Three-dimensional measurement system and method for correcting thereof
By capturing images of reference objects and calculating the phase height conversion coefficient using a phase error model, the problem of complex and time-consuming calibration in 3D measurement systems is solved, achieving rapid and efficient 3D detection and improved accuracy.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- IND TECH RES INST
- Filing Date
- 2022-03-22
- Publication Date
- 2026-06-23
AI Technical Summary
Existing 3D measurement systems have complex, time-consuming, and inconvenient calibration methods, making it difficult to quickly and efficiently perform accurate testing on different types of products on the production line.
By capturing images of a reference object and using a phase error model for phase compensation, the phase height conversion coefficient is calculated, thus achieving adaptive phase compensation for the three-dimensional measurement system.
It simplifies the calibration process, improves detection accuracy and efficiency, is suitable for full-screen calibration, reduces phase error, and enhances the accuracy of 3D point cloud models.
Smart Images
Figure CN116255929B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to three-dimensional measurement based on phase shift, and particularly to a three-dimensional measurement system and its calibration method. Background Technology
[0002] In high-precision inspection in industrial manufacturing, phase-shift-based three-dimensional measurement systems are often used because they have advantages such as non-contact operation and high measurement accuracy. Therefore, they can be applied in fields such as semiconductor measurement, industrial product quality inspection, or workpiece three-dimensional measurement.
[0003] To balance production efficiency and quality, a 3D measurement system can be installed on the production line for real-time product inspection. Due to the diverse product types, the 3D measurement system needs to be calibrated for different product types to ensure accuracy. However, current calibration methods involve sending the 3D measurement system back to the laboratory or the original manufacturer for calibration. Calibration requires a high-precision fine-tuning platform and special standard parts, and most known methods correct system errors by modifying the equipment's gamma value or using lookup tables. The entire calibration process is complex, time-consuming, and inconvenient. Summary of the Invention
[0004] In view of this, the present invention proposes a three-dimensional measurement system based on phase shift and a correction method thereof. The three-dimensional measurement system of the present invention captures a reference object having a first correction plane and a second correction plane to generate at least one reference object image, performs phase compensation on the phase data in the at least one reference object image according to a phase error model, and then calculates the phase height conversion coefficient based on the compensated phase data.
[0005] A calibration method for a three-dimensional measurement system according to an embodiment of the present invention includes: a projection device projecting a structured light onto a reference object, wherein the reference object includes a first calibration plane and a second calibration plane; a camera device capturing images of the reference object to obtain at least one image of the reference object; a processor performing decoding operations based on the at least one image of the reference object to obtain a plurality of phase data of the at least one image of the reference object; the processor calculating a first phase corresponding to the first calibration plane and a second phase corresponding to the second calibration plane based on the phase data; and the processor calculating a planar phase difference between the first phase and the second phase, and performing operations based on the planar phase difference and the height of the second calibration plane relative to the first calibration plane to obtain a phase height conversion coefficient.
[0006] According to an embodiment of the present invention, a phase-shift-based three-dimensional measurement system includes: a reference object including a first correction plane and a second correction plane; a projection device for projecting structured light onto the reference object; a camera device for capturing images of the reference object to obtain at least one image of the reference object; and a processor electrically connected to the camera device and the projection device, the processor being configured to perform: performing decoding operations based on the at least one image of the reference object to obtain a plurality of phase data of the at least one image of the reference object; calculating a first phase corresponding to the first correction plane and a second phase corresponding to the second correction plane based on the phase data; and calculating a planar phase difference between the first phase and the second phase, and performing operations based on the planar phase difference and the height of the second correction plane relative to the first correction plane to obtain a phase height conversion coefficient.
[0007] In summary, this invention proposes a phase-shift-based 3D measurement system, a height measurement method, and a method for generating phase-height conversion coefficients. This invention only requires capturing an image of a single height block as a reference object, and the phase information within it can be used to generate the phase-height conversion coefficients needed for subsequent correction. The generation of phase-height conversion coefficients is both fast and applicable to the entire image. Regardless of the placement of the height block, the phase-height conversion coefficients generated by this invention are stable. Furthermore, the phase error model proposed in this invention can adaptively compensate the phase of the 3D measurement system, reducing phase errors and improving the accuracy of 3D point cloud model creation.
