Method and device for improving GNSS positioning precision, computer equipment and medium

The non-line-of-sight signals are identified through infrared fisheye cameras and sky segmentation algorithms, and an equivalent weight matrix is ​​constructed to process GNSS data, which solves the problem of reduced GNSS positioning accuracy and significantly improves positioning accuracy and reliability in urban environments.

CN118981032APending Publication Date: 2024-11-19Chinese People's Liberation Army Cyberspace Force Information Engineering University
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202410975588.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-19
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

In complex urban environments, the reduction in GNSS positioning accuracy is mainly due to the influence of non-line-of-sight signals, resulting in a significant increase in positioning errors, which is difficult to effectively handle in the prior art, affecting the availability and reliability of GNSS.

Method used

Images of the sky and occlusion areas are obtained through infrared fisheye cameras, and the non-horizontal signal is identified using sky segmentation algorithm and satellite coordinate projection technology, and the non-horizontal signal equivalence weight matrix is ​​constructed, and the GNSS positioning data is downweighted to correct positioning result.

Benefits of technology

The accurate identification and processing of non-line-of-sight signals is realized, which weakens its impact on GNSS positioning, improves positioning accuracy and reliability, and significantly improves position errors in urban occlusion environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118981032A_ABST
    Figure CN118981032A_ABST
Patent Text Reader

Abstract

The embodiment of the invention provides a method and a device for improving GNSS (Global Navigation Satellite System) positioning accuracy, computer equipment and a medium, and relates to the technical field of satellite navigation, and the method comprises the following steps: dividing an infrared fisheye picture into a sky region and a shelter region by adopting a sky segmentation algorithm, and generating a sky segmentation map; obtaining a satellite position of a satellite under the earth-centered earth-fixed coordinate system, converting the satellite position from the earth-centered earth-fixed coordinate system to a pixel coordinate system, generating satellite pixel points under the pixel coordinate system, projecting the satellite pixel points into the infrared fisheye image, and generating a satellite coordinate projection drawing; calculating the proportion of the shielding object area in a set range, and judging whether the satellite is a non-line-of-sight satellite or not according to the proportion; and performing weight reduction processing on the observation value through a non-line-of-sight signal equivalent weight matrix, and correcting an output GNSS positioning result. According to the scheme, the observation value of the non-line-of-sight satellite is subjected to weight reduction processing, so that the precision of GNSS positioning is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of satellite navigation technology, and in particular to a method, device, computer equipment and medium for improving GNSS positioning accuracy. Background Art

[0002] In recent years, with the full completion of my country's BeiDou Navigation Satellite System (BDS), the continued advancement of the modernization of the US GPS (Global Positioning System) and Russia's GLONASS (Global Navigation Satellite System), and the continuous development and improvement of the EU Galileo, the global navigation satellite system GNSS has entered a new era of multi-constellation integration and complementarity. At present, the number of satellites in orbit of the four major satellite navigation systems has reached more than 140, and more than 40 satellites can be observed simultaneously around the world. The integration of multiple GNSS constellations has significantly improved the observation redundancy and optimized the constellation space configuration. Especially in harsh and complex urban occlusion environments, multi-constellation GNSS can provide navigation users with more continuous, robust and reliable PNT services.

[0003] However, in the face of increasingly diverse and complex application requirements, GNSS is severely degraded or even unavailable in complex urban environments such as urban canyons, viaducts, and tree-shaded areas, which greatly affects the GNSS precision positioning performance and cannot guarantee the availability, continuity, and reliability of GNSS PNT services. The multipath effect is an important factor affecting GNSS positioning performance in urban obstruction scenarios, which can cause positioning errors of tens or even hundreds of meters.

[0004] In obstructed scenarios, GNSS signals are easily diffracted and reflected by surrounding buildings, forming line-of-sight (LOS) signals, non-line-of-sight (NLOS) signals, and multipath signals. NLOS signals and multipath signals are collectively referred to as multipath effects. The difference between the two is that multipath signals are the superposition of indirect signals and direct signals caused by reflection or diffraction, while NLOS only contains indirect signals such as diffraction or reflection. The ranging error caused by NLOS signals is generally much larger than that of multipath signals, and in severe cases can cause positioning errors of hundreds of meters or even unbounded. Therefore, studying NLOS signal processing under the influence of strong interference, low cost, multi-maneuverability, and high effectiveness is a difficult problem and key to the high-precision and high-reliability applications of GNSS, and is a breakthrough to further expand its scope of application.

