Light center extraction method and system based on polarization improved threshold segmentation

Abstract

The invention relates to a light center extraction method and system based on polarization improved threshold segmentation. The method comprises the following steps: acquiring a laser image and a polarization degree image which are acquired by a polarization camera and modulated by linear polarization; calculating a polarization degree value and a gray value according to the acquired polarization degree image and the laser image, amplifying the resolution ratio of the polarization degree image to be equal to that of the laser image by adopting an interpolation method, and respectively carrying out normalization processing on the polarization degree value and the gray value to obtain a normalized polarization degree value and a normalized gray value; classifying corresponding pixel points of the laser image and the polarization degree image based on the size relation between the normalized gray value and the polarization degree value and the combined threshold value, determining an optimal segmentation threshold value through maximizing between-class variance, and further extracting a gray image of the laser region; obtaining a first-order derivative of the grayscale image through a filter, and positioning a zero point of the first-order derivative by adopting a bicubic interpolation method to obtain a light center point coordinate; the method disclosed by the invention can realize high-precision light center coordinate positioning.

Description

Technical Field

[0001] The present disclosure relates to the field of computer vision technology, and in particular, to a method and system for extracting the optical center based on polarization-improved threshold segmentation. Background Art

[0002] The line structured light three-dimensional detection technology is an important research direction in the field of three-dimensional detection in recent years. This technology is particularly suitable for fields with strict requirements for accuracy and anti-interference, such as industrial precision part detection, optical free-form surface measurement, and robot navigation. The accuracy of the line structured light three-dimensional detection technology depends on the accuracy of optical center extraction. However, when measuring a reflective metal surface, laser reflection will cause the loss of image texture information, making it difficult to extract the projected laser stripe. Therefore, in the presence of noise caused by laser reflection, accurately extracting the laser center line is the key to achieving accurate measurement of the entire system.

[0003] Traditional optical center extraction techniques. Traditional optical stripe center extraction algorithms mainly achieve sub-pixel positioning by analyzing the gray distribution or geometric features of laser lines. These methods usually include: The gray centroid method is based on the symmetry feature of the gray distribution of the cross-section of the optical stripe. It scans the laser line image row by row, extracts the gray values of each row of pixels, and then uses the gray value as the weight to calculate the weighted average of the pixel abscissa to obtain the gray centroid coordinates of each row. By calculating and connecting all centroid points row by row, the complete optical stripe center line is fitted; The direction template method enhances the linear feature of the optical stripe by designing a direction-sensitive filter (such as a Gabor filter or a Sobel direction operator). The algorithm presets a set of direction templates with different angles, performs multi-directional convolution filtering on the image, and selects the direction with the strongest response as the local direction of the optical stripe. Search for the extreme points (such as the maximum value or the zero point of the first derivative) of the gray distribution along this direction as the sub-pixel center position, and finally connect the points along the extension direction of the optical stripe to form the center line; The Steger algorithm is based on the geometric structure characteristics of the optical stripe. It uses the second derivative of the image to construct the Hessian matrix, analyzes the curvature characteristics of the local area, and determines the normal direction of the optical stripe by solving the direction of the maximum eigenvalue of the Hessian matrix. Perform a second-order Taylor expansion of the gray distribution along the normal direction, and analytically solve the point where the first derivative is zero to achieve sub-pixel center positioning. Finally, connect the discrete center points into a smooth curve through polynomial fitting or spline interpolation. However, when facing a highly reflective surface, it still faces problems such as noise, surface gloss, and complex reflection, which may lead to a decrease in the accuracy of laser line extraction. Summary of the Invention

[0004] To solve the problems that the traditional method causes the loss of image texture information due to noise caused by laser reflection, resulting in a decrease in the accuracy of laser center line extraction, etc. The present disclosure proposes a method for extracting the optical center based on polarization-improved threshold segmentation to solve the above problems.

