Method of calibrating a digital image recording device

The method uses a focused energy beam to create calibration markers on a reference plane, eliminating the need for separate targets and enabling efficient, continuous camera calibration in industrial processes, aligning the camera's coordinate system with the manufacturing process for improved monitoring and integration.

GB2636399APending Publication Date: 2025-06-18WAYLAND ADDITIVE LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
GB2023018954
Authority / Receiving Office
GB · GB
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-12-12
Publication Date
2025-06-18

AI Technical Summary

Technical Problem

Existing camera calibration methods in industrial processes, particularly in additive layer manufacturing, require downtime for fiducial marker installation and frequent recalibration, which is inefficient and disruptive.

Method used

A method using a focused energy beam to generate discrete points on a reference plane, which serve as calibration markers, eliminating the need for a separate calibration target and allowing in-process calibration by correlating the camera's coordinate system with the beam deflection system.

Benefits of technology

Enables flexible, rapid, and integrated camera calibration without downtime, aligning the camera's coordinate system with the manufacturing process, facilitating continuous monitoring and improved machine integration.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000000_0000_ABST
    Figure 00000000_0000_ABST
Patent Text Reader

Abstract

A method of calibrating a digital image recording device, for example a thermal imaging or visible light camera (23, 24), comprises generating a beam (13) from a focused energy source (14a) and direct
Need to check novelty before this filing date? Find Prior Art

