A self-increment order-based encoding structured light order correction algorithm

Abstract

The application discloses a kind of based on self-increment order encoding structured light order correction algorithm, the present application can effectively inhibit phase jump error, it can more accurately reconstruct the mutation part of object surface, effectively improve the precision of measurement;The method of the present application is not only suitable for smooth, simple object, but also can effectively deal with surface mutation and complex texture object, it has strong adaptability and wide applicability, by the existing grating realization order self-increment and order noise removal, both can guarantee the data precision of three-dimensional measurement, and can guarantee running processing speed.

Description

Technical Field

[0001] This invention belongs to the field of optical three-dimensional measurement, specifically relating to a coding structured light order correction algorithm based on self-increasing order. Background Technology

[0002] Structured light 3D measurement, with its advantages of non-contact operation, high precision, and real-time performance, occupies an important position in the field of optical measurement. Among them, coded structured light technology, due to its simple structure and strong robustness, is widely used in industrial inspection, human-computer interaction, and 3D printing. Existing systems use a projector to project two sets of digital gratings—one for phase shift and one for encoding—combined with camera acquisition and computer processing to achieve 3D reconstruction of objects. However, this technology faces two major challenges: First, obtaining the fringe order is difficult, especially in the phase-jump region. Due to instability near the phase boundary, order errors are easily caused, affecting reconstruction accuracy. This difficulty stems from the complex correspondence between the coded digital grating and the phase period, as well as the dynamic changes in the phase-jump region. Second, measurement speed is limited. Current phase-coded structured light technology is still immature and cannot meet the needs of high-speed measurement.

[0003] To address the aforementioned issues, numerous solutions exist. Based on whether the removal of order errors occurs before or after obtaining the reconstructed 3D data, these solutions can be categorized into post-correction methods and pre-avoidance methods. Regarding post-correction methods, Professor Zhang Song of Purdue University proposed detecting and eliminating errors by judging the monotonicity of the unwrapped phase; however, this method increases computational load and may cause neighbor contamination. The research group of Da Feipeng at Southeast University proposed an adaptive median filtering method, which can mitigate the impact of noise and reduce neighbor contamination; however, this method is time-consuming, and further research is needed on the real-time performance of fringe projection contour measurements.

[0004] Regarding the pre-avoidance method, Zhang Qican et al. proposed the complementary Gray code technique [3], which obtains two interlaced fringe orders by projecting additional fringe patterns. Although this method can effectively prevent fringe order errors, it is based on sacrificing measurement speed. Wu Zhoujie et al. pre-avoided the generation of fringe order errors by constructing two additional interlaced phases from the original fringe sequence and dividing each period of the fringe order into three parts. However, when the phase quality of the wrapped phase is low, error propagation problems will occur. Overall, there is currently no good method that can solve the problem of decreased reconstruction accuracy caused by order noise due to boundary instability in 3D reconstruction of coded structured light while ensuring measurement speed. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a coded structured light order correction algorithm based on auto-incrementing order. This algorithm achieves order increment and order noise removal using existing gratings without requiring additional projection gratings, filtering, or phase shift order swapping. This ensures both the accuracy of 3D measurement data and the processing speed.

[0006] To achieve the above technical effects, this invention designs a code structured light order correction algorithm based on self-increasing order, specifically including the following steps:

[0007] S1: Construct a structured light 3D measurement system consisting of a projection system and a camera;

[0008] S2: Project the sinusoidal pattern and phase-coded pattern onto the object surface in the form of an image sequence;

[0009] S3: The wrapping phase φ(x,y) of the sinusoidal pattern and the fringe order k(x,y) of the coded pattern are solved by the three-step phase shift method and the related decoding algorithm, respectively.

[0010] S4: By adjusting the fringe modulation intensity b(x,y) and the threshold T for region segmentation h The region of origin (ROI) is obtained by comparing (x,y), and then the wrapping phase in R(x,y) is solved based on the region segmentation. and stripe order k R (x,y);

[0011] S5: Applying the area threshold function to k R Small-area noise points in (x,y) are eliminated to obtain k. RZ (x,y); Next, find the position of the minimum phase point in each row within this order range, and finally obtain the phase mask M(x,y) by comparing the position size of each row;

[0012] S6: For the stripe order k RZ The order of the interlaced fringe Z(x,y) is obtained by combining the phase mask M(x,y) and the phase mask M(x,y).

