Method for eliminating integer ambiguity of RTK positioning by using image positioning and altitude

Abstract

The present invention provides a method for eliminating the integer ambiguity of RTK positioning by using image positioning and altitude. By installing an image positioning device on the vehicle body, continuous shooting is carried out when the vehicle travels near the feature points for image positioning. The RTK measurement value is corrected by using the image positioning result. During the movement process, incorrect RTK fixed solutions can be detected in a timely manner and corrected with high precision. Compared with the traditional method for correcting the integer ambiguity error, this method does not require the prior setting of reference points and control points, does not require the use of additional equipment such as total stations to measure the control points, and does not require multiple measurements. The method for eliminating the integer ambiguity of RTK positioning is simpler, has strong anti-interference ability and higher measurement accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of satellite positioning, and in particular to a method for eliminating the integer ambiguity of RTK positioning by using image positioning and altitude. Background Art

[0002] Satellite positioning systems, such as GPS, Beidou, etc., can achieve an accuracy of approximately 10 meters using CA messages for positioning. Using RTK carrier phase, theoretically, an accuracy of millimeter level can be achieved. However, RTK positioning has high environmental requirements. For example, when there are high-rise buildings, high walls, or trees near the site, and a vehicle passes through these positions, these obstacles will block or even reflect the satellite signals received by the satellite positioning receiver, thus affecting positioning. The signals received by the receiver may converge to an incorrect fixed solution after calculation, resulting in unexpected errors. At the same time, this error is relatively difficult to detect. In actual surveying and mapping projects and positioning processes, dual-frequency satellite signals are used to mutually verify to eliminate the influence of the ionosphere, and the method of using a measurement reference station is used to detect errors in a timely manner;

[0003] Detecting errors in a timely manner is a prerequisite for correction. Currently, there are relatively few methods for detecting errors. For example, by setting reference points and control points in advance and using a method that does not require additional equipment such as a total station to measure the control points to detect errors;

[0004] Another method for detecting errors is multiple measurements. Since satellites are moving at high speeds, different fixed solutions are obtained using different satellite distributions, and the correct solution is selected from them;

[0005] The above methods are more applicable to fixed-point measurements, such as building deformation monitoring. For mobile positioning, such as vehicle positioning during movement, they are basically inapplicable because each positioning at this time is single-point positioning, and it is difficult to obtain the condition of multiple measurements of a point.

[0006] Engineers have developed methods for integrated positioning, such as integrating inertial navigation and integrating vision, to assist in detecting incorrect RTK fixed solutions. However, the accuracy of these methods does not match the accuracy of the RTK fixed solution. Larger offsets, such as meter-level offsets, can be detected, but it is difficult to detect small offsets, such as centimeter-level offsets, and it is also difficult to perform subsequent corrections. Summary of the Invention

[0007] The object of the present invention is to overcome the deficiencies of the prior art and provide a method for accurately eliminating the integer ambiguity of RTK positioning by using image positioning, altitude, and the mathematical model established therefrom.

[0008] To achieve the above object, the present invention provides a method for eliminating the integer ambiguity of RTK positioning, including a method for eliminating the integer ambiguity of RTK positioning by using image positioning, and the steps are as follows:

[0009] S1. First, survey the site containing the garage to obtain longitude, latitude, and altitude coordinates. Fix and install an image positioning device on the vehicle. The image positioning device includes multiple cameras C1 and a data processing device, and calibrate the cameras C1. Install a satellite positioning receiver and pre-position the satellite positioning receiver; drive the vehicle into the garage of the site, and use the calibrated cameras C1 to capture pictures in real time. After passing through the data processing device for distortion removal and inverse perspective, obtain the top view of the vehicle. The pre-positioning includes recording the longitude and latitude of the lower left corner O (olat, olon, oheight), the lower right corner B (blat, blon, bheight), and the upper left corner A (alat, alon, aheight) of the garage of the site. Establish a rectangular coordinate system of OAB with OA as the y-axis and OB as the x-axis; measure the actual line length of one of the storage lines of the garage of the site, and the data processing device calculates the number of pixel points of this storage line in the top view of the vehicle, and then obtains the ratio r of the number of pixel points in the top view of the vehicle to the actual line length;

[0010] S2. The satellite positioning receiver performs RTK real-time solution once per second, and the solution obtains the longitude and latitude P1 (p1lat, p1lon, plheight) of the satellite positioning receiver. By substituting P1 (p1lat, p1lon) into the rectangular coordinate system of OAB, the position P1 (x1, y1) of the satellite positioning receiver in this coordinate system is solved:

[0011] Let the equatorial radius of the earth be Ea, the polar radius be Eb, the polar circumference be Db, the longitude and latitude of the origin O of the coordinate system be (olat, olon), and the included angle between the y-axis and the due north be angle,

[0012] Equation for converting longitude and latitude (lat, lon) to (x, y) coordinates:

[0013] ,

[0014] ,

[0015] Displacement in the due east direction: ,

[0016] Displacement in the due north direction: ,

[0017] x coordinate: ,

[0018] y coordinate: ;

[0019] S3. Use the top view of the vehicle to perform image color filtering analysis to obtain the set of all pixel points of the site garage. Filter out the colors other than the parking space lines, and use the least squares method to fit straight lines to the point sets of the parts of the left parking space line and the lower parking space line of the site garage that are closest to the image center, respectively, to identify the parking space lines of the left parking space line and the lower parking space line.

