in the scan data set relative to the distances of the calibration objects Ko arranged on the calibration body KK.

[0085] The first error function can be Fl = Eij (AK(i,j) - AS(i,j)) 2 be used with

[0086] AK(i,j) equals the distance of the i-th calibration object from the j-th calibration object of the calibration body and

[0087] AS(i,j) equals the distance of the i-th calibration object from the j-th calibration object in the scan data set

[0088] It is advantageous if, in each iteration, the point clouds are re-extracted from the scan dataset that has been adjusted according to the previous iteration, and the unique points are re-determined for the newly extracted point clouds. This can be repeated until the sum of squares of the errors according to the above error function falls below a certain threshold or becomes minimal.

[0089] For further optimization of the correction values, it may be provided to carry out the procedure described above with the calibration body KK several times and to generate final correction values ​​from the correction values ​​thus determined several times.

[0090] The correction values ​​can now be saved for further use.

[0091] A second embodiment of the method according to the invention is described in more detail below with reference to Fig. 6.

[0092] In a first step S1.1 (scanning step) of the method, a calibration body KK is scanned using the 3D line scanner S. The 3D line scanner S and the calibration body KK are moved relative to one another along the coordinate axis YT. According to the second embodiment of the method according to the invention, the calibration body KK has at least one calibration object Ko arranged on the calibration body KK, as described with reference to Fig. 2. However, several calibration objects Ko can also be arranged on the calibration body KK, which can increase the accuracy of the subsequently determined correction values. In principle, the method according to the second embodiment can also be carried out with a single calibration object Ko.

[0093] The method according to the second embodiment is described below using a calibration object Ko arranged on the calibration body KK.

[0094] In step S2.1, a scan data set is generated and made available for further processing in the method according to the invention. The scan data set contains scan data relating to the calibration object Ko. If the alignment of the first coordinate axis Ys deviates from the second coordinate axis YT, the scan data contained in the scan data set are distorted, as explained with reference to figure (b) of Fig. 4, wherein figure (b) of Fig. 4 shows scan data relating to three calibration objects Ko. If the scan data were visualized, the visualized calibration body would then be displayed distorted compared to the physical calibration body KK.

[0095] In the subsequent step S2.2 (evaluation step) of the method, correction values ​​are determined from the scan data contained in the scan data set for the calibration object Ko. The determined correction values ​​are indicative of a deviation of the orientation of the first coordinate axis Ys from the orientation of the second coordinate axis YT. Using these correction values, the scan data set can be modified such that the shape of the calibration object in the scan data set corresponds to the shape of the calibration object Ko arranged on the calibration body KK, as shown with reference to illustration (c) of Fig. 4. The determined correction values ​​can therefore be used to rectify the scan data contained in the scan data set in order to eliminate the distortion of the scan data caused by the different orientation of the two coordinate axes.

[0096] If the alignment of the two coordinate axes is not changed for a further scanning process, for example for scanning another object, the scan data for this object will also show a distortion, which can be corrected using the correction values ​​determined for the calibration body KK.

[0097] As a result, the correction values ​​are used to perform a translational correction of the positions of the points in the scan data set, which ultimately depends only on the coordinate of the transport axis. Each scanned point is thus shifted in such a way that the resulting distortion is eliminated, resulting in the shape of the calibration object in the scan data set corresponding to the shape of the calibration object Ko on the calibration body KK.

[0098] Mathematically, this translation corresponds to a matrix multiplication with a shear matrix M (shear parameter), which offsets the X, Y, and Z coordinates of the points in the scan data set depending on Y: with cx, cy, cz = correction values ​​of the respective coordinate.

[0099] The evaluation step S2.2 can include an extraction step in which the scan data for the calibration object is extracted from the scan data set. The extracted scan data is assigned to the calibration object Ko arranged on the calibration body KK. From the scan points contained in the scan data, those points that can be assigned to the calibration object Ko are extracted and assigned to the calibration object Ko. This has the advantage that the entire scan data set is no longer used to determine the correction values, but only those scan data that can be assigned to the calibration object Ko. This enables faster determination of the correction values ​​and leads to more precise correction values.

[0100] The assignment itself can be performed by segmenting the extracted scan data. Segmentation is particularly advantageous when the calibration body KK has multiple calibration objects Ko, but can also be performed for a calibration body KK with only one calibration object Ko.