[0005] According to one aspect of the present disclosure, there is provided a method for extracting the optical center based on polarization-improved threshold segmentation, including: S10. Obtain the laser image and the polarization degree image modulated by meridional polarization collected by a polarization camera; S20. Calculate the polarization degree value and the gray value according to the obtained polarization degree image and the laser image, use the interpolation method to magnify the resolution of the polarization degree image to be equal to that of the laser image, and perform normalization processing on the polarization degree value and the gray value respectively to obtain the normalized polarization degree value and the gray value; S30. Based on the magnitude relationship between the normalized gray value and the polarization degree value and the joint threshold, classify the corresponding pixel points of the laser image and the polarization degree image, determine the optimal segmentation threshold by maximizing the between-class variance, and further extract the gray image of the laser region; S40. Obtain the first derivative of the gray image through a filter, and use the bicubic interpolation method to locate the zero point of the first derivative to obtain the coordinates of the optical center point.

[0006] Preferably, the polarization degree value is calculated according to the obtained polarization degree image, expressed as: , wherein, S0 represents the total light intensity, S1 represents the variance of the light intensity of horizontal polarization relative to vertical polarization, S2 represents the variance of the light intensity of 45° polarization relative to 135° polarization, S3 represents the change in the light intensity of the right-handed circular polarization flux relative to the left-handed circular polarization flux, and Dolp is the polarization degree value describing the light in the polarized light state.

[0007] Preferably, the polarization degree value and the gray value are respectively normalized to obtain the normalized polarization degree value and the gray value, expressed as: , , wherein, and are the original gray image and the polarization degree image respectively, and are the maximum value and the minimum value of the image polarization degree respectively, and are the maximum value and the minimum value of the image gray value respectively, is the normalized polarization degree value, is the normalized gray value.

[0008] Preferably, based on the magnitude relationship between the normalized gray value and the polarization degree value and the joint threshold, classify the corresponding pixel points of the laser image and the polarization degree image, including: setting the joint threshold , where Represents the threshold of the laser image, Represents the threshold of the corresponding degree-of-polarization image. For each pixel, according to the relationship between its normalized gray value and degree-of-polarization value and the combined threshold, the pixel is divided into four categories: When the gray value of the pixel is greater than , and the degree-of-polarization value at the corresponding pixel in the degree-of-polarization image is greater than , the pixel is of category k1; when the gray value of the pixel is greater than , and the degree-of-polarization value at the corresponding pixel in the degree-of-polarization image is less than or equal to , the pixel is of category k2; when the gray value of the pixel is less than or equal to , and the degree-of-polarization value at the corresponding pixel in the degree-of-polarization image is greater than , the pixel is of category k3; when the gray value of the pixel is less than or equal to , and the degree-of-polarization value at the corresponding pixel in the degree-of-polarization image is less than or equal to , the pixel is of category k4.

[0009] Preferably, the optimal segmentation threshold is determined by maximizing the between-class variance, including: calculating the between-class variance of each category according to the classification results of the corresponding pixels of the laser image and the degree-of-polarization image . If the current between-class variance is greater than the historical maximum value, then update the combined threshold to obtain the optimal segmentation threshold .

[0010] Preferably, the first derivative of the gray image is obtained through a filter, expressed as: , wherein, is the gray image, H is the expression of the filter, ( x , y ) is the coordinate of any target point, and D( x ) is the first derivative image.

[0011] Preferably, the bicubic interpolation method is used to locate the zero point of the first derivative, including: using the bicubic interpolation method to fit the derivative values near the zero point of the first derivative, smoothing and reconstructing the gradient distribution in the zero point region by constructing a two-dimensional cubic polynomial function, and finding the zero intersection point at the sub-pixel scale.

[0012] According to one aspect of the present disclosure, there is provided an optical center extraction system based on polarization-improved threshold segmentation, including: A laser image and degree of polarization image acquisition module, which acquires a laser image and a degree of polarization image modulated by meridional polarization collected by a polarization camera; A degree of polarization value and grayscale value normalization processing module, which calculates the degree of polarization value and the grayscale value according to the acquired degree of polarization image and laser image, uses an interpolation method to magnify the resolution of the degree of polarization image to be equal to that of the laser image, and respectively performs normalization processing on the degree of polarization value and the grayscale value to obtain the normalized degree of polarization value and grayscale value; A grayscale image extraction module, which classifies the corresponding pixel points of the laser image and the degree of polarization image based on the magnitude relationship between the normalized grayscale value and the degree of polarization value and the joint threshold, determines the optimal segmentation threshold by maximizing the between-class variance, and further extracts the grayscale image of the laser region; A light center point coordinate acquisition module, which obtains the first derivative of the grayscale image through a filter, uses bicubic interpolation to locate the zero point of the first derivative, and obtains the light center point coordinate.