[0020] S4. The pixel distances from the actual position P2 point of the satellite positioning receiver to the left parking space line and to the lower parking space line, combined with the ratio r of the number of pixel points to the actual line length in the top view of the vehicle, can calculate the actual distance from the P2 point to the outside of the left parking space line and the actual distance from the P2 point to the outside of the lower parking space line. Finally, the coordinates P2(x2, y2) of P2 are obtained. The longitude and latitude of the coordinate system origin and the angle between the y-axis of the coordinate system and the true north are known. Let, using the coordinate conversion formula from longitude and latitude (lat, lon) to (x, y), we can get:

[0021] The variation relationship between the longitude and the distance in the due east direction ;

[0022] The variation relationship between the latitude and the distance in the due north direction ;

[0023] The offset distance of the coordinate (x, y) in the due north direction is ;

[0024] The offset distance in the due east direction is

[0025] So

[0026]

[0027] The longitude and latitude P2(p2lat, p2lon) corresponding to the coordinates P2(x2, y2) can be obtained;

[0028] S5. Use P2(p2lat, p2lon) to correct P1(p1lat, p1lon), calculate the offset amount, and then correct the coordinates collected each time with a fixed offset amount. The offset amount is (p2lat - p1lat, p2lon - p1lon). After that, the longitude and latitude collected each time are fixed plus the offset amount for correction to obtain Rn(Xn, Yn).

[0029] Further, the site garage can be any standard garage, and the length and width of the standard garage are known.

[0030] Further, when the image positioning accuracy is insufficient and the library location lines cannot be recognized or only part of the library location lines are recognized during the initial image filtering, the calculation of the value of P2 (p2lat, p2lon) is abandoned. Under the interference of uncertain factors such as dim light, wet and reflective ground, and object occlusion of the library location lines, the library location lines cannot be recognized or only part of the library location lines are recognized during the initial image filtering, thus affecting the fitting result.

[0031] Further, when the offset in step S5 is greater than the critical value, under the loose coupling condition, the satellite positioning receiver can be selected to be restarted for satellite searching, positioning, and convergence again; under the tight coupling condition, the current star map can be selected to be trusted and converged again. Loose coupling means that the rtk algorithm is loosely coupled with other algorithms and the results are coupled; tight coupling means that the rtk algorithm is tightly coupled with other algorithms, and the parameters + algorithm process are coupled.

[0032] Further, the step of obtaining all pixel point sets of the site garage by image color filtering in step S3 includes calculating the rgb value of each pixel in the vehicle top view, filtering the rbg values other than the library location lines in the site garage, and filtering the pixel points with rgb values close to (255, 255, 0) if the library location lines in the site garage are yellow.

[0033] Further, the data processing device includes a memory and a data processor for executing the method of eliminating the RTK positioning integer ambiguity, and the data processing device can be a mobile phone, a computer, or a vehicle-mounted terminal.

[0034] The technical solution of the present invention also proposes a method for eliminating the RTK positioning integer ambiguity by using altitude, including the following steps:

[0035] S1. Model the site: Intercept a circle with an area of S on the earth's surface, and establish an XYZ three-dimensional coordinate system with longitude, latitude, and altitude;

[0036] S2. Fix and install a satellite positioning receiver on the top of the vehicle and perform RTK calculation. The calculation results are R1 (X1, Y1, Z1)... Rn (Xn, Yn, Zn);

[0037] S3. The vehicle randomly travels on the road, continuously collects the coordinates Rn, eliminates the Rn with non-fixed solutions by using the statistically collected multiple groups of coordinates Rn, and takes the 95% Zn value interval of the statistics as the normal range of Zn in this site;

[0038] S4. Logically judge the altitude Zn of the subsequently collected coordinates Rn (Xn, Yn, Zn) to determine whether it is within the normal range. If it is within the normal range, the collected coordinate information is adopted; if it exceeds the normal range, it is determined that the collected coordinate information is untrustworthy and this coordinate information is not used in subsequent calculations;

[0039] S5. If it is found that Zn in Rn(Xn, Yn, Zn) in the fixed solution is not within the normal range, under loose coupling conditions, the satellite positioning receiver can be selected to be restarted for re-positioning and convergence; under tight coupling conditions, re-convergence can be selected.