[0005] Antenna and receiver improvement methods, such as the use of choke antennas and dual-polarized antennas, can only partially reduce the impact of multipath errors, and will greatly increase hardware costs. They are mainly used in the field of high-precision geodetic measurements and are not suitable for dynamic navigation and positioning of low-cost mass users in urban environments. With the rapid popularization of multi-sensors, new opportunities have been provided for the efficient recognition of GNSS NLOS signals. Among them, visual sensors can perceive rich environmental information around them and achieve effective detection of NLOS satellites; inertial navigation systems (INS) can continuously and accurately provide position, speed and attitude information to the carrier. Under harsh conditions such as large changes in lighting conditions, insufficient lighting or smoke, the imaging performance of visible light cameras is severely reduced or even unable to operate normally. Infrared cameras have the advantages of all-weather and all-day, and can achieve high-quality imaging at the same time during the day and at night. Therefore, how to fuse the observation data of infrared fisheye vision / inertial sensors to identify NLOS signals and improve the positioning accuracy of GNSS in urban occlusion environments is a key problem that needs to be solved urgently. Summary of the invention

[0006] In view of this, an embodiment of the present invention provides a method for improving GNSS positioning accuracy to solve the technical problem of reduced GNSS positioning accuracy caused by non-line-of-sight signals in the prior art. The method includes:

[0007] An infrared fisheye camera is used to obtain an infrared fisheye image facing the sky, and a sky segmentation algorithm is used to divide the infrared fisheye image into a sky area and an occlusion area to generate a sky segmentation map;

[0008] Obtain the satellite position of the satellite in the geocentric earth-fixed coordinate system, convert the satellite position from the geocentric earth-fixed coordinate system to the pixel coordinate system, generate satellite pixel points in the pixel coordinate system, project the satellite pixel points into the infrared fisheye image, and generate a satellite coordinate projection map;

[0009] The sky segmentation map and satellite coordinate projection map are used to calculate the proportion of the obstruction area within the set range, and the proportion is used to determine whether the satellite is a non-line-of-sight satellite.

[0010] If the satellite is a non-line-of-sight satellite, obtain the observation value of the non-line-of-sight satellite in the GNSS positioning data, construct a non-line-of-sight signal equivalent weight matrix, downweight the observation value through the non-line-of-sight signal equivalent weight matrix, and correct the output GNSS positioning result.

[0011] The embodiment of the present invention also provides a device for improving GNSS positioning accuracy to solve the technical problem of reduced GNSS positioning accuracy caused by non-line-of-sight signals in the prior art. The device includes:

[0012] A sky segmentation map generation module is used to obtain an infrared fisheye image facing the sky through an infrared fisheye camera, and use a sky segmentation algorithm to divide the infrared fisheye image into a sky area and an occlusion area to generate a sky segmentation map;

[0013] A satellite coordinate projection map generation module is used to obtain the satellite position of the satellite in the geocentric earth-fixed coordinate system, convert the satellite position from the geocentric earth-fixed coordinate system to the pixel coordinate system, generate satellite pixel points in the pixel coordinate system, project the satellite pixel points into the infrared fisheye image, and generate a satellite coordinate projection map;

[0014] A non-line-of-sight satellite judgment module is used to calculate the proportion of the obstruction area within the set range through the sky segmentation map and the satellite coordinate projection map, and judge whether the satellite is a non-line-of-sight satellite based on the proportion;

[0015] The positioning data correction module is used to obtain the observation value of the non-line-of-sight satellite in the GNSS positioning data if the satellite is a non-line-of-sight satellite, construct the non-line-of-sight signal equivalent weight matrix, down-weight the observation value through the non-line-of-sight signal equivalent weight matrix, and correct the output GNSS positioning result.

[0016] An embodiment of the present invention further provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any of the above-mentioned methods for improving GNSS positioning accuracy when executing the computer program, thereby solving the technical problem in the prior art of reduced GNSS positioning accuracy caused by non-line-of-sight signals.

[0017] An embodiment of the present invention further provides a computer-readable storage medium storing a computer program for executing any of the above-mentioned methods for improving GNSS positioning accuracy, so as to solve the technical problem in the prior art that the GNSS positioning accuracy is reduced due to non-line-of-sight signals.

[0018] Compared with the prior art, the at least one technical solution adopted in the embodiments of this specification can achieve the following beneficial effects:

[0019] Through hardware equipment (inertial navigation system and infrared fisheye camera) and three main steps of sky segmentation, satellite coordinate projection and non-line-of-sight signal recognition, accurate recognition of non-line-of-sight signals is achieved. According to the results of non-line-of-sight signal recognition, the impact of non-line-of-sight signals on GNSS positioning is weakened, and the accuracy of GNSS positioning is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 is a flow chart of a method for improving GNSS positioning accuracy provided by an embodiment of the present invention;

[0022] Figure 2 is a flow chart of a method for improving GNSS positioning accuracy provided by an embodiment of the present invention;

[0023] Figure 3 is a flow chart of satellite coordinate projection according to an embodiment of the present invention;

[0024] Figure 4 is a schematic diagram of identifying a non-line-of-sight signal according to an embodiment of the present invention;

[0025] Figure 5 is a structural block diagram of a computer device provided by an embodiment of the present invention;

[0026] Figure 6 It is a structural block diagram of a device for improving GNSS positioning accuracy provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0028] The following describes the implementation methods of the present application through specific examples, and those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The present application can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, in the absence of conflict, the following embodiments and the features in the embodiments can be combined with each other. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work belong to the scope of protection of the present application.

[0029] The hardware devices used in the present invention include:

[0030] (1) GNSS (GNSS receiver), which can receive signals from satellite navigation systems such as GPS, BDS, Galileo and GLONASS in real time.