[0013] According to one aspect of the present disclosure, an electronic device is provided, including: a processor; a memory for storing instructions executable by the processor; wherein, the processor is configured to: execute the above-mentioned light center extraction method based on polarization-improved threshold segmentation.

[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 light center extraction method based on polarization-improved threshold segmentation is implemented.

[0015] Compared with the prior art, the beneficial effects of the present disclosure are as follows: 1) By introducing the degree of polarization as a factor affecting threshold calculation, the present disclosure classifies the corresponding pixel points of the laser image and the degree of polarization image, and expands the maximum between-class variance method based on these categories, which can quantify the consistency of the vibration direction of the light wave electric field in the high-reflection region, so as to achieve stronger discrimination ability of the algorithm in laser segmentation.

[0016] 2) The present disclosure uses a filter to approximately obtain its first derivative and combines a bicubic interpolation model to locate the zero point of the first derivative, so that the approximation of the first derivative and the gradient fitting can make full use of the surrounding pixels, thereby achieving high-precision light center coordinate positioning.

[0017] 3) The present disclosure can significantly improve the laser discrimination ability of high-reflection objects, segment a more accurate laser region, and can also make full use of the surrounding pixel information in the light center coordinate positioning to avoid the influence of light pollution caused by the high-reflection surface.

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

[0019] Other features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Description of the Drawings

[0020] The accompanying drawings herein are incorporated into 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 method for extracting the optical center based on polarization-improved threshold segmentation; Figure 2 Block diagram showing the structure of a system for extracting the optical center based on polarization-improved threshold segmentation in an embodiment of the present disclosure. Detailed Embodiments

[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 or better than other embodiments.

[0024] The term "and / or" herein is merely a description of the associated relationship of the associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the term "at least one" herein means any one of a plurality or any combination of at least two of a plurality. For example, including at least one of A, B, and C may represent including any one or more elements selected from the set composed of A, B, and C.

[0025] In addition, for a better description of the present disclosure, numerous specific details are given in the following detailed embodiments. Those skilled in the art should understand that the present disclosure can be implemented without some 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 following will clearly and completely describe the technical solutions in the embodiments of the present invention 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 an optical center extraction method based on polarization-improved threshold segmentation. Figure 1 The flowchart of an optical center extraction method based on polarization-improved threshold segmentation is shown. The method includes: S10. Obtain the laser image modulated by meridional polarization and the polarization degree image collected by a polarization camera; S20. Calculate the polarization degree value and the gray value according to the obtained polarization degree image and laser image, use the interpolation method to enlarge the resolution of the polarization degree image to be equal to that of the laser image, and perform normalization processing on the polarization degree value and the gray value respectively to obtain the normalized polarization degree value and gray value; S30. Based on the magnitude relationship between the normalized gray value and polarization degree value and the joint threshold, classify the corresponding pixel points of the laser image and the polarization degree image, determine the optimal segmentation threshold by maximizing the between-class variance, and further extract the gray image of the laser region; S40. Obtain the first-order derivative of the gray image through a filter, use the bicubic interpolation method to locate the zero point of the first-order derivative, and obtain the coordinates of the optical center point.

[0028] The embodiments of the present disclosure provide an optical center extraction method based on polarization-improved threshold segmentation, which specifically includes the following steps: S10. Obtain the laser image modulated by meridional polarization and the polarization degree image collected by a polarization camera.

[0029] In this embodiment, the laser image collected by the polarization camera is modulated into linearly polarized light by a polarization lens, and the polarization state images at 4 angles of the sub-focal plane of the polarization camera are obtained.

[0030] S20. Calculate the polarization degree value and the gray value according to the obtained polarization degree image and laser image, use the interpolation method to enlarge the resolution of the polarization degree image to be equal to that of the laser image, and perform normalization processing on the polarization degree value and the gray value respectively to obtain the normalized polarization degree value and gray value.

