A dynamic three-dimensional measurement error correction method

Abstract

The application discloses a dynamic three-dimensional measurement error correction method, which can correct residual high-order harmonic components in a reverse error average method by using an iterative Hilbert transform method, and effectively improves dynamic three-dimensional measurement precision.

Description

Technical Field

[0001] This invention belongs to the field of three-dimensional measurement technology. Specifically, this invention relates to a dynamic three-dimensional measurement error correction method. Background Technology

[0002] Phase-shifting profilometry, as an important optical 3D measurement technique, has been widely used in many fields such as industrial inspection, aerospace, biomedicine, and scientific experiments. However, dynamic 3D measurement remains a key challenge for this technology: object motion causes non-ideal phase shifts between continuously acquired fringe images, introducing periodic phase errors that directly affect the accuracy of 3D measurements. To suppress phase errors caused by motion, the academic community has proposed several technical approaches. One approach is high-speed image acquisition, which uses high-speed cameras and projectors to improve image acquisition and processing speed to suppress phase shifts, but this approach has high hardware costs. Another approach is object motion estimation, which estimates the motion information of objects through object tracking and then compensates for and corrects phase-shifting fringes. However, this approach relies on the accuracy of motion estimation, and achieving high-precision estimation for non-rigid, complex, or unpredictable motion remains a challenge, limiting its applicability in practical scenarios.

[0003] In addition, another approach is the inverse error averaging method, which attempts to directly suppress or eliminate dynamic 3D measurement errors at the algorithmic level. For example, some works generate a set of virtual phase-shifting fringes by performing a Hilbert transform on the actual phase-shifting fringes, constructing a virtual phase with a phase difference of π / 2 from the actual phase, and then canceling the first-order phase error by calculating the average phase of the two. Other works divide the four-step phase-shifting fringes into two sets of non-isochronous phase-shifting fringes, and weaken the first-order phase error by fusing the phase results of the two sets of specific inverse errors. Although such approaches can eliminate the first-order fundamental frequency component of the phase error, higher-order harmonic residues still affect the accuracy of dynamic 3D measurements. How to further suppress or eliminate these higher-order residual errors is of great significance for improving the accuracy of dynamic 3D measurements using phase-shifting profilometry. Summary of the Invention

[0004] This invention provides a dynamic three-dimensional measurement error correction method to solve the problems existing in the background art.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: a dynamic three-dimensional measurement error correction method, specifically including the following steps:

[0006] Step S1: Construct a dynamic 3D measurement system, including a projector, a camera, and a dynamic object. The three form a triangulation relationship, and the dynamic object is always within the common field of view of the projector and the camera. In addition, the projector and the camera are synchronously triggered through external hardware circuits to ensure the correspondence between the stripe projection and the acquisition.

[0007] Step S2: Simultaneously turn on the projector and camera. The projector projects the ideal three-step phase-shift fringe pattern onto the surface of the dynamic object in a loop. Due to the modulation effect of the three-dimensional contour of the dynamic object, the phase-shift fringe pattern will be distorted and deformed. The camera synchronously captures the deformed three-step phase-shift fringe image at time t. ;

[0008] Step S3: Process each original phase-shifted fringe image separately. Perform a Hilbert transform to obtain a set of phase-shifted fringe images from the first transform. The original phase-shifted fringe images were extracted using a three-step phase-shifting algorithm. The phase-shifted fringe image of the first transformation Corresponding truncation phase and And calculate the average cutoff phase of both. ;

[0009] Step S4: The average cutoff phase obtained after step S3 Higher-order harmonic components still remain and need to be further eliminated; therefore, the average phase is truncated. Encoded into a set of virtual phase-shifted fringe images Then, for each virtual phase-shifting fringe image... Perform a Hilbert transform to obtain the phase-shifted fringe image of the second transform. ;

[0010] Step S5: Extract the phase-shifted fringe image of the second transformation using the three-step phase-shift algorithm. truncated phase And the continuous absolute phase is recovered using the geometric constraint method. ; Calibrating the geometric parameters of the dynamic three-dimensional measurement system will correct the absolute phase. Mapped to three-dimensional physical space, the three-dimensional shape of the dynamic object is reconstructed.

[0011] Preferably, in step S2, the camera captures a three-step phase-shift fringe image. It can be represented as:

[0012] ;

[0013] In the formula: The sequence number of the phase-shifted fringe image; Represents the pixel coordinates of the phase-shifted fringe image; and These represent background intensity and modulation intensity, respectively. Represents the original phase-shifted fringe image The truncated phase; This indicates the additional phase shift caused by the motion of the object.

[0014] Furthermore, three-step phase-shifting fringe images in a short time The offset is small, so the background intensity can be assumed to be small. and modulation intensity The three-step phase-shifted fringe image remains almost unchanged. It can be represented as:

[0015] ;

[0016] In the formula: and Representing the background intensity and modulation intensity, respectively, for a set of phase-shifted fringe images. The background intensity can be considered as and modulation intensity It remains unchanged.