[0031] (2) Inertial Navigation System (INS), including Inertial Measurement Unit (IMU), which consists of accelerometers and gyroscopes, provides the position and attitude of the carrier. The inertial navigation system can continuously provide the carrier with high-precision position, velocity and attitude information;

[0032] (3) Infrared fisheye cameras can work in all-weather and all-day conditions to achieve imaging of the sky and obstructed areas. Infrared fisheye cameras have a wide viewing angle and can perceive rich surrounding environment information both during the day and at night.

[0033] The present invention combines the advantages of an inertial navigation system and an infrared fisheye camera to realize the recognition of GNSS non-line-of-sight signals throughout the day.

[0034] In an embodiment of the present invention, a method for improving GNSS positioning accuracy is provided, such as Figure 1 and Figure 2 As shown, the method includes:

[0035] Step S101: obtaining an infrared fisheye image facing the sky through an infrared fisheye camera, dividing the infrared fisheye image into a sky area and an occlusion area using a sky segmentation algorithm, and generating a sky segmentation map;

[0036] Step S102: obtaining the satellite position of the satellite in the geocentric earth-fixed coordinate system, converting the satellite position from the geocentric earth-fixed coordinate system to the pixel coordinate system, generating satellite pixel points in the pixel coordinate system, projecting the satellite pixel points into the infrared fisheye image, and generating a satellite coordinate projection map;

[0037] Step S103: Calculate the proportion of the obstruction area within the set range through the sky segmentation map and the satellite coordinate projection map, and determine whether the satellite is a non-line-of-sight satellite based on the proportion;

[0038] Step S104: If the satellite is a non-line-of-sight satellite, obtain the observation value of the non-line-of-sight satellite in the GNSS positioning data, construct a non-line-of-sight signal equivalent weight matrix, down-weight the observation value through the non-line-of-sight signal equivalent weight matrix, and correct the output GNSS positioning result.

[0039] In the specific implementation, in order to determine whether the satellite is in the sky area or the obstruction area, the following steps are used to implement the sky segmentation algorithm to divide the infrared fisheye image into the sky area and the obstruction area:

[0040] The maximum inter-class variance method is used to extract the grayscale difference between the sky and the occluder in the infrared fisheye photo, and the grayscale difference is compared with a set threshold. The part of the infrared fisheye picture where the grayscale difference is greater than the set threshold is segmented into a first sky area; through a morphological opening operation, the holes in the first sky area are removed to generate a second sky area; through a morphological closing operation, the second sky area is optimized and denoised to generate a sky area corresponding to the infrared fisheye picture, and the area in the infrared fisheye picture except the sky area is used as the occluder area.

[0041] Specifically, a wide-angle infrared fisheye camera facing the sky is used to shoot, and the captured infrared fisheye image is subjected to sky segmentation processing to extract the sky area and the occlusion area. The specific steps are as follows:

[0042] (I) Rough sky segmentation based on the maximum inter-class variance method.

[0043] By using the difference in grayscale between the target object to be extracted and the background in the image, a suitable threshold is selected, and the image is segmented into two categories, background and target, by judging the characteristic attributes of each pixel. It can be seen that determining an optimal threshold is the key to image segmentation. The specific steps are as follows:

[0044] Assume an image of size m×n pixels, with a total of L different gray levels. The target and background segmentation threshold is T, the proportion of target points in the image is P0, the average gray level is μ0, and the proportion of background points in the image is P B , the average grayscale is μ B The total average grayscale of the image is:

[0045] μ=P O ×μ O +P B ×μ B .

[0046] Between-class variance σ 2 It can be expressed as:

[0047] σ 2 =P O ×(μ O -μ) 2 +P B ×(μ B -μ) 2 .

[0048] From the minimum gray level 0 to the maximum gray level L of the image, the segmentation threshold T is traversed. When the inter-class variance of the objective function σ 2 When the maximum is taken, the segmentation threshold T is the required line-of-sight signal segmentation threshold.

[0049] The sky image (infrared fisheye sky image) is binarized and segmented according to the line-of-sight signal segmentation threshold.

[0050] (ii) Mis-segmentation removal based on morphological opening operation.

[0051] After the sky image (infrared fisheye sky image) is roughly segmented using the maximum inter-class variance method, there will be isolated small holes in the image, making the sky area no longer continuous. The opening operation in the morphological operation (first corrosion and then expansion) can remove isolated points, burrs and small points connecting two areas in the image, so it can be used to eliminate the mis-segmented areas in the image after the rough segmentation of the Otsu method.

[0052] Grayscale morphological expansion is centered at a certain pixel (x, y) and takes the local grayscale maximum value. The expression is:

[0053] dst(x, y) = max (x′,y′):element(x′,y′)≠0 src(x+x′,y+y′).

[0054] The operation of erosion is opposite to that of expansion. To find the local minimum value of the image grayscale, the expression is:

[0055] dst(x, y) = min (x′,y′):element(x′,y′)≠0 src(x+x′,y+y′).

[0056] Where src(x, y) is the original pixel value of the image, dst(x, y) is the pixel value after morphological processing, and element is the m×n template.

[0057] First, grayscale erosion is performed on the image to remove small isolated noise areas, and then grayscale dilation is performed to fill smaller holes.