[0040] Further, (Xn, Yn) in Rn in step S2 can be the longitude and latitude corrected by the method of eliminating the RTK positioning cycle ambiguity using image positioning.

[0041] The technical solution of the present invention also proposes a method for simultaneously using image positioning and altitude to eliminate the RTK positioning cycle ambiguity; when (Zn) in Rn is incorrect and (Xn, Yn) is accurate, or when (Xn, Yn) in Rn is incorrect and (Zn) is accurate, re-convergence calculation should also be performed.

[0042] The beneficial effects of the present invention are as follows: using the plane coordinate longitude and latitude obtained by image positioning, or the altitude measured in advance and the mathematical model established therefrom to provide a method for accurately eliminating the RTK positioning cycle ambiguity, mapping the characteristic points of the site in advance to obtain longitude, latitude, and altitude, installing an image positioning device on the vehicle body, continuously shooting when the vehicle travels near the characteristic points for image positioning, using the image positioning result to correct the RTK measurement value, promptly detecting incorrect RTK fixed solutions during the movement process and performing high-precision correction;

[0043] Continuously collect and correct the coordinates Rn of the longitude and latitude, use the multiple groups of coordinates Rn obtained by statistics, eliminate the Rn of non-fixed solutions, and use the 95% Zn value interval obtained by statistics as the normal range of Zn in this site; and filter the altitude Zn of the coordinates Rn(Xn, Yn, Zn) collected subsequently. Compared with the traditional method for correcting the cycle ambiguity error, this method does not require setting benchmark points and control points in advance, does not require using additional equipment such as total stations to measure control points, and does not require multiple measurements. The method for eliminating the RTK positioning cycle ambiguity is simpler, has strong anti-interference ability, and higher measurement accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] For ease of explanation, the present invention is described in detail by the following preferred embodiments and accompanying drawings.

[0045] Figure 1 is a flowchart of the method for eliminating the RTK positioning cycle ambiguity using image positioning according to the present invention;

[0046] Figure 2 is a schematic diagram of establishing a coordinate system for the top view of the vehicle of the method for eliminating the RTK positioning cycle ambiguity according to the present invention;

[0047] Figure 3 It is a schematic diagram of the position of a satellite positioning receiver in a coordinate system for a method of eliminating RTK positioning cycle slips using image positioning according to the present invention;

[0048] Figure 4 It is a schematic diagram of the actual position of a satellite positioning receiver after fitting a straight line to the pixel points of a bin line after image color filtering for a method of eliminating RTK positioning cycle slips using image positioning according to the present invention;

[0049] Figure 5 It is a schematic diagram of an image positioning device for a method of eliminating RTK positioning cycle slips using image positioning according to the present invention;

[0050] Figure 6 It is a flowchart of a method of eliminating RTK positioning cycle slips using altitude according to the present invention;

[0051] Figure 7 It is a schematic diagram of the principle of fixed solution error for a method of eliminating RTK positioning cycle slips using altitude according to the present invention;

[0052] Figure 8 It is a schematic diagram of judging a correct fixed solution for a method of eliminating RTK positioning cycle slips using altitude according to the present invention. Detailed implementation mode

[0053] The following further describes a method of eliminating RTK positioning cycle slips according to the present invention with reference to the accompanying drawings:

[0054] RTK positioning principle: When a satellite is at a certain point on the positioning plane, it will first send a string of pulse signals. For example, the satellite carrier phase for high-precision positioning is within the range of L1 carrier (λ1 = 19.03 cm) to L2 carrier (λ2 = 24.42 cm) ± 2 - 3 mm. The problem of discontinuous modulation wave and phase when receiving the satellite signal needs to perform a binary phase modulation on the satellite signal through demodulation methods such as a modem to remove the ranging code and navigation message and reconstruct the carrier signal. Reconstruct the carrier by the squaring method: The self-product of the satellite signal

[0055] .

[0056] Usually, multiple satellites are used to locate a point. Since the carrier phases between satellites are not the same, the points located by multiple satellites are discretely distributed on the mapped plane. Moreover, when the satellite signal reaches near the plane, there is an error of λsin(θ) between the signal and the plane. When the satellite is exactly on the positive z-axis of the point, θ approaches 0, and at this time, the error between the satellite pulse signal and the plane is at most ±λ of one wavelength. Usually, we take the satellites within the range of ±45 degrees with the zenith of the point to be measured as the 0-degree reference. Therefore, when the satellite pulse signal reaches the plane, there is at least error. And due to the different pulse frequencies between satellites, the discrete points are all within the range of the measured altitude H0±λ, and the satellite pulse signals within this range can be approximately regarded as reliable. Assume that n satellites of a certain satellite navigation system are used to observe the same observation station (the point to be located), and let

[0057]

[0058]

[0059]

[0060]

[0061]

[0062]

[0063]

[0064] Then the single-frequency single-difference positioning mathematical model can be expressed as:

[0065]

[0066]

[0067] Suppose s satellite navigation systems are used for joint positioning of the same observation station, and corresponding reference satellites are selected for double-differencing respectively. Then the corresponding multi-system multi-frequency double-difference non-combination positioning mathematical model can be expressed as:

[0068]

[0069]

[0070] In summary, the unknown parameter vector form of the multi-system multi-frequency non-combination differential positioning model is as follows:

[0071]

[0072] Let the parameter list be

[0073] A new parameter list is obtained by linear combination:

[0074]

[0075] The result matrix after satellite signal fusion is:

[0076] (where )

[0077] A fusion method for integer ambiguity resolution based on geometry for the positioning information of multiple satellites at the same point: When the navigation solution is overdetermined (i.e., the number of double-difference ADR observations between different satellites is more than the number of positions to be estimated and bias states), the signal geometry configuration can be used to assist in ambiguity resolution, or other external positioning methods can be used to assist in ambiguity resolution.

[0078] Example: As Figure 1-2 shown

[0079] A method for eliminating the integer ambiguity of RTK positioning using image positioning, including the following steps:

[0080] S1. Conduct a survey of the site containing the garage in advance to obtain longitude, latitude, and altitude coordinates. Fix and install an image positioning device on the vehicle. The image positioning device includes multiple cameras C1 and a data processing device, and calibrate the cameras C1. Install a satellite positioning receiver and perform pre-positioning on the satellite positioning receiver; ensure that the satellite positioning receiver can correctly receive satellite signals and the power supply is stable. Through the data processing device, sniff the longitude and latitude of the satellite positioning receiver, and confirm whether the signal value of the received GPS positioning quality status is 4. Fixed solution. The positioning quality GPS status is divided into 0, Initialization, 1. Single-point positioning, 2. Code differential, 3. Invalid PPS, 4. Fixed solution, 5. Float solution, 6. Estimating, 7. Manually input fixed value, 8. Simulation mode, 9. WAAS differential; if not, inform the user that the longitude and latitude are unavailable; Drive the vehicle into the garage of the site, and use the calibrated cameras C1 to take real-time pictures. After passing through the data processing device for distortion removal and inverse perspective, obtain the top view of the vehicle. The pre-positioning includes recording the longitude and latitude of the lower left corner O (olat, olon, oheight), the lower right corner B (blat, blon, bheight), and the upper left corner A (alat, alon, aheight) of the garage in the site. Since the site is flat, the altitude can be ignored during the calculation process. Take OA as the y-axis and OB as the x-axis to establish a rectangular coordinate system of OAB; Manually measure the actual line length of one of the parking lines in the garage of the site. If it is a standard garage, the length is known and no manual measurement is required. According to the camera parameters, the number of pixels in the taken photo is known. The data processing device calculates the number of pixel points of this parking line in the top view of the vehicle, and then obtains the ratio r of the number of pixel points in the top view of the vehicle to the actual line length; In this embodiment, the cameras C1 adopt 4 fisheye cameras, which are respectively arranged around the vehicle body. Before use, the cameras C1 need to be calibrated to normally obtain the top view of the vehicle through distortion removal and inverse perspective; The data processing device is arranged inside the vehicle, and the satellite positioning receiver is installed inside the vehicle; The garage of the site is a standard garage for reverse parking in a driving training ground, and the length and width are known. For example, we measure that the length of a parking line parallel to the vehicle body is 543 cm, and from the top view of the vehicle, we can see that the length of this parking line occupies 290 pixel points. So r = 290 / 543 pixels per centimeter, and the correction accuracy is relatively high, with a typical value of 1 cm.

[0081] S2. As Figure 3 shown,

[0082] The satellite positioning receiver performs RTK real-time solution once per second, can detect errors relatively quickly, with a typical value of 33 ms, and obtains the longitude and latitude P1 (p1lat, p1lon, plheight) of the satellite positioning receiver. By substituting P1(p1lat, p1lon) into the rectangular coordinate system of OAB, the position P1(x1, y1) of the satellite positioning receiver in this coordinate system is obtained:

[0083] Let the equatorial radius of the earth be Ea, the polar radius be Eb, the polar circumference be Db, the longitude and latitude of the coordinate system origin O be (olat, olon), and the angle between the y-axis and the due north be angle.

[0084] Equation for converting longitude and latitude (lat, lon) to (x, y) coordinates:

[0085] ,

[0086] ,

[0087] Displacement in the due east direction: ,

[0088] Displacement in the due north direction: ,

[0089] x coordinate: ,

[0090] y coordinate: ;

[0091] S3. As Figure 4 shown, all pixel point sets of the site garage are obtained by performing image color filtering analysis using the vehicle top view, filtering out colors other than the parking space lines, and respectively performing least squares fitting on the point sets closest to the image center of the left and lower parking space lines of the site garage to identify the parking space lines of the left and lower parking space lines;

[0092] The image color filtering calculates the rgb value of each pixel in the vehicle top view, filtering out rbg values other than the parking space lines in the site garage. If the parking space lines in the site garage are yellow, pixel points with rgb values close to (255, 255, 0) are filtered out. If the parking space lines in the site garage are white, pixel points with rgb values close to (255, 255, 225) are filtered out.