[0031] In this embodiment, the polarization degree value of the image is calculated according to the obtained polarization degree image by using the Stokes vector, which is expressed as: , Wherein, S0 represents the total light intensity, S1 represents the variance of the light intensity of horizontal polarization relative to vertical polarization, S2 represents the variance of the light intensity of 45° polarization relative to 135° polarization, S3 represents the change in the light intensity of the right-handed circular polarization flux relative to the left-handed circular polarization flux, and Dolp is the polarization degree value describing the light in the polarized light state. 。

[0032] Further, the resolution of the polarization degree image is magnified to be equal to that of the laser image by using the interpolation method, and the polarization degree value and the gray value are respectively normalized to obtain the normalized polarization degree value and gray value, which are expressed as: , , Wherein, and are the original gray image and polarization degree image respectively, and are the maximum and minimum values of the image polarization degree respectively, and are the maximum and minimum values of the image gray value respectively, is the normalized polarization degree value, is the normalized gray value, both of which are in [0,1].

[0033] S30. Based on the magnitude relationship between the normalized gray value and polarization degree value and the joint threshold, the corresponding pixel points of the laser image and the polarization degree image are classified, and the optimal segmentation threshold is determined by maximizing the between-class variance, and further the gray image of the laser region is extracted.

[0034] In this embodiment, the pixel points of the laser image are classified in combination with the polarization degree image. Based on the magnitude relationship between the normalized gray value and polarization degree value and the joint threshold, the corresponding pixel points of the laser image and the polarization degree image are classified, including: setting the joint threshold , where represents the threshold of the laser image, represents the threshold of the corresponding polarization degree image. For each pixel point, according to the magnitude relationship between its normalized gray value and polarization degree value and the joint threshold, the pixel point is divided into four categories: When the gray value of the pixel point is greater than , and the polarization degree value at the corresponding pixel point in the polarization degree image is greater than , the pixel point is of category k1; when the gray value of the pixel is greater than , and the polarization degree value at the corresponding pixel point in the polarization degree image is less than or equal to , the pixel point is of category k2; when the gray value of the pixel is less than or equal to , when the polarization degree value at the corresponding pixel point in the polarization degree image is greater than , the pixel point is of category k3; when the gray value of the pixel is less than or equal to , and the polarization degree value at the corresponding pixel point in the polarization degree image is less than or equal to , the pixel point is of category k4.

[0035] In this embodiment, the between-class variance of each class is calculated, and the optimal segmentation threshold is obtained by maximizing the between-class variance that measures different features of different classes. Among them, determining the optimal segmentation threshold by maximizing the between-class variance includes: calculating the between-class variance of each class according to the classification results of the corresponding pixel points of the laser image and the polarization degree image , expressed as: , In the formula, k represents the class, represents the probability of the occurrence of pixels of class k, and respectively represent the average gray value and polarization degree in the pixels of class k, and respectively represent the average gray value and polarization degree in the whole image. The calculation formulas of these parameters are shown in Table 1 below.

[0036] Table 1:

[0037] The obtaining of the optimal segmentation threshold by maximizing the between-class variance that measures different features of different classes includes: if the current between-class variance is greater than the historical maximum value, then update the joint threshold , and obtain the optimal segmentation threshold , expressed as: , S40. Obtain the first derivative of the gray image through a filter, and use bicubic interpolation to locate the zero point of the first derivative to obtain the coordinates of the light center point.

[0038] In this embodiment, for the gray image of the laser region, using a fir filter to approximately obtain its first derivative and combining with a bicubic interpolation model to locate the zero point of the first derivative includes: Applying the fir filter H to each pixel in the gray image to obtain the first derivative , expressed as: , In the formula, is the gray image, H is the expression of the filter, ( x , y) is the coordinate of any target point, D( x ) is the first derivative image.

[0039] Furthermore, the bicubic interpolation method is used to locate the zero point of the first derivative, including: using the bicubic interpolation method to fit the derivative values near the zero point of the first derivative, and by constructing a two-dimensional cubic polynomial function, the gradient distribution within the zero point region is smoothly reconstructed, and the zero intersection point is found at the sub-pixel scale, expressed as: , , In the formula, is the interpolation coefficient, determined by the derivative values in the field, is the upper left corner coordinate of the interpolation region, is the coordinate of any target point, represents the coefficient of the interpolation polynomial, determined by the interpolation process.