[0013] S7: By... The ROI region is divided into three areas based on the assessment.

[0014] S8: Solve for the absolute phase based on the order of the interlaced fringes and the divided regions.

[0015] Furthermore, in step S4, the wrapping phase in R(x,y) is solved. and stripe order k R The method for (x,y) is as follows:

[0016]

[0017] k R (x,y)=k(x,y)*R(x,y)

[0018] Where b(x,y) is the modulation intensity of the grating, I i (i = 1, 2, 3) represents the intensity of the sinusoidal grating, T h This is the threshold for region segmentation.

[0019] Furthermore, the method for obtaining the phase mask M(x,y) in S5 is as follows:

[0020] S5.1: Applying a binary area thresholding algorithm to k R Small-area noise points in (x,y) are eliminated to obtain k. RZ (x,y);

[0021] S5.2: Traversing corresponding to k RZ The position of the minimum phase value in each row within the (x,y) order range. For example, the position of the j-th order is denoted as... Make the row position within each order range greater than Set the region to 1 and the other regions to 0; after traversing each order, obtain the composite region mask M(x,y).

[0022] Furthermore, in step S6, the fringe order k is... RZ The method for obtaining the order Z(x,y) of the interlaced fringe by combining (x,y) and the phase mask M(x,y) is as follows:

[0023]

[0024] The round[·] function is a rounding function.

[0025] Furthermore, in step S7, by... The method for determining and dividing the ROI region into three regions is as follows:

[0026]

[0027] Furthermore, in step S8, the method for calculating the absolute phase based on the solved order of the interlaced fringes and the divided regions is as follows:

[0028] Φ(x,y)=[φ(x,y)+2πZ(x,y)]R1(x,y)+[φ(x,y)+2πk R (x,y)]R2(x,y)+[φ(x,y)+2π(Z(x,y)-1)]R3(x,y)

[0029] Where φ(x,y) is the wrapping phase of the sinusoidal coding pattern.

[0030] Furthermore, the structured light 3D measurement system includes an industrial CCD camera, a structured light projection system, and a computer system, wherein the industrial CCD camera and the structured light projection system are both communicatively connected to the computer system.

[0031] The beneficial effects of this invention are:

[0032] 1. This invention does not require additional projection gratings, filtering, or phase shift order swapping. It avoids phase jump regions from appearing at the edge of the order by simply incrementing the existing order, thereby simplifying the experimental operation and improving the ease of use.

[0033] 2. This invention achieves order increment and order noise removal using existing gratings without the need to project additional gratings or perform filtering operations. This ensures both the accuracy of 3D measurement data and the processing speed, and effectively avoids the phenomenon of neighborhood contamination. Attached Figure Description

[0034] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below.

[0035] Figure 1 This is a block diagram of the structured light three-dimensional measurement system used in the self-increasing order-based coded structured light order correction algorithm of the present invention.

[0036] Figure 2 This is a flowchart of a coding structured light order correction algorithm based on self-increasing order according to the present invention;

[0037] Figure 3 The measurement target for sudden surface changes in Embodiment 2 of the present invention;

[0038] Figure 4 These are the Gray code and phase-coded modulation fringes captured by the camera in Embodiment 2 of the present invention, where a is the Gray code (GC) modulation fringes and b is the phase-coded (PC) modulation fringes.

[0039] Figure 5 This is an absolute phase diagram of Gray code and phase encoding in Embodiment 2 of the present invention; in the figure, a is the absolute phase diagram of Gray code (GC) and b is the absolute phase diagram of phase encoding (PC);

[0040] Figure 6 This is a comparison diagram of the stripe order of Gray code and phase code in Embodiment 2 of the present invention; in the figure, a is a comparison of the stripe order of (a) Gray code (GC) and b is a comparison of the stripe order of phase code (PC);

[0041] Figure 7 The figures show the absolute phase diagrams of GC and PC after correction using the method of this invention; in the figure, a is the absolute phase diagram of GC after correction, and b is the absolute phase diagram of PC after correction.