[0008] The foregoing description of the contents of this disclosure and the following description of the embodiments are used to demonstrate and explain the spirit and principles of the present invention, and to provide a further explanation of the claims of the present invention. Attached Figure Description
[0009] Figure 1 This is a block diagram of a three-dimensional measurement system based on phase shift according to an embodiment of the present invention;
[0010] Figure 2 This is a flowchart of a three-dimensional measurement method based on phase shift according to an embodiment of the present invention;
[0011] Figure 3A This is an example of a target image;
[0012] Figure 3B It is an example of an absolute phase diagram;
[0013] Figure 3C This is an example of a phase difference diagram of a target object;
[0014] Figure 4 This is a flowchart of establishing a phase error model according to an embodiment of the present invention;
[0015] Figure 5A It is an example of an error-corrected planar image;
[0016] Figure 5B It is an example of modeling data;
[0017] Figure 5C It is an example of ideal phase data;
[0018] Figure 5D This is an example of phase error data;
[0019] Figure 5E It is an example of a phase error model;
[0020] Figure 6 This is a flowchart of a method for generating phase height conversion coefficients according to an embodiment of the present invention;
[0021] Figure 7 It is an example of a reference object;
[0022] Figure 8 This is another example of a reference object;
[0023] Figure 9 This is an example of a phase difference diagram of a reference object;
[0024] Figure 10 It is a comparison chart of the height error heatmaps of the stepped blocks at different positions before and after compensation; and Figures 11A to 11C This is an example of phase error data before and after compensation.
[0025] [Symbol Explanation]
[0026] 100: Three-dimensional measurement system
[0027] 10: Projection device
[0028] 20: Camera device
[0029] 30: Processor
[0030] 40: Target object
[0031] 42: Error Correction Plane
[0032] 52, 52', 521, 522: Rank blocks
[0033] 52a, 52b, 52c: Position of the order block
[0034] 54, 54': base plate
[0035] 50, 50': Reference object
[0036] 60: Reference plane
[0037] A1, A2: Average phase values
[0038] B1: First Phase
[0039] B2: Second Phase
[0040] C1: Modeling Data
[0041] C2: Ideal phase data
[0042] F: Vision
[0043] I1: Target image
[0044] I2: Error Correction Plane Image
[0045] L1, L2: Scan lines
[0046] SL: Structured Light
[0047] S1~S5, P1~P7, U1~U6: Steps
[0048] X 52 Y 52 : Dimensions of the upper surface of the stepped block
[0049] X 54 Y 54 Dimensions of the upper surface of the substrate
[0050] Z 52 Z 521 Z 522 The height difference between the stepped block and the substrate
[0051] phase Detailed Implementation
[0052] The following detailed description of the features and characteristics of the present invention in the embodiments is sufficient to enable those skilled in the art to understand the technical content of the present invention and to implement it accordingly. Based on the disclosure of this specification, the claims, and the accompanying drawings, those skilled in the art can easily understand the relevant concepts and features of the present invention. The following embodiments further illustrate the viewpoints of the present invention in detail, but are not intended to limit the scope of the present invention in any way.
[0053] Figure 1 This is a block diagram of a phase-shift-based three-dimensional measurement system according to an embodiment of the present invention. This system is suitable for measuring the height of a target object 40, wherein the target object 40 is disposed on a reference surface 60. Figure 1As shown, the phase-shift-based three-dimensional measurement system 100 includes a projection device 10, a camera device 20, and a processor 30. The projection device 10 projects structured light SL onto the target object 40. The camera device 20 captures images of the target object 40 to generate a target image. The field of view (FOV) F of the camera device 20 must include the reference surface 60 and the target object 40. The shooting angle of the camera device 20 can be adaptively adjusted according to the position and size of the target object 40. In this embodiment, the reference surface 60 is a plane; however, in other embodiments, the reference surface can also be a surface or curved surface with non-uniform height, and is not limited to the plane in this embodiment.
[0054] Generally speaking, the principle of phase shift-based three-dimensional measurement is to project structured light SL (such as texture image, phase stripe or grating image) onto the surface of target object 40, then collect the texture information reflected from the surface of target object 40, and then calculate the height information of the surface of target object 40 using trigonometric geometry. The height information can be calculated from the phase information in the target image by using the phase height conversion coefficient k, as shown in Method 1 below.