[0101] The segmentation results in a point cloud (if there are multiple calibration objects Ko, a number of point clouds are obtained), which is assigned to the calibration object Ko arranged on the calibration body KK (if there are multiple calibration objects Ko, each point cloud is assigned to exactly one calibration object Ko arranged on the calibration body KK). Each point cloud therefore represents a calibration object Ko of the calibration body KK, albeit in a distorted form (if the alignment of the two transport axes differs).

[0102] The point cloud obtained through segmentation can be optimized as needed, for example, by removing outliers to smooth the surface of the point cloud. If multiple point clouds are present, all point clouds can be optimized accordingly.

[0103] In one embodiment, the correction values ​​can be determined iteratively by minimizing a predetermined second error function F2 with respect to the deviation of the shape of the calibration object contained in the scan data set from the predetermined shape of the calibration object (Ko) arranged on the calibration body (KK).

[0104] If several calibration objects Ko are arranged on the calibration body KK, the second error function F2 = Ei (Form(Ki) - Form(Si)) 2 be used with

[0105] Form(Ki) equals the shape of the i-th calibration object of the calibration body and Form(Si) equals the shape of the i-th calibration object in the scan data set

[0106] It is advantageous if, in each iteration, the point cloud (or the point clouds in the case of multiple calibration objects Ko) is extracted again from the scan data set, which was adjusted according to the previous iteration, and the shape of the calibration object Ko(s) in the scan data set is re-determined for the newly extracted point cloud (or point clouds). This can be repeated until the second error function, or in the case of multiple calibration objects Ko, the error sum of squares according to the above error function, falls below a certain threshold or becomes minimal.

[0107] For a further optimization of the correction values, it can be provided to carry out the method described above according to the second embodiment with the calibration body KK several times and to generate final correction values ​​from the correction values ​​thus determined several times.

[0108] The correction values ​​can now be saved for further use.

[0109] A third embodiment of the method according to the invention is described in more detail below with reference to Fig. 7.

[0110] According to a third embodiment of the method according to the invention, the two embodiments described above can be combined. In the first embodiment of the method according to the invention, in addition to both steps S1.1 and S1.2, step S2.2 of the method according to the second embodiment is also carried out.

[0111] In the second embodiment of the method according to the invention, in addition to both steps S2.1 and S2.2, step S1.2 of the method according to the first embodiment is also carried out, but only if at least two calibration objects Ko are arranged on the calibration body KK, between which the distance can be corrected according to the first error function.

[0112] According to the third embodiment of the method according to the invention, correction values ​​are determined with which the scan data set can be changed so that both

[0113] - the shape of the calibration objects Ko in the scan data set with predetermined shapes of the calibration objects Ko arranged on the calibration body KK, as well as

[0114] - the distances between the calibration objects Ko in the scan data set correspond to the predetermined distances between the calibration objects Ko arranged on the calibration body KK.

[0115] The two error functions Fl and F2 mentioned above can be minimized, for example by executing the procedure iteratively.

[0116] It may be advantageous to determine an overall error function from the first error function F1 and the second error function F2, which is then minimized. The overall error function can, for example, be a weighted average of the first error function F1 and the second error function F2.

[0117] In all embodiments of the method according to the invention described above, a transformation step S3 can be performed after the evaluation step, in which a transformation matrix is ​​determined. Using the transformation matrix, the scan data set modified or corrected with the correction values ​​can be aligned into a coordinate system defined by the calibration body KK.

[0118] Reference symbol:

