Binocular stereoscopic vision three-dimensional reconstruction method and system based on polarization state

Abstract

The invention relates to a binocular stereoscopic vision three-dimensional reconstruction method and system based on a polarization state. The method comprises the following steps: acquiring polarization images of a polarization camera under different polarization degrees; according to the obtained polarization image, the angle of a normal vector is calculated based on the Fresnel theory, and the angle of the normal vector comprises the zenith angle and the azimuth angle of the surface of the object; the ambiguity of the azimuth angle is eliminated through the projection included angle of the normal vector on the YOZ plane, and mismatching is eliminated by using the angle constraint of the normal vector; and performing gradient integration on the angle of the normal vector after mismatching elimination to complete three-dimensional reconstruction. According to the method, the projection included angle is utilized to assist in judging the quadrant of the normal, so that mismatching of the corresponding points is eliminated, and the measurement efficiency is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of computer vision technology, and particularly to a binocular stereo vision three-dimensional reconstruction method and system based on polarization state. Background Art

[0002] The application of non-contact three-dimensional shape measurement technology in industrial inspection has gradually become one of the core technologies with the rapid development of intelligent manufacturing in recent years. Especially for the requirements of high robustness and high efficiency, it has become a key factor in improving production quality and ensuring product qualification rate. However, traditional binocular vision reconstruction methods often encounter difficulties when dealing with low-texture and highly reflective objects. Especially when there are no obvious texture features on the target surface, it often leads to problems such as data holes or inaccurate depth information. To solve this problem, a three-dimensional reconstruction method based on polarization information has emerged. The polarization technology utilizes the polarization characteristics of reflected light and combines with Fresnel's theorem. When parallel incident light is reflected by the object surface, it causes differences in the polarization information of the reflected light. Polarization three-dimensional reconstruction is to use the polarization information of the reflected light to inversely calculate the object shape, effectively suppressing the interference of strong light reflection and improving the accuracy and robustness of three-dimensional reconstruction, especially suitable for the reconstruction of low-texture and highly reflective surfaces.

[0003] Maeda R et al. proposed a method for decomposing the reflection component based on the rotation direction of the polarization plane, which solved the problem of specular mutual reflection on metal objects and promoted the application development of the polarization state of light in three-dimensional measurement. Wolff L B et al. studied the polarization-based light reflection model, proposed to use the Fresnel reflection coefficient for light reflection and transmission analysis, and explored the application of polarization technology in computer vision. Ping Qianqian et al. proposed a three-dimensional reconstruction method based on polarization binocular vision. By combining the world coordinates of a small number of feature points obtained by binocular stereo vision, the point cloud data in the pixel coordinate system obtained by polarization imaging was converted into absolute data in the world coordinate system, effectively solving the three-dimensional reconstruction problem of smooth and highly reflective targets. Tian X et al. innovatively improved the accuracy and quality of three-dimensional reconstruction by fusing polarization imaging and binocular stereo vision, correcting azimuth ambiguity and introducing low-rank matrix decomposition constraints. However, it is insufficient for high-reflection and low-texture objects, and has a high computational complexity and insufficient real-time performance. However, in polarization binocular vision reconstruction, the ambiguity problem of the azimuth angle and zenith angle of the normal vector is still an important challenge. Summary of the Invention

[0004] To solve the problem that there are multiple equivalent solutions for the definitions of the azimuth angle and zenith angle in traditional methods, such that the same normal vector may correspond to different angle combinations, especially in a complex binocular vision system. The present disclosure proposes a binocular stereo vision three-dimensional reconstruction method based on polarization state to solve the above problems.

[0005] According to one aspect of the present disclosure, a three-dimensional reconstruction method for binocular stereoscopic vision based on polarization state is provided, including: S10. Obtain polarization images at different polarization degrees of a polarization camera; S20. Based on the obtained polarization images, calculate the angle of the normal vector according to Fresnel theory, where the angle of the normal vector includes the zenith angle and the azimuth angle of the object surface; S30. Eliminate the ambiguity of the azimuth angle through the projection angle of the normal vector in the YOZ plane, and eliminate false matches by using the angle constraint of the normal vector; S40. Perform gradient integration on the angle of the normal vector after eliminating false matches to complete three-dimensional reconstruction.