[0017] Preferably, in step S3, the phase-shifted fringe image after the first Hilbert transform is... It can be represented as:

[0018] ;

[0019] In the formula: and Representing the background intensity and modulation intensity, respectively, for a set of phase-shifted fringe images. The background intensity can be considered as and modulation intensity It remains unchanged.

[0020] Preferably, in step S3, the original phase-shifted fringe image The phase-shifted fringe image of the first transformation Corresponding truncation phase and as follows:

[0021] ;

[0022] ;

[0023] In the formula: This represents the arctangent operation function, which truncates the phase. and The range of values ​​is limited to ;

[0024] Furthermore, the average truncation phase as follows:

[0025] ;

[0026] In the formula: This represents the modulo operation function, hence the average truncation phase. The range of values ​​for is also limited to .

[0027] Preferably, in step S4, the virtual phase-shifted fringe image It can be represented as:

[0028] ;

[0029] Furthermore, after the Hilbert transform, the phase-shifted fringe image of the second transform... It can be represented as:

[0030] .

[0031] Preferably, in step S5, the phase-shifted fringe image of the second transformation truncated phase as follows:

[0032] ;

[0033] Furthermore, the absolute phase It can be represented as:

[0034] ;

[0035] In the formula: The stripe level is indicated by the geometric constraint method, which can be found in Optics Express, 2016, 24(16): 18445.

[0036] Preferably, in step S5, the geometric parameter calibration method of the dynamic three-dimensional measurement system adopts the traditional phase-height mapping method. For details, please refer to Optics and Lasers in Engineering, 2021, 143:106622.

[0037] The beneficial effects of adopting the above technical solutions are:

[0038] 1. The present invention provides a dynamic three-dimensional measurement error correction method, which utilizes the iterative Hilbert transform method to correct the residual high-order harmonic components in the reverse error averaging method, thereby effectively improving the accuracy of dynamic three-dimensional measurement. Attached Figure Description

[0039] Figure 1 Original phase-shifted fringe image truncated phase Phase-shifted fringe image of the first transformation truncated phase and average cutoff phase ;

[0040] Figure 2 The phase-shifted fringe image of the first transformation and the phase-shifted fringe image of the second transformation ;

[0041] Figure 3 The phase-shifted fringe image of the second transformation truncated phase stripe levels Phase-shifted fringe image of the second transformation absolute phase ;

[0042] Figure 4 To utilize the original phase-shifted fringe image truncated phase Mean cutoff phase and the phase-shifted fringe image of the second transformation truncated phase The reconstructed three-dimensional shape and its errors. Detailed Implementation

[0043] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings, in order to help those skilled in the art to have a more complete, accurate and in-depth understanding of the concept and technical solutions of the present invention, and to facilitate its implementation.

[0044] like Figures 1 to 4 As shown, this invention is a dynamic three-dimensional measurement error correction method. By using the iterative Hilbert transform method, it can correct the high-order harmonic components remaining in the reverse error averaging method, effectively improving the accuracy of dynamic three-dimensional measurement.

[0045] Example 1:

[0046] Step S1: Build a dynamic 3D measurement system, including a projector, a camera, and a dynamic object. The three form a triangulation relationship, and the dynamic object is always within the common field of view of the projector and the camera. In addition, the projector and the camera are synchronously triggered through external hardware circuits to ensure the correspondence between stripe projection and acquisition.

[0047] Step S2: Simultaneously turn on the projector and camera. The projector projects the ideal three-step phase-shift fringe pattern onto the surface of the dynamic object in a loop. Due to the modulation effect of the three-dimensional contour of the dynamic object, the phase-shift fringe pattern will be distorted and deformed. The camera synchronously captures the deformed three-step phase-shift fringe image at time t. ;like Figure 1 As shown in the first line.

[0048] Step S3: Process each original phase-shifted fringe image separately. Perform a Hilbert transform to obtain a set of phase-shifted fringe images from the first transform. ,like Figure 1 As shown in the second row; the original phase-shifted fringe image is extracted using a three-step phase-shifting algorithm. The phase-shifted fringe image of the first transformation Corresponding truncation phase and And calculate the average cutoff phase of both. ;like Figure 1 As shown in the third row.

[0049] Step S4: The average cutoff phase obtained after step S3 Higher-order harmonic components still remain and need to be further eliminated; therefore, the average phase is truncated. Encoded into a set of virtual phase-shifted fringe images ,like Figure 2 As shown in the first row, each virtual phase-shifting fringe image is then analyzed separately. Perform a Hilbert transform to obtain the phase-shifted fringe image of the second transform. ;like Figure 2 As shown in the second row.

