Polarized sun vector resolving method and device based on confidence mask

Abstract

This application relates to the field of polarization navigation technology, and particularly to a method and apparatus for calculating polarized solar vectors based on confidence masks. The method includes: acquiring polarization angle images of the sky region; determining the mask features of the polarization angle images; constructing azimuth and elevation masks of the polarization angle images using the mask features; obtaining corresponding pixel data from the polarization angle images using the azimuth and elevation masks respectively; and constructing a three-dimensional E-vector corresponding to the azimuth and elevation masks based on the pixel data, and calculating the solar vector of the polarization angle images based on the three-dimensional E-vector. This solves the problems of related technologies that employ full-pixel sampling mode and full-field-of-view equal-weight fusion strategies, resulting in huge computational resource consumption, and low-quality observation data interfering with global optimization, leading to random fluctuations in the calculation results and limiting the robustness of the system in complex environments.

Description

Technical Field

[0001] This application relates to the field of polarized light navigation technology, and in particular to a method and apparatus for calculating polarized solar vectors based on a confidence mask. Background Technology

[0002] In the field of navigation technology, the demand for unmanned system navigation in GNSS (Global Navigation Satellite System) denied environments is becoming increasingly urgent. Bionic polarization navigation technology, with its unique advantages of being completely autonomous, passive, resistant to electromagnetic interference, and free from cumulative errors, has become a highly promising solution in this field. Among them, sky polarization navigation, as the core branch of bionic polarization navigation technology, has received widespread attention, while imaging polarization sensors, which can acquire richer sky polarization information, have gradually become the mainstream sensing hardware in this field.

[0003] Among related technologies, navigation research based on imaging polarization observation includes a method for extracting the symmetry axis of the solar meridian. This method utilizes the symmetry characteristics of the sky polarization mode to obtain the solar azimuth angle relatively well, providing a basis for navigation orientation. However, this method is essentially a two-dimensional image geometric feature extraction and cannot independently perceive the solar altitude angle. It requires a multi-field stitching scheme to solve the altitude angle, which leads to a significant increase in system volume, weight, and power consumption, an increase in axis installation error, and a limitation on the system's ultimate accuracy. In addition, there is a three-dimensional E-vector reconstruction algorithm under a single-lens scheme. This algorithm can solve the complete solar vector without multi-lens stitching, simplifying the system structure.

[0004] However, in related technologies, the three-dimensional E-vector reconstruction algorithm adopts a full-pixel sampling mode and a full-field equal-weight fusion strategy, which results in huge consumption of computational resources. In addition, low-quality observation data interferes with global optimization, and the solution results are prone to random fluctuations, which limits the robustness of the system in complex environments and urgently needs to be solved. Summary of the Invention

[0005] This application provides a method and apparatus for calculating polarization solar vectors based on confidence masks, in order to solve the problems of related technologies that adopt full-pixel sampling mode and full-field equal-weight fusion strategy, resulting in huge consumption of computing resources, low-quality observation data interfering with global optimization, and the calculation results being prone to random fluctuations, which limits the robustness of the system in complex environments.

[0006] The first aspect of this application provides a method for calculating the polarization solar vector based on a confidence mask, comprising the following steps: acquiring a polarization angle image of a sky region; determining the mask features of the polarization angle image, and constructing an azimuth mask and an elevation mask of the polarization angle image using the mask features; obtaining corresponding pixel data from the polarization angle image using the azimuth mask and the elevation mask, and constructing a three-dimensional E-vector corresponding to the azimuth mask and the elevation mask based on the pixel data, so as to calculate the solar vector of the polarization angle image based on the three-dimensional E-vector.

[0007] Based on the above technical means, this embodiment of the application acquires polarization angle images, constructs azimuth and elevation angle masks using mask features, extracts corresponding pixel data from the polarization angle images and reconstructs them into three-dimensional E vectors, and then calculates the solar vector. This enables differentiated screening and sparse sampling of observed pixels, effectively reducing the interference of low-quality observation data on the global optimization process, suppressing random fluctuations in the solution results, and significantly reducing the consumption of computing resources and the amount of data processing. In addition, through the dimensional processing of azimuth and elevation angle masks, the accuracy and stability of the solar vector solution are improved, significantly enhancing the robustness and engineering applicability of the system in complex environments.

[0008] Optionally, in one embodiment of this application, before determining the mask features of the polarization angle image, the method further includes: determining the orthogonality between the solar vector and the three-dimensional E vector of the target polarization angle image; constructing an orthogonal constraint equation between the solar vector of the target polarization angle image and the polarization angle measurement value of the target polarization angle image based on the orthogonal relationship; determining the error propagation coefficient of the polarization measurement noise of the target polarization angle image to the solar vector of the target polarization angle image based on the orthogonal constraint equation; determining the azimuth score and elevation score of the target polarization angle image based on the error propagation coefficient; and constructing a confidence score field based on the azimuth score and elevation score of the target polarization angle image.

[0009] Based on the above technical means, this application embodiment takes the target polarization angle image with a known solar vector as the analysis object, first constructs an orthogonal constraint equation between the solar vector and the measured polarization angle value, then determines the error propagation coefficient of polarization measurement noise on the solar vector, and then quantifies the noise resistance capability of each pixel, obtains the azimuth angle score and elevation angle score, and constructs a confidence score field, realizing an accurate evaluation of the confidence of the full field-of-view pixel solution of the target polarization angle image, providing a scientific and accurate prior basis for subsequent construction of sampling masks and extraction of mask features, and laying the foundation for subsequent high-precision solar vector solution based on the acquired polarization angle image.

[0010] Optionally, in one embodiment of this application, before determining the mask features of the polarization angle image, the method further includes: determining the azimuth score of the target polarization angle image based on the confidence score field; comparing the azimuth score with a preset azimuth score threshold to obtain a first comparison result of the target polarization angle image; determining pixels that meet a first preset screening condition based on the first comparison result; and constructing an azimuth mask of the target polarization angle image based on the pixels that meet the first preset screening condition.