[0055] Formula 1:
[0056]
[0057] Assuming the reference plane 60 is located on the xy plane, then h(x, y) represents the height at coordinates (x, y), P represents the period length of the structured light SL, d represents the distance between the projection device 10 and the camera device 20, L represents the distance between the camera device 20 and the reference plane 60, and k represents the phase height conversion coefficient. This represents the phase shift between the target object 40 and the reference surface 60 in coordinates (x, y). This invention proposes a method for rapidly generating the phase height conversion coefficient k, and a method and system for calculating the object height using the phase height conversion coefficient k.
[0058] like Figure 1 As shown, the processor 30 is electrically connected to the projection device 10 and the camera device 20, thereby controlling the projection device 10 to emit structured light SL and controlling the camera device 20 to capture a target image. The processor 30 is configured to execute instructions, such as executing multiple first instructions based on the target image to calculate the height of the target object 40. The flow corresponding to these first instructions is as follows: Figure 2 As shown.
[0059] Figure 2This is a flowchart of a phase-shift-based three-dimensional measurement method according to an embodiment of the present invention. Step S1 is "obtaining multiple target phase data based on the target image", step S2 is "compensating the target phase data based on the error phase model", step S3 is "obtaining basic phase data", step S4 is "calculating the phase difference based on the basic phase data and the target phase data", and step S5 is "calculating the height of the target object based on the phase difference and the phase height conversion coefficient".
[0060] Figure 3A This is an example of target image I1. In step S1, processor 30 can generate multiple scan lines, one of which is Figure 3A The scan line L1 shown passes through the reference plane 60 and the target object 40. The processor 30 obtains multiple target phase data as described in step S1 from the scan line L1. Based on this target phase data, a plot can be drawn. Figure 3B The absolute phase diagram shown has the horizontal axis representing the X-axis coordinate and the vertical axis representing the phase, with the unit being radians.
[0061] After step S1, that is, after the processor 30 obtains the target phase data based on the target image I1, step S2 is executed. In step S2, the processor 30 compensates for these target phase data according to a phase error model. The phase error model reflects the error value between the measured target phase and the ideal phase. This phase error model needs to be established in advance before step S2, and the method of establishment will be described later. Alternatively, in one embodiment, step S2 can be omitted, and the process can proceed directly from step S1 to step S3.
[0062] After step S2, that is, after the processor 30 compensates for the target phase data according to the phase error model, step S3 is executed. In one embodiment of step S3, the processor 30 obtains pre-stored basic phase data. Specifically, before the target object 40 is placed on the reference surface 60, the camera device 20 pre-captures the reference surface 60 to generate a reference surface image. The processor 30 performs decoding operations based on this reference surface image to obtain multiple phase data of the reference surface image. The processor 30 then compensates the phase data of the reference surface image using an error correction model to generate basic phase data. Therefore, in step S3, the processor 30 can obtain the pre-generated basic phase data. In another embodiment of step S3, the reference surface 60 can also be captured according to the above process to generate the latest basic phase data in real time, and the present invention does not limit this.
[0063] Step S4 is "calculating the phase difference based on the basic phase data and the target phase data". Specifically, multiple phase data points from the basic phase data and multiple data points from the target phase data are subtracted based on the same pixel coordinates to obtain multiple phase difference data points, which are then plotted as follows: Figure 3CA schematic diagram. Based on the magnitude of the phase difference, Figure 3C It can be roughly divided into two parts: coordinates 0 to 900 and coordinates 1600 to 2500 correspond to the areas in reference image I1 where the target object 40 does not exist, and coordinates 900 to 1600 correspond to the areas in reference image I1 where the target object 40 exists.
[0064] In step S5, processor 30 calculates the phase difference. The product of the phase height conversion coefficient k and the height h of the target object 40 is given by equation 1. Figure 3C As shown. According to the process of steps S1 to S5, the processor 30 can calculate the height of each point in the target image I1, and thus establish a point cloud model of the target object 40 based on these height information.
[0065] Regarding the method for establishing the phase error model described in step S2, in detail, the processor 30 is configured to execute instructions, such as executing multiple second instructions to generate the phase error model. The flow corresponding to these second instructions is as follows: Figure 4 As shown.