[0050] Step S5: Extract the phase-shifted fringe image of the second transformation using the three-step phase-shift algorithm. truncated phase And the continuous absolute phase is recovered using the geometric constraint method. ,like Figure 3 As shown; calibrating the geometric parameters of the dynamic three-dimensional measurement system will correct the absolute phase. Mapped to three-dimensional physical space, the three-dimensional shape of the dynamic object is reconstructed. Figure 4 Comparison using phase cutoff , and The reconstructed three-dimensional shape and its errors.

[0051] It can be seen that, using the original phase-shifted fringe image truncated phase The reconstructed 3D topography exhibits significant periodic phase errors; using the average truncation phase... The reconstructed 3D topography can eliminate the first-order phase error, but some higher-order harmonic errors remain; the phase-shifted fringe image obtained from the second transformation is used. truncated phase The reconstructed three-dimensional morphology is relatively smooth, and the higher harmonic errors are greatly reduced, demonstrating the effectiveness of the method of the present invention.

[0052] The present invention has been described above by way of example with reference to the accompanying drawings. Obviously, the specific implementation of the present invention is not limited to the above-described manner. Any non-substantial improvements made using the inventive concept and technical solution of the present invention, or the direct application of the inventive concept and technical solution of the present invention to other situations without modification, are all within the protection scope of the present invention.

Claims

1. A dynamic three-dimensional measurement error correction method, characterized in that: Specifically, the following steps are included: Step S1: Construct a dynamic 3D measurement system, including a projector, a camera, and a dynamic object. The three form a triangulation relationship, and the dynamic object is always within the common field of view of the projector and the camera. In addition, the projector and the camera are synchronously triggered through external hardware circuits to ensure the correspondence between the stripe projection and the acquisition. Step S2: Simultaneously turn on the projector and camera. The projector projects the ideal three-step phase-shift fringe pattern onto the surface of the dynamic object in a loop. Due to the modulation effect of the three-dimensional contour of the dynamic object, the phase-shift fringe pattern will be distorted and deformed. The camera synchronously captures the deformed three-step phase-shift fringe image at time t. ; Step S3: Process each original phase-shifted fringe image separately. Perform a Hilbert transform to obtain a set of phase-shifted fringe images from the first transform. The original phase-shifted fringe images were extracted using a three-step phase-shifting algorithm. The phase-shifted fringe image of the first transformation Corresponding truncation phase and And calculate the average cutoff phase of both. ; Step S4: The average cutoff phase obtained after step S3 Higher-order harmonic components still remain and need to be further eliminated; therefore, the average phase is truncated. Encoded into a set of virtual phase-shifted fringe images Then, for each virtual phase-shifting fringe image... Perform a Hilbert transform to obtain the phase-shifted fringe image of the second transform. ; Step S5: Extract the phase-shifted fringe image of the second transformation using the three-step phase-shift algorithm. truncated phase And the continuous absolute phase is recovered using the geometric constraint method. ; Calibrate the geometric parameters of the dynamic three-dimensional measurement system to correct the absolute phase. Mapped to three-dimensional physical space, the three-dimensional shape of the dynamic object is reconstructed.

2. The dynamic three-dimensional measurement error correction method according to claim 1, characterized in that: In step S2, the camera captures a three-step phase-shift fringe image. It can be represented as: ； In the formula: The sequence number of the phase-shifted fringe image; Represents the pixel coordinates of the phase-shifted fringe image; and These represent background intensity and modulation intensity, respectively. Represents the original phase-shifted fringe image The truncated phase; This indicates the additional phase shift caused by the motion of the object.

3. The dynamic three-dimensional measurement error correction method according to claim 1, characterized in that: In step S3, after the Hilbert transform, the phase-shifted fringe image of the first transformation is... It can be represented as: ； In the formula: and Representing the background intensity and modulation intensity, respectively, for a set of phase-shifted fringe images. The background intensity can be considered as and modulation intensity It remains unchanged.

4. The dynamic three-dimensional measurement error correction method according to claim 1, characterized in that: In step S3, the original phase-shifted fringe image The phase-shifted fringe image of the first transformation Corresponding truncation phase and as follows: ； ； In the formula: This represents the arctangent operation function, which truncates the phase. and The range of values ​​is limited to ; Furthermore, the average truncation phase as follows: ； In the formula: This represents the modulo operation function, hence the average truncation phase. The range of values ​​for is also limited to .

5. The dynamic three-dimensional measurement error correction method according to claim 1, characterized in that: In step S4, the virtual phase-shifted fringe image It can be represented as: ； Furthermore, after the Hilbert transform, the phase-shifted fringe image of the second transform... It can be represented as: 。 6. The dynamic three-dimensional measurement error correction method according to claim 1, characterized in that: In step S5, the phase-shifted fringe image of the second transformation. truncated phase as follows: ； Furthermore, the absolute phase It can be represented as: ； In the formula: Indicates the stripe level.

Images