[0011] Based on the above technical means, this application embodiment determines the azimuth score of the target polarization angle image and compares it with a preset azimuth score threshold to filter out high-confidence pixels that meet the first preset screening conditions, thereby constructing an azimuth mask. This can accurately select pixels that contribute highly to the azimuth calculation and have strong resistance to measurement noise, while removing low-confidence pixels, thus providing a basis for the subsequent extraction of reliable mask features, improving the targeting and rationality of the mask, and ensuring the accuracy and robustness of the azimuth calculation in the subsequent solar vector calculation process.

[0012] Optionally, in one embodiment of this application, before determining the mask features of the polarization angle image, the method further includes: determining the elevation angle score of the target polarization angle image based on the confidence score field; comparing the elevation angle score with a preset elevation angle score threshold to obtain a second comparison result of the target polarization angle image; determining pixels that meet a second preset screening condition based on the second comparison result; and constructing an elevation angle mask of the target polarization angle image based on the pixels that meet the second preset screening condition.

[0013] Based on the above technical means, this application embodiment determines the elevation angle score of the target polarization angle image and compares it with a preset elevation angle score threshold to select high-confidence pixels that meet the second preset screening conditions to construct an elevation angle mask. This can accurately select pixels that contribute highly to the elevation angle calculation and have strong resistance to measurement noise, while removing low-confidence pixels. This provides a basis for extracting reliable mask features in the future, improves the targeting and rationality of the mask, and ensures the accuracy and robustness of elevation angle calculation in the subsequent solar vector calculation process.

[0014] Optionally, in one embodiment of this application, before determining the mask features of the polarization angle image, the method further includes: determining a corresponding first mask feature based on the elevation angle mask of the target polarization angle image; and determining a corresponding second mask feature based on the azimuth angle mask of the target polarization angle image, so as to determine the mask features of the polarization angle image based on the first mask feature and the second mask feature.

[0015] Based on the above technical means, this application embodiment extracts the first mask features and the second mask features from the elevation angle mask and azimuth angle mask of the target polarization angle image respectively, and determines the mask features of the polarization angle image, forming a set of universal configuration rules that can be directly reused. This provides a reliable prior basis for quickly generating targeted sparse sampling masks on polarization angle images with unknown solar vectors, and improves mask configuration efficiency.

[0016] Optionally, in one embodiment of this application, the step of calculating the solar vector of the polarization angle image based on the three-dimensional E-vector includes: determining the three-dimensional E-vector corresponding to the azimuth mask of the polarization angle image, and determining the three-dimensional E-vector corresponding to the elevation mask of the polarization angle image; establishing a corresponding azimuth observation matrix based on the three-dimensional E-vector corresponding to the azimuth mask of the polarization angle image, and establishing a corresponding elevation observation matrix based on the three-dimensional E-vector corresponding to the elevation mask of the polarization angle image; calculating the optimized azimuth angle of the solar vector based on the azimuth observation matrix, and calculating the optimized elevation angle of the solar vector based on the elevation observation matrix; and determining the solar vector based on the optimized azimuth angle and the optimized elevation angle.

[0017] Based on the above technical means, the embodiments of this application establish corresponding azimuth and elevation observation matrices according to the three-dimensional E vectors corresponding to the azimuth and elevation masks of the polarization angle image, respectively. Then, the optimized azimuth and optimized elevation angles of the solar vector are calculated respectively, and the solar vector is synthesized. This effectively realizes the differentiated configuration of the observation matrix, maximizes the retention of the contribution of each effective pixel of the azimuth and elevation angles, and significantly improves the accuracy, stability and robustness of the solar vector solution.

[0018] A second aspect of this application provides a polarization solar vector calculation device based on a confidence mask, comprising: an acquisition module for acquiring a polarization angle image of a sky region; a determination module for determining the mask features of the polarization angle image and constructing an azimuth mask and an elevation mask of the polarization angle image using the mask features; and a calculation module for obtaining corresponding pixel data from the polarization angle image using the azimuth mask and the elevation mask, constructing a three-dimensional E-vector corresponding to the azimuth mask and the elevation mask based on the pixel data, and calculating the solar vector of the polarization angle image based on the three-dimensional E-vector.

[0019] Based on the above technical means, this embodiment of the application acquires polarization angle images, constructs azimuth and elevation angle masks using mask features, extracts corresponding pixel data from the polarization angle images and reconstructs them into three-dimensional E vectors, and then calculates the solar vector. This enables differentiated screening and sparse sampling of observed pixels, effectively reducing the interference of low-quality observation data on the global optimization process, suppressing random fluctuations in the solution results, and significantly reducing the consumption of computing resources and the amount of data processing. In addition, through the dimensional processing of azimuth and elevation angle masks, the accuracy and stability of the solar vector solution are improved, significantly enhancing the robustness and engineering applicability of the system in complex environments.

[0020] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the program to implement the polarization solar vector calculation method based on confidence mask as described in the above embodiments.

[0021] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for calculating polarization solar vectors based on confidence masks.

[0022] A fifth aspect of this application provides a computer program product, including a computer program that, when executed, implements the above-described method for calculating polarization solar vectors based on confidence masks.

[0023] This application's embodiments acquire polarization angle images, construct azimuth and elevation angle masks using mask features, extract corresponding pixel data from the polarization angle images, and reconstruct a three-dimensional E-vector to calculate the solar vector. This enables differentiated filtering and sparse sampling of observed pixels, effectively reducing the interference of low-quality observation data on the global optimization process, suppressing random fluctuations in the solution results, and significantly reducing computational resource consumption and data processing volume. Furthermore, the dimensional processing of the azimuth and elevation angle masks improves the accuracy and stability of the solar vector calculation, significantly enhancing the system's robustness and engineering applicability in complex environments. Therefore, this solves the problems of related technologies that employ full-pixel sampling and full-field-of-view equal-weight fusion strategies, resulting in huge computational resource consumption and low-quality observation data interfering with global optimization, leading to random fluctuations in the solution results and limiting the system's robustness in complex environments.