[0093] S4. The pixel distances from the actual position P2 of the satellite positioning receiver to the left storage position line and to the lower storage position line, combined with the ratio r of the number of pixel points to the actual line length in the vehicle top view, can be used to calculate the actual distance from point P2 to the outside of the left storage position line and the actual distance from point P2 to the outside of the lower storage position line. Finally, the coordinates P2(x2, y2) of P2 are obtained. The longitude and latitude of the coordinate system origin and the angle between the y-axis of the coordinate system and the true north are known. Let and using the formula for converting longitude and latitude (lat, lon) to (x, y) coordinates, we can get:

[0094] The variation relationship between longitude and distance in the due east direction ;

[0095] The variation relationship between latitude and distance in the due north direction ;

[0096] The offset distance of the coordinate (x, y) in the due north direction is ;

[0097] The offset distance in the due east direction is

[0098] Therefore

[0099]

[0100] the longitude and latitude P2(p2lat, p2lon) corresponding to the coordinates P2(x2, y2) can be obtained;

[0101] S5. Use P2(p2lat, p2lon) to correct P1(p1lat, p1lon), calculate the offset amount, and then correct each collected coordinate with a fixed offset amount. The offset amount is (p2lat - p1lat, p2lon - p1lon). After that, each collected longitude and latitude is fixed by adding the offset amount for correction to obtain Rn(Xn, Yn).

[0102] When the image positioning accuracy is insufficient and the image is initially filtered and the storage position line cannot be recognized or only part of the storage position line is recognized, the value of P2(p2lat, p2lon) is not calculated. Under the interference of uncertain factors such as dim light, wet and reflective ground, and object occlusion of the storage position line, the storage position line cannot be recognized or only part of the storage position line is recognized during the initial image filtering, which affects the fitting result. Therefore, it is necessary to give up calculating the value of P2(p2lat, p2lon).

[0103] When the offset in step S5 is greater than the critical value of 150 cm, under the loose coupling condition, the satellite positioning receiver can be selected to be restarted, and satellite searching, positioning, and convergence can be performed again; under the tight coupling condition, the current star map can be selected to be trusted and convergence can be performed again. Loose coupling means that the RTK algorithm is loosely coupled with other algorithms and the results are coupled; tight coupling means that the RTK algorithm and other algorithms are tightly coupled, and the parameters + algorithm process are coupled.

[0104] As Figure 5 shown, the camera C1 is connected to the data processing device. The data processing device required for the method of eliminating the RTK positioning cycle ambiguity using image positioning includes a memory and a data processor. The data processing device can be an intelligent device that executes programs, such as a mobile phone, a computer, or a vehicle-mounted terminal. The calibration, distortion removal, and inverse perspective technology of the camera C1 adopt industry-standard processing methods.

[0105] Embodiment 2:

[0106] As Figure 7 shown, the distances from the ground P0 to the 4 positioning satellites are BD1: d1, BD2: d2, BD3: d3, BD4: d4 respectively; due to environmental interference, the coordinates of P0 (X0, Y0, Z0) become the wrong convergent coordinate fixed solution P0' (X0', Y0', Z0').

[0107] Assume that the vehicle is moving on the ground and the altitude should be Z0. By the difference between Z0' and Z0, the error of the current fixed solution can be known. Since the radio wavelengths currently used for satellite positioning are about 19 cm to 24 cm, during RTK positioning, the cycle ambiguity is a discrete value in a space, such as Figure 8 . Thus, it can be seen that when the cycle ambiguity solution is incorrect, the elevation difference between the probably incorrect coordinates and the correct coordinates is an integer multiple of the wavelength, that is, an integer multiple of ~20 cm. According to this characteristic, we can use the elevation difference to judge whether the fixed solution is correct.

[0108] As Figure 6 shown, a method for eliminating the RTK positioning cycle ambiguity using altitude is provided, including the following steps:

[0109] S1. Model the site. The surface of the Earth can be approximately regarded as a sphere. We intercept a circle with an area of S on the surface of the sphere and establish a mathematical model for it. The established model is a three-dimensional model. Each point R on this three-dimensional model has R(X, Y, Z), where X is the longitude of the Earth, Y is the latitude of the Earth, and Z is the altitude. Let the surface area of the Earth be S_earth. Then, when taking the limit of (S / S_earth), when (S / S_earth) approaches 0, the radian of the circle with an area of S can be ignored, that is, it is approximately regarded as a plane. Then the altitude of each point falling on this plane should approach a certain specific value. When a satellite locates a certain point on the positioning plane, it will first send a string of pulse signals. The carrier phase of a satellite for high-precision positioning is in the range of L1 carrier (λ1 = 19.03 cm) to L2 carrier (λ2 = 24.42 cm) ±2 to 3 mm. The problem of modulation wave and discontinuous phase appearing when receiving this satellite signal needs to perform a binary phase modulation on the satellite signal through demodulation methods such as a modem to remove the ranging code and navigation message and reconstruct the carrier signal. Reconstruct the carrier by the squaring method: The satellite signal multiplies itself.