[0066] Figure 4 This is a flowchart of establishing a phase error model according to an embodiment of the present invention. Step P1 is "projecting structured light onto the error correction plane", step P2 is "capturing the error correction plane to generate an image of the error correction plane", step P3 is "performing decoding operations to obtain multiple phase data of the error correction plane", step P4 is "obtaining multiple modeling data located on the error correction line from the multiple phase data", step P5 is "performing linear fitting on the modeling data to generate multiple ideal phase data", step P6 is "calculating multiple phase error data based on the modeling data and the ideal phase data", and step P7 is "establishing a phase error model based on the phase error data".
[0067] In steps P1 and P2, the processor 30 executing the second instruction controls the projection device 10 and the camera device 20 to perform corresponding operations. In step P1, the structured light is periodic structured light, and the phase of the structured light increases along an extension direction.
[0068] Figure 5A This is an example of an error-corrected planar image I2. Figure 5B This is an example of modeling data. In one embodiment of step P3, processor 30 performs a decoding operation based on the error correction plane image I2 to obtain multiple phase data corresponding to the error correction plane. In step P4, processor 30 obtains multiple modeling phase data located on the error correction line from the multiple phase data of the error correction plane image I2, based on an error correction line parallel to the extension direction in the error correction plane image I2. Processor 30 in Figure 5AA scan line L2 (error correction line) is generated in the error correction plane image I2 shown. This scan line L2 (error correction line) must pass through the error correction plane 42 and have the following characteristics: Figure 5B The slope characteristic shown is that, along this scan line L2, the phase distribution direction of multiple modeling data is increasing (equivalent to the extension direction parallel to the phase of the structured light). In other words, as the pixel coordinates of a modeling data point increase, its phase value also increases. Overall, step P3 is: in the error correction plane image I2, the processor 30 obtains multiple modeling data points according to a straight line (scan line L2, i.e., the error correction line) with an increasing phase distribution direction.
[0069] Since the error correction plane 42 is a flat plane with no height variation, ideally, the phase data corresponding to this error correction plane 42 should show a linear increase. However, hardware components in the projection device 10 or the camera device 20 may cause measurement errors, making the phase data corresponding to the error correction plane 42 not show a perfect linear increase. Therefore, in step P5, the processor 30 performs linear fitting. In one embodiment, the processor 30, for example, uses the least squares method to generate ideal phase data based on the modeling data. Figure 5C This is an example of ideal phase data generated using the least squares method, where curve C1 with slight up-and-down oscillations is the modeling data, and straight line C2 is the ideal phase data. It should be noted that in this embodiment, the error correction plane 42 and the reference plane 60 in steps S1-S7 are the same plane; however, in other embodiments, the error correction plane 42 and the reference plane 60 in steps S1-S7 may be different planes.
[0070] In one embodiment of step P6, for each pixel coordinate, the processor 30 calculates the difference between the ideal phase and the actual phase as a phase error data point. Figure 5D This is an example of phase error data generated using the above method.
[0071] In one embodiment of step P7, the processor 30 performs Fourier analysis and low-pass filtering based on multiple phase error data to establish a phase error model. Figure 5E This is an example of a phase error model established in the above manner, which can be used to compensate for the target phase data in step S2.
[0072] In one embodiment, the process of steps P1 to P7 described above further includes a horizontal calibration step. Specifically, the processor 30's calculation of the first phase corresponding to the first calibration plane and the second phase corresponding to the second calibration plane based on the phase data further includes: the processor 30 obtaining multiple reference phase data located on a reference straight line passing through the first and second calibration planes in at least one reference object image; the processor obtaining at least one set of phase data from these reference phase data, wherein the at least one set of phase data includes at least two of the reference phase data, and the at least one set of phase data corresponds to the first or second calibration plane; the processor 30 performing horizontal calibration on these phase data based on the at least one set of phase data; and the processor 30 calculating the first phase corresponding to the first calibration plane and the second phase corresponding to the second calibration plane based on the reference phase data after horizontal calibration.
[0073] Regarding the generation method of the phase height conversion coefficient k in step S5, in detail, the processor 30 is configured to execute instructions, such as executing multiple third instructions to generate the phase height conversion coefficient k. The process corresponding to these third instructions is as follows: Figure 6 As shown. Figure 6 This is a flowchart of a calibration method for a three-dimensional measurement system according to an embodiment of the present invention. This method is used to generate a phase height conversion coefficient k. Step U1 is "projecting structured light onto a reference object", step U2 is "taking pictures of the reference object to generate an image of the reference object", step U3 is "obtaining multiple raw phase data based on the image of the reference object", step U4 is "compensating the raw phase data based on a phase error model", step U5 is "performing horizontal calibration on the raw phase data", and step U6 is "calculating the phase difference and phase height conversion coefficient based on these raw phase data".