[0024] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0025] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart of a polarization solar vector calculation method based on a confidence mask, according to an embodiment of this application. Figure 2 This is a schematic diagram illustrating the orthogonality between the three-dimensional E-vector and the solar vector and its reconstruction principle according to an embodiment of this application; Figure 3 This is a schematic diagram of a region where the polarization angle and pixel polar angle are collinear according to an embodiment of this application; Figure 4 This is a schematic diagram of an azimuth sparse sampling mask provided according to an embodiment of this application; Figure 5 This is a schematic diagram of an elevation angle sparse sampling mask provided according to an embodiment of this application; Figure 6 This is a schematic diagram illustrating the principle of a polarization solar vector calculation method based on a confidence mask according to an embodiment of this application; Figure 7 This is a block diagram of a polarization solar vector solving device based on a confidence mask according to an embodiment of this application; Figure 8 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application.

[0026] Figure label: 70-A polarization solar vector calculation device based on confidence mask; 100-Acquisition module; 200-First determination module; 300-Calculation module; 801-Memory; 802-Processor; 803-Communication interface. Detailed Implementation

[0027] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0028] Before going into detail, the coordinate system used in the embodiments of this application will be explained. The embodiments of this application use a camera coordinate system, the origin of which is set at the optical center of the polarization sensor. The optical axis of the axial polarization sensor points outside the field of view (i.e., the imaging direction). shaft and The axes are all parallel to the image plane of the polarization sensor, and axis, shaft and The axes satisfy the right-hand rule, and this camera coordinate system is used to represent all observation vectors and solution vectors in the embodiments of this application, providing a unified coordinate reference.

[0029] The following describes, with reference to the accompanying drawings, a method and apparatus for calculating polarized solar vectors based on confidence masks, according to embodiments of this application.

[0030] Figure 1 This is a flowchart of a polarization solar vector calculation method based on a confidence mask, according to an embodiment of this application.

[0031] like Figure 1 As shown, the method for solving the polarization solar vector based on confidence masks includes the following steps: In step S101, a polarization angle image of the sky region is acquired.

[0032] In actual implementation, the embodiments of this application use a polarization sensor to image the sky area. It can be understood that the polarization sensor uses a focal plane polarization CMOS (Complementary Metal-Oxide-Semiconductor) unit to acquire light intensity images of multiple polarization direction channels under the same field of view, and then calculates the Stokes vector to generate a polarization angle image.

[0033] Optionally, in one embodiment of this application, before determining the mask features of the polarization angle image, the method further includes: determining the orthogonality between the solar vector and the three-dimensional E vector of the target polarization angle image; constructing an orthogonal constraint equation between the solar vector of the target polarization angle image and the polarization angle measurement value of the target polarization angle image based on the orthogonal relationship; determining the error propagation coefficient of the polarization measurement noise of the target polarization angle image to the solar vector of the target polarization angle image based on the orthogonal constraint equation; determining the azimuth angle score and elevation angle score of the target polarization angle image based on the error propagation coefficient; and constructing a confidence score field based on the azimuth angle score and elevation angle score of the target polarization angle image.

[0034] In the embodiments of this application, the target polarization angle image refers to the reference polarization angle image used to construct a confidence scoring field, analyze the mask distribution pattern, and extract mask features. The solar vector corresponding to this image is a known quantity. The embodiments of this application can establish orthogonal constraint equations based on the known solar vector and polarization angle observations, determine the error propagation coefficient, construct a confidence scoring field, and then determine the azimuth and elevation angle masks corresponding to the image. The morphological differences and distribution patterns of the two masks are analyzed, providing a priori basis for the mask configuration of the actual acquired polarization angle image.

[0035] like Figure 2As shown, in some cases, embodiments of this application establish geometric constraints for single-lens imaging based on the principle of atmospheric Rayleigh scattering. For any pixel within the field of view (i.e., a pixel in the target polarization angle image), the three-dimensional E-vector of the sky-scattered light in the direction of the observation line corresponding to that pixel is... Perpendicular to the solar vector With line of sight vector The scattering surface formed.

[0036] Sun vector in camera coordinate system The expression can be, but is not limited to: , in, The sun vector in the camera coordinate system. The solar azimuth angle in the camera coordinate system. This is the solar altitude angle in the camera coordinate system.

[0037] Line of sight vector in camera coordinate system The expression can be, but is not limited to, as: , in, The line-of-sight vector in the camera coordinate system. The focal length of the polarization sensor. This represents the radial distance of a pixel in the polar coordinate system of the image plane. The polar angle of the pixel in the polar coordinate system of the image plane.

[0038] 3D E-vector in camera coordinate system The expression can be, but is not limited to: , in, The three-dimensional E-vector in the camera coordinate system. For the three-dimensional E vector in Components on the axis, For the three-dimensional E vector in Components on the axis, For the three-dimensional E vector in Components on the axis.

[0039] 3D E-vector in camera coordinate system The components satisfy the polarization angle observation relationship, and the expression of this observation relationship can be, but is not limited to, as: , in, This is the measured value of the polarization angle from the polarization sensor.

[0040] 3D E-vector in camera coordinate system With line of sight vector The orthogonality constraint is satisfied, and the expression of the orthogonality constraint can be, but is not limited to, as follows: .

[0041] Based on the above geometric constraints, the three-dimensional E-vector in the camera coordinate system Reconfigurable can be represented as: , in, As an intermediate quantity, .

[0042] To quantify polarization measurement noise The spatial impact on the accuracy of solar vector calculation is addressed in this application's embodiments, which construct a system for calculating the solar vector (azimuth angle). Altitude angle ) and polarization angle measurement value The orthogonal constraint equations between them can be, but are not limited to, the following: , in, These are orthogonal constraint equations. This is the azimuth angle corresponding to the solar vector. This is the altitude angle corresponding to the solar vector.