[0040] The embodiment of the present disclosure proposes an optical center extraction algorithm based on polarization-improved threshold segmentation. By adding a polarization information to act on the threshold segmentation of the laser image and combining with a high-precision zero point positioning model, the segmentation threshold has stronger discrimination ability for polarized lasers, which can effectively improve the measurement accuracy.

[0041] Embodiment 2 As another aspect of the embodiment of the present disclosure, an optical center extraction system 100 based on polarization-improved threshold segmentation is further provided, as Figure 2 shown, including: A laser image and polarization degree image acquisition module 1, which acquires a laser image and a polarization degree image modulated by meridional polarization collected by a polarization camera; A polarization degree value and gray value normalization processing module 2, which calculates the polarization degree value and gray value according to the acquired polarization degree image and laser image, uses the interpolation method to magnify the resolution of the polarization degree image to be equal to that of the laser image, and respectively normalizes the polarization degree value and gray value to obtain the normalized polarization degree value and gray value; A gray image extraction module 3, which classifies the corresponding pixel points of the laser image and the polarization degree image based on the size relationship between the normalized gray value and polarization degree value and the joint threshold, determines the optimal segmentation threshold by maximizing the between-class variance, and further extracts the gray image of the laser region; An optical center point coordinate acquisition module 4, which obtains the first derivative of the gray image through a filter, uses the bicubic interpolation method to locate the zero point of the first derivative, and obtains the optical center point coordinate.

[0042] Without contradiction, the above modules in the system of the embodiment of the present disclosure can implement any implementation manner of the above method.

[0043] Based on the description of the above embodiments, the embodiments of the present disclosure can achieve the following technical effects: 1) By introducing the degree of polarization as a factor affecting threshold calculation, the present disclosure classifies the corresponding pixel points of the laser image and the polarization degree image, and extends the maximum inter-class variance method based on these categories, enabling the quantification of the consistency of the vibration direction of the light wave electric field in the high-reflection region, thereby achieving stronger discrimination ability of the algorithm in laser segmentation.

[0044] 2) The present disclosure uses a filter to approximately obtain its first derivative and combines it with a bicubic interpolation model to locate the zero point of the first derivative, enabling the approximation of the first derivative and gradient fitting to make full use of surrounding pixels, thereby achieving high-precision positioning of the optical center coordinates.

[0045] 3) The present disclosure can significantly improve the laser discrimination ability of high-reflection objects, segment a more accurate laser region, and can also make full use of the surrounding pixel information in the positioning of the optical center coordinates to avoid the influence of light pollution caused by high-reflection surfaces.

[0046] 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 optical center extraction method based on polarization-improved threshold segmentation. The electronic device can be provided as a terminal, a server or other forms of devices.

[0047] The embodiments of the present disclosure also propose a computer-readable storage medium, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the above optical center extraction method based on polarization-improved threshold segmentation is implemented. The computer-readable storage medium can be a non-volatile computer-readable storage medium.

[0048] Those skilled in the art can understand that in the above optical center extraction method and system based on polarization-improved threshold segmentation in the specific implementation manner, the writing order of each step does not mean a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined according to its function and possible internal logic.

[0049] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations 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 portion of an instruction, which contains one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions noted in the blocks may occur in a different order than noted in the accompanying drawings. For example, two consecutive blocks may actually be executed substantially in parallel, or 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 diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented by a dedicated hardware-based system that performs the specified functions or actions, or by a combination of dedicated hardware and computer instructions.

[0050] The 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 will be apparent to those of ordinary skill in the art in the 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 art in the field to understand the embodiments disclosed herein.

Claims

1. An optical center extraction method based on polarization-improved threshold segmentation, characterized in that, It includes the following steps: S10. Obtain the laser image and polarization degree image modulated by meridional polarization collected by a polarization camera; S20. Calculate the polarization degree value and gray value based on the obtained polarization degree image and laser image, use the interpolation method to magnify the resolution of the polarization degree image to be equal to that of the laser image, and perform normalization processing on the polarization degree value and gray value respectively to obtain the normalized polarization degree value and gray value; S30. Classify the corresponding pixel points of the laser image and the polarization degree image based on the magnitude relationship between the normalized gray value and polarization degree value and the joint threshold, determine the optimal segmentation threshold by maximizing the between-class variance, and further extract the gray image of the laser region; S40. Obtain the first-order derivative of the gray image through a filter, and use the bicubic interpolation method to locate the zero point of the first-order derivative to obtain the coordinates of the light center point.