[0074] In steps U1 and U2, the processor 30 executing the third instruction controls the projection device 10 and the camera device 20 to perform corresponding operations. Figure 7 This is an example of a reference object 50. In step U1, the reference object 50 includes a first correction plane and a second correction plane. In one embodiment, the reference object 50 includes a step block 52 and a substrate 54, with the step block 52 disposed on the substrate 54. The first correction plane is the upper surface (top surface) of the step block 52, and the second correction plane is the upper surface (top surface) of the substrate 54. In step U2, the imaging device 20 captures images of the reference object 50 to obtain at least one reference object image. The present invention does not limit the number of reference object images. In one embodiment, the same structured light pattern can be presented with different phases. Therefore, in steps U1 to U2, the projection device 10 can project structured light with different phases, and the imaging device 20 then captures images one by one.
[0075] In one embodiment, several properties of the step height block 52 and the substrate 54 can be adjusted as needed. These properties include: the upper surface dimension of the step height block 52 (which can be adjusted using X...). 52 Y 52 (represented by) the height difference Z between the step block 52 and the substrate 54 52 The number of step height blocks 52, the position of the step height blocks 52 on the substrate 54, and the upper surface dimensions of the substrate 54 (which can be expressed as X). 54 Y 54 (to indicate).
[0076] Regarding the upper surface dimension X of step block 52 52 Y 52 In one embodiment of the reference object image in step U2, the length X of the upper surface of the step block 52 is... 52 and width Y 52 All are at least 20 pixels. However, the present invention is not limited to the above values. In practice, the distance between the lens of the imaging device 20 and the step block 52 and the shooting angle can be adaptively adjusted according to multiple parameters (such as the focal length of the camera and the resolution of the reference object image) so that the upper surface size X of the step block 52 in the reference object image is... 52 Y 52 Exceeding the preset value.
[0077] Regarding the height difference Z between the step block 52 and the substrate 54 52 Its value depends on the Z-axis measurement accuracy required by the three-dimensional measurement system 100. In one embodiment, the height difference Z... 52 It is more than 10 times more accurate than the measurement. For example, if the accuracy is 1 micrometer (μm), then the height difference Z... 52 It must be greater than or equal to 10 micrometers.
[0078] Figure 8 This is an example of a reference object 50' with two level blocks 521 and 522. For example... Figure 8 As shown, step height block 522 is disposed on substrate 54', and step height block 521 is disposed on step height block 522. The upper surface dimension of step height block 521 is smaller than the upper surface dimension of step height block 522. In other words, the two step height blocks 521 and 522 can be considered as a single second-order step height block 52'. Therefore, step height block 52 has two height differences Z relative to substrate 54. 521 Z 522 This invention does not limit the number (or order) of the order blocks. Without significantly affecting the correction speed, a greater number (order) of order blocks can increase the correction accuracy.
[0079] The present invention does not impose any particular restrictions on the position of the step block 52 on the substrate 54.
[0080] Regarding the upper surface dimension X of substrate 54 54 Y 54 Please refer to Figure 7 Specifically, the processor 30 can locate an image block corresponding to the substrate 54 in the reference object image. In one embodiment, the size of this image block is at least [a certain percentage] of the size of the reference object image. In other words, the length and width of the substrate 54 are at least equal to the field of view F of the three-dimensional measurement system 100. However, the present invention is not limited to the above values, nor does it limit the shape of the substrate 54 to be rectangular.
[0081] In step U3, the processor 30 performs a decoding operation based on at least one reference object image to obtain multiple phase raw data of at least one reference object image. The method is basically similar to the way the processor 30 obtains multiple modeling data based on the error correction plane image I2 in step P3, and will not be described again here.
[0082] In step U4, the processor 30 uses the phase error model established in step P7 to compensate for multiple original phase data of at least one reference object image. In other words, step P7 needs to be completed before step U4. In one embodiment, step U4 can be omitted, and the process can proceed directly from step U3 to step U5.
[0083] In step U5, the processor 30 performs horizontal calibration on the raw phase data in a manner that is essentially similar to the processor 30 performs horizontal calibration on the target phase data in step S3, and will not be described again here.