[0110] 。

[0111] Usually, multiple satellites are used to locate a point. The carrier phases between satellites are not the same. Therefore, when multiple satellites locate this point and map it to this plane, it is discretely distributed. Moreover, when the satellite signal falls near this plane, the error between it and this plane is an error of λsin(θ). When the satellite is exactly in the positive z-axis direction of this point, θ approaches 0. At this time, when the satellite pulse signal reaches this plane, the error from this plane is at most one wavelength ±λ. And due to the different pulse frequencies between satellites, the discrete points are all within the range of the measured altitude H0 ±λ. The satellite pulse signals within this range can be approximately regarded as credible. We use

[0112]

[0113]

[0114] the least squares +

[0115] method to select the positioning information for a pile of discrete points. This specific value can be regarded as the average altitude of this plane. If a point has normal longitude and latitude but its altitude deviates from the specific value by a certain range, then this point itself is not credible.

[0116] S2. Fix and install a satellite positioning receiver on the top of the vehicle, and perform RTK solution. The solution results are R1(X1, Y1, Z1)…Rn(Xn, Yn, Zn); (Xn, Yn) in Rn can be the longitude and latitude corrected by the method of using image positioning to eliminate the RTK positioning cycle ambiguity.

[0117] S3. The vehicle randomly travels on the road, continuously collects the coordinates Rn, uses the multiple groups of collected coordinates Rn, eliminates the Rn with non-fixed solutions, statistically obtains the average value H0 of Zn, and takes 95% of the statistically obtained Zn values or the range from H0 - 5 cm to H0 + 5 cm as the normal range of Zn in this site.

[0118] S4. When using the final result coordinates Rn(Xn, Yn, Zn) subsequently, perform a logical judgment on the altitude to determine whether it is within the normal range (usually the average altitude of the site). If it is within the normal range, adopt the longitude and latitude information of this message; if it exceeds the normal range, it is determined that the longitude and latitude of this message are not credible, and this longitude and latitude are not adopted in subsequent calculation methods. For example, Z1 = H0 + 2 cm, within the allowable range of the site altitude error, Z2 = H0 - 3 cm, also within the allowable range of the altitude error, Z3 = H0 + 20 cm, exceeding the site altitude error range.

[0119] S5. If it is found that Zn in Rn(Xn, Yn, Zn) of the fixed solution is not within the normal range, under the loose coupling condition, the satellite positioning receiver can be selected to be restarted for re-positioning and convergence; under the tight coupling condition, re-convergence can be selected.

[0120] Embodiment 3: The method of simultaneously using image positioning and altitude to eliminate the RTK positioning cycle ambiguity includes the following steps:

[0121] S1. Conduct a survey of the site containing the garage in advance to obtain longitude, latitude, and altitude coordinates. Fix and install an image positioning device on the vehicle. The image positioning device includes multiple cameras C1 and a data processing device, and calibrate the cameras C1. Install a satellite positioning receiver and perform pre-positioning on the satellite positioning receiver. Drive the vehicle into the garage of the site, use the calibrated cameras C1 to capture pictures in real time, and after passing through the data processing device for distortion removal and inverse perspective, obtain the top view of the vehicle. The pre-positioning includes recording the longitude and latitude of the lower left corner O (olat, olon, oheight), the lower right corner B (blat, blon, oheight), and the upper left corner A (alat, alon, oheight) of the garage in the site. Establish a rectangular coordinate system of OAB with OA as the y-axis and OB as the x-axis. Measure the actual length of one of the parking lines in the garage of the site, and the data processing device calculates the number of pixel points of this parking line in the top view of the vehicle, and then obtains the ratio r of the number of pixel points to the actual line length in the top view of the vehicle.

[0122] S2. The satellite positioning receiver performs RTK real-time solution once per second, and the obtained longitude and latitude of the satellite positioning receiver are P1 (p1lat, p1lon, plheight). By substituting P1 (p1lat, p1lon) into the rectangular coordinate system of OAB, the position P1 (x1, y1) of the satellite positioning receiver in this coordinate system is solved:

[0123] S3. Use the top view of the vehicle to perform image color filtering analysis to obtain all pixel point sets of the garage in the site. Filter out colors other than the parking lines, and perform least squares fitting of straight lines on the point sets of the parts of the left and lower parking lines in the garage in the site that are closest to the image center respectively, and identify the parking lines of the left and lower parking lines.