Description

The present invention relates to a method of calibrating a digital image recording device and to a process of additive layer manufacture including such a method. Digital image recording devices, in particular digital cameras, are used in many technical fields for process monitoring and control and in order to achieve appropriate levels of accuracy it is normally essential for the camera to be calibrated with respect to both intrinsic and extrinsic parameters. The intrinsic parameters include focal length, distortion, skew and image centre and the extrinsic parameters include position and orientation of the camera relative to a specific reference co-ordinate system. The intrinsic and extrinsic parameters are represented by the matrices Mh and Mext in the equation below. The combination of these two matrices forms the camera calibration or projection matrix, which relates the camera pixel location (u, v) to any three-dimensional point (x, y, z) on the reference coordinate system. u SV ^int^ext 111 z -1 r / x o 0 fy LO 0 Ox Oy 1 r12 r22 r32 0 r13 r23 r33 0 tx ty ^Z 1. X y z -1 In order to estimate the parameters in these matrices, an image must be taken of an artefact with features of known location in the reference co-ordinate system. To facilitate identification of these features, it has been usual practice to use an artefact with a planar calibration grid having well-defined features with high contrast such as black circles or lines on a white ground or a checkerboard pattern. The centre point, corners and edges of these patterns, often referred to as fiducial features, are then easily identified by suitable image analysis algorithms. Such a mathematical analysis is capable of transforming the image to a real-world co-ordinate system, the result being an undistorted image of known scale regardless of the position and orientation of the camera. An example of such an imaging and analysis procedure is described in United States Patent Specification US 6437823 B1. A problem with this approach, particularly in the case of an industrial process in which the process is monitored by the camera, is that downtime is required to perform the calibration procedure. Moreover, if the camera is moved or adjusted or if any other configuration changes are made then the procedure will need to be repeated. It is also advisable to perform camera calibration relatively frequently to take account of any drift over longer periods of time. In an industrial context and particularly in the case of multiple cameras, this is undesirable. Accordingly, it would be desirable to free the calibration procedure from reliance on fiducial markers present on an artefact and achieve a more flexible procedure that has scope to be performed by, especially, equipment present in a machine or apparatus involved in a monitored process, for example an additive layer manufacturing machine. To this end, United States Patent Specification US 8049779 B2, which relates to camera calibrating apparatus and an associated method, discloses use of a laser to illuminate reference points of known position and then operation of a camera, which is to be calibrated, to capture an image of the illuminated points. These illuminated points are light reflecting members of a frame, which constitutes a calibration target and has to be located in a camera calibration position. Consequently, the calibration method is still dependent on use and positioning of a separate marker device. Camera calibration in the context of additive layer manufacture involves indexing the camera pixels to a geometrical co-ordinate system so that meaningful spatial measurements can be made from image data. In powder-bed additive layer manufacture an article is produced by selectively melting fusible powder material through irradiation by a scanning laser or electron beam. Irradiation and fusion are carried out in relation to successively deposited layers of the powder material so that material in each layer is melted in accordance with a predefined layer shape and fused not only to itself, but also to any previously fused material of an underlying layer, whereby the article shape is progressively created on a layer-by-layer or additive basis. The layer shapes are normally based on a three-dimensional virtual model which is processed by software to resolve the model into two-dimensional sequential cross-sectional slices. Data defining each slice are then used to direct the beam along a scanning path for melting the powder material within each deposited layer in an area determined by the slice data. Since the size, shape and location of article features as well as the scanning path of the beam are known for each powder layer from the slice data, it would be particularly helpful if the co-ordinate system of the camera could be mapped onto that of the beam deflection system. This would offer significant benefits when using camera data for in-process monitoring and closed loop control. It is therefore the principal object of the present invention to provide a calibration method for a digital image recording device, in particular with use of an energy beam, without requiring positioning of a separate calibration target in a calibration zone. A further object is creation of a calibration method able to exploit an energy beam generated by equipment to perform a process intended to be monitored by the image recording device to be calibrated. Other objects and advantages of the invention will be apparent from the following description. According to a first aspect of the present invention there is provided a method of calibrating a digital image recording device, comprising the steps of generating a beam from a focused energy source and directing the beam towards a surface lying in a reference plane, deflecting the beam to successively produce on the surface a plurality of discrete points of visible or invisible light cumulatively forming a pattern in which each point has known coordinates, obtaining an image of the pattern by a digital image recording device to be calibrated, and calibrating the device with use of the co-ordinates of the points in the image of the pattern. A method exemplifying the invention has the advantage that it dispenses with the need for a separate calibration target and thus, when the image recording device is used for in-process monitoring, eliminates downtime arising from installation and removal of such a target. The points are produced by a deflectable beam, which can be a beam for performing a process monitored by the device, from a focused energy source. In the case of, for example, additive layer manufacturing with a laser or electron beam source it is expected that the beam deflection system will already be calibrated and can be deflected to demanded positions with relatively high accuracy. Thus, instead of using a dedicated calibration target such as a plate with defined features or light reflecting members as in the prior art, a point, especially a spot, of visible or invisible light produced by the beam in the reference plane, particularly on an already present surface lying in that plane, can be used as a marker. Depending on type, the image recording device can detect the marker as light or infrared radiation. The contrast between the spot and, in relative terms, its darker background is sufficient to enable spot position identification by, for example, simple thresholding algorithms. The beam can be progressively deflected to several different locations in emulation of a calibration plate or other target with features functioning as markers. The advantages of this approach are twofold: firstly, it removes the need for a dedicated calibration plate, thus leading to a simpler, quicker