Coordinate calibration method based on image recognition and Beidou positioning

Abstract

The invention relates to the field of image recognition technology, and in particular to a coordinate calibration method based on image recognition and Beidou positioning, comprising: constructing a cloud database, storing a city bird's-eye view image based on the cloud database, constructing an axis grid coordinate system, placing the city bird's-eye view image in the constructed axis grid coordinate system according to a collection position of the city bird's-eye view image, and splicing the city bird's-eye view images placed based on the axis grid coordinate system to form a complete city bird's-eye view image; the invention combines image recognition technology to provide a coordinate calibration reference for a Beidou positioning result, thereby providing a Beidou positioning coordinate calibration distance and calibration direction, and realizing the final calibration of the Beidou positioning coordinates. Moreover, due to the limitation of the calibration reference, this calibration logic enables the calibration of the Beidou positioning coordinates to be always kept within a relatively small error range, and is not prone to the problem of excessive or excessive deviation between the calibrated Beidou positioning coordinates and the actual coordinates.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to a coordinate calibration method based on image recognition and Beidou positioning. Background Art

[0002] Beidou positioning is a satellite navigation system independently developed by China. It consists of multiple satellites and can provide high-precision positioning, navigation and timing services to users around the world. Beidou positioning has the following advantages: high positioning accuracy, which can achieve meter-level or even higher positioning accuracy on a global scale; wide coverage, with service areas covering the entire world; short message communication function, which can achieve two-way communication under special circumstances. Whether it is transportation, marine fisheries, surveying and mapping geographic information and other industries, or personal travel navigation, Beidou positioning plays an important role and brings great convenience to people's lives and work.

[0003] The invention patent with application number 202010893644.3 discloses a speed calibrator based on Beidou BDS high-precision positioning, which is characterized by comprising: a vehicle-mounted machine and a BDS background; wherein the vehicle-mounted machine comprises: a housing, a BDS receiving chip arranged in the housing for receiving Beidou satellite signals, a GPS receiving chip for receiving GPS satellite signals, a control system, a communication module and an LED display screen: the BDS receiving chip, the GPS receiving chip, the communication module and the LED display screen are all controlled and connected with the control system: the communication module is used to send the captured satellite positioning information and the number of positioning satellites to the BDS background, and receive feedback information from the BDS background: the control system The control system processes the information from the BDS receiving chip and the GPS receiving chip, and converts it into positioning longitude and latitude and captured BDS satellite number information. The control system receives feedback information from the communication module, and performs corresponding speed calibration according to the procedural steps based on the feedback information; the LED display screen displays the vehicle speed value calculated by the control system: the BDS background includes: a signal collection module, which is used to receive satellite positioning information and positioning satellite number information sent from the communication module of the vehicle-mounted machine; a computer, which is used to process the satellite positioning information and positioning satellite number information transmitted by the signal collection module, and then stores the satellite positioning information and positioning satellite number information in a storage module connected to it in communication.

[0004] This application aims to solve the problem that "existing vehicle speed calibration usually uses a single GPS sampling module for sampling, and a single-chip microcontroller implements vehicle speed calibration and display. Its accuracy depends entirely on whether the corresponding GPS sampling module is accurately positioned, which has a high degree of dependence and may result in speed measurement deviations."

[0005] However, when the Beidou positioning technology used in the above applications does not have a calibration function, due to the wide application of Beidou positioning technology, how to calibrate the positioning results while using Beidou positioning is a technical problem that needs to be solved urgently;

[0006] To this end, we proposed a coordinate calibration method based on image recognition and Beidou positioning. Summary of the invention

[0007] In view of the above-mentioned shortcomings of the prior art, the present invention provides a coordinate calibration method based on image recognition and Beidou positioning, which solves the technical problems raised in the above-mentioned background technology.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0009] The coordinate calibration method based on image recognition and Beidou positioning includes:

[0010] Construct a cloud database, store the city's bird's-eye view image based on the cloud database, construct an axis grid coordinate system, place the city's bird's-eye view image in the constructed axis grid coordinate system according to the acquisition position of the city's bird's-eye view image, and splice the city's bird's-eye view images placed based on the axis grid coordinate system to form a complete city's bird's-eye view image; receive Beidou positioning coordinates, synchronously acquire Beidou positioning coordinate area images when collecting Beidou positioning coordinates, scale the Beidou positioning coordinate area images based on the axis grid coordinate system to make the axis grid coordinate system and the Beidou positioning coordinate image have the same scale; obtain the Beidou positioning coordinate area images after scaling, place the images in the axis grid coordinate system where the complete city's bird's-eye view image is placed, and segment the edges in the complete city's bird's-eye view image based on the boundaries of the Beidou positioning coordinate area images. A local city overhead image with overlapping boundaries and center points of Beidou positioning coordinates; obtaining the Beidou positioning coordinate regional image and the local city overhead image in the axis grid coordinate system, identifying the building image offset in the two groups of images, and predicting the Beidou positioning coordinate correction value based on the building image offset identification result; selecting the building image representing the same building in the Beidou positioning coordinate regional image and the local city overhead image, selecting a group of corresponding points on the two groups of building images, connecting the two groups of corresponding points, and the endpoints on the line segment representing the corresponding points on the building image in the local city overhead image, the direction to the other end point of the obtained line segment and limited by the obtained line segment, that is, the Beidou positioning coordinate calibration direction; calibrating the Beidou positioning coordinates according to the predicted Beidou positioning coordinate correction value and Beidou positioning coordinate calibration direction.

[0011] Furthermore, the city bird's-eye view image is based on an acquisition operation performed by an unmanned aerial vehicle equipped with an image acquisition device over the city. The construction ratio of the axis grid coordinate system is the same as the ratio of the city bird's-eye view image and the scene entities in the city bird's-eye view image. When the city bird's-eye view image is placed in the axis grid coordinate system, the placement position of the city bird's-eye view image is consistent with the acquisition position, direction, and angle of the city bird's-eye view image, and the city bird's-eye view images placed in the axis grid coordinate system are adjacent to each other.

[0012] Furthermore, after the Beidou positioning coordinate area image is acquired and the scaling process is performed, the Beidou positioning coordinate area image is further intercepted;

[0013] Among them, the captured Beidou positioning coordinate area image is a circular image with the Beidou positioning coordinate as the center. In the capture stage of the Beidou positioning area image, the capture logic is synchronously set, and the Beidou positioning area image is captured based on the capture logic to obtain the captured Beidou positioning coordinate area image.

[0014] Furthermore, the interception logic set in the interception stage of the Beidou positioning area image, that is, the setting of the interception radius in the Beidou positioning area image interception stage, the setting of the interception radius is subject to: ;

[0015] Where: R is the interception radius; L is the length of the short side of the complete city overhead image; ω is the weight; n is the set of surrounding buildings of the Beidou positioning coordinates; d i is the distance from the i-th group of buildings to the Beidou positioning coordinates; d max ,d min The maximum and minimum values ​​of the distance from the surrounding buildings to the Beidou positioning coordinates;

[0016] in, surface The average of , the judgment of the buildings around the Beidou positioning coordinates is: on the straight line connecting the Beidou positioning coordinates and the building location coordinates, the location coordinates of a group of buildings closest to the Beidou positioning coordinates correspond to the building, d max i.e. min The corresponding buildings are all from the surrounding building set n of Beidou positioning coordinates.

[0017] Furthermore, the weight ω takes a value of (0, 1]. If the area included in the Beidou positioning coordinate area image after scaling is larger than that included in the complete city overhead image, the larger the Beidou positioning coordinate area image after scaling is, the smaller the weight value is; otherwise, the larger the weight value is. If the area included in the Beidou positioning coordinate area image after scaling is smaller than that included in the complete city overhead image, the larger the Beidou positioning coordinate area image after scaling is, the larger the weight value is; otherwise, the smaller the weight value is.

[0018] Furthermore, the operation of placing the scaled Beidou positioning coordinate area image into the axis grid coordinate system where the complete city overhead image is placed, that is, placing the intercepted Beidou positioning coordinate area image into the axis grid coordinate system where the complete city overhead image is placed;

[0019] The operation of obtaining the Beidou positioning coordinate area image in the axis grid coordinate system is the operation of obtaining the intercepted Beidou positioning coordinate area image in the axis grid coordinate system.

[0020] Furthermore, the building images in the Beidou positioning coordinate area image and the local city overhead image are obtained from the Beidou positioning coordinate area image and the local city overhead image based on an image edge detection algorithm, and each group of building images is marked with a building name;

[0021] The building image offset recognition logic is expressed as: ;

[0022] Where: H(a,b) is the offset of the building image in the Beidou positioning coordinate area image a and the local city overhead image b; m is the set based on the building name; u j is the set of building image corner points in the Beidou positioning coordinate regional image a and the building image in the local city overhead image b with building name j; d(p a ,p b ) is the distance between the corner point p in the building image in the Beidou positioning coordinate area image a and the corner point p in the building image in the local city overhead image b;

[0023] in, Table pair Seeking average.