[0124] S4. The pixel distances from the actual position P2 point of the satellite positioning receiver to the left parking line and to the lower parking line, combined with the ratio r of the number of pixel points to the actual line length in the top view of the vehicle, can calculate the actual distance from the P2 point to the outside of the left parking line and the actual distance from the P2 point to the outside of the lower parking line, and finally obtain the two-dimensional coordinates P2 (x2, y2) of P2.

[0125] S5. Use the two-dimensional coordinates (p2lat, p2lon) of P2 to correct the two-dimensional coordinates (p1lat, p1lon) of P1, calculate the offset, and then perform fixed offset correction on each collected coordinate. The offset is (p2lat - p1lat, p2lon - p1lon). After that, each collected longitude and latitude are fixed plus the offset for correction to obtain Rn (Xn, Yn).

[0126] S6. Model the site: Taking the vehicle as the center, establish a three-dimensional coordinate system of longitude, latitude, and altitude XYZ;

[0127] S7. Fix and install a satellite positioning receiver on the top of the vehicle and perform RTK solution. The solution results are R1(X1, Y1, Z1)…Rn(Xn, Yn, Zn);

[0128] S8. The vehicle randomly travels on the road, continuously collects the coordinates Rn, uses the multiple groups of coordinates Rn obtained through statistics, eliminates the Rn with non-fixed solutions, and takes the range of 95% of the Zn values obtained through statistics as the normal range of Zn in this site;

[0129] S9. Conduct a logical judgment on the altitude Zn of the subsequently collected coordinates Rn(Xn, Yn, Zn) to determine whether it is within the normal range. If it is within the normal range, the collected coordinate information is adopted; if it exceeds the normal range, it is determined that the collected coordinate information is not credible, and this coordinate information is not used in subsequent calculations;

[0130] S10. If it is found that Zn in Rn(Xn, Yn, Zn) among the fixed solutions is not within the normal range, under the loose coupling condition, the satellite positioning receiver can be selected to be restarted for repositioning and convergence; under the tight coupling condition, re-convergence can be selected.

[0131] In order to obtain more accurate results, the following three different schemes can also be adopted:

[0132] 1. The satellite can be made to lose lock multiple times, then refixed, and the value can be taken after waiting for stability. Repeat the above process for clustering analysis. Within a certain probability, the closest true RTK fixed solution can be obtained through the clustering results;

[0133] 2. When taking the RTK fixed solution points, take values multiple times, and at least take more than 3 points as RTK calibration points. The calibration points should not be on the same straight line;

[0134] 3. Placing the satellite positioning receiver at a high point can also reduce the RTK false fixation phenomenon. This ensures that the satellite signals received by the satellite positioning receiver are reflected and blocked by surrounding obstacles as little as possible, and eliminates the RTK false fixation influence by reducing the frequency of bit errors caused by noise to digital signals during transmission;