and more flexible calibration routine, and, secondly, the device calibration will correspond with that of the beam deflection system, which has benefits for machine integration and especially for augmented reality. In one example of the calibration method, the step of obtaining the image comprises recording an individual image of each of the points by the image recording device and combining the images of the points to form the image of the pattern. This represents a simple procedure, but is subject to limitation of the rate of movement of the beam by the frame rate of the image recording device. Alternatively, the step of obtaining the image comprises recording a single image of all the points by the image recording device with use of a recording time at least equal to the time for deflection of the beam to produce all the points. This procedure provides a faster rate of deflection of the beam in return for a longer exposure time of the image recording device to allow the beam to generate all the points of the pattern. For preference, the deflection of the beam is carried out in accordance with a predetermined co-ordinate system for beam deflection and the step of calibrating the device is carried out to bring a co-ordinate system of the image recording device into agreement with the beam deflection co-ordinate system. The predetermined co-ordinate system can be, in particular, that of equipment for performing a process to be monitored by the image recording device, for example an additive layer manufacturing machine. In that case the equipment system may be a beam deflection system, so that the co-ordinate systems of image recording device and machine will be coincident as a result of the calibration. This is particularly advantageous with regard to integration of the operation of the image recording device into the machine operation. The step of calibrating the image recording device preferably comprises setting intrinsic and extrinsic parameters thereof, so that the device is comprehensively calibrated with respect to itself and its environment. The intrinsic parameters can comprise focal length, distortion, skew and image centre of the image recording device, whilst the extrinsic parameters can comprise the spatial position of the image recording device and the orientation of the device relative to the reference plane. In a convenient and simple example of the method the pattern has the form of a grid, so that the points of the pattern lie at easily definable locations and deflection of the beam can be tailored to a relatively straightforward scanning path. The scanning path is particularly simple if the points are equidistantly spaced in the grid, but irregular spacing is equally possible and may be advantageous in some circumstances. The beam can be an electron beam or a laser beam, the method having the advantage that it is compatible with either type of beam. Thus, the image recording device can be, for example, a camera producing images based on light or on infrared radiation. According to a second aspect of the present invention there is provided an additive layer manufacturing process comprising the steps of successively depositing layers of powder material on a deposit surface of a support in a build area, selectively melting powder material in each layer by a beam from a focused energy source so as to fuse the melted material together and to already fused material of any layer immediately thereunder for layer-by-layer manufacture of a three-dimensional article, the melting being carried out over the build area by deflection of the beam, monitoring the build area by a digital image recording device during at least one of the step of depositing and the step of melting, and calibrating the device by the method according to the first aspect of the invention, in which the beam used in the method of calibrating is subsequently used for the step of melting in the manufacturing process and the surface in the reference plane is the deposit surface or the top surface of a layer or layers on the deposit surface. An additive layer manufacturing process carried out in that way has the advantage that a critical step or critical steps of the process, namely powder deposition and / or powder melting, can be monitored by the image recording device calibrated with respect to a reference plane where the actual deposition or melting is carried out, so that the calibration is linked to a manufacturing procedure or manufacturing procedures. Since the beam deflection will itself normally be calibrated with respect to the powder deposit surface or the top of a layer on that surface, the device calibration and beam calibration can be the same, i.e. based on the same co-ordinate system. Advantageously, the calibration of the device can be carried out in conjunction with raising or lowering the support to position the deposit surface or the layer top surface in the reference plane. The surface can thus be repeatedly returned to a defined position for both device calibration and manufacturing. In a preferred example of the process, in which the layers of powder material correspond with predetermined cross-sectional slices of the article to be manufactured, calibration of the device is by the above-mentioned method option in which the deflection of the beam is carried out in accordance with a predetermined co-ordinate system for beam deflection and the device calibration brings a co-ordinate system of the device into agreement with the beam deflection co-ordinate system, in which case the latter system is predetermined on the basis of co-ordinate data of the slices. The calibration of the image recording device is thus directly based on data relating to the article to be produced, especially data derived from slicing a computer model of the article. A preferred example of a calibration method exemplifying the present invention will now be more particularly described with reference to the accompanying drawings, in which: Fig. 1 is a highly diagrammatic elevation of additive layer manufacturing apparatus in which process monitoring can be carried out by a digital image recording device able to be calibrated by a method exemplifying the invention; and Fig. 2 is a diagram of a pattern of points produced in a reference plane by an energy beam of the apparatus and able to be imaged by the image recording device to enable calibration of the device by way of a method exemplifying the invention. Referring now to the drawings there is shown in Fig. 1 in diagrammatic form additive layer manufacturing apparatus 10 for manufacturing three-dimensional articles of predefinable shapes by selective melting and fusion of powder material, particularly a metallic material, in successively deposited layers in a build zone, the layers forming a powder material bed. The apparatus provides melting by the action of an electron beam directed downwardly along a vertical neutral axis from the top of the apparatus. The beam can be deflected relative to the axis in X and Y directions of a co-ordinate system of the apparatus to provide movement of the point of incidence of the beam on an uppermost powder material layer so as to scan and melt an area of predetermined shape corresponding with an individual cross-sectional layer of an article undergoing manufacture. Apparatus of this kind, which is an example of electron beam scanning apparatus in general, are well-known. Accordingly, more detailed illustration and explanation of parts not essential to an understanding of additive layer manufacturing apparatus are unnecessary. The environment for operation with an electron beam is provided by a housing 11 bounding a vacuum chamber 12 in which a vacuum, as a precondition for