High dynamic range three-dimensional measurement method for binary fringe differential demodulation

Abstract

The invention discloses a binary fringe differential demodulation high dynamic range three-dimensional measurement method, which comprises the following steps: generating a sine gray fringe, decomposing the sine gray fringe into a binary bit plane, and generating a binary bit fringe based on a four-step phase shift algorithm; projecting binary fringes generated by complementation and carrying out differential processing to acquire an image; synchronously triggering projection and camera acquisition, synthesizing an 8-bit grayscale modulation image and generating a phase diagram; parallax is calculated based on phase diagram matching, calibration parameters are combined to be converted into three-dimensional point cloud, and HDR surface reconstruction measurement is achieved. According to the method, three-dimensional reconstruction of a high dynamic range (HDR) surface is realized by utilizing binary fringe conversion and complementary fringe differential demodulation technologies, so that the problem that traditional gray-level fringes are easy to saturate or insufficient in signal in high reflectivity and reflectivity abrupt change areas is effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional measurement, and particularly to a high-dynamic-range three-dimensional measurement method based on binary fringe differential demodulation. Background Art

[0002] Regarding the three-dimensional reconstruction problem of high-dynamic-range (HDR) surfaces, traditional gray-level fringe projection methods are prone to interference when measuring surfaces with high reflectivity or abrupt changes in reflectivity, resulting in reconstruction failure or large errors. Specifically, the existing methods mainly face the problems of fringe distortion caused by abrupt changes in reflectivity of high-reflectivity surfaces and the limitations of traditional demodulation methods.

[0003] When measuring high-reflectivity surfaces such as smooth metals and polished ceramics, strong specular reflections can cause overexposure or loss of the fringe pattern, making it difficult for the camera to capture accurate fringe information. Since the brightness of some areas exceeds the dynamic range of the camera, phase calculation is distorted, thereby affecting the final three-dimensional reconstruction accuracy. When projecting fringes onto an object with local high-brightness or low-brightness regions (such as a surface with a pattern or a component with material changes), the fringe contrast will change due to abrupt changes in reflectivity. Traditional methods are difficult to simultaneously capture effective fringe information in both high-brightness and dark regions under a single exposure setting, resulting in measurement errors. Traditional phase-shifting methods and gray-level fringe projection rely on continuous gray-scale changes, but due to the non-linear response of the camera and the Gamma correction error of the projector, the reconstruction accuracy is limited. The method of multi-exposure fusion requires additional computing and storage resources and is difficult to implement in dynamic measurement scenarios.

[0004] There are many deficiencies in the prior art in the three-dimensional measurement of HDR surfaces. First, traditional gray-level fringe projection methods are prone to saturation in high-reflectivity regions, resulting in loss of fringe information, and the signal is insufficient in low-reflectivity regions, reducing the fringe contrast and thus affecting the demodulation accuracy. In addition, although the multi-exposure fusion technology can alleviate this problem to a certain extent, this method requires collecting multiple sets of images under different exposure conditions and performing complex fusion calculations, increasing the time cost of data processing and making it difficult to meet the requirements of real-time measurement. The region adaptive method can optimize the exposure of different regions, but due to the need to perform point-by-point calculations and adjustments on the measurement object, the computational complexity is greatly increased, making it difficult to be efficiently applied in dynamic measurement environments. At the same time, some existing methods rely on high-dynamic-range cameras or complex optical devices, such as polarization filters, adaptive optical systems, etc. The high cost of these devices limits their popularization and application in industrial measurement. Summary of the Invention

[0005] The object of the present invention is to overcome the deficiencies of the prior art. To achieve the above object, a high-dynamic-range three-dimensional measurement method based on binary fringe differential demodulation is adopted to solve the problems raised in the above background art.

[0006] A high-dynamic-range three-dimensional measurement method based on binary fringe differential demodulation, comprising the following steps:

[0007] Step S1: Decompose the generated sine gray-scale fringe into binary bit planes, and generate binary bit fringes based on the four-step phase-shift algorithm;

[0008] Step S2: Project the generated complementary binary bit fringes and perform differential processing to collect images;

[0009] Step S3: Synchronously trigger projection and camera acquisition, synthesize an 8-bit gray-scale modulated image and generate a phase map;

[0010] Step S4: Calculate the disparity based on the phase map matching, and convert it into a three-dimensional point cloud in combination with the calibration parameters to realize HDR surface reconstruction measurement.

[0011] As a further solution of the present invention: The specific steps in the step S1 include:

[0012] Step S11: Decompose the 8-bit sine gray-scale fringe according to the binary bit planes, convert the gray-scale value of each pixel into an 8-bit binary code, and generate 8 binary bit planes, where each plane corresponds to a bit, and the formula is:

[0013]

[0014] where, I n (i, j) is an eight-bit sine fringe, and its gray level is 0-255; is the generated binary bit fringe; m represents the number of phase shifts;

[0015] Step S12: Process the generated 8 binary bit planes based on the four-step phase-shift algorithm to generate 32 binary bit fringe modulation patterns.