[0024] Furthermore, the prediction logic of the Beidou positioning coordinate correction value is expressed as:

[0025] ;

[0026] Where: c is the correction value of Beidou positioning coordinates; q is the total number of buildings around Beidou positioning coordinates; A is the correction; λ is a constant; H(a,b) is the offset of the building image in the Beidou positioning coordinate area image a and the local city overhead image b;

[0027] Among them, the correction .

[0028] Furthermore, the value of the constant λ is customized by the user, and the value range of the constant λ is always between 1% and 5% of q / A.

[0029] Furthermore, after calibrating the Beidou positioning coordinates according to the predicted Beidou positioning coordinate correction value and the Beidou positioning coordinate calibration direction, a new set of Beidou positioning coordinates is obtained. The length of the line connecting the new Beidou positioning coordinates and the Beidou positioning coordinates before calibration is equal to the Beidou positioning coordinate correction value, and the connecting line is consistent with the Beidou positioning coordinate calibration direction.

[0030] Compared with the known public technology, the technical solution provided by the present invention has the following beneficial effects:

[0031] The present invention provides a coordinate calibration method based on image recognition and Beidou positioning. During the execution of the method, the image recognition technology is combined to provide a coordinate calibration reference for the Beidou positioning result, thereby providing a Beidou positioning coordinate calibration distance and a calibration direction, and realizing the final calibration of the Beidou positioning coordinates. Moreover, due to the limitation of the calibration reference, this calibration logic enables the calibration of the Beidou positioning coordinates to always be kept within a relatively small error range, and it is not easy to have the problem that the calibrated Beidou positioning coordinates deviate too much or too little from the actual coordinates. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0033] Figure 1 It is a flowchart of the coordinate calibration method based on image recognition and Beidou positioning. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] The present invention will be further described below in conjunction with the embodiments.

[0036] Embodiment 1:

[0037] The coordinate calibration method based on image recognition and Beidou positioning in this embodiment is as follows: Figure 1 As shown, including:

[0038] Step 1: Build a cloud database, store the city's overhead image based on the cloud database, build an axis grid coordinate system, place the city's overhead image in the constructed axis grid coordinate system according to the collection position of the city's overhead image, and splice the city's overhead images placed based on the axis grid coordinate system to form a complete city's overhead image;

[0039] Step 2: Receive Beidou positioning coordinates, synchronously collect Beidou positioning coordinate area images when Beidou positioning coordinates are collected, and scale the Beidou positioning coordinate area images based on the axis grid coordinate system to make the axis grid coordinate system consistent with the scale of the Beidou positioning coordinate image;

[0040] Step 3: Obtain the scaled Beidou positioning coordinate area image, place the image in the axis grid coordinate system where the complete city overhead view image is placed, and based on the Beidou positioning coordinate area image boundary, segment the local city overhead view image whose boundaries overlap and whose center points are all Beidou positioning coordinates from the complete city overhead view image;

[0041] Step 4: Obtain the Beidou positioning coordinate regional image and the local city overhead image in the axis grid coordinate system, identify the building image offset in the two sets of images, and predict the Beidou positioning coordinate correction value based on the building image offset identification result;

[0042] The building images in the Beidou positioning coordinate regional images and the local city overhead images are obtained from the Beidou positioning coordinate regional images and the local city overhead images based on the image edge detection algorithm, and each group of building images is marked with the building name;

[0043] The logic of building image offset recognition is expressed as: ;

[0044] Where: H(a,b) is the offset of the building image in the Beidou positioning coordinate area image a and the local city overhead image b; m is the set based on the building name; u j is the set of building image corner points in the Beidou positioning coordinate regional image a and the building image in the local city overhead image b with building name j; d(p a ,p b ) is the distance between the corner point p in the building image in the Beidou positioning coordinate area image a and the corner point p in the building image in the local city overhead image b;

[0045] in, Table pair ;

[0046] Step 5: Select the building image representing the same building in the Beidou positioning coordinate regional image and the local city overhead image, select a set of corresponding points on the two sets of building images, connect the two sets of corresponding points, and the end point on the line segment obtained by connecting the lines represents the corresponding point on the building image in the local city overhead image, and the direction defined by the other end point of the obtained line segment is the Beidou positioning coordinate calibration direction;

[0047] Step 6: Calibrate the Beidou positioning coordinates according to the predicted Beidou positioning coordinate correction value and Beidou positioning coordinate calibration direction.