[0042] Figure 8 The measurement target with complex texture in Embodiment 3 of the present invention;

[0043] Figure 9 This is the absolute phase diagram of GC and PC in Embodiment 3 of the present invention; a in the figure is the absolute phase diagram of Gray code (GC); b is the absolute phase of phase code (PC);

[0044] Figure 10 The figures show the absolute phases of GC and PC after correction processing according to the method of the present invention in Embodiment 3 of the present invention; in the figure, a is the absolute phase diagram of GC after correction processing, and b is the absolute phase diagram of PC after correction processing. Detailed Implementation

[0045] Example 1

[0046] This invention discloses a code structured light order correction algorithm based on auto-incrementing order, comprising the following steps:

[0047] S1: Construct a structured light 3D measurement system consisting of an industrial CCD camera, a structured light projection system, and a computer system. The structural diagram of the structured light 3D measurement system is shown below. Figure 1 As shown, specifically, the structured light projection system includes a projector and a mounting bracket. The projector is mounted on the mounting bracket, which allows adjustment of the projector's height, angle, etc. The mounting bracket is existing technology and is not shown in the figure. Two industrial CCD cameras are provided, symmetrically arranged around the projector's optical axis. The industrial CCD cameras output raw data with a wide spectral range, and also possess advantages such as low power consumption, high sensitivity, and fast response speed, making them suitable for high-quality image processing algorithms. Both the industrial CCD cameras and the structured light projection system are communicatively connected to a computer system. The computer system can control the structured light projection system, and the images acquired by the industrial CCD cameras are transmitted to the computer system for processing. The computer system includes a storage module, a communication module, an image acquisition module, and a 3D reconstruction module. The modules included in the computer system are existing technologies and can be directly selected and used; further explanation is not provided here.

[0048] S2: Project the sinusoidal pattern and phase-coded pattern onto the object surface in the form of an image sequence.

[0049] S3: The wrapping phase φ(x,y) of the sinusoidal pattern and the fringe order k(x,y) of the coded pattern are solved by the three-step phase shift method and the related decoding algorithm, respectively.

[0050] S4: By adjusting the fringe modulation intensity b(x,y) and the threshold T for region segmentation h By comparing (x,y), the ROI region is obtained, and then the wrapping phase in R(x,y) is solved based on the region segmentation. and stripe order k R (x,y). Obtain the ROI region, and then solve for the wrapping phase in R(x,y) based on the region segmentation. and stripe order k R The method for (x,y) is as follows:

[0051]

[0052] k R (x,y)=k(x,y)*R(x,y)

[0053] Where b(x,y) is the modulation intensity of the grating, I i (i = 1, 2, 3) represents the intensity of the sinusoidal grating, T h This is the threshold for region segmentation.

[0054] S5: Applying the area threshold function to k R Small-area noise points in (x,y) are eliminated to obtain k. RZ (x,y). Next, find the position of the minimum phase point in each row within this order range, and finally obtain the phase mask M(x,y) by comparing the position of each row. Obtain k RZ The method for obtaining the phase mask M(x,y) by finding the minimum phase value point of each row within this order range and then comparing the position of each row is as follows:

[0055] The binary area thresholding algorithm is used to evaluate k. R Eliminate small-area noise points in (x,y); traverse the area corresponding to k RZ The position of the minimum phase value in each row within the (x,y) order range. For example, the position of the j-th order is denoted as... Make the row position within each order range greater than Set the region to 1 and the other regions to 0. After traversing each order, obtain the composite region mask M(x,y).

[0056] S6: For the stripe order k RZ The order of the interlaced fringe Z(x,y) is obtained by combining the phase mask M(x,y) and the phase mask M(x,y). The method for obtaining the order of the interlaced fringe Z(x,y) is as follows:

[0057]

[0058] The round[·] function is a rounding function.