propagation of an electron beam 13, can be established. The beam 13 is generated by beam generating and scanning means comprising an electron optical column 14 with an electron gun 14a, the column being located at the top of the housing and oriented to direct the generated beam along the neutral axis towards a target region. The beam 13 can be deflected in the X and Y directions by a beam deflecting arrangement comprising electromagnetic beam deflection coils and can be adjusted with respect to focus and beam spot shape at the target region by a beam influencing arrangement comprising electromagnetic influencing coils, which are positioned around the beam path and electromagnetically influence different parameters of the beam. The various coils are represented schematically by a coil assembly 15 and are individually controllable by a control unit 16, which on a software basis determines desired values of energising current for each of the coils. The apparatus co-ordinate system mentioned above is here a system governing beam deflection by the beam deflecting arrangement. Provided in the vacuum chamber 12 at a spacing below the column 14 and in the target region of the generated beam is a raisable and lowerable support 17, thus a support movable in opposite directions on a Z axis. The support 17 has the form of a table which is guided for vertical stepped movement in a shaft 18, the walls of which confine the powder material bed when present to the table. Above the shaft walls, the table is enclosed by a planar surface 19 which is at the interior side of the housing 11 and with which the top surface of the table or top surface of a powder material bed on the table is generally alignable to substantially lie in a common plane. The area of the top surface of the table represents a planar build zone for start of production of the article from a molten, solidifiable layer of the powder material. The build zone lies in a reference plane 20 for the co-ordinate system of the beam deflecting arrangement and the arrangement is calibrated with respect to that plane, in particular the surface in the build zone area. Further components of the apparatus 10 include a depositor 21 for depositing powder material for the layers forming the bed and a reciprocating spreader 22 for spreading the deposited material to form each of the layers. Associated with or integrated in the apparatus 10 is monitoring system comprising a digital image recording device, such as a thermal imaging camera 23 oriented to cover the build zone in the reference plane. The camera 23 is responsive to infrared radiation produced in the build zone by the electron beam 13. The monitoring system also includes a further camera 24 oriented in the same manner, but responsive to visible light to allow, in particular, imaging of an article undergoing manufacture. The two cameras, which for convenience are schematically shown as halves of a single block, but which may be separately arranged at different locations, serve to monitor different aspects of the manufacturing process and are controlled by and have feedback to the control unit 16 by way of a control and feedback line. It is critical for accurate monitoring that each of the cameras 23 and 24 is accurately calibrated by indexing the camera pixels to a geometrical co-ordinate system, in particular a system associated with the reference plane 20. The camera co-ordinate system can thus be identical with that of the deflecting arrangement for the beam 13. Accurate calibration is carried out by the method explained in the following with reference to Fig. 2. For the purpose of explanation of the calibration method, Fig. 2 shows the reference plane 20, specifically the build zone area of the table surface in that plane, a camera symbol representing either one or both of the cameras 23 and 24, which are both directed towards the reference plane. Also shown in Fig. 2 are the electron gun 14a of the electron optical column 14 and a mirror symbol 15a representing the beam deflection coils forming the beam deflecting arrangement within the coil assembly 15. The electron gun is energised in a low-power mode to generate the beam 13 which propagates in the vacuum in the vacuum chamber 12 towards the reference plane 20 along the neutral axis, in relation to which the beam is deflectable by the beam deflecting arrangement. Deflection is carried out so that the beam 13 scans the plane to successively mark or illuminate discrete points 25 of visible or invisible light on the table surface in the plane, in particular points cumulatively forming a grid pattern in which the points are preferably, but not necessarily, equidistantly spaced. The points 25, which as a consequence of the low beam power are in the form of spots of very small size, have co-ordinates corresponding with a co-ordinate system of the beam deflecting arrangement, which is itself already calibrated on the basis of co-ordinates established by slice data of a computer model of the article to be manufactured, as described generally in the introduction. Each camera 23, 24 directed at the reference plane 20 is activated conjunctively with the electron gun to record an individual image of each point 25, as it is produced, in the pattern or, with a longer exposure time, but a faster rate of beam scanning, a single image of all of the points of the pattern. If individual images are taken these are subsequently merged to produce a single image of the pattern. In the case of an electron beam the images are more readily obtained by the thermal imaging camera 23 responsive to the infrared radiation of the beam 13 at the surface in the reference plane, since this results in a particularly clear contrast between the spot forming each point and the surrounding part of the table surface. However, the generated spots also create at the surface a signature in visible light which is detectable by the visible light camera 24. The pattern image of the points 25, which have known co-ordinates within the co-ordinate system of the beam deflecting arrangement, can then, with use of suitable algorithms, be correlated with a co-ordinate system of the respective camera and corrected so that the systems are in agreement and the camera thereby calibrated. The calibration thus brings the co-ordinate system of the camera into correspondence with that of the apparatus 10 and, by extension, the cross-sectional slice data of an article that the apparatus is programmed to manufacture in a subsequent manufacturing cycle. In an additive manufacturing application this means that the beam and camera systems are all calibrated to one, unified co-ordinate system, which opens up opportunities for overlaying data from multiple sources and permits sophisticated analysis techniques, such as augmented reality. Further, the calibration method entirely dispenses with the need to use a separately installed calibration plate or similar target with own reference markers having known co-ordinates. In the event of subsequent recalibration of the camera due to drift or other factors, a fresh calibration can be quickly and easily performed entirely within the apparatus and with avoidance of additional downtime resulting from installation and removal of a separate target plate or similar artefact providing markers. Repeated calibration can be carried out with minimal interruption to manufacturing time and thus performed more frequently. Although the calibration method has been described by reference to additive layer manufacturing apparatus for article manufacture by an electron beam, the method is equally applicable to apparatus carrying out manufacture by laser beam and also to other apparatus in which a focused energy source is employed.