[0058] (iii) Segmentation optimization based on morphological closing operation.

[0059] After the sky is roughly segmented and the false segmentation is removed, the white target pixels in the image basically represent the sky area, and the black background pixels represent buildings and other large obstacles. The main goal of this paper is to identify occlusions caused by large obstacles, so small occlusions (such as traffic signs) will be regarded as noise. The closing operation in the morphological operation (dilation followed by erosion) can connect adjacent objects and smooth the boundaries of the sky area to achieve the purpose of removing noise.

[0060] First, the image is grayscale expanded to eliminate noise points such as small occlusions; then grayscale erosion is performed to restore the main form of the image.

[0061] In specific implementation, the satellite position in the Earth-centered Earth-fixed coordinate system is obtained through the following steps:

[0062] The satellite position based on the Earth-centered Earth-fixed coordinate system is obtained through GNSS positioning data and satellite ephemeris.

[0063] In a specific implementation, in order to project the satellite into the infrared fisheye image, the satellite position is converted from the Earth-centered Earth-fixed coordinate system to the pixel coordinate system to generate the satellite pixel points in the pixel coordinate system through the following steps:

[0064] The satellite position based on the Earth-centered Earth-fixed coordinate system is converted into the navigation satellite position based on the navigation coordinate system; the position and attitude of the infrared fisheye camera are obtained through the inertial navigation system, the camera extrinsic parameters and position and attitude of the infrared fisheye camera are used as parameters of the inertial navigation system, and the navigation satellite position is converted into a satellite camera coordinate point based on the camera coordinate system through the inertial navigation system; through the fisheye distortion correction of the infrared fisheye camera, the satellite camera coordinate points are projected into satellite image plane coordinate points in the image plane coordinate system; through the camera intrinsic parameters of the infrared fisheye camera, the satellite image plane coordinate points are converted into satellite pixel points in the pixel coordinate system.

[0065] Specifically, satellite coordinate projection is essentially the process of projecting the satellite's three-dimensional spatial coordinates in the Earth-centered fixed system to two-dimensional pixel points in the pixel coordinate system through a series of coordinate rotations, translations, and scaling transformations, and finally forming a satellite sky projection map.

[0066] The process of satellite coordinate projection is as follows Figure 3 As shown, it can be summarized as:

[0067] 1) Calculate the satellite position in the Earth-centered Earth-fixed coordinate system using observation data and multi-system precise orbit clock products, and convert the station-satellite vector to the navigation coordinate system in combination with the real-time calculated receiver position;

[0068] 2) The point to be projected in the navigation coordinate system is converted to the camera coordinate system after the position and attitude correction of the IMU sensor, combined with the external parameters of the infrared fisheye camera;

[0069] 3) The spatial coordinate points in the camera coordinate system are projected into plane coordinate points in the image physical coordinate system after being corrected for infrared fisheye distortion;

[0070] 4) The projection points in the image plane coordinate system are converted to the pixel coordinate system in combination with the internal parameters of the infrared fisheye camera, and finally form the satellite's sky projection map.

[0071] In the specific implementation, in order to determine the non-line-of-sight satellite, the ranging value of the non-line-of-sight satellite with a relatively large error is downgraded. The following steps are implemented to calculate the proportion of the obstruction area within the set range through the sky segmentation map and the satellite coordinate projection map, and determine whether the satellite is a non-line-of-sight satellite based on the proportion:

[0072] A disk is generated with the projection point of the satellite on the satellite coordinate projection map as the center of the circle and the set value as the radius, and the disk is used as the set range; it is determined whether all the pixels in the sky segmentation map are within the set range, and the pixels within the set range are obtained as the target pixels; the proportion of the occluded area within the set range is calculated through the grayscale value of the target pixel, and the proportion is used as the confidence score; the confidence score is compared with the confidence threshold. If the confidence score is greater than or equal to the confidence threshold, the satellite is a non-line-of-sight satellite.

[0073] In the specific implementation, in order to calculate the confidence score, the following steps are performed to calculate the proportion of the occluded area within the set range through the grayscale value of the target pixel, and the proportion is used as the confidence score:

[0074] Get the grayscale value of each pixel within the set range, determine whether the corresponding pixel is an occluded pixel by the grayscale value, assign values ​​to the occluded pixel and the non-occluded pixel respectively, and generate the occlusion value corresponding to the pixel; add the occlusion values ​​of all pixels within the set range to get the sum of the occlusion values; calculate the confidence score by the sum of the occlusion values ​​and the total number of pixels within the set range.

[0075] Specifically, non-line-of-sight (NLOS) signals are identified. After sky image segmentation and satellite coordinate projection processing, the satellite is projected as a two-dimensional pixel point in the sky segmentation image. It is necessary to further combine certain recognition criteria to give the judgment conditions of whether the satellite is blocked by buildings or other obstacles, so as to analyze and determine whether the satellite signal is a non-line-of-sight (NLOS) signal.