[0084] In step U6, the processor 30 calculates the first phase corresponding to the first correction plane and the second phase corresponding to the second correction plane based on the (corrected and horizontally calibrated) raw phase data. In one embodiment, the processor 30 obtains multiple reference phase data located on a reference line passing through the first and second correction planes in at least one reference object image. The processor 30 calculates the first phase corresponding to the first correction plane and the second phase corresponding to the second correction plane based on these reference phase data. In practice, the aforementioned reference line is conceptually similar to... Figure 3A The scan line L1 shown passes through the error correction plane 42 and the target object 40. In one embodiment, the first phase is the average of a plurality of reference phase data of the first correction plane, and the second phase is the average of a plurality of reference phase data of the second correction plane.
[0085] Processor 30 calculates the planar phase difference between the first phase and the second phase, and then performs calculations based on the planar phase difference and the height of the second correction plane relative to the first correction plane to obtain the phase height conversion coefficient k. The method for calculating the phase difference is similar to step S4. Please refer to... Figure 9 It is the phase difference of the higher-order block 52. A schematic diagram. Phase difference. Associated with the first phase B1 and the second phase B2, the first phase B1 is associated with the phase data corresponding to the substrate 54 among multiple phase data, and the second phase B2 is associated with multiple phase data corresponding to the step height block 52 among multiple phase data, and the phase height conversion coefficient k is associated with the height Z of the step height block 52. 52 and phase difference
[0086] In one embodiment, the phase difference is The difference between the first phase B1 and the second phase B2, i.e.
[0087] In one embodiment, the first phase B1 is the average value of multiple phase data corresponding to the substrate 54, such as phase data with coordinates 0 to 900 and phase data with coordinates 1600 to 2500. The second phase B2 is the average value of multiple phase data corresponding to the step block 52, such as phase data with coordinates 900 to 1600.
[0088] In one embodiment, the phase height conversion factor k is the quotient of the height of the second correction plane relative to the first correction plane divided by the plane phase difference. In another embodiment, the phase height conversion factor k is the height Z of the step block 52. 52 Divide by phase difference The quotient, that is After calculating the phase height conversion coefficient k, it is possible to... Figure 2 The height of the target object 40 is calculated in step S5.
[0089] exist Figure 6 To illustrate the effectiveness of the phase error model in the illustrated process, please refer to [the relevant documentation / reference]. Figure 10 It is a comparison chart of the height error heatmaps of the stepped blocks at different positions before and after compensation. Figure 10 The unit for medium-height error is micrometer (μm), while Figures 11A to 11C This is an example of phase error data before and after compensation.
[0090] Figure 10The substrate 54 before and after compensation is presented in a top view. The squares in the substrate 54 represent multiple positions where the step-height blocks 52 can be set. Examples of the phase diagrams before and after compensation when the step-height blocks 52 are set at positions 52a, 52b, and 52c on the substrate 54 are as follows: Figure 11A , Figure 11B , Figure 11C As shown. From Figure 10 The two height error heatmaps show that: before using phase error model compensation in step S2 or step U4, if the height block 52 is placed at the corner of the substrate 54, the final calculated height of the target object 40 has a large measurement error; however, after using phase error model compensation in step S2 or step U4, the final calculated height error of the target object 40 is consistent regardless of where the height block 52 is placed on the substrate 54. This indicates that the phase height conversion coefficient k generated by an embodiment of the present invention is applicable to the entire field of view F that the camera device 20 can capture.
[0091] Please refer to Figure 1 and Figure 7 In one embodiment, the present invention proposes a three-dimensional measurement system based on phase shift, the system comprising: a reference object 50, a projection device 10, a camera device 20, and a processor 30. The reference object 50 includes a first correction plane (e.g., the upper surface of a step block 52) and a second correction plane (e.g., the upper surface of a substrate 54). The projection device 10 projects structured light SL onto the reference object 50. The camera device 20 captures images of the reference object 50 to obtain at least one image of the reference object. The processor 30 is electrically connected to the camera device 20 and the projection device 10, and is configured to perform: performing decoding operations based on at least one image of the reference object to obtain multiple phase data of the at least one image of the reference object; calculating a first phase corresponding to the first correction plane and a second phase corresponding to the second correction plane based on these phase data; and calculating the planar phase difference between the first phase and the second phase, and performing operations based on the planar phase difference and the height of the second correction plane relative to the first correction plane to obtain a phase height conversion coefficient k.