[0048] In this embodiment, by executing the steps of the method in the above embodiment, a calibration method based on image recognition technology is provided for the Beidou positioning coordinates, thereby improving the accuracy of the Beidou positioning coordinates and enabling the Beidou positioning technology to be applied more reliably.

[0049] Embodiment 2:

[0050] In terms of specific implementation, based on Example 1, this example refers to Figure 1 The coordinate calibration method based on image recognition and Beidou positioning in Example 1 is further described in detail:

[0051] The city bird's-eye view image is based on the acquisition operation performed by the drone equipped with image acquisition equipment over the city. The construction ratio of the axis grid coordinate system is the same as the ratio of the city bird's-eye view image and the scene entities in the city bird's-eye view image. When the city bird's-eye view image is placed in the axis grid coordinate system, the placement position of the city bird's-eye view image is consistent with the acquisition position, direction and angle of the city bird's-eye view image. The city bird's-eye view images placed in the axis grid coordinate system are adjacent to each other.

[0052] Through the above settings, the collection source and logic of urban bird's-eye view image data are further limited.

[0053] like Figure 1 As shown, after the Beidou positioning coordinate area image is acquired and zoomed, the Beidou positioning coordinate area image is further intercepted;

[0054] The intercepted Beidou positioning coordinate area image is a circular image with the Beidou positioning coordinate as the center. The Beidou positioning area image is intercepted, and the interception logic is synchronously set. The Beidou positioning area image is intercepted based on the interception logic to obtain the intercepted Beidou positioning coordinate area image;

[0055] The interception logic set in the interception stage of the Beidou positioning area image, that is, the setting of the interception radius in the Beidou positioning area image interception stage, the setting of the interception radius obeys: ;

[0056] Where: R is the interception radius; L is the length of the short side of the complete city overhead image; ω is the weight; n is the set of surrounding buildings of the Beidou positioning coordinates; d i is the distance from the i-th group of buildings to the Beidou positioning coordinates; d max ,d min The maximum and minimum values ​​of the distance from the surrounding buildings to the Beidou positioning coordinates;

[0057] in, surface The average of , the judgment of the buildings around the Beidou positioning coordinates is: on the straight line connecting the Beidou positioning coordinates and the building location coordinates, the location coordinates of a group of buildings closest to the Beidou positioning coordinates correspond to the building, d max i.e. min The corresponding buildings are all from the surrounding building set n of Beidou positioning coordinates;

[0058] The weight ω takes the value of (0, 1]. If the area of ​​the Beidou positioning coordinate area image after scaling is larger than that of the complete city overhead image, the larger the Beidou positioning coordinate area image after scaling is, the smaller the weight value is. Otherwise, the larger the weight value is. If the area of ​​the Beidou positioning coordinate area image after scaling is smaller than that of the complete city overhead image, the larger the Beidou positioning coordinate area image after scaling is, the larger the weight value is. Otherwise, the smaller the weight value is.

[0059] Through the above settings, the interception logic of the Beidou positioning area image is further limited. For the calibration of the Beidou positioning coordinates in the method in Example 1, the reference data is limited to a specified range, thereby improving the efficiency and accuracy of the Beidou positioning coordinates in the calibration based on the method in Example 1.

[0060] like Figure 1 As shown, the operation of placing the scaled Beidou positioning coordinate area image into the axis grid coordinate system where the complete city overhead image is placed, that is, the operation of placing the intercepted Beidou positioning coordinate area image into the axis grid coordinate system where the complete city overhead image is placed;

[0061] The operation of obtaining the Beidou positioning coordinate area image in the axis grid coordinate system is the operation of obtaining the intercepted Beidou positioning coordinate area image in the axis grid coordinate system.