2. The method according to claim 1, wherein The polarization degree value is calculated based on the obtained polarization degree image, which is expressed as: ， In the formula, S0 represents the total light intensity, S1 represents the variance of the light intensity of horizontal polarization relative to vertical polarization, S2 represents the variance of the light intensity of 45° polarization relative to 135° polarization, S3 represents the change in the light intensity of the right-handed circular polarization flux relative to the left-handed circular polarization flux, and Dolp is the polarization degree value describing the light in the polarized light state.

3. The method according to claim 1, wherein The polarization degree value and gray value are respectively normalized to obtain the normalized polarization degree value and gray value, which are expressed as: ， ， In the formula, and are the original grayscale image and the polarization degree image respectively, and are the maximum value and the minimum value of the image polarization degree respectively, and are the maximum value and the minimum value of the image grayscale value respectively, is the normalized polarization degree value, is the normalized grayscale value.

4. The method according to claim 1, wherein Based on the magnitude relationship between the normalized grayscale value and the polarization degree value and the joint threshold, the corresponding pixel points of the laser image and the polarization degree image are classified, including: setting the joint threshold , where represents the threshold of the laser image, represents the threshold of the corresponding polarization degree image. For each pixel point, according to the magnitude relationship between its normalized grayscale value and polarization degree value and the joint threshold, the pixel point is divided into four categories: When the gray value of the pixel is greater than , and the polarization degree value at the corresponding pixel in the polarization degree image is greater than , this pixel belongs to category k1; when the gray value of this pixel is greater than , and the polarization degree value at the corresponding pixel in the polarization degree image is less than or equal to , this pixel belongs to category k2; when the gray value of this pixel is less than or equal to , and the polarization degree value at the corresponding pixel in the polarization degree image is greater than , this pixel belongs to category k3; when the gray value of this pixel is less than or equal to , and the polarization degree value at the corresponding pixel in the polarization degree image is less than or equal to , this pixel belongs to category k4.

5. The method according to claim 4, characterized in that Determining the optimal segmentation threshold by maximizing the between-class variance, including: calculating the between-class variance of each category according to the classification results of the corresponding pixel points of the laser image and the polarization degree image , if the current between-class variance is greater than the historical maximum value, then update the joint threshold to obtain the optimal segmentation threshold .

6. The method according to claim 1, wherein The first-order derivative of the gray image is obtained through a filter, which is expressed as: ， In the formula, is a grayscale image, H is the expression of the filter, ( x , y ) is the coordinate of any target point, D( x ) is the first derivative image.

7. The method according to any one of claims 1 or 6, characterized in that Using the bicubic interpolation method to locate the zero point of the first-order derivative includes: using the bicubic interpolation method to fit the derivative values near the zero point of the first-order derivative, constructing a two-dimensional cubic polynomial function to smoothly reconstruct the gradient distribution within the zero point region, and finding the zero intersection point at the sub-pixel scale.

8. An optical center extraction system based on polarization-improved threshold segmentation, characterized in that It includes: A laser image and polarization degree image acquisition module, which acquires the laser image and polarization degree image modulated by meridional polarization collected by a polarization camera; A polarization degree value and gray value normalization processing module, which calculates the polarization degree value and gray value based on the obtained polarization degree image and laser image, uses the interpolation method to magnify the resolution of the polarization degree image to be equal to that of the laser image, and performs normalization processing on the polarization degree value and gray value respectively to obtain the normalized polarization degree value and gray value; A gray image extraction module, which classifies the corresponding pixel points of the laser image and the polarization degree image based on the magnitude relationship between the normalized gray value and polarization degree value and the joint threshold, determines the optimal segmentation threshold by maximizing the between-class variance, and further extracts the gray image of the laser region; A light center point coordinate acquisition module, which obtains the first-order derivative of the gray image through a filter, uses the bicubic interpolation method to locate the zero point of the first-order derivative, and obtains the coordinates of the light center point.

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 light center extraction method based on polarization-improved threshold segmentation 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 light center extraction method based on polarization-improved threshold segmentation according to any one of claims 1 to 7.

Images