[0092] In summary, this invention proposes a phase-shift-based 3D measurement system, a height measurement method, and a method for generating phase-height conversion coefficients. This invention only requires capturing an image of a single height block as a reference object, and the phase information within it can be used to generate the phase-height conversion coefficients needed for subsequent correction. The generation of phase-height conversion coefficients is both fast and applicable to the entire image. Regardless of the placement of the height block, the phase-height conversion coefficients generated by this invention are stable. Furthermore, the phase error model proposed in this invention can adaptively compensate the phase of the 3D measurement system, reducing phase errors and improving the accuracy of 3D point cloud model creation.
Claims
1. A calibration method for a three-dimensional measurement system, comprising: The projection device projects structured light onto a reference object, wherein the reference object includes a first correction plane and a second correction plane; The camera device captures an image of the reference object to obtain at least one image of the reference object; The processor performs decoding operations based on the at least one reference object image to obtain multiple phase data of the at least one reference object image; The processor calculates the first phase corresponding to the first correction plane and the second phase corresponding to the second correction plane based on these phase data; as well as The processor calculates the planar phase difference between the first phase and the second phase, and performs calculations based on the planar phase difference and the height of the second correction plane relative to the first correction plane to obtain the phase height conversion coefficient; The processor, based on these phase data, calculates the first phase corresponding to the first correction plane and the second phase corresponding to the second correction plane, and further includes: The processor obtains multiple reference phase data located on the reference line in the at least one reference object image, based on the reference line passing through the first correction plane and the second correction plane. as well as The processor calculates the first phase corresponding to the first correction plane and the second phase corresponding to the second correction plane based on these reference phase data.
2. The calibration method for the three-dimensional measurement system as described in claim 1, wherein the processor calculates the first phase corresponding to the first calibration plane and the second phase corresponding to the second calibration plane based on the phase data, further comprising: The processor obtains multiple reference phase data located on the reference line in the at least one reference object image, based on the reference line passing through the first correction plane and the second correction plane. The processor obtains at least one set of phase data from these reference phase data, wherein the at least one set of phase data includes at least two of the reference phase data, and the at least one set of phase data corresponds to the first correction plane or the second correction plane; The processor performs horizontal calibration on these phase data based on at least one set of phase data; as well as The processor calculates the first phase corresponding to the first correction plane and the second phase corresponding to the second correction plane based on these reference phase data after performing horizontal calibration.
3. The calibration method for the three-dimensional measurement system as described in claim 1, wherein the structured light is a periodic structured light, and the phase of the structured light increases along the extension direction, the calibration method further includes: The projection device projects the structured light onto the error correction plane; The camera device captures the error correction plane to generate an image of the error correction plane; The processor performs decoding operations based on the error-corrected plane image to obtain multiple phase data of the error-corrected plane image; The processor obtains multiple modeling phase data located on the error correction line in the error correction plane image based on the error correction line parallel to the extension direction in the error correction plane image; The processor performs a linear fit on these modeling phase data to generate multiple ideal phase data; The processor calculates multiple phase error data based on these modeled phase data and these ideal phase data; as well as The processor builds a phase error model based on these phase error data.
4. The calibration method of the three-dimensional measurement system as claimed in claim 1, wherein the reference object includes a substrate and a stepped block disposed on the substrate, and the first calibration plane is the top surface of the substrate, and the second calibration plane is the top surface of the stepped block.
5. The calibration method for the three-dimensional measurement system as described in claim 1, wherein... The first phase is the average value of these reference phase data for the first correction plane; The second phase is the average of these reference phase data for the second correction plane; and The phase height conversion factor is the quotient of the height of the second correction plane relative to the first correction plane divided by the phase difference between the planes.
6. The calibration method for the three-dimensional measurement system as described in claim 3, wherein the processor performs decoding operations based on the at least one reference object image to obtain the phase data, comprising: The processor performs decoding operations based on the image of at least one reference object to generate several phase raw data; as well as The processor performs phase compensation on these raw phase data according to the phase error model to obtain these phase data of the reference object image.
7. The calibration method for the three-dimensional measurement system as described in claim 3, wherein the linear fitting is the least squares method.
8. The calibration method for the three-dimensional measurement system as described in claim 3, wherein the processor establishes the phase error model based on these phase error data, comprising: The processor performs Fourier analysis and low-pass filtering based on these phase error data to establish the phase error model.