[0062] Embodiment 3:

[0063] In terms of specific implementation, based on Example 1, this example refers to Figure 1 The coordinate calibration method based on image recognition and Beidou positioning in Example 1 is further described in detail:

[0064] The prediction logic of Beidou positioning coordinate correction value is expressed as:

[0065] ;

[0066] Where: c is the correction value of Beidou positioning coordinates; q is the total number of buildings around Beidou positioning coordinates; A is the correction; λ is a constant;

[0067] Among them, the correction ;

[0068] The value of constant λ is defined by the user, and the value range of constant λ is always between 1% and 5% of q / A;

[0069] After calibrating the Beidou positioning coordinates according to the predicted Beidou positioning coordinate correction value and the Beidou positioning coordinate calibration direction, a new set of Beidou positioning coordinates is obtained. The length of the line connecting the new Beidou positioning coordinates and the Beidou positioning coordinates before calibration is equal to the Beidou positioning coordinate correction value, and the connecting line is consistent with the Beidou positioning coordinate calibration direction.

[0070] In this embodiment, the prediction logic of the Beidou positioning coordinate correction value is limited by the above-mentioned logical formula, and the correction is referenced in the logical formula. For the value of correction A, the smaller the value, the more it can reflect to a certain extent whether the shape of the buildings around the Beidou positioning coordinates is more circular. The more circular it is, the more valuable the correction is for reference, which in turn affects the Beidou positioning coordinate correction value c and achieves the purpose of outputting the final Beidou positioning coordinate correction value c.

[0071] In summary, during the execution of the method in the above embodiment, combined with image recognition technology, a coordinate calibration reference is provided for the Beidou positioning result, thereby providing the Beidou positioning coordinate calibration distance and calibration direction, and realizing the final calibration of the Beidou positioning coordinates. Moreover, due to the limitation of the calibration reference, this calibration logic enables the calibration of the Beidou positioning coordinates to always be kept within a small error range, and it is not easy for the calibrated Beidou positioning coordinates to deviate too much or too little from the actual coordinates.

[0072] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A coordinate calibration method based on image recognition and Beidou positioning, characterized in that: include: Step 1: Build a cloud database, store the city's overhead image based on the cloud database, build an axis grid coordinate system, place the city's overhead image in the constructed axis grid coordinate system according to the collection position of the city's overhead image, and splice the city's overhead images placed based on the axis grid coordinate system to form a complete city's overhead image; Step 2: Receive Beidou positioning coordinates, synchronously collect Beidou positioning coordinate area images when Beidou positioning coordinates are collected, and scale the Beidou positioning coordinate area images based on the axis grid coordinate system to make the axis grid coordinate system consistent with the scale of the Beidou positioning coordinate image; Step 3: Obtain the scaled Beidou positioning coordinate area image, place the image in the axis grid coordinate system where the complete city overhead view image is placed, and based on the Beidou positioning coordinate area image boundary, segment the local city overhead view image whose boundaries overlap and whose center points are all Beidou positioning coordinates from the complete city overhead view image; Step 4: Obtain the Beidou positioning coordinate regional image and the local city overhead image in the axis grid coordinate system, identify the building image offset in the two sets of images, and predict the Beidou positioning coordinate correction value based on the building image offset identification result; Step 5: Select the building image representing the same building in the Beidou positioning coordinate regional image and the local city overhead image, select a set of corresponding points on the two sets of building images, connect the two sets of corresponding points, and the end point on the line segment obtained by connecting the lines represents the corresponding point on the building image in the local city overhead image, and the direction defined by the other end point of the obtained line segment is the Beidou positioning coordinate calibration direction; Step 6: Calibrate the Beidou positioning coordinates according to the predicted Beidou positioning coordinate correction value and Beidou positioning coordinate calibration direction.

2. The coordinate calibration method based on image recognition and Beidou positioning according to claim 1, characterized in that: The city bird's-eye view image is based on an unmanned aerial vehicle equipped with an image acquisition device performing an acquisition operation over the city. The construction ratio of the axis grid coordinate system is the same as the ratio of the city bird's-eye view image and the scene entities in the city bird's-eye view image. When the city bird's-eye view image is placed in the axis grid coordinate system, the placement position of the city bird's-eye view image is consistent with the acquisition position, direction, and angle of the city bird's-eye view image, and the city bird's-eye view images placed in the axis grid coordinate system are adjacent to each other.

3. The coordinate calibration method based on image recognition and Beidou positioning according to claim 1, characterized in that: After the Beidou positioning coordinate area image is acquired and zoomed, the Beidou positioning coordinate area image is further intercepted; Among them, the captured Beidou positioning coordinate area image is a circular image with the Beidou positioning coordinate as the center. In the capture stage of the Beidou positioning area image, the capture logic is synchronously set, and the Beidou positioning area image is captured based on the capture logic to obtain the captured Beidou positioning coordinate area image.