[0006] Preferably, calculating the angle of the normal vector based on Fresnel theory includes: calculating the Stokes parameters according to the obtained polarization images, and further obtaining the polarization degree and the polarization azimuth angle, expressed as: , , , , , In the formula, , , and respectively represent the polarization images of the linear polarization degree in the directions of 0°, 45°, 90°, and 135°, is the polarization degree, is the polarization azimuth angle.

[0007] Preferably, calculating the angle of the normal vector based on Fresnel theory further includes: obtaining the azimuth angle of the object surface according to the relationship between the polarization azimuth angle and the azimuth angle, expressed as: , In the formula, is the azimuth angle.

[0008] Preferably, calculating the angle of the normal vector based on Fresnel theory further includes: according to the relationship formula between the polarization degree and the zenith angle, numerically iteratively solve the relationship formula to obtain the zenith angle of the object surface, and the relationship formula between the polarization degree and the zenith angle is expressed as: , In the formula, is the zenith angle, is the complex refractive index, =n(1 + ik), where i is the imaginary unit and k is the attenuation index.

[0009] Preferably, the ambiguity of the azimuth angle is eliminated by the projection angle of the normal vector on the YOZ plane, including: according to the relationship between the polarization azimuth angle and the azimuth angle, introducing the projection angle of the normal vector on the YOZ plane, simulating rotation through a binocular vision system, judging the sign of the projection angle, and determining the quadrant where the azimuth angle is located.

[0010] Preferably, the three-dimensional reconstruction is completed by performing gradient integration on the angles of the normal vectors after eliminating false matches, including: converting the zenith angle and the azimuth angle into three-dimensional normal vectors, expressed as: , wherein, is the normal vector of any point on the surface.

[0011] Preferably, the three-dimensional reconstruction is completed by performing gradient integration on the angles of the normal vectors after eliminating false matches, and further includes: Converting the three-dimensional normal vector into a discrete approximate gradient through local linear approximation; Fitting the discrete approximate gradient to the surface gradient field by minimizing the cost function, and reconstructing the object surface shape through numerical integration.

[0012] According to one aspect of the present disclosure, a binocular stereo vision three-dimensional reconstruction system based on polarization state is provided, including: A polarization image acquisition module for acquiring polarization images at different polarization degrees of a polarization camera; A Fresnel theory calculation module for calculating the angle of the normal vector based on the acquired polarization images according to the Fresnel theory, wherein the angle of the normal vector includes the zenith angle and the azimuth angle of the object surface; A false match elimination module for eliminating the ambiguity of the azimuth angle by the projection angle of the normal vector on the YOZ plane, and eliminating false matches by using the angle constraint of the normal vector; A three-dimensional reconstruction module for completing three-dimensional reconstruction by performing gradient integration on the angles of the normal vectors after eliminating false matches.

[0013] According to one aspect of the present disclosure, an electronic device is provided, including: a processor; a memory for storing processor-executable instructions; wherein, the processor is configured to: execute the above-mentioned binocular stereo vision three-dimensional reconstruction method based on polarization state.

[0014] According to one aspect of the present disclosure, a computer-readable storage medium is provided, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the above-mentioned binocular stereo vision three-dimensional reconstruction method based on polarization state is implemented.

[0015] Compared with the prior art, the beneficial effects of the present disclosure are: 1) The present disclosure proposes a new polarization-based binocular ambiguity elimination (PBAE) method to solve the ambiguity problem in normal integration in high-precision surface reconstruction.

[0016] 2) The present disclosure uses the projection angle to assist in determining the normal quadrant, eliminates point mismatches, improves the measurement efficiency, and effectively solves the point cloud hole problem.

[0017] 3) The present disclosure provides a simple and effective solution to correct normal integration and mitigate missing data in the point cloud, with high robustness and accuracy.

[0018] It should be understood that the above general description and the following detailed description are merely exemplary and explanatory, and do not limit the present disclosure.

[0019] Other features and aspects of the present disclosure will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are incorporated herein and constitute a part of this specification. These drawings illustrate embodiments consistent with the present disclosure and, together with the specification, are used to explain the technical solutions of the present disclosure.