9. The calibration method for the three-dimensional measurement system as claimed in claim 4, wherein the reference object image has an image block corresponding to the substrate, and the size of the image block is at least 2 / 3 of the size of the reference object image.
10. A three-dimensional measurement system based on phase shift, the system comprising: Reference objects, including a first calibration plane and a second calibration plane; A projection device for projecting structured light onto the reference object; A camera device is used to photograph the reference object to obtain an image of at least one reference object; as well as A processor, electrically connected to the camera device and the projection device, is configured to execute: Decoding operations are performed based on the at least one reference object image to obtain multiple phase data of the at least one reference object image; Based on these phase data, calculate the first phase corresponding to the first correction plane and the second phase corresponding to the second correction plane; and Calculate the planar phase difference between the first phase and the second phase, and perform an operation based on the planar phase difference and the height of the second correction plane relative to the first correction plane to obtain the phase height conversion coefficient; The processor, configured to perform calculations based on these phase data, including the first phase corresponding to the first correction plane and the second phase corresponding to the second correction plane, further includes: The processor obtains multiple reference phase data located on the reference line in the at least one reference object image, based on the reference line passing through the first correction plane and the second correction plane. as well as The processor calculates the first phase corresponding to the first correction plane and the second phase corresponding to the second correction plane based on these reference phase data.
11. The phase-shift-based three-dimensional measurement system of claim 10, wherein the processor is configured to calculate, based on the phase data, the first phase corresponding to the first correction plane and the second phase corresponding to the second correction plane, further comprising: The processor obtains multiple reference phase data located on the reference line in the at least one reference object image, based on the reference line passing through the first correction plane and the second correction plane. The processor obtains at least one set of phase data from these reference phase data, wherein the at least one set of phase data includes at least two of the reference phase data, and the at least one set of phase data corresponds to the first correction plane or the second correction plane; The processor performs horizontal calibration on these phase data based on at least one set of phase data; as well as The processor calculates the first phase corresponding to the first correction plane and the second phase corresponding to the second correction plane based on these reference phase data after performing horizontal calibration.
12. The phase-shift-based three-dimensional measurement system of claim 10, wherein the structured light is a periodic structured light, and the phase of the structured light increases along the extension direction; The projection device is also used to project the structured light onto the error correction plane; The camera device is also used to capture the error correction plane to generate an image of the error correction plane; and The processor is also configured to perform: Decoding operations are performed based on the error-corrected plane image to obtain multiple phase data of the error-corrected plane image; Based on the error correction line parallel to the extension direction in the error correction plane image, obtain multiple modeling phase data located on the error correction line from multiple phase data of the error correction plane image; These modeling phase data are linearly fitted to generate multiple ideal phase data; Based on these modeling phase data and these ideal phase data, multiple phase error data are calculated; as well as A phase error model is established based on these phase error data.
13. The phase-shift-based three-dimensional measurement system of claim 10, wherein the reference object includes a substrate and a step block disposed on the substrate, and the first correction plane is the top surface of the substrate and the second correction plane is the top surface of the step block.
14. The phase-shift-based three-dimensional measurement system as described in claim 10, wherein... The processor calculates the average of these reference phase data of the first correction plane to obtain the first phase; The processor calculates the average of these reference phase data of the second correction plane to obtain the second phase; and The processor calculates the quotient of the height of the second correction plane relative to the first correction plane divided by the phase difference of the planes to obtain the phase height conversion coefficient.
15. The phase-shift-based three-dimensional measurement system of claim 12, wherein the processor is configured to perform decoding operations based on the at least one reference object image to obtain the phase data, comprising: The processor performs decoding operations based on the image of at least one reference object to generate several phase raw data; as well as The processor performs phase compensation on these raw phase data according to the phase error model to obtain these phase data of the reference object image.
16. The phase-shift-based three-dimensional measurement system of claim 12, wherein the processor is configured to perform a least-squares linear fit.
17. The phase-shift-based three-dimensional measurement system of claim 12, wherein the processor is configured to perform the function of establishing the phase error model based on the phase error data, comprising: The processor performs Fourier analysis and low-pass filtering based on these phase error data to establish the phase error model.
18. The phase-shift-based three-dimensional measurement system of claim 13, wherein the reference object image has an image block corresponding to the substrate, and the size of the image block is at least 2 / 3 of the size of the reference object image.