[0016] As a further solution of the present invention: The specific steps in the step S2 include:

[0017] Project the preset generated binary bit fringes onto the target object, and then synchronously collect them by the camera;

[0018] And binarize the collected modulated image by using the binarization formula. Among the 4 groups of four-step phase-shift pattern sequences projected, the first group and the third group are complementary, and the second group and the fourth group are complementary;

[0019] The binarization formula is:

[0020]

[0021] where, I Bk (x, y) iand are respectively the images acquired by the camera when projecting complementary fringe patterns; is the matrix obtained after taking the difference of the corresponding pixel values.

[0022] As a further solution of the present invention: The specific steps in the step S3 include:

[0023] After obtaining the binary modulation information according to step S2, then synthesize through an 8-bit synthesis according to the synthesis formula, covering the 0-255 gray scale range, and finally generate the left and right phase diagrams from the synthesized modulation pattern;

[0024] The synthesis formula is:

[0025]

[0026] wherein, is the extracted binary bit information.

[0027] As a further solution of the present invention: The specific steps in the step S4 include:

[0028] After generating the left and right phase diagrams, through the phase diagram matching of the left and right cameras, calculate the disparity map through gray-scale similarity;

[0029] Then, according to the pre-calibrated system parameters, convert the disparity map into a three-dimensional point cloud, and generate high-quality and high-dynamic-range three-dimensional information through multi-exposure synthesis or adaptive weight adjustment.

[0030] Compared with the prior art, the present invention has the following technical effects:

[0031] By adopting the above technical solution, by decomposing the gray-scale fringes into binary fringes and using complementary fringes for differential demodulation, the fringe contrast is effectively improved, the dependence on the dynamic range of the camera is reduced, and at the same time, the complex calculations and expensive hardware requirements in the traditional method are avoided, so that the three-dimensional topography of the HDR surface can be accurately restored, the measurement efficiency is improved, and the stability and applicability of the system are enhanced. Aiming at the above problems, the present invention proposes a method based on binary fringe differential demodulation, which can accurately restore the three-dimensional topography of the HDR surface, while reducing the computational complexity and improving the measurement efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The following will describe in detail the specific embodiments of the present invention with reference to the drawings:

[0033] Figure 1 is a schematic diagram of the steps of the high-dynamic-range three-dimensional measurement method of the disclosed embodiment of the present application;

[0034] Figure 2Schematic diagram of the sine gray - scale stripe and its binary bit - planes according to the disclosed embodiments of the present application;

[0035] Figure 3 Schematic diagram of the modulation pattern binarized by the differential binary method according to the disclosed embodiments of the present application;

[0036] Figure 4 Flow chart of the high - dynamic - range three - dimensional measurement method according to the disclosed embodiments of the present application. Detailed implementation manners

[0037] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the protection scope of the present invention.

[0038] Please refer to Figure 1 and Figure 2 , in the embodiments of the present invention, a high - dynamic - range three - dimensional measurement method based on binary stripe differential demodulation includes the following steps:

[0039] Step S1: Generate sine gray - scale stripes decomposed into binary bit - planes, and generate binary bit - stripes based on the four - step phase - shifting algorithm. The specific steps include:

[0040] Step S11: Decompose the 8 - bit sine gray - scale stripe into binary bit - planes, convert the gray - scale value of each pixel into an 8 - bit binary code, and generate 8 binary bit - planes, where each plane corresponds to one bit. The formula is:

[0041]

[0042] where, I n (i,j) is an 8 - bit sine stripe with a gray - scale level of 0 - 255; is the generated binary bit - stripe; m represents the number of phase - shifts;

[0043] Step S12: Process the generated 8 binary bit - planes based on the four - step phase - shifting algorithm to generate 32 binary bit - stripe modulation patterns.

[0044] In this embodiment, as Figure 2 shown, the figure shows an 8 - bit sine gray - scale stripe and its 8 binary bit - planes. The specific generation method of this kind of binary bit - plane

[0045] Step S2: Project the complementary generated binary bit - stripes and perform differential processing to collect images. The specific steps include:

[0046] Project the preset generated binary fringe pattern onto the target object, and then synchronously collect it with the camera;

[0047] And binarize the collected modulated image using the binarization formula. Among the 4 groups of four-step phase-shift pattern sequences projected, the first group and the third group are complementary, and the second group and the fourth group are complementary;

[0048] The binarization formula is:

[0049]

[0050] Where, I Bk (x,y) i And Are the images collected by the camera when projecting complementary fringe patterns respectively; Is the matrix obtained after taking the difference of their corresponding pixel values.