[0021] Figure 1 Flowchart showing a binocular 3D reconstruction method based on polarization state; Figure 2 Schematic diagram showing polarization images at different polarization degrees of a polarization camera in an example of the present disclosure; Figure 3 Schematic diagram showing binocular ambiguity elimination assisted by the projection angle in an example of the present disclosure; Figure 4 Schematic diagram showing 3D reconstruction by gradient integration of the normal vector information after eliminating false matches in an example of the present disclosure; Figure 5 Schematic block diagram showing a binocular 3D reconstruction system based on polarization state in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0022] Various exemplary embodiments, features, and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. Like reference numerals in the drawings denote functionally identical or similar elements. Although various aspects of the embodiments are shown in the drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0023] The term "exemplary" used herein means "serving as an example, embodiment, or illustration". Any embodiment described herein as "exemplary" is not necessarily to be construed as superior to or better than other embodiments.

[0024] As used herein, the term "and / or" is merely a description of the relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent three cases: A exists alone, A and B exist simultaneously, and B exists alone. Additionally, the term "at least one" as used herein means any one of multiple items or any combination of at least two of multiple items. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set composed of A, B, and C.

[0025] In addition, for a better illustration of the present disclosure, numerous specific details are given in the following detailed implementation manners. Those skilled in the art should understand that the present disclosure can still be implemented without certain specific details. In some instances, methods, means, elements, and circuits well-known to those skilled in the art are not described in detail so as to highlight the gist of the present disclosure.

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0027] Embodiment 1 Based on the above idea, the present invention proposes a binocular stereo vision three-dimensional reconstruction method based on polarization state. Figure 1 The flowchart of a binocular stereo vision three-dimensional reconstruction method based on polarization state is shown. The method includes: S10. Obtain polarization images at different polarization degrees of a polarization camera; S20. Based on the obtained polarization images, calculate the angle of the normal vector according to the Fresnel theory, where the angle of the normal vector includes the zenith angle and the azimuth angle of the object surface; S30. Eliminate the ambiguity of the azimuth angle through the projection angle of the normal vector in the YOZ plane, and eliminate false matches using the angle constraint of the normal vector; S40. Perform gradient integration on the angle of the normal vector after eliminating false matches to complete three-dimensional reconstruction.

[0028] The embodiments of the present disclosure provide a binocular stereo vision three-dimensional reconstruction method based on polarization state, which specifically includes the following steps: S10. Obtain polarization images at different polarization degrees of a polarization camera.

[0029] In this embodiment, light waves belong to transverse waves, and their vibration direction is perpendicular to the propagation direction. In nature, the vibration direction of natural light is evenly distributed in all directions, so it belongs to unpolarized light. For any plane light wave, its optical vector can be decomposed into two mutually orthogonal components, namely the s component vibrating perpendicular to the incident plane and the p component parallel to the incident plane. When the distribution of the s component and the p component in a beam of light is uneven, the polarization phenomenon of light will occur.

[0030] According to the Fresnel equation, it can be known that the reflectivity R will change accordingly with the change of the vibration direction of the optical vector. For vertically polarized light (s-polarization), the reflectivity is related to the incident angle , and the calculation formula is: , wherein, is the refractive index of the incident medium, is the refractive index of the transmitted medium, is the angle of the incident light, is the angle of the refracted light.

[0031] For parallel polarized light (p-polarization), the reflectivity The calculation formula is: , According to Malus' law, by detecting the light intensity received by the detector placed after the linear polarizer, the polarization state of light can be inferred, which is expressed as: , wherein, is the angle between the transmission axis of the polarizer and the starting position (reference position), and respectively represent the maximum and minimum light intensities observed at a certain point during the rotation of the polarizer, is the polarization azimuth angle, which can be defined as when , the maximum brightness is observed, and when , the minimum brightness is observed.

[0032] S20. According to the obtained polarization image, calculate the angle of the normal vector based on the Fresnel theory, wherein the angle of the normal vector includes the zenith angle and the azimuth angle of the object surface.

[0033] In this embodiment, the polarization information of the reflected light on the object surface can be obtained through the polarization imaging technology. The polarization information of the reflected light contains the normal vector information of the object surface. The schematic diagram of the polarization images at different polarization degrees of the polarization camera is as Figure 2As shown. Specifically, by analyzing the polarization state of the reflected light, according to the acquired polarization image, the Stokes parameters are calculated, and further the degree of polarization and the polarization azimuth angle are obtained, which are expressed as: , , , , , In the formula, , , and respectively represent the polarization images of the degree of linear polarization in the directions of 0°, 45°, 90° and 135°, is the degree of polarization, is the polarization azimuth angle.