[0059] S7: Through the The ROI region is then divided into three regions based on the following criteria:

[0060]

[0061] S8: Solve for the absolute phase based on the obtained order of the interlaced fringes and the divided regions. The method for solving for the absolute phase is as follows:

[0062] Φ(x,y)=[φ(x,y)+2πZ(x,y)]R1(x,y)+[φ(x,y)+2πk R (x,y)]R2(x,y)+[φ(x,y)+2π(Z(x,y)-1)]R3(x,y)

[0063] Where φ(x,y) is the wrapping phase of the sinusoidal coding pattern.

[0064] Example 2

[0065] In this embodiment, based on the structured light three-dimensional measurement system and method constructed in Embodiment 1, further investigation and verification are carried out by conducting experiments on objects with abrupt surface changes. The specific process is as follows:

[0066] like Figure 3 As shown, the selected mutation object consists of a sculpture, a turbine blade, and a section of standard railway track with a surface mutation:

[0067] First, a stripe projection system is constructed, consisting of a projector and a camera. The specific structure is as follows: Figure 1 As shown, stripes are projected onto the target object using a projector, and then the pattern modulated on the surface of the object is captured by a camera. During the experimental measurement, the measuring platform is placed 500-600mm in front of the target object.

[0068] Then, two representative phase analysis methods in structured light coding technology, namely Gray coding (GC) and Phase coding (PC), were selected for encoding. In this embodiment, GC and PC were used to calculate the wrapped phase and fringe order. GC used a three-step phase-shifting method with a frequency of 57 to solve the wrapped phase and 7 binary fringes for phase unwrapping, while PC used a three-step phase-shifting method with a frequency of 10 to solve the wrapped phase and 3 coded fringes for experimentation. The final Gray code and phase-coded modulation fringes captured by the camera are shown below. Figure 4 As shown in a and b in the figure;

[0069] Phase unrolling was performed on both methods to obtain the corresponding absolute phases. The absolute phase maps for GC and PC are shown below. Figure 5 As shown in a and b; then the captured edge parts that are prone to phase jumps are processed using the correction method proposed in this invention, and the phases of the two methods are re-expanded to obtain the corresponding corrected absolute phase and corrected fringe order.

[0070] Specifically, the stripe orders of GC and PC after correction using the method described in this paper are as follows: Figure 6 As shown in Figures a and b, the absolute phases of GC and PC after correction processing using the method of this invention are as follows: Figure 7 As shown in a and b. From Figure 5 As can be seen from Figures a and b, both the GC and PC methods are affected by phase jump errors, and the number of phase jump error points is relatively large at the edges of both figures. Figure 6 It can be seen that the jump points of the original fringe order were successfully eliminated after correction by both methods, and the number of jump point errors was reduced. Meanwhile, from Figure 7 and Figure 5 The comparison shows that, within the same region, the phase jump errors obtained by the two original methods are significantly improved to a certain extent after being corrected by the present invention. In summary, the results indicate that the method of the present invention has a strong ability to suppress phase jump errors when measuring objects with abrupt surface changes.

[0071] Example 3

[0072] In this embodiment, experiments are conducted on objects with complex textures. The selected measurement target is a statue with complex textures placed in front of the panel, specifically as follows: Figure 8 As shown;

[0073] A fringe projection system as described in Example 2 is constructed, and the measurement target is encoded using Gray-coding (GC) and Phase-coding (PC). The absolute phases of both methods are then obtained through fringe projection and phase unwrapping. The absolute phase diagrams of GC and PC in this example are shown below. Figure 9 As shown in a and b

[0074] Similarly, the captured portion prone to phase jumps is processed using the correction method proposed in this invention, and the phases corrected by the two methods are re-expanded to obtain the corrected absolute phase.

[0075] The absolute phases of GC and PC after correction using the method in this invention are as follows: Figure 10 As shown in a and b. From Figure 10 As shown in Figures a and b, both the GC and PC methods are affected by phase jump errors, and the PC method is more prone to decoding errors when measuring complex textures. Figure 10 As can be seen from the two figures, after correction using this method, most of the phase transition errors are corrected, and... Figure 9 Compared to method b, this method is less affected by decoding errors in the PC method. In summary, the results show that the method in this invention also has a strong ability to suppress phase jump errors when measuring objects with complex surface textures.