[0076] like Figure 4 As shown in the figure, the idea of ​​counting the same-value areas of the origin in the SUSAN corner detection operator is borrowed. The projection point (x, y) of the satellite on the sky segmentation image is taken as the center and the disk with radius r is taken as the template atom (i.e., the set range). By counting the proportion of the non-sky area within the template atom N, it is determined whether the satellite is a non-line-of-sight (NLOS) type satellite.

[0077] By N = {(x′, y′)|(x′-x) 2 +(y′-y) 2 ≤r 2} Determine whether all pixel points in the sky segmentation map are within the set range, where c′ and y′ are the position and posture of the pixel points.

[0078]

[0079] Where x', y' are the position and posture of any pixel, g(x', y') is the gray value of the pixel. c(x', y') is the occlusion value corresponding to the pixel. When the gray value of the pixel is 0, c(x', y') is 1, representing an occluded pixel; when the gray value of the pixel is 255, c(x', y') is 0, representing a non-occluded pixel.

[0080] Add the occlusion values ​​together to get the total of the results:

[0081]

[0082] Where n(x, y) is the sum of the occlusion values ​​(i.e., the number of pixels in the non-sky area (occluder area)).

[0083] Calculate the confidence score:

[0084]

[0085] Where n max It is the maximum value that n can take (the total number of pixels in the template atom (set range)).

[0086] Compare the confidence score score with the preset confidence threshold θ to determine whether the satellite is a non-line-of-sight (NLOS) type:

[0087]

[0088] That is, if the confidence score is greater than or equal to the confidence threshold, the satellite is a non-line-of-sight satellite, and if the confidence score is less than the confidence threshold, the satellite is a line-of-sight satellite.

[0089] In one embodiment, θ=0.8, r=10 pixels.

[0090] In the specific implementation, in order to improve the accuracy of GNSS positioning, the following steps are used to construct the non-line-of-sight signal equivalent weight matrix, and the observation values ​​are down-weighted by the non-line-of-sight signal equivalent weight matrix to correct the output GNSS positioning results:

[0091] The maximum constant and the minimum constant are set respectively, and the weight of the observations obtained from the non-line-of-sight satellite is used as the weight of the observations before optimization. The standardized residual is calculated by the residual in the observations obtained from the non-line-of-sight satellite and the mean error corresponding to the residual; if the absolute value of the standardized residual is less than or equal to the minimum constant, the weight of the observation after optimization is set to the weight of the observation before optimization; if the absolute value of the standardized residual is less than the maximum constant and greater than the minimum constant, the weight of the observation after optimization is calculated by the weight of the observation before optimization, the maximum constant, the minimum constant and the standardized residual; if the absolute value of the standardized residual is greater than or equal to the maximum constant, the weight of the observation after optimization is set to 0; the observations obtained from the non-line-of-sight satellite are downgraded by the weight of the optimized observations, and the output GNSS positioning results are corrected.

[0092] Specifically, the GNSS positioning model is optimized based on non-line-of-sight (NLOS) recognition to improve the GNSS positioning accuracy.

[0093] In urban canyons with severe obstruction, if the detected non-line-of-sight (NLOS) signals are directly eliminated, the already insufficient number of available satellites in the urban canyons will be further reduced, causing serious distortion of the satellite space geometry, greatly reducing the positioning accuracy, and even making it impossible to position, while losing the useful information in the non-line-of-sight (NLOS) signals. Therefore, it is necessary to explore an effective processing strategy for non-line-of-sight (NLOS) observations to further improve the positioning accuracy and reliability of GNSS.

[0094] In order to reduce the impact of non-line-of-sight (NLOS) observations on GNSS positioning solutions, it is considered to expand the variance of the identified non-line-of-sight (NLOS) observations, and to downweight the NLOS observations by constructing an equivalent weight function to improve the performance of the Kalman filter solution in resisting NLOS gross error signals. The equivalent weight function used in the embodiment of the present invention is the IGGⅢ model.

[0095] In the IGGⅢ model:

[0096]

[0097] In the formula, p i The observation weights are determined by combining the signal-to-noise ratio and the satellite's elevation angle before optimization processing. To construct the weights of the observations after optimization of the equivalent weight function, is the standardized residual, v i is the residual of the observation value of the i-th satellite, v i The corresponding mean error k0 and k1 are constants, preferably, k0 = 1.0-1.5, k0 = 2.5-8.0.

[0098] After the weight reduction process, the influence of NLOS observations on GNSS positioning solution is weakened, and the positioning residual v before optimization is i The larger it is, the more obvious the weakening effect will be.

[0099] In one embodiment, the GNSS positioning accuracy is tested using the method of an embodiment of the present invention. The experiment used two sets of vehicle-mounted dynamic data collected in Wuhan and Zhengzhou, and both sets of data covered daytime and nighttime periods. The proportion of GNSS non-line-of-sight (NLOS) signals in different scenarios varies greatly, and the occlusion of trees and buildings is the main factor causing the occurrence of non-line-of-sight (NLOS) signals. Using the INS and infrared fisheye camera proposed in the present invention, all GNSS non-line-of-sight (NLOS) signals that occur can be identified. The GNSS positioning accuracy and stability after weakening the influence of non-line-of-sight (NLOS) signals are improved to a certain extent. Among them, the improvement effect of bridge occlusion and shaded environment is the most obvious, and the position error is reduced from 21.26m and 17.12m to 11.74m and 9.30m respectively, and the accuracy improvement is 20.8% and 19.5%.