[0034] In specular reflected light, the angle of incidence is equal to the zenith angle. According to the relationship between the degree of polarization and the zenith angle, by numerically iteratively solving the relationship, the zenith angle of the object surface is obtained, and the specularly reflected linear polarization degree of the metal target surface has the following geometric expression: , In the formula, is the zenith angle, is the complex refractive index, =n(1 + ik), where i is the imaginary unit and k is the attenuation index.

[0035] The polarization phase angle is the angle between the vibration direction of the linear polarization component in the partially polarized light and the starting axis. In specular reflected light, the reflectivity of the vertical component is higher than that of the horizontal component, and the vibration direction of the linear polarization component in the generated partially polarized light is perpendicular to the reflection plane, that is, the azimuth angle and the polarization phase angle differ by 90°.

[0036] Azimuth angle is directly related to the polarization azimuth angle . According to the relationship between the polarization azimuth angle and the azimuth angle, the azimuth angle of the object surface is obtained, which is expressed as: , In the formula, is the azimuth angle.

[0037] The disambiguation of the azimuth angle is to add a constraint condition, and select a more reliable one from two that differ by π as the final solution. The normal vector in the three-dimensional coordinate system has a directionality. The angle between the projection of the normal vector on the YOZ plane and the Z axis is defined as the projection angle , which is expressed as: , In the formula, is the projection angle.

[0038] The projection angle between the normal vector and the Z-axis may have different signs in different quadrants. Taking the XOZ plane as the dividing quadrant, when the normal vector points to the right it is defined as positive, and when the normal vector points to the left it is defined as negative. Therefore, the quadrant where the normal vector is located can be determined by combining the polarization azimuth angle and the positive and negative of the projection angle . The XYZ coordinate system divides the space into 8 quadrants. The normal vector in this embodiment only exists in the four quadrants on the positive semi-axis of the Z-axis. The four quadrants where the positive Z-axis is located are divided into quadrants 1, 2, 3, and 4 in a clockwise direction.

[0039] Polarization phase angle . When , the normal vector falls in the second or fourth quadrant, ; when , the normal vector falls in the first or third quadrant, .

[0040] As long as the magnitude of the polarization phase angle and the positive and negative of are determined, the quadrant where the normal vector is located can be determined, and the azimuth angle can also be determined according to the relationship between the above and . Specifically, when the normal line falls in the second or fourth quadrant, if it is positive, it falls in the second quadrant, ; if it is negative, it falls in the second quadrant, .

[0041] S30. Eliminate the ambiguity of the azimuth angle through the projection angle of the normal vector in the YOZ plane, and eliminate false matching by using the angle constraint of the normal vector.

[0042] Using rotational motion to judge the positive and negative of is difficult to control whether rotating the target object by a certain angle or rotating the camera around the target object by a certain angle. This embodiment combines a binocular vision system to replace the rotational motion. Specifically, eliminating the ambiguity of the azimuth angle through the projection angle of the normal vector in the YOZ plane includes: according to the relationship between the polarization azimuth angle and the azimuth angle, introducing the projection angle of the normal vector in the YOZ plane, simulating rotation through the binocular vision system, judging the sign of the projection angle, and determining the quadrant where the azimuth angle is located.

[0043] Assume that the target point is measured before rotation , the rotation angle , then the measured value of the target point after rotation . If , then . If , then . As long as the value on the target point after rotation is found, the positive and negative nature of this point before rotation can be determined.

[0044] In this embodiment, the schematic diagram of using the projection angle to assist binocular disambiguation is as shown in Figure 3 . First, synchronously collect multi-angle polarization images, calculate Stokes parameters and polarization degree information; then, calculate the zenith angle and azimuth angle based on Fresnel's reflection law, and use the binocular disambiguation constrained by the projection angle to further complete 3D reconstruction. The core idea of the binocular disambiguation algorithm can be summarized in the following steps: First, establish a list of potential matching points based on the angle between normal vectors , and preliminarily determine possible corresponding point pairs. Then, by introducing a series of geometric and physical constraints, filter out those incorrect point pairs that do not meet the matching conditions, and finally obtain the correct corresponding relationship and the angle value of the same target point in two images from different perspectives. According to the foregoing theory, further determine the sign of for pairing.

[0045] Assume that the angle between the normal vectors of all reference points in the image before rotation is within the range of , where is the maximum acceptable angle range defined according to the camera rotation angle and field of view limitation. In the rotated image, the angle of the target point changes to , that is, by performing a translation transformation on the rotation angle, the potential matching points in the rotated image can be searched.