[0051] In this embodiment, as Figure 3 Shown in the figure, the preset generated binary fringe pattern is projected onto the target object, and then synchronously collected with the camera. Finally, refer to formula 3 to binarize the collected modulated image. Among the 4 groups of four-step phase-shift pattern sequences projected, the first group and the third group, and the second group and the fourth group are all complementary. Using this relationship can make the binarization effect more accurate;

[0052] Step S3, synchronously trigger the projection and camera collection, synthesize an 8-bit gray-scale modulated image and generate a phase map. The specific steps include:

[0053] After obtaining the binary modulation information in step S2, then synthesize through 8-bit synthesis according to the synthesis formula, covering the 0-255 gray-scale range, and finally generate the left and right phase maps from the synthesized modulation pattern;

[0054] The synthesis formula is:

[0055]

[0056] Where, Is the extracted binary bit information.

[0057] In this embodiment, as Figure 4 Shown in the figure, it is the experimental specific flow chart of the method of this embodiment, and also includes the basic experimental device: consisting of a projector, two cameras and a target. Among them, the projector and the left and right cameras are synchronously triggered to improve the efficiency of image collection. After obtaining the binary modulation information by using the provided differential binary method, then synthesize through 8-bit synthesis according to the synthesis formula, covering the 0-255 gray-scale range to avoid saturation in the high-reflection area. Finally, generate the left and right phase maps from the synthesized modulation pattern;

[0058] Step S4: Calculate the disparity based on phase map matching, and convert it into a 3D point cloud in combination with the calibration parameters to achieve HDR surface reconstruction measurement. The specific steps include:

[0059] After generating the left and right phase maps, through the phase map matching of the left and right cameras, calculate the disparity map by gray similarity.

[0060] Then, according to the pre-calibrated system parameters, convert the disparity map into a 3D point cloud, and generate high-quality and high-dynamic-range 3D information through multi-exposure synthesis or adaptive weight adjustment.

[0061] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents, and all of them should be included within the protection scope of the present invention.

Claims

1. A high-dynamic-range three-dimensional measurement method based on binary fringe differential demodulation, characterized in that It includes the following steps: Step S1: Decompose the generated sine gray - scale stripes into binary bit - planes, and generate binary bit - stripes based on the four - step phase - shifting algorithm; Step S2: Project the complementary - generated binary bit - stripes and perform differential processing to collect images; Step S3: Synchronously trigger the projection and camera acquisition, synthesize an 8 - bit gray - scale modulated image and generate a phase map; Step S4: Calculate the disparity based on the phase - map matching, combine the calibration parameters to convert it into a three - dimensional point cloud, and realize the HDR surface reconstruction measurement.

2. The high-dynamic-range three-dimensional measurement method for binary fringe differential demodulation according to claim 1, wherein The specific steps in step S1 include: Step S11: Decompose the 8 - bit sine gray - scale stripes according to the binary bit - planes. The gray - scale value of each pixel is converted into an 8 - bit binary code, generating 8 binary bit - planes, where each plane corresponds to one bit. The formula is: Among them, I n (i, j) is an eight-bit sine stripe, and its gray level is 0 - 255; is the generated binary stripe; m represents the number of phase shifts; Step S12: Process the 8 generated binary bit - planes based on the four - step phase - shifting algorithm to generate 32 binary bit - stripe modulation patterns.

3. The high-dynamic-range three-dimensional measurement method for binary fringe differential demodulation according to claim 1, wherein The specific steps in step S2 include: Project the pre - generated binary bit - stripes onto the target object, and then synchronously collect them by the camera; And binarize the collected modulated image using the binarization formula. Among the 4 groups of four - step phase - shifting pattern sequences projected, the first group and the third group are complementary, and the second group and the fourth group are complementary; The binarization formula is: Among them, I Bk (x, y) i and are respectively the images collected by the camera when projecting complementary bit fringes; is the matrix obtained after taking the difference of the corresponding pixel values.

4. The high-dynamic-range three-dimensional measurement method for binary fringe differential demodulation according to claim 1, characterized in that The specific steps in step S3 include: After obtaining the binary modulation information in step S2, then synthesize it through an 8 - bit synthesis according to the synthesis formula, covering the 0 - 255 gray - scale range, and finally generate the left - and - right phase maps from the synthesized modulation patterns; The synthesis formula is: Among them, is the extracted binary bit information.

5. The high-dynamic-range three-dimensional measurement method based on binary fringe differential demodulation according to claim 1, characterized in that The specific steps in step S4 include: After generating the left - and - right phase maps, through the phase - map matching of the left - and - right cameras, calculate the disparity map by gray - scale similarity; Then, according to the pre - calibrated system parameters, convert the disparity map into a three - dimensional point cloud, and generate high - quality and high - dynamic - range three - dimensional information through multi - exposure synthesis or adaptive weight adjustment.

Images