[0100] In this embodiment, a computer device is provided, such as Figure 5 As shown, it includes a memory 501, a processor 502 and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, any of the above methods for improving GNSS positioning accuracy is implemented.

[0101] Specifically, the computer device may be a computer terminal, a server or a similar computing device.

[0102] In this embodiment, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program for executing any of the above methods for improving GNSS positioning accuracy.

[0103] Specifically, computer-readable storage media include permanent and non-permanent, removable and non-removable media, and information storage can be achieved by any method or technology. Information can be computer-readable instructions, data structures, modules of programs or other data. Examples of computer-readable storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable storage media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0104] Based on the same inventive concept, an apparatus for improving GNSS positioning accuracy is also provided in an embodiment of the present invention, as described in the following embodiments. Since the principle of solving the problem by the apparatus for improving GNSS positioning accuracy is similar to that of the method for improving GNSS positioning accuracy, the implementation of the apparatus for improving GNSS positioning accuracy can refer to the implementation of the method for improving GNSS positioning accuracy, and the repeated parts will not be repeated. As used below, the term "unit" or "module" can be a combination of software and / or hardware that implements a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0105] Figure 6 is a structural block diagram of a device for improving GNSS positioning accuracy according to an embodiment of the present invention, such as Figure 6 As shown, it includes: a sky segmentation map generation module 601, a satellite coordinate projection map generation module 602, a non-line-of-sight satellite judgment module 603 and a positioning data correction module 604. The structure is described below.

[0106] A sky segmentation map generation module 601 is used to obtain an infrared fisheye picture facing the sky through an infrared fisheye camera, and use a sky segmentation algorithm to divide the infrared fisheye picture into a sky area and an occlusion area to generate a sky segmentation map;

[0107] The satellite coordinate projection map generation module 602 is used to obtain the satellite position of the satellite in the geocentric earth-fixed coordinate system, convert the satellite position from the geocentric earth-fixed coordinate system to the pixel coordinate system, generate satellite pixel points in the pixel coordinate system, project the satellite pixel points into the infrared fisheye image, and generate a satellite coordinate projection map;

[0108] The non-line-of-sight satellite determination module 603 is used to calculate the proportion of the obstruction area within the set range through the sky segmentation map and the satellite coordinate projection map, and determine whether the satellite is a non-line-of-sight satellite based on the proportion;

[0109] The positioning data correction module 604 is used to obtain the observation value of the non-line-of-sight satellite in the GNSS positioning data if the satellite is a non-line-of-sight satellite, construct a non-line-of-sight signal equivalent weight matrix, down-weight the observation value through the non-line-of-sight signal equivalent weight matrix, and correct the output GNSS positioning result.

[0110] In one embodiment, the sky segmentation map generation module includes:

[0111] A coarse segmentation unit is used to extract the grayscale difference between the sky and the obstruction in the infrared fisheye photo by using the maximum inter-class variance method, compare the grayscale difference with a set threshold, and segment the part of the infrared fisheye photo where the grayscale difference is greater than the set threshold as a first sky area;

[0112] The mis-segmentation removal unit is used to remove holes in the first sky region through a morphological opening operation to generate a second sky region;

[0113] The final sky region segmentation unit is used to optimize and denoise the second sky region through morphological closing operations, generate a sky region corresponding to the infrared fisheye image, and use the area other than the sky area in the infrared fisheye image as the occlusion area.

[0114] In one embodiment, the satellite coordinate projection map generation module includes:

[0115] A navigation coordinate system conversion unit, used for converting a satellite position based on an earth-centered earth-fixed coordinate system into a navigation satellite position based on a navigation coordinate system;

[0116] The camera coordinate system conversion unit is used to obtain the position and attitude of the infrared fisheye camera through the inertial navigation system, use the camera external parameters and position and attitude of the infrared fisheye camera as parameters of the inertial navigation system, and convert the navigation satellite position into a satellite camera coordinate point based on the camera coordinate system through the inertial navigation system;

[0117] An image plane coordinate conversion unit is used to project the satellite camera coordinate points into satellite image plane coordinate points in an image plane coordinate system by correcting the fisheye distortion of the infrared fisheye camera;

[0118] The satellite pixel point generation unit is used to convert the satellite image plane coordinate points into satellite pixel points in the pixel coordinate system through the camera internal parameters of the infrared fisheye camera.

[0119] In one embodiment, the satellite coordinate projection map generation module further includes:

[0120] The satellite position acquisition unit is used to acquire the satellite position based on the Earth-centered Earth-fixed coordinate system through GNSS positioning data and satellite ephemeris.