[0043] Furthermore, embodiments of this application perform total differential analysis on the orthogonal constraint equations to derive the polarization measurement noise. For solar vector (azimuth angle) Altitude angle The first-order error propagation coefficient is defined, and the reciprocal of the magnitude of the error propagation coefficient is defined as the solution confidence level, thereby generating a confidence score field covering the entire field of view.

[0044] The expression for total differential analysis of orthogonal constraint equations can be, but is not limited to, as follows: , in, For polarization measurement noise, This refers to the azimuth angle calculation error. This represents the error in calculating the elevation angle.

[0045] Therefore, the expression for the first-order error propagation coefficient can be, but is not limited to, as follows: , , in, This is a mold taking operation.

[0046] Furthermore, the embodiments of this application define pixels. Azimuth rating and elevation angle rating These are the reciprocals of the magnitude of the corresponding error propagation coefficient, used to characterize the pixel's ability to resist polarization measurement noise in angle calculation. The higher the score, the stronger the pixel's ability to resist polarization measurement noise; the lower the score, the weaker the pixel's ability to resist polarization measurement noise.

[0047] Azimuth rating and elevation angle rating The expression can be, but is not limited to: , , in, For pixels Azimuth rating, For pixels Altitude angle rating This is a norm operation.

[0048] This application embodiment uses a target polarization angle image with a known solar vector as the analysis object. First, it constructs an orthogonal constraint equation between the solar vector and the measured polarization angle value. Then, it determines the error propagation coefficient of polarization measurement noise on the solar vector, thereby quantifying the noise resistance of each pixel, obtaining azimuth and elevation angle scores, and constructing a confidence score field. This enables accurate evaluation of the confidence of pixel solution for the entire field of view of the target polarization angle image, providing a scientific and accurate prior basis for subsequent construction of sampling masks and extraction of mask features, and laying the foundation for high-precision solar vector solution based on the acquired polarization angle image.

[0049] Optionally, in one embodiment of this application, before determining the mask features of the polarization angle image, the method further includes: determining the azimuth score of the target polarization angle image based on the confidence score field; comparing the azimuth score with a preset azimuth score threshold to obtain a first comparison result of the target polarization angle image; determining the pixels that meet the first preset screening conditions based on the first comparison result; and constructing an azimuth mask of the target polarization angle image based on the pixels that meet the first preset screening conditions.

[0050] In the embodiments of this application, the preset azimuth angle scoring threshold refers to the confidence score threshold set in advance when constructing the azimuth angle mask of the target polarization angle image. It is used to determine whether each pixel in the target polarization angle image has a high azimuth angle calculation confidence. The preset azimuth angle scoring threshold can be adaptively set by those skilled in the art according to the accuracy requirements of the actual application scenario, noise level, or proportion of high-resolution pixels (such as the top 10% of high-resolution pixels). This application does not impose specific limitations.

[0051] The first comparison result refers to the result obtained by comparing the azimuth score of each pixel in the target polarization angle image with the preset azimuth score threshold. It is used to determine whether the azimuth score of each pixel reaches the preset azimuth score threshold, and provides a criterion for selecting pixels and constructing the azimuth mask of the target polarization angle image.

[0052] The first preset filtering condition refers to the pre-set pixel selection rules. For example, the first preset filtering condition is that the azimuth score is greater than or equal to the preset azimuth score threshold. Pixels that meet the first preset filtering condition are considered as high-confidence pixels, and pixels that do not meet the first preset filtering condition are considered as low-confidence pixels.

[0053] The azimuth mask of the target polarization angle image refers to a binary mask constructed from high-confidence pixels selected according to the first preset screening conditions based on the first comparison result.

[0054] In some cases, embodiments of this application first determine the azimuth score of each pixel in the target polarization angle image based on the confidence score field, then compare the azimuth score with a preset azimuth score threshold (such as taking the score threshold corresponding to the top 10% of high-scoring pixels in the azimuth score) to obtain a first comparison result, and then select pixels that meet the first preset screening condition (i.e., the azimuth score is greater than or equal to the preset azimuth score threshold) based on the first comparison result, thereby constructing an azimuth mask of the target polarization angle image based on such pixels.

[0055] This application embodiment determines the azimuth score of the target polarization angle image and compares it with a preset azimuth score threshold to filter out high-confidence pixels that meet the first preset screening condition. This constructs an azimuth mask, which can accurately select pixels that contribute highly to the azimuth calculation and have strong resistance to measurement noise, while removing low-confidence pixels. This provides a basis for extracting reliable mask features in the future, improves the targeting and rationality of the mask, and ensures the accuracy and robustness of the azimuth calculation in the subsequent solar vector calculation process.

[0056] Optionally, in one embodiment of this application, before determining the mask features of the polarization angle image, the method further includes: determining the elevation angle score of the target polarization angle image based on a confidence score field; comparing the elevation angle score with a preset elevation angle score threshold to obtain a second comparison result of the target polarization angle image; determining pixels that meet the second preset screening conditions based on the second comparison result; and constructing an elevation angle mask of the target polarization angle image based on the pixels that meet the second preset screening conditions.

[0057] In the embodiments of this application, the preset elevation angle scoring threshold refers to the confidence score threshold set in advance when constructing the elevation angle mask of the target polarization angle image. It is used to determine whether each pixel in the target polarization angle image has a high confidence in elevation angle calculation. The preset elevation angle scoring threshold can be adaptively set by those skilled in the art according to the accuracy requirements, noise level or proportion of high-resolution pixels (such as the top 10% of high-resolution pixels) of the actual application scenario. This application does not impose any specific limitations.