Claims

1. A method of calibrating a digital image recording device comprises the steps ofgenerating a beam from a focused energy source and directing the beam towards a surface lying in a reference plane,deflecting the beam to successively produce on the surface a plurality of discrete points of visible or invisible light cumulatively forming a pattern in which each point has known coordinates,obtaining an image of the pattern by a digital image recording device to be calibratedand calibrating the device with use of the co-ordinates of the points in the image of the pattern.

2. A method according to claim 1, wherein the step of obtaining the image of the pattern comprises recording an individual image of each of the points by the image recording device and combining the images of the points to form the image of the pattern.

3. A method according to claim 1, wherein the step of obtaining the image of the pattern comprises recording a single image of all the points by the image recording device with use of a recording time at least equal to the time for deflection of the beam to produce all the points.

4. A method according to any one of the preceding claims, wherein the deflection of the beam is carried out in accordance with a predetermined co-ordinate system for beam deflection and the step of calibrating the device is carried out to bring a co-ordinate system of the image recording device into agreement with the beam deflection co-ordinate system.

5. A method according to any one of the preceding claims, wherein the step of calibrating the device comprises setting intrinsic and extrinsic parameters thereof.

6. A method according to claim 5, wherein the intrinsic parameters comprise focal length, distortion, skew and image centre of the image recording device.

7. A method according to claim 5 or claim 6, wherein the extrinsic parameters comprise the position in three dimensions of the image recording device and the orientation thereof relative to the reference plane.

8. A method according to any one of the preceding claims, wherein the pattern is a grid.

9. A method according to claim 8, wherein the points are equidistantly spaced in thegrid.

10. A method according to any one of the preceding claims, wherein the beam is an electron beam or a laser beam.

11. An additive layer manufacturing process comprising the steps ofsuccessively depositing layers of powder material on a deposit surface of a support in a build area,selectively melting powder material in each layer by a beam from a focused energy source so as to fuse the melted material together and to already fused material of any layer immediately thereunder for layer-by-layer manufacture of a three-dimensional article, the melting being carried out over the build area by deflection of the beam,monitoring the build area by a digital image recording device during at least one of the step of depositing and the step of meltingand calibrating the device by the method according to any one of the preceding claims, in which the beam used in the method of calibrating is subsequently used for the step of melting in the manufacturing process and the surface in the reference plane is the deposit surface or the top surface of a layer or layers on the deposit surface.

12. A process according to claim 11, wherein the calibration of the device comprises raising or lowering the support to position the deposit surface or the layer top surface in the reference plane.

13. A method according to claim 11 or 12, wherein the layers of powder material correspond with predetermined cross-sectional slices of the article to be manufactured and calibration of the device is by the method according to claim 4, the beam deflection coordinate system being predetermined on the basis of co-ordinate data of the slices.13

Citation Information

Patent Citations

  • Optical fiducial generation for galvanometric scanner calibration

    US20180281067A1

  • Techniques for optical control calibration in additive fabrication and related systems and methods

    US20220032547A1

  • Calibrating beam generation systems and imaging systems for additive manufacturing

    US20220080668A1

  • Calibration of a camera provided for monitoring an additive manufacturing process

    US20220157346A1