[0121] In one embodiment, the non-line-of-sight satellite determination module includes:

[0122] A range setting unit, used to generate a disk with the projection point of the satellite on the satellite coordinate projection map as the center of the circle and the set value as the radius, and use the disk as the set range;

[0123] A target pixel point determination unit is used to determine whether all the pixel points in the sky segmentation map are within a set range, and obtain the pixel points within the set range as the target pixel points;

[0124] A confidence score calculation unit is used to calculate the proportion of the occluded area within the set range through the grayscale value of the target pixel point, and use the proportion as the confidence score;

[0125] The non-line-of-sight satellite judgment unit is used to compare the confidence score and the confidence threshold. If the confidence score is greater than or equal to the confidence threshold, the satellite is a non-line-of-sight satellite.

[0126] In one embodiment, a confidence score calculation unit is used to obtain the grayscale value of each pixel within a set range, determine whether the corresponding pixel is an occluded pixel by the grayscale value, assign values ​​to the occluded pixel and the non-occluded pixel respectively, and generate an occlusion value corresponding to the pixel; add the occlusion values ​​of all pixels within the set range to obtain the sum of the occlusion values; and calculate the confidence score by the sum of the occlusion values ​​and the total number of pixels within the set range.

[0127] In one embodiment, the positioning data correction module includes:

[0128] The standardized residual calculation unit is used to set the maximum constant and the minimum constant respectively, take the weight of the observation value obtained by the non-line-of-sight satellite as the weight of the observation value before optimization, and calculate the standardized residual through the residual in the observation value obtained by the non-line-of-sight satellite and the mean error corresponding to the residual;

[0129] A first weight determination unit, configured to set the weight of the observation value after optimization to the weight of the observation value before optimization if the absolute value of the standardized residual is less than or equal to the minimum constant;

[0130] A second weight judgment unit is used to calculate the optimized observation value weight by using the observation value weight before optimization, the maximum constant, the minimum constant and the standardized residual if the absolute value of the standardized residual is less than the maximum constant and greater than the minimum constant;

[0131] A third weight judgment unit is used to set the weight of the optimized observation value to 0 if the absolute value of the compared standardized residual is greater than or equal to the maximum constant;

[0132] The positioning data correction unit is used to correct the observation values ​​obtained from the non-line-of-sight satellites through the optimized observation value weights, and to correct the output GNSS positioning results.

[0133] The embodiments of the present invention achieve the following technical effects:

[0134] Accurate recognition of non-line-of-sight signals is achieved through hardware equipment (inertial navigation system and infrared fisheye camera) and three main steps of sky segmentation, satellite coordinate projection and non-line-of-sight signal recognition. The infrared fisheye camera has a large viewing angle and can perceive rich surrounding environment information both during the day and at night. The inertial navigation system can continuously provide position, speed and attitude information to the carrier with high precision. The method of the embodiment of the present invention combines the advantages of the two types of sensors to achieve the recognition of GNSS non-line-of-sight signals throughout the day. According to the results of non-line-of-sight signal recognition, an update scheme for the equivalent weight matrix of non-line-of-sight signals based on observation value residuals is designed to weaken the influence of non-line-of-sight signals on GNSS precise positioning. Experimental data show that after being processed by the method of the present invention, the accuracy of GNSS positioning is improved.

[0135] Obviously, those skilled in the art should understand that the modules or steps of the above-mentioned embodiments of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. In this way, the embodiments of the present invention are not limited to any specific combination of hardware and software.

[0136] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the embodiments of the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for improving GNSS positioning accuracy, characterized in that: include: An infrared fisheye camera is used to obtain an infrared fisheye picture facing the sky, and a sky segmentation algorithm is used to divide the infrared fisheye picture into a sky area and an occlusion area to generate a sky segmentation map; Obtaining a satellite position of a satellite in an earth-centered earth-fixed coordinate system, converting the satellite position from the earth-centered earth-fixed coordinate system to a pixel coordinate system, generating satellite pixel points in the pixel coordinate system, and projecting the satellite pixel points into the infrared fisheye image to generate a satellite coordinate projection map; Calculate the proportion of the obstruction area within a set range through the sky segmentation map and the satellite coordinate projection map, and determine whether the satellite is a non-line-of-sight satellite through the proportion; If the satellite is the non-line-of-sight satellite, the observation value of the non-line-of-sight satellite in the GNSS positioning data is obtained, a non-line-of-sight signal equivalent weight matrix is ​​constructed, the observation value is downgraded by the non-line-of-sight signal equivalent weight matrix, and the output GNSS positioning result is corrected.

2. The method for improving GNSS positioning accuracy according to claim 1, characterized in that: The infrared fisheye image is divided into a sky area and an occlusion area using a sky segmentation algorithm, including: Extracting the grayscale difference between the sky and the obstruction in the infrared fisheye photo by the maximum inter-class variance method, comparing the grayscale difference with a set threshold, and segmenting the portion of the infrared fisheye photo where the grayscale difference is greater than the set threshold as a first sky area; Removing holes in the first sky region through a morphological opening operation to generate a second sky region; The second sky area is optimized and denoised through a morphological closing operation to generate a sky area corresponding to the infrared fisheye image, and an area other than the sky area in the infrared fisheye image is used as an occlusion area.