[0058] The second comparison result refers to the result obtained by comparing the elevation angle score of each pixel in the target polarization angle image with the preset elevation angle score threshold. This result is used to determine whether the elevation angle score of each pixel reaches the preset elevation angle score threshold, providing a criterion for selecting pixels and constructing an elevation angle mask for the target polarization angle image.

[0059] The second preset filtering condition refers to the pre-set pixel selection rules. For example, the second preset filtering condition is that the elevation angle score is greater than or equal to the preset elevation angle score threshold. Pixels that meet the second preset filtering condition are considered as high-confidence pixels, and pixels that do not meet the second preset filtering condition are considered as low-confidence pixels.

[0060] The elevation angle mask of the target polarization angle image refers to a binary mask constructed from high-confidence pixels selected according to the second preset screening conditions based on the second comparison result.

[0061] In some cases, embodiments of this application first determine the elevation angle score of the target polarization angle image based on the confidence score field, then compare the elevation angle score with a preset elevation angle score threshold (such as taking the score threshold corresponding to the top 10% of high-scoring pixels in the elevation angle score) to obtain a second comparison result, and then determine the pixels that meet the second preset screening conditions (i.e., the elevation angle score is greater than or equal to the preset elevation angle score threshold) based on the second comparison result, thereby constructing an elevation angle mask of the target polarization angle image based on such pixels.

[0062] This application embodiment determines the elevation angle score of the target polarization angle image and compares it with a preset elevation angle score threshold. High-confidence pixels that meet the second preset screening conditions are selected to construct an elevation angle mask. This can accurately select pixels that contribute highly to the elevation angle calculation and have strong resistance to measurement noise, while removing low-confidence pixels. This provides a basis for extracting reliable mask features in the future, improves the targeting and rationality of the mask, and ensures the accuracy and robustness of elevation angle calculation in the subsequent solar vector calculation process.

[0063] Based on the description of the above embodiments, it can be understood that both the azimuth mask and the elevation mask are binary masks, and their construction logic is basically the same. Therefore, the embodiments of this application provide a unified and simplified description of the binary expressions corresponding to the two types of masks. The expressions for the binary masks corresponding to the azimuth mask and the elevation mask can be, but are not limited to, as follows: , in, For binary masking, For the corresponding pixel, an azimuth angle score or an elevation angle score is given. The preset scoring threshold for matching. When the binarized mask... A value of 1 indicates that the pixel is a high-confidence pixel; when the binarized mask... A value of 0 indicates that the pixel is a low-confidence pixel.

[0064] Optionally, in one embodiment of this application, before determining the mask features of the polarization angle image, the method further includes: determining a corresponding first mask feature based on the elevation angle mask of the target polarization angle image; and determining a corresponding second mask feature based on the azimuth angle mask of the target polarization angle image, so as to determine the mask features of the polarization angle image based on the first mask feature and the second mask feature.

[0065] In the embodiments of this application, the first mask feature can be understood as the morphological distribution pattern obtained from the height angle mask of the target polarization angle image. Specifically, the height angle mask of the target polarization angle image is mainly distributed in the collinear region (i.e., the region with large radial distance) that expands towards the edge of the field of view, and forms a hole in the center of the field of view after removing the low confidence singular region.

[0066] The second mask feature can be understood as the morphological distribution pattern obtained from the azimuth mask of the target polarization angle image. Specifically, the azimuth mask of the target polarization angle image is mainly distributed in the collinear region (i.e., the region with small radial distance) that converges towards the center of the field of view.

[0067] The mask features of the polarization angle image can be understood as mask configuration rules determined based on the first mask features and the second mask features, which can be used in the actual solution stage, providing a standard for quickly generating sparse sampling masks on polarization angle images with unknown solar vectors.

[0068] like Figure 3 As shown, after obtaining the elevation angle mask and azimuth angle mask of the target polarization angle image in this embodiment of the application, analysis revealed that the elevation angle mask and azimuth angle mask of the target polarization angle image exhibit significant morphological distribution patterns under full daylight conditions: both types of masks are based on the pixel position polar angle. With polarization angle measurement value Collinear regions that tend to be consistent share a common morphological feature. Based on this common morphological feature, the two types of masks exhibit different radial distance distribution patterns: such as Figure 4 As shown, the azimuth mask is mainly distributed in the collinear region converging towards the center of the field of view (i.e., the region with small radial distance); as Figure 5 As shown, the elevation angle mask is mainly distributed in the collinear region (i.e., the region with large radial distance) that expands towards the edge of the field of view, and forms a hole in the center of the field of view after removing the low confidence singular region.

[0069] This application embodiment extracts first mask features and second mask features from the elevation angle mask and azimuth angle mask of the target polarization angle image respectively, and determines the mask features of the polarization angle image, forming a set of universal configuration rules that can be directly reused. This provides a reliable prior basis for quickly generating targeted sparse sampling masks on polarization angle images with unknown solar vectors, and improves mask configuration efficiency.

[0070] In step S102, the mask features of the polarization angle image are determined, and the azimuth and elevation angle masks of the polarization angle image are constructed using the mask features.

[0071] In the embodiments of this application, the azimuth mask of the polarization angle image refers to a binary mask generated on the actual polarization angle image for solving based on the determined mask features. It is used to mark pixel areas with high confidence in solving the solar azimuth angle and to select pixels that are less affected by noise and have high solution accuracy to participate in the azimuth angle solution.

[0072] The elevation angle mask of a polarization angle image refers to a binary mask generated on the actual polarization angle image used for calculation based on the determined mask features. It is used to mark pixel regions with high confidence in the solar elevation angle calculation and to select pixels that are less affected by noise and have high calculation accuracy to participate in the elevation angle calculation.

[0073] In actual implementation, the embodiments of this application can utilize the mask features extracted from the target polarization angle image to generate corresponding azimuth and elevation angle masks on the polarization angle image to be solved by a polarization sensor.