[0046] S40. Perform gradient integration on the angles of the normal vectors after eliminating false matches to complete 3D reconstruction.

[0047] In this embodiment, the schematic diagram of 3D reconstruction by performing gradient integration on the normal vector information after eliminating false matches is as shown in Figure 4 . Use the three-dimensional coordinates of the normal to obtain a local point cloud map, and further obtain a 3D reconstruction map. Performing gradient integration on the angles of the normal vectors after eliminating false matches to complete 3D reconstruction includes: converting the three-dimensional normal vector into a discrete approximate gradient through local linear approximation; fitting the discrete approximate gradient to the surface gradient field by minimizing the cost function, and reconstructing the surface shape of the object through numerical integration.

[0048] Assume that the target surface is a Cartesian surface , is the normal vector of any point on the surface, which can be expressed by the zenith angle and the azimuth angle . Converting the zenith angle and the azimuth angle into a three-dimensional normal vector, it is expressed as: , wherein, is the normal vector of any point on the surface.

[0049] Specifically, the discrete approximate gradient (p, q) can be calculated using the polarization normal vector , where , , The direction of the normal vector is the same as the direction of the surface gradient, and the magnitude of the normal vector is proportional to the rate of change of the surface. The change of the normal vector can be expressed by the partial derivative of the gradient as: , , The shape can be calculated by minimizing the following cost function, expressed as: , After obtaining the surface normal vector N and the surface gradient ( Z x , Z y ) and the cost function of the discrete approximate gradient (p, q), the following integration method is used to integrate the gradient field.

[0050] , , , When ( u , v ) ≠ (0, 0), solving the above equation, we get: .

[0051] A binocular stereo vision three-dimensional reconstruction method based on polarization state proposed in the embodiments of the present disclosure can more effectively obtain high-precision three-dimensional reconstruction results in complex low-texture and high-reflection environments by combining polarization state information and binocular vision. It provides a new idea for solving the azimuth ambiguity problem and provides technical support for the popularization of polarization binocular vision in practical industrial applications. The method of the present disclosure can greatly improve the quality of the point cloud, reduce the normal error, and effectively solve the problem of holes in the point cloud data.

[0052] Example 2 As another aspect of the embodiments of the present disclosure, a binocular stereo vision three-dimensional reconstruction system 100 based on polarization states is further provided. As shown in Figure 5 the following, it includes: A polarization image acquisition module 1, which acquires polarization images at different polarization degrees of a polarization camera; A Fresnel theory calculation module 2, which calculates the angle of the normal vector based on the acquired polarization images according to the Fresnel theory. Among them, the angle of the normal vector includes the zenith angle and the azimuth angle of the object surface; A false matching elimination module 3, which eliminates the ambiguity of the azimuth angle through the projection angle of the normal vector in the YOZ plane and eliminates false matching by using the angle constraint of the normal vector; A three-dimensional reconstruction module 4, which completes three-dimensional reconstruction by performing gradient integration on the angles of the normal vectors after false matching elimination.

[0053] Without contradiction, the above modules in the system of the embodiments of the present disclosure can implement any of the above implementation manners of the method.

[0054] Based on the description of the above embodiments, the embodiments of the present disclosure can achieve the following technical effects: 1) The present disclosure proposes a new polarization-based binocular stereo vision disambiguation (PBAE) method to solve the ambiguity problem in normal integration in high-precision surface reconstruction.

[0055] 2) The present disclosure uses the projection angle to assist in determining the normal quadrant, eliminates point mismatches, improves the measurement efficiency, and effectively solves the problem of point cloud holes.

[0056] 3) The present disclosure provides a simple and effective solution to correct normal integration and reduce missing data in the point cloud, and has high robustness and accuracy.

[0057] The embodiments of the present disclosure also propose an electronic device, including: a processor; a memory for storing instructions executable by the processor; wherein, the processor is configured to perform the above-mentioned binocular stereo vision three-dimensional reconstruction method based on polarization states. Among them, the electronic device can be provided as a terminal, a server or other forms of devices.

[0058] The embodiments of the present disclosure also propose a computer-readable storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the above-mentioned binocular stereo vision three-dimensional reconstruction method based on polarization states is implemented. The computer-readable storage medium can be a non-volatile computer-readable storage medium.

[0059] Those skilled in the art can understand that in the above-described binocular stereo vision three-dimensional reconstruction method and system based on polarization state in the specific implementation manner, the writing order of each step does not mean a strict execution order and does not impose any limitation on the implementation process. The specific execution order of each step should be determined according to its function and possible internal logic.