3. The method for improving GNSS positioning accuracy according to claim 1, characterized in that: Converting the satellite position from the Earth-centered Earth-fixed coordinate system to a pixel coordinate system to generate satellite pixel points in the pixel coordinate system includes: Converting the satellite position based on the Earth-centered Earth-fixed coordinate system into a navigation satellite position based on a navigation coordinate system; The position and attitude of the infrared fisheye camera are obtained by an inertial navigation system, the camera external parameters of the infrared fisheye camera and the position and attitude are used as parameters of the inertial navigation system, and the navigation satellite position is converted into a satellite camera coordinate point based on a camera coordinate system by the inertial navigation system; Projecting the satellite camera coordinate points into satellite image plane coordinate points in an image plane coordinate system by correcting the fisheye distortion of the infrared fisheye camera; The satellite image plane coordinate points are converted into satellite pixel points in a pixel coordinate system through the camera internal parameters of the infrared fisheye camera.

4. The method for improving GNSS positioning accuracy according to claim 1, characterized in that: Calculating the proportion of the obstruction area within a set range through the sky segmentation map and the satellite coordinate projection map, and judging whether the satellite is a non-line-of-sight satellite through the proportion, including: A disk is generated with the projection point of the satellite on the satellite coordinate projection map as the center of the circle and the set value as the radius, and the disk is used as the set range; Determine whether all the pixels in the sky segmentation map are within the set range, and obtain the pixels within the set range as target pixels; Calculate the proportion of the occluded area within the set range by the grayscale value of the target pixel, and use the proportion as a confidence score; The confidence score is compared with a confidence threshold, and if the confidence score is greater than or equal to the confidence threshold, the satellite is a non-line-of-sight satellite.

5. The method for improving GNSS positioning accuracy as claimed in claim 4, characterized in that: Calculating the proportion of the occluded area within the set range by the grayscale value of the target pixel, and using the proportion as a confidence score, includes: Obtaining the grayscale value of each pixel point within the set range, determining whether the corresponding pixel point is an occluded pixel according to the grayscale value, assigning values ​​to the occluded pixel and the non-occluded pixel respectively, and generating an occlusion value corresponding to the pixel point; Adding the occlusion values ​​of all pixels within the set range to obtain a sum of the occlusion values; The confidence score is calculated by the sum of the occlusion values ​​and the total number of pixels within the set range.

6. The method for improving GNSS positioning accuracy as claimed in claim 1, characterized in that: Constructing a non-line-of-sight signal equivalent weight matrix, down-weighting the observation value by using the non-line-of-sight signal equivalent weight matrix, and correcting the output GNSS positioning result, including: The maximum constant and the minimum constant are set respectively, the weight of the observation value obtained by the non-line-of-sight satellite is used as the weight of the observation value before optimization, and the standardized residual is calculated by the residual in the observation value obtained by the non-line-of-sight satellite and the mean error corresponding to the residual; If the absolute value of the standardized residual is less than or equal to the minimum constant, the weight of the observation value after optimization is set to the weight of the observation value before optimization; If the absolute value of the standardized residual is less than the maximum constant and greater than the minimum constant, the optimized observation weight is calculated by the observation weight before optimization, the maximum constant, the minimum constant and the standardized residual; If the absolute value of the compared standardized residual is greater than or equal to the maximum constant, the weight of the optimized observation value is set to 0; The observation values ​​obtained by the non-line-of-sight satellites are re-weighted by using the optimized observation value weights, and the output GNSS positioning results are corrected.

7. The method for improving GNSS positioning accuracy according to any one of claims 1 to 6, characterized in that: Get the satellite position in the Earth-centered Earth-fixed coordinate system, including: The satellite position based on the Earth-centered Earth-fixed coordinate system is obtained through GNSS positioning data and satellite ephemeris.

8. A device for improving GNSS positioning accuracy, characterized in that: include: A sky segmentation map generation module is used to obtain an infrared fisheye picture facing the sky through an infrared fisheye camera, divide the infrared fisheye picture into a sky area and an occlusion area using a sky segmentation algorithm, and generate a sky segmentation map; A satellite coordinate projection map generation module is used to obtain the satellite position of the satellite in the geocentric earth-fixed coordinate system, convert the satellite position from the geocentric earth-fixed coordinate system to a pixel coordinate system, generate satellite pixel points in the pixel coordinate system, project the satellite pixel points into the infrared fisheye image, and generate a satellite coordinate projection map; A non-line-of-sight satellite determination module is used to calculate the proportion of the obstruction area within a set range through the sky segmentation map and the satellite coordinate projection map, and determine whether the satellite is a non-line-of-sight satellite through the proportion; The positioning data correction module is used to obtain the observation value of the non-line-of-sight satellite in the GNSS positioning data if the satellite is the non-line-of-sight satellite, construct a non-line-of-sight signal equivalent weight matrix, down-weight the observation value through the non-line-of-sight signal equivalent weight matrix, and correct the output GNSS positioning result.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method for improving GNSS positioning accuracy according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program for executing the method for improving GNSS positioning accuracy according to any one of claims 1 to 7.

Citation Information

Cited By

  • Signal processing method and device, electronic equipment and storage medium

    CN117671520A

  • Method, system and device for improving alignment precision of Beidou coordinate system and medium

    CN120044565A

  • A method, system, device and medium for improving alignment accuracy of a Beidou coordinate system

    CN120044565B