[0074] In step S103, the corresponding pixel data is obtained from the polarization angle image using the azimuth mask and elevation mask, respectively. A three-dimensional E vector corresponding to the azimuth mask and elevation mask is constructed based on the pixel data, so as to calculate the solar vector of the polarization angle image based on the three-dimensional E vector.

[0075] In practical implementation, the embodiments of this application can obtain corresponding high-confidence pixel data from the polarization angle image using the azimuth and elevation angle masks of the polarization angle image, and then, through the orthogonal geometric constraints in Rayleigh scattering theory, combine the polarization angle measurement value obtained by the polarization sensor. Following the derivation in step S101, the two-dimensional observation information is reconstructed into a three-dimensional E-vector in the camera coordinate system. 3D E-vector The expression can be, but is not limited to: , in, .

[0076] Further, in one embodiment of this application, calculating the solar vector of a polarization angle image based on a three-dimensional E-vector includes: determining the three-dimensional E-vector corresponding to the azimuth mask of the polarization angle image, and determining the three-dimensional E-vector corresponding to the elevation mask of the polarization angle image; establishing a corresponding azimuth observation matrix based on the three-dimensional E-vector corresponding to the azimuth mask of the polarization angle image, and establishing a corresponding elevation observation matrix based on the three-dimensional E-vector corresponding to the elevation mask of the polarization angle image; calculating the optimized azimuth angle of the solar vector based on the azimuth observation matrix, and calculating the optimized elevation angle of the solar vector based on the elevation observation matrix; and determining the solar vector based on the optimized azimuth angle and the optimized elevation angle.

[0077] Since the azimuth and elevation angles in this embodiment of the application each use corresponding sparse sampling masks, two independent observation matrices need to be constructed during the solution process. , This is to obtain optimized azimuth and optimized elevation angles. Azimuth observation matrix It has higher confidence in azimuth angle solution and elevation angle observation matrix. It has higher confidence in solving for elevation angles. The unified construction expression for the observation matrix can be, but is not limited to, the following: , in, Here is the observation matrix, representing the azimuth observation matrix. or elevation angle observation matrix ; For the first The three-dimensional E-vector of each effective pixel in the camera coordinate system For the first The three-dimensional E-vector of each effective pixel in the camera coordinate system For the first The three-dimensional E-vector of each effective pixel in the camera coordinate system.

[0078] Furthermore, embodiments of this application can calculate the solar vector based on discrete observation matrices. Based on Rayleigh scattering geometric constraints, the solar vector in the camera coordinate system... The least squares objective function should be established based on the 3D E-vector corresponding to all valid pixels. The least squares objective function can be, but is not limited to, the following: , in, The sun vector in the camera coordinate system. This is the least squares solution operation.

[0079] Furthermore, embodiments of this application utilize the observation matrix... and By performing the above optimization solutions, the optimized azimuth and elevation angles in the camera coordinate system can be obtained analytically, and finally, a high-precision solar vector can be synthesized. Solar Vector The expression can be, but is not limited to: .

[0080] This application embodiment establishes corresponding azimuth and elevation observation matrices based on the 3D E-vectors corresponding to the azimuth and elevation masks of the polarization angle image, respectively. Then, it calculates the optimized azimuth and optimized elevation angles of the solar vector and synthesizes the solar vector. This effectively realizes the differentiated configuration of the observation matrix, maximizes the retention of the contribution of each effective pixel of the azimuth and elevation angles, and significantly improves the accuracy, stability and robustness of the solar vector solution.

[0081] The principle of the polarization solar vector calculation method based on confidence mask proposed in this application is illustrated below with a specific embodiment.

[0082] like Figure 6 As shown on the left, the embodiments of this application can first perform sparse sampling strategy modeling and analysis: For the target polarization angle image, an orthogonal constraint model of the solar vector and the three-dimensional E vector is first established, then the propagation coefficient of polarization measurement noise on the solar vector solution error is analyzed, and a full-field confidence scoring function (which can be understood as a confidence scoring field) is constructed to quantify the reliability level of each pixel in the image plane for solar vector solution. Based on the confidence scoring field, two types of binary sparse sampling masks, azimuth and elevation angle, are constructed. High-confidence pixel regions suitable for azimuth and elevation angle solutions are selected respectively. Subsequently, the distribution patterns of the two types of masks are analyzed. It is found that the azimuth mask exhibits the characteristics of converging at the center of the field of view and being distributed with a small radial distance, while the elevation angle mask exhibits the characteristics of extending along the edge of the field of view and being distributed with a large radial distance. Moreover, the two have a common collinear morphological basis. It should be noted that, based on the above distribution patterns, the embodiments of this application can configure discrete binary sparse sampling masks in the solution stage.

[0083] like Figure 6As shown on the right, this embodiment of the application uses a polarization sensor to acquire a sky polarization angle image, and configures a binary sparse sampling mask for the polarization angle image (i.e., an azimuth angle mask and an elevation angle mask for the polarization angle image) based on the aforementioned distribution pattern. Furthermore, this embodiment uses the azimuth angle mask to filter effective pixels, reconstructing the corresponding 3D E-vector based on the polarization angle measurements of the effective pixels, and uses the elevation angle mask to filter effective pixels, reconstructing the corresponding 3D E-vector based on the polarization angle measurements of the effective pixels. Then, corresponding azimuth angle observation matrices and elevation angle observation matrices are established respectively. Subsequently, the azimuth angle (including the optimal azimuth angle) is calculated and optimized based on the azimuth angle observation matrix, and the elevation angle (including the optimal elevation angle) is calculated and optimized based on the elevation angle observation matrix. The optimized azimuth angle and optimized elevation angle are then combined to obtain a complete solar vector (including the optimal solar vector) in the camera coordinate system.