[0060] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a part of an instruction, and the module, segment of a program, or part of an instruction contains one or more executable instructions for implementing the specified logical function. In some alternative implementations, the functions marked in the block may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0061] The various embodiments of the present disclosure have been described above. The above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations are obvious to those of ordinary skill in the art in the technical field without departing from the scope and spirit of the described embodiments. The choice of terms used herein is intended to best explain the principles of the embodiments, the practical application, or the technical improvement of the technology in the market, or to enable other ordinary skilled persons in the technical field to understand the embodiments disclosed herein.

Claims

1. A binocular stereo vision three-dimensional reconstruction method based on polarization state, characterized in that It includes the following steps: S10. Obtain polarization images of a polarization camera at different degrees of polarization; S20. Based on the obtained polarization images, calculate the angle of the normal vector according to Fresnel's theory, where the angle of the normal vector includes the zenith angle and the azimuth angle of the object surface; S30. Eliminate the ambiguity of the azimuth angle through the projection angle of the normal vector in the YOZ plane, and eliminate false matches using the angle constraint of the normal vector; S40. Perform gradient integration on the angle of the normal vector after eliminating false matches to complete three-dimensional reconstruction.

2. The method according to claim 1, wherein Calculating the angle of the normal vector based on Fresnel's theory includes: calculating the Stokes parameters according to the obtained polarization images, and further obtaining the degree of polarization and the polarization azimuth angle, expressed as: ， ， ， ， ， In the formula, , , and respectively represent the polarization images of the degree of linear polarization in the directions of 0°, 45°, 90° and 135°, is the degree of polarization, is the polarization azimuth angle.

3. The method according to claim 2, wherein Calculating the angle of the normal vector based on Fresnel's theory further includes: obtaining the azimuth angle of the object surface according to the relationship between the polarization azimuth angle and the azimuth angle, expressed as: ， In the formula, is the azimuth.

4. The method according to claim 3, characterized in that, Calculating the angle of the normal vector based on Fresnel's theory further includes: according to the relationship formula between the degree of polarization and the zenith angle, numerically iterate to solve the relationship formula to obtain the zenith angle of the object surface, and the relationship formula between the degree of polarization and the zenith angle is expressed as: ， In the formula, is the zenith angle, is the complex refractive index, =n(1 + ik), where i is the imaginary unit and k is the attenuation index.

5. The method according to claim 1, wherein Eliminating the ambiguity of the azimuth angle through the projection angle of the normal vector in the YOZ plane includes: according to the relationship between the polarization azimuth angle and the azimuth angle, introduce the projection angle of the normal vector in the YOZ plane, simulate rotation through a binocular vision system, judge the sign of the projection angle, and determine the quadrant where the azimuth angle is located.

6. The method according to claim 1, characterized in that, Performing gradient integration on the angle of the normal vector after eliminating false matches to complete three-dimensional reconstruction includes: converting the zenith angle and the azimuth angle into a three-dimensional normal vector, expressed as: ， In the formula, is the normal vector of any point on the surface.

7. The method according to claim 6, characterized in that, Performing gradient integration on the angle of the normal vector after eliminating false matches to complete three-dimensional reconstruction further includes: Converting the three-dimensional normal vector into a discrete approximate gradient through local linear approximation; Fitting the discrete approximate gradient to the surface gradient field by minimizing the cost function, and reconstructing the object surface shape through numerical integration.

8. A binocular stereo vision three-dimensional reconstruction system based on the polarization state, characterized in that It includes: A polarization image acquisition module that acquires polarization images of a polarization camera at different degrees of polarization; A Fresnel theory calculation module that calculates the angle of the normal vector based on Fresnel's theory according to the obtained polarization images, where the angle of the normal vector includes the zenith angle and the azimuth angle of the object surface; A false match elimination module that eliminates the ambiguity of the azimuth angle through the projection angle of the normal vector in the YOZ plane and eliminates false matches using the angle constraint of the normal vector; A three-dimensional reconstruction module that performs gradient integration on the angle of the normal vector after eliminating false matches to complete three-dimensional reconstruction.

9. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the binocular stereo vision three-dimensional reconstruction method based on polarization state according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the binocular stereo vision three-dimensional reconstruction method based on polarization state according to any one of claims 1 to 7.

Images