4. The coordinate calibration method based on image recognition and Beidou positioning according to claim 3 is characterized in that: The interception logic set in the interception stage of the Beidou positioning area image, that is, the setting of the interception radius in the Beidou positioning area image interception stage, the setting of the interception radius is subject to: ； Where: R is the interception radius; L is the length of the short side of the complete city overhead image; ω is the weight; n is the set of surrounding buildings of the Beidou positioning coordinates; d i is the distance from the i-th group of buildings to the Beidou positioning coordinates; d max d min The maximum and minimum values ​​of the distance from the surrounding buildings to the Beidou positioning coordinates; in, surface The average of , the judgment of the buildings around the Beidou positioning coordinates is: on the straight line connecting the Beidou positioning coordinates and the building location coordinates, the location coordinates of a group of buildings closest to the Beidou positioning coordinates correspond to the building, d max i.e. min The corresponding buildings are all from the surrounding building set n of Beidou positioning coordinates.

5. The coordinate calibration method based on image recognition and Beidou positioning according to claim 4 is characterized in that: The weight ω takes a value of (0, 1]. If the area included in the Beidou positioning coordinate area image after scaling is larger than that included in the complete city overhead image, the larger the Beidou positioning coordinate area image after scaling is, the smaller the weight value is. Otherwise, the larger the weight value is. If the area included in the Beidou positioning coordinate area image after scaling is smaller than that included in the complete city overhead image, the larger the Beidou positioning coordinate area image after scaling is, the larger the weight value is. Otherwise, the smaller the weight value is.

6. The coordinate calibration method based on image recognition and Beidou positioning according to claim 1, characterized in that: The operation of placing the scaled Beidou positioning coordinate area image into the axis grid coordinate system where the complete city overhead image is placed, that is, placing the intercepted Beidou positioning coordinate area image into the axis grid coordinate system where the complete city overhead image is placed; The operation of obtaining the Beidou positioning coordinate area image in the axis grid coordinate system is the operation of obtaining the intercepted Beidou positioning coordinate area image in the axis grid coordinate system.

7. The coordinate calibration method based on image recognition and Beidou positioning according to claim 1, characterized in that: The building images in the Beidou positioning coordinate area image and the local city overhead image are obtained from the Beidou positioning coordinate area image and the local city overhead image based on an image edge detection algorithm, and each group of building images is marked with a building name; The building image offset recognition logic is expressed as: ; Where: H(a,b) is the offset of the building image in the Beidou positioning coordinate area image a and the local city overhead image b; m is the set based on the building name; u j is the set of building image corner points in the Beidou positioning coordinate regional image a and the building image in the local city overhead image b with building name j; d(p a ,p b ) is the distance between the corner point p in the building image in the Beidou positioning coordinate area image a and the corner point p in the building image in the local city overhead image b; in, Table pair Seeking the average.

8. The coordinate calibration method based on image recognition and Beidou positioning according to claim 1, characterized in that: The prediction logic of the Beidou positioning coordinate correction value is expressed as: ； Where: c is the correction value of Beidou positioning coordinates; q is the total number of buildings around Beidou positioning coordinates; A is the correction; λ is a constant; H(a,b) is the offset of the building image in the Beidou positioning coordinate area image a and the local city overhead image b; Among them, the correction , where: n is the set of surrounding buildings of Beidou positioning coordinates; d i is the distance from the i-th group of buildings to the Beidou positioning coordinates; d max d min It is the maximum and minimum value of the distance from the surrounding buildings of the Beidou positioning coordinates to the Beidou positioning coordinates.

9. The coordinate calibration method based on image recognition and Beidou positioning according to claim 8, characterized in that: The value of the constant λ is customized by the user, and the value range of the constant λ is always between 1% and 5% of q / A.

10. The coordinate calibration method based on image recognition and Beidou positioning according to claim 1, characterized in that: After calibrating the Beidou positioning coordinates according to the predicted Beidou positioning coordinate correction value and the Beidou positioning coordinate calibration direction, a new set of Beidou positioning coordinates is obtained. The length of the line connecting the new Beidou positioning coordinates and the Beidou positioning coordinates before calibration is equal to the Beidou positioning coordinate correction value, and the connecting line is consistent with the Beidou positioning coordinate calibration direction.

Images