[0084] The polarization solar vector calculation method based on confidence masks proposed in this application acquires polarization angle images, constructs azimuth and elevation angle masks using mask features, extracts corresponding pixel data from the polarization angle images, reconstructs a three-dimensional E-vector, and then calculates the solar vector. This method enables differentiated filtering and sparse sampling of observed pixels, effectively reducing the interference of low-quality observation data on the global optimization process, suppressing random fluctuations in the calculation results, and significantly reducing computational resource consumption and data processing volume. Furthermore, the dimensional processing of azimuth and elevation angle masks improves the accuracy and stability of solar vector calculation, significantly enhancing the system's robustness and engineering applicability in complex environments. Therefore, this method solves the problems of related technologies that employ full-pixel sampling and full-field equal-weight fusion strategies, resulting in huge computational resource consumption and low-quality observation data interfering with global optimization, leading to random fluctuations in the calculation results and limiting the system's robustness in complex environments.

[0085] Next, referring to the accompanying drawings, a polarization solar vector solving device based on a confidence mask is described according to an embodiment of this application.

[0086] Figure 7 This is a block diagram of a polarization solar vector solving device based on a confidence mask according to an embodiment of this application.

[0087] like Figure 7 As shown, the polarization solar vector solving device 70 based on confidence mask includes: an acquisition module 100, a first determination module 200, and a calculation module 300.

[0088] The acquisition module 100 is used to acquire polarization angle images of the sky region.

[0089] The first determining module 200 is used to determine the mask features of the polarization angle image, and to construct the azimuth mask and elevation mask of the polarization angle image using the mask features.

[0090] The calculation module 300 is used to obtain the corresponding pixel data from the polarization angle image using the azimuth angle mask and the elevation angle mask respectively, and to construct a three-dimensional E vector corresponding to the azimuth angle mask and the elevation angle mask based on the pixel data, so as to calculate the solar vector of the polarization angle image based on the three-dimensional E vector.

[0091] Optionally, in one embodiment of this application, it further includes: a second determining module, a first constructing module, a third determining module, a fourth determining module, and a second constructing module.

[0092] The second determining module is used to determine the orthogonality between the solar vector and the three-dimensional E vector of the target polarization angle image.

[0093] The first construction module is used to construct an orthogonal constraint equation between the solar vector of the target polarization angle image and the measured polarization angle value of the target polarization angle image based on the orthogonality relationship.

[0094] The third determining module is used to determine the error propagation coefficient of the polarization measurement noise of the target polarization angle image to the solar vector of the target polarization angle image based on the orthogonal constraint equation.

[0095] The fourth determination module is used to determine the azimuth and elevation angle scores of the target polarization angle image based on the error propagation coefficient.

[0096] The second construction module is used to construct a confidence score field based on the azimuth and elevation angle scores of the target polarization angle image.

[0097] Optionally, in one embodiment of this application, it further includes: a fifth determining module, a first comparison module, a sixth determining module, and a third constructing module.

[0098] The fifth determining module is used to determine the azimuth angle score of the target polarization angle image based on the confidence score field.

[0099] The first comparison module is used to compare the azimuth angle score with a preset azimuth angle score threshold to obtain the first comparison result of the target polarization angle image.

[0100] The sixth determining module is used to determine the pixels that meet the first preset filtering conditions based on the first comparison result.

[0101] The third construction module is used to construct an azimuth mask of the target polarization angle image based on the pixels that meet the first preset screening conditions.

[0102] Optionally, in one embodiment of this application, it further includes: a seventh determining module, a second comparison module, an eighth determining module, and a fourth constructing module.

[0103] The seventh determination module is used to determine the elevation angle score of the target polarization angle image based on the confidence score field.

[0104] The second comparison module is used to compare the elevation angle score with a preset elevation angle score threshold to obtain a second comparison result of the target polarization angle image.

[0105] The eighth determining module is used to determine the pixels that meet the second preset filtering conditions based on the second comparison result.

[0106] The fourth construction module is used to construct an elevation angle mask of the target polarization angle image based on the pixels that meet the second preset screening conditions.

[0107] Optionally, in one embodiment of this application, it further includes a ninth determining module and a tenth determining module.

[0108] The ninth determining module is used to determine the corresponding first mask feature based on the elevation angle mask of the target polarization angle image.

[0109] The tenth determining module is used to determine the corresponding second mask features based on the azimuth mask of the target polarization angle image, so as to determine the mask features of the polarization angle image based on the first mask features and the second mask features.

[0110] Optionally, in one embodiment of this application, the calculation module 300 includes: a first determining unit, an establishing unit, a calculation unit, and a second determining unit.

[0111] The first determining unit is used to determine the three-dimensional E vector corresponding to the azimuth mask of the polarization angle image and to determine the three-dimensional E vector corresponding to the elevation mask of the polarization angle image.

[0112] The unit is used to establish the corresponding azimuth observation matrix based on the three-dimensional E-vector corresponding to the azimuth mask of the polarization angle image, and to establish the corresponding elevation angle observation matrix based on the three-dimensional E-vector corresponding to the elevation mask of the polarization angle image.

[0113] The calculation unit is used to calculate the optimized azimuth angle of the solar vector based on the azimuth observation matrix and the optimized altitude angle of the solar vector based on the altitude observation matrix.

[0114] The second determining unit is used to determine the solar vector based on the optimized azimuth angle and the optimized elevation angle.

[0115] It should be noted that the foregoing explanation of the embodiment of the polarization solar vector calculation method based on confidence mask also applies to the polarization solar vector calculation device based on confidence mask in this embodiment, and will not be repeated here.

[0116] The polarization solar vector calculation device based on confidence masks proposed in this application acquires polarization angle images, constructs azimuth and elevation angle masks using mask features, extracts corresponding pixel data from the polarization angle images, reconstructs a three-dimensional E-vector, and then calculates the solar vector. This enables differentiated filtering and sparse sampling of observed pixels, effectively reducing the interference of low-quality observation data on the global optimization process, suppressing random fluctuations in the calculation results, and significantly reducing computational resource consumption and data processing volume. Furthermore, the dimensional processing of azimuth and elevation angle masks improves the accuracy and stability of the solar vector calculation, significantly enhancing the system's robustness and engineering applicability in complex environments. Therefore, this solves the problems of related technologies that employ full-pixel sampling and full-field equal-weight fusion strategies, resulting in huge computational resource consumption and low-quality observation data interfering with global optimization, leading to random fluctuations in the calculation results and limiting the system's robustness in complex environments.