[0076] In summary, the results of Examples 2 and 3 demonstrate that the present invention can effectively suppress phase jump errors and more accurately reconstruct abrupt changes on the surface of an object, thereby improving measurement accuracy. The method of the present invention is not only applicable to objects with smooth and simple surfaces, but also effectively addresses objects with abrupt changes and complex textures. It has strong adaptability and wide applicability. By using existing gratings to achieve order increment and order noise removal, it can ensure both the data accuracy of three-dimensional measurements and the processing speed.

[0077] The preferred embodiments of the present invention disclosed above are only for the purpose of illustrating the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation described.

Claims

1. A code structured light order correction algorithm based on auto-increasing order, characterized in that, Includes the following steps: S1: Construct a structured light 3D measurement system consisting of a projection system and a camera; S2: Project the sinusoidal pattern and phase-coded pattern onto the object surface in the form of an image sequence; S3: The wrapping phase φ(x,y) of the sinusoidal pattern and the fringe order k(x,y) of the coded pattern are solved by the three-step phase shift method and the related decoding algorithm, respectively. S4: By adjusting the fringe modulation intensity b(x,y) and the threshold T for region segmentation h By comparing (x,y) to obtain the ROI region, and then solving for the wrapping phase in R(x,y) based on the region segmentation. and stripe order k R (x,y); S5: Applying the area threshold function to k R Small-area noise points in (x,y) are eliminated to obtain k. RZ (x,y); Next, find the position of the minimum phase point in each row within this order range, and finally obtain the phase mask M(x,y) by comparing the position size of each row; S6: For the stripe order k RZ The order of the interlaced fringe Z(x,y) is obtained by combining the phase mask M(x,y) and the phase mask M(x,y). S7: By... The ROI region is divided into three areas based on the assessment. S8: Solve for the absolute phase based on the order of the interlaced fringes and the divided regions.

2. The coded structured light order correction algorithm based on self-increasing order as described in claim 1, characterized in that, In step S4, the wrapping phase in R(x,y) is solved. and stripe order k R The method for (x,y) is as follows: k R (x,y)=k(x,y)*R(x,y) Where b(x,y) is the modulation intensity of the grating, I i (i = 1, 2, 3) represents the intensity of the sinusoidal grating, T h This is the threshold for region segmentation.

3. The coded structured light order correction algorithm based on self-increasing order as described in claim 2, characterized in that, The method for obtaining the phase mask M(x,y) in S5 is as follows: S5.1: Applying a binary area thresholding algorithm to k R Small-area noise points in (x,y) are eliminated to obtain k. RZ (x,y); S5.2: Traversing corresponding to k RZ The position of the minimum phase value in each row within the (x,y) order range. For example, the position of the j-th order is denoted as... Make the row position within each order range greater than Set the region to 1 and the other regions to 0; after traversing each order, obtain the composite region mask M(x,y).

4. The coded structured light order correction algorithm based on self-increasing order as described in claim 1, characterized in that, In step S6, the fringe order k RZ The method for obtaining the order Z(x,y) of the interlaced fringe by combining (x,y) and the phase mask M(x,y) is as follows: The round[·] function is a rounding function.

5. The coded structured light order correction algorithm based on self-increasing order as described in claim 1, characterized in that, In step S7, by... The method for determining and dividing the ROI region into three regions is as follows:

6. The coded structured light order correction algorithm based on self-increasing order as described in claim 1, characterized in that, In step S8, the method for solving the absolute phase based on the solved order of the interlaced fringes and the divided regions is as follows: Φ(x,y)=[φ(x,y)+2πZ(x,y)]R1(x,y)+[φ(x,y)+2πk R (x,y)]R2(x,y)+[φ(x,y)+2π(Z(x,y)-1)]R3(x,y) Where φ(x,y) is the wrapping phase of the sinusoidal coding pattern.

7. The coded structured light order correction algorithm based on self-increasing order as described in claim 1, characterized in that, The structured light 3D measurement system includes an industrial CCD camera, a structured light projection system, and a computer system. Both the industrial CCD camera and the structured light projection system are communicatively connected to the computer system.

Images