[0119] KK calibration body Ko calibration objects

[0120] L laser lines

[0121] O real object to be scanned

[0122] S 3D line scanner

[0123] SD scan data set Sl.l, S1.2, S2.1, S2.2, S3

[0124] Steps of the procedure

[0125] T transport system

[0126] V Offset between two scan profiles

[0127] Ys first coordinate axis Ys YT second coordinate axis YT

[0128] X, Z further axes of the scanner S or the transport system T

Claims

Claims 1. Method for determining a deviation of the position of a first coordinate axis (Ys) of a 3D line scanner (S) from a second coordinate axis (YT) of a transport system (T) corresponding to the first coordinate axis (Ys) in a 3D scanning arrangement comprising the 3D line scanner and the transport system, wherein - in one scanning step, a calibration body (KK) is scanned with the 3D line scanner (S), wherein in the scanning step - the scanned profiles of the 3D line scanner (S) in a scan data set along the first coordinate axis (Ys) and - the calibration body (KK) and the 3D line scanner (S) are moved relative to each other along the second coordinate axis (YT), wherein the calibration body (KK) has at least two calibration objects (Ko) arranged at a predetermined distance from each other on the calibration body (KK), and wherein the scanning step generates and provides the scan data set containing scan data on the calibration objects (Ko), and - in an evaluation step, a number of correction values ​​are determined from the scan data contained in the scan data set relating to the calibration objects (Ko), which correction values ​​are indicative of the deviation of the position of the first coordinate axis (Ys) from the second coordinate axis (YT), wherein the scan data set can be changed using the determined correction values ​​such that the distance between the calibration objects (Ko) in the scan data set corresponds to the predetermined distance between the calibration objects (Ko) arranged on the calibration body (KK).

2. Method according to the preceding claim, wherein the correction values ​​comprise shear parameters with which the scan data set can be equalized such that the distance between the calibration objects (Ko) in the scan data set corresponds to the predetermined distance between the calibration objects (Ko) arranged on the calibration body (KK).

3. Method according to one of the two preceding claims, wherein the evaluation step comprises an extraction step in which the scan data relating to the calibration objects (Ko) are extracted from the scan data set, a part of the extracted scan data being assigned to each calibration object (Ko) arranged on the calibration body (KK).

4. Method according to the preceding claim, wherein the assignment is carried out by means of a segmentation of the extracted scan data, wherein a number of point clouds are obtained by the segmentation, wherein each point cloud is assigned to exactly one calibration object (Ko) arranged on the calibration body (KK).

5. Method according to one of the preceding claims 3 or 4, wherein in the evaluation step a unique point is determined for each part of the extracted scan data, each point representing a calibration object (Ko) arranged on the calibration body (KK).

6. Method according to the preceding claim, wherein the unique point represents a center point of the respective calibration object (Ko) arranged on the calibration body (KK).

7. Method according to one of the preceding claims, wherein the correction values ​​are determined iteratively by a predetermined first error function tion Fl is minimized with regard to the distances of the calibration objects (Ko) in the scan data set relative to the distances of the calibration objects (Ko) arranged on the calibration body (KK).

8. Method according to the preceding claim, wherein the predetermined first error function Fl = Sij (AK(i,j) - AS(i,j)) 2 is with AK(i,j) equal to the distance of the i-th calibration object from the j-th calibration object of the calibration body and with AS(i,j) equal to the distance of the i-th calibration object from the j-th calibration object in the scan data set.

9. Method according to one of the preceding claims, wherein the scan data set can be changed with the determined correction values ​​such that the shape of the calibration objects (Ko) in the scan data set corresponds to predetermined shapes of the calibration objects (Ko) arranged on the calibration body (KK).

10. Method according to the preceding claim, wherein, when determining the correction values, a predetermined second error function F2 is minimized with regard to the deviations of the shape of the calibration objects contained in the scan data set from the respective predetermined shape of the calibration object (Ko) arranged on the calibration body (KK).

11. Method according to the preceding claim, wherein the second error function F2 = Ei (Form(Ki) - Form(Si)) 2 is, with Form(Ki) equal to the shape of the i-th calibration object of the calibration body and with Form(Si) equal to the shape of the i-th calibration object in the scan data set.

12. Method according to one of the two preceding claims, wherein an overall error function is determined from the first error function Fl and from the second error function F2, preferably a weighted mean of the first error function Fl and the second error function F2.

13. Method for determining a deviation of the position of a first coordinate axis (Ys) of a 3D line scanner (S) from a second coordinate axis (YT) of a transport system (T) corresponding to the first coordinate axis (Ys) in a 3D scanning arrangement comprising the 3D line scanner and the transport system, wherein - in one scanning step, a calibration body (KK) is scanned with the 3D line scanner (S), wherein in the scanning step - the scanned profiles of the 3D line scanner (S) along the first coordinate axis (Ys) and - the calibration body (KK) and the 3D line scanner (S) are moved relative to each other along the second coordinate axis (YT), wherein the calibration body (KK) has at least one calibration object (KO) arranged on the calibration body (KK), wherein the calibration object (KO) has a predetermined shape, and wherein the scanning step generates and provides a scan data set containing scan data on the at least one calibration object (KO), and - in an evaluation step, a number of correction values ​​are determined from the scan data contained in the scan data set relating to the calibration object (Ko), which correction values ​​are indicative of the deviation of the position of the first coordinate axis (Ys) from the second coordinate axis (YT), wherein the scan data set can be changed using the determined correction values ​​such that the shape of the calibration object (Ko) in the scan data set corresponds to the shape of the calibration object (Ko) arranged on the calibration body (KK).