[0117] Figure 8 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application. The electronic device may include: The memory 801, the processor 802, and the computer program stored on the memory 801 and capable of running on the processor 802.

[0118] When the processor 802 executes the program, it implements the polarization solar vector calculation method based on confidence mask provided in the above embodiments.

[0119] Furthermore, electronic devices also include: Communication interface 803 is used for communication between memory 801 and processor 802.

[0120] The memory 801 is used to store computer programs that can run on the processor 802.

[0121] The memory 801 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0122] If the memory 801, processor 802, and communication interface 803 are implemented independently, then the communication interface 803, memory 801, and processor 802 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized into address buses, data buses, control buses, etc. For ease of representation, Figure 8 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0123] Optionally, in a specific implementation, if the memory 801, processor 802, and communication interface 803 are integrated on a single chip, then the memory 801, processor 802, and communication interface 803 can communicate with each other through an internal interface.

[0124] The processor 802 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0125] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for calculating polarization solar vectors based on confidence masks.

[0126] This application also provides a computer program product, including a computer program that, when executed, implements the above-described method for calculating polarization solar vectors based on confidence masks.

[0127] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0128] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0129] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0130] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0131] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, it can be implemented using any one or more of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0132] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0133] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0134] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for calculating polarization solar vectors based on confidence masks, characterized in that, Includes the following steps: Acquire polarization angle images of the sky region; Determine the mask features of the polarization angle image, and use the mask features to construct the azimuth angle mask and elevation angle mask of the polarization angle image respectively; The corresponding pixel data is obtained from the polarization angle image using the azimuth and elevation angle masks, respectively. A three-dimensional E-vector corresponding to the azimuth and elevation angle masks is constructed based on the pixel data, and the solar vector of the polarization angle image is calculated based on the three-dimensional E-vector.

2. The method according to claim 1, characterized in that, Before determining the mask features of the polarization angle image, the method further includes: Determine the orthogonality between the solar vector and the three-dimensional E vector in the target polarization angle image; Based on the orthogonality relationship, construct an orthogonal constraint equation between the solar vector of the target polarization angle image and the measured polarization angle value of the target polarization angle image; The error propagation coefficient of the polarization measurement noise of the target polarization angle image to the solar vector of the target polarization angle image is determined according to the orthogonal constraint equation. The azimuth and elevation scores of the target polarization angle image are determined based on the error propagation coefficient. A confidence score field is constructed based on the azimuth and elevation scores of the target polarization angle image.

3. The method according to claim 2, characterized in that, Before determining the mask features of the polarization angle image, the method further includes: Based on the confidence score field, the azimuth score of the target polarization angle image is determined; The azimuth angle score is compared with a preset azimuth angle score threshold to obtain a first comparison result of the target polarization angle image; Pixels that meet the first preset filtering conditions are determined based on the first comparison result; An azimuth mask for the target polarization angle image is constructed based on the pixels that meet the first preset screening conditions.

4. The method according to claim 3, characterized in that, Before determining the mask features of the polarization angle image, the method further includes: Based on the confidence score field, the elevation angle score of the target polarization angle image is determined; The elevation angle score is compared with a preset elevation angle score threshold to obtain a second comparison result of the target polarization angle image; The pixels that meet the second preset filtering conditions are determined based on the second comparison result; Construct an elevation angle mask for the target polarization angle image based on the pixels that meet the second preset filtering conditions.

5. The method according to claim 4, characterized in that, Before determining the mask features of the polarization angle image, the method further includes: The corresponding first mask feature is determined based on the elevation angle mask of the target polarization angle image; The corresponding second mask feature is determined based on the azimuth mask of the target polarization angle image, so as to determine the mask feature of the polarization angle image based on the first mask feature and the second mask feature.

6. The method according to claim 1, characterized in that, The step of calculating the solar vector of the polarization angle image based on the three-dimensional E vector includes: Determine the three-dimensional E-vector corresponding to the azimuth mask of the polarization angle image, and determine the three-dimensional E-vector corresponding to the elevation mask of the polarization angle image; A corresponding azimuth observation matrix is ​​established based on the three-dimensional E-vector corresponding to the azimuth mask of the polarization angle image, and a corresponding elevation angle observation matrix is ​​established based on the three-dimensional E-vector corresponding to the elevation mask of the polarization angle image. The optimal azimuth angle of the solar vector is calculated based on the azimuth observation matrix, and the optimal altitude angle of the solar vector is calculated based on the altitude observation matrix. The solar vector is determined based on the optimized azimuth angle and the optimized altitude angle.

7. A polarization solar vector solving device based on confidence mask, characterized in that, include: The acquisition module is used to acquire polarization angle images of the sky region; A determination module is used to determine the mask features of the polarization angle image, and to construct the azimuth mask and elevation mask of the polarization angle image using the mask features respectively; The calculation module is used to obtain corresponding pixel data from the polarization angle image using the azimuth mask and the elevation mask, respectively, and to construct a three-dimensional E vector corresponding to the azimuth mask and the elevation mask based on the pixel data, so as to calculate the solar vector of the polarization angle image based on the three-dimensional E vector.

8. An electronic device, characterized in that, include: The memory, the processor, and the computer program stored in the memory and capable of running on the processor, the processor executing the program to implement the polarization solar vector calculation method based on confidence mask as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the polarization solar vector calculation method based on confidence mask as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, The computer program is executed to implement the polarization solar vector calculation method based on confidence mask as described in any one of claims 1-6.

Images