[0135] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for eliminating the integer ambiguity in RTK positioning, characterized in that: it includes a method for eliminating the integer ambiguity in RTK positioning by using image positioning, and the steps are as follows: S1. Prior to surveying the site containing the garage to obtain longitude, latitude, and altitude coordinates, fixedly installing an image positioning device on the vehicle. The image positioning device includes multiple cameras C1 and a data processing device, calibrating the cameras C1, installing a satellite positioning receiver, and pre-positioning the satellite positioning receiver; driving the vehicle into the garage of the site, using the calibrated cameras C1 to capture pictures in real time, and performing de-distortion and inverse perspective through the data processing device to obtain a top view of the vehicle; the pre-positioning includes recording the longitude and latitude of the lower left corner O (olat, olon, oheight), the lower right corner B (blat, blon, bheight), and the upper left corner A (alat, alon, aheight) of the garage in the site, establishing a rectangular coordinate system of OAB with OA as the y-axis and OB as the x-axis; measuring the actual line length of one of the parking lines in the garage of the site, and the data processing device calculating the number of pixel points of this parking line in the top view of the vehicle, and further obtaining the ratio r of the number of pixel points to the actual line length in the top view of the vehicle; S2. The satellite positioning receiver performs RTK real-time solution once per second, and the solution obtains the longitude and latitude P1 (p1lat, p1lon, plheight) of the satellite positioning receiver. By substituting P1 (p1lat, p1lon) into the rectangular coordinate system of OAB, the position P1 (x1, y1) of the satellite positioning receiver in this coordinate system is solved: Let the equatorial radius of the earth be Ea, the polar radius be Eb, the polar circumference be Db, the longitude and latitude of the coordinate system origin O be (olat, olon), and the included angle between the y-axis and the due north be angle, Equation for converting longitude and latitude (lat, lon) to (x, y) coordinates: pD = R × 2πcos(olat), Displacement in the due east direction: Displacement in the due north direction: x coordinate: x = dE × cos(angle) - dN × sin(angle), y coordinate: y = dN × cos(angle) + dE × sin(angle); S3. Using the top view of the vehicle to perform image color filtering analysis to obtain all pixel point sets of the garage in the site, filtering out colors other than the parking lines, respectively performing least squares fitting of straight lines on the point sets of the parts of the left and lower parking lines in the garage of the site that are closest to the image center, and identifying the parking lines of the left and lower parking lines; The pixel distances from the actual position P2 point of the satellite positioning receiver to the left parking line and to the lower parking line, combined with the ratio r of the number of pixel points to the actual line length in the top view of the vehicle, can calculate the actual distance from the P2 point to the outside of the left parking line and the actual distance from the P2 point to the outside of the lower parking line, and finally obtain the two-dimensional coordinates P2 (x2, y2) of P2. The longitude and latitude of the coordinate system origin and the included angle angle between the y-axis of the coordinate system and the due north are known. Let angle = 0, and using the formula for converting longitude and latitude (lat, lon) to (x, y) coordinates, it can be obtained that: Variation relationship between longitude and distance in the due east direction Variation relationship between latitude and distance in the due north direction The offset distance of the coordinate (x, y) in the due north direction is Δn = -xsin(angle) + ycos(angle); The offset distance in the due east direction is Δe = xcos(angle) + ysin(angle) Therefore The longitude and latitude P2(p2lat, p2lon) corresponding to the coordinate P2(x2, y2) can be obtained; S5. Use the two-dimensional coordinates (p2lat, p2lon) of P2 to correct the two-dimensional coordinates (p1lat, p1lon) of P1, calculate the offset, and then correct the coordinates collected each time with a fixed offset. The offset is (p2lat - p1lat, p2lon - p1lon). After that, the longitude and latitude collected each time are fixed plus the offset for correction to obtain Rn(Xn, Yn).

2. The method for eliminating the RTK positioning cycle ambiguity according to claim 1, wherein, the site garage can be any standard garage.

3. The method for eliminating the RTK positioning cycle ambiguity according to claim 2, wherein, When the image positioning accuracy is insufficient and the library position line cannot be recognized or only part of the library position line is recognized during the initial image filtering, the value of P2(p2lat, p2lon) is abandoned.

4. The method for eliminating the RTK positioning cycle ambiguity according to claim 3, wherein, When the offset in step S5 is greater than the critical value, under the loose coupling condition, select to restart the satellite positioning receiver and perform satellite search, positioning, and convergence again; under the tight coupling condition, select to trust the current star map and converge again.

5. The method for eliminating the RTK positioning cycle ambiguity according to claim 4, wherein, All the pixel point sets of the site garage obtained by image color filtering in step S3 include calculating the rgb values of each pixel in the vehicle top view and filtering out the rbg values other than the library position lines in the site garage.

6. The method for eliminating the RTK positioning cycle ambiguity according to claim 5, wherein, The data processing device includes a memory and a data processor for executing the method for eliminating the RTK positioning cycle ambiguity according to any one of claims 1 to 5.

7. The method for eliminating the RTK positioning cycle ambiguity according to claim 1, characterized in that: There is also provided a method for eliminating the RTK positioning cycle ambiguity by using altitude, including the following steps: S1. Model the site: Intercept a circle with an area of S on the earth's surface and establish an XYZ three-dimensional coordinate system with longitude, latitude, and altitude; S2. Fix and install a satellite positioning receiver on the vehicle top and perform RTK solution. The solution results are R1(X1, Y1, Z1)... Rn(Xn, Yn, Zn); S3. The vehicle randomly travels on the road, continuously collects the coordinates Rn, uses the multiple groups of coordinates Rn obtained by statistics to eliminate the Rn with non-fixed solutions, and uses the 95% Zn value interval obtained by statistics as the normal range of Zn in this site; S4. Logically judge the altitude Zn of the subsequently collected coordinates Rn(Xn, Yn, Zn) to determine whether it is within the normal range. If it is within the normal range, the collected coordinate information is adopted; if it exceeds the normal range, the collected coordinate information is considered untrustworthy and this coordinate information is not used in subsequent calculations. S5. If it is found that Zn in Rn(Xn, Yn, Zn) in the fixed solution is not within the normal range, under the loose coupling condition, select to restart the satellite positioning receiver and re - perform positioning and convergence; under the tight coupling condition, select to converge again.

8. The method for eliminating the RTK positioning integer ambiguity according to claim 1, characterized in that, it further includes a method for eliminating the RTK positioning integer ambiguity by simultaneously using the image positioning and altitude described in claims 1 and 7; when (Zn) in Rn is incorrect and (Xn, Yn) is accurate; when (Xn, Yn) in Rn is incorrect and (Zn) is accurate, re - perform the convergence calculation.

Images