14. The method according to claim 13, wherein the evaluation step comprises an extraction step in which the scan data relating to the calibration object (Ko) are extracted from the scan data set, the extracted scan data being assigned to the calibration object (Ko) arranged on the calibration body (KK).

15. Method according to the preceding claim, wherein extracted scan data contain a point cloud, wherein the point cloud is assigned to the calibration object (Ko) arranged on the calibration body (KK).

16. Method according to one of claims 13 to 15, wherein, when determining the correction values, a predetermined second error function F2 is minimized with regard to the deviation of the shape of the calibration object contained in the scan data set from the predetermined shape of the calibration object (Ko) arranged on the calibration body (KK).

17. Method according to the preceding claim, wherein, in the case of several calibration objects (Ko) arranged on the calibration body, the second error function F2 = Ei (Form(Ki) - Form(Si)) 2 is, with Form(Ki) equal to the shape of the i-th calibration object of the calibration body and with Form(Si) equal to the shape of the i-th calibration object in the scan data set.

18. Method according to one of claims 13 to 17, wherein the calibration body (KK) has at least two calibration objects (Ko) arranged at a predetermined distance from one another, wherein the scan data set can be changed with the determined correction values ​​such that the distance from one another of the calibration objects (Ko) in the scan data set corresponds to the predetermined distance from one another of the calibration objects (Ko) arranged on the calibration body (KK).

19. Method according to the preceding claim, wherein the correction values ​​comprise shear parameters with which the scan data set can be equalized such that the distance between the calibration objects (Ko) in the scan data set corresponds to the predetermined distance between the calibration objects (Ko) arranged on the calibration body (KK).

20. Method according to one of the two preceding claims, wherein in the extraction step a part of the extracted scan data is assigned to each calibration object (Ko) arranged on the calibration body (KK).

21. Method according to the preceding claim, wherein the assignment is carried out by means of a segmentation of the extracted scan data, wherein a number of point clouds are obtained by the segmentation, wherein each point cloud is assigned to exactly one calibration object (Ko) arranged on the calibration body (KK).

22. Method according to one of the preceding claims 18 to 21, wherein in the evaluation step a unique point is determined for each part of the extracted scan data, each point representing a calibration object (Ko) arranged on the calibration body (KK).

23. Method according to the preceding claim, wherein the unique point represents a center point of the respective calibration object (Ko) arranged on the calibration body (KK).

24. Method according to one of the preceding claims 18 to 23, wherein the correction values ​​are determined iteratively by minimizing a predetermined first error function Fl with respect to the distances of the calibration objects (Ko) in the scan data set relative to the distances of the calibration objects (Ko) arranged on the calibration body (KK).

25. Method according to the preceding claim, wherein the predetermined first error function Fl = Sij (AK(i,j) - AS(i,j)) 2is with AK(i,j) equal to the distance of the i-th calibration object from the j-th calibration object of the calibration body and with AS(i,j) equal to the distance of the i-th calibration object from the j-th calibration object in the scan data set.

26. Method according to one of the two preceding claims, wherein an overall error function is determined from the first error function Fl and from the second error function F2, preferably a weighted mean of the first error function Fl and the second error function F2.

27. Method according to one of the preceding claims, wherein after the evaluation step a transformation step is carried out, wherein in the transformation step a transformation matrix is ​​determined, wherein with the transformation matrix the scan data set modified with the correction values ​​is aligned in a coordinate system defined by the calibration body (KK).

28. Method for scanning an object with a 3D scanning arrangement comprising a 3D line scanner (S) and a transport system (T), wherein during the scanning process the object is moved along the second coordinate axis (YT) of the transport system (T) relative to the first coordinate axis (Ys) of the 3D line scanner (S), wherein the scan data of the object obtained during the scanning process are adjusted with correction values, and wherein the correction values ​​are determined using a method according to one of claims 1 to 27.

Images