Positioning analysis method based on multi-source image

Abstract

The invention discloses a positioning analysis method based on a multi-source image, and the method comprises the following steps: 1), collecting images, including buildings, shops, factories, roads and traffic signs, with identification objects around a to-be-positioned position; 2) calculating contour features of the marker according to the acquired image, and comparing the contour features with the contour and feature library of the established identified object; and 3) calculating the real-time fusion position of the information multi-source measurement data. By using the method of the invention, the position of the intelligence target object can be determined more accurately, and the high-precision positioning is helpful for improving the accuracy of the intelligence position.

Description

Technical Field

[0001] The invention relates to image processing technology, and in particular to a positioning analysis method based on multi-source images. Background Art

[0002] In actual location positioning applications, GPS or base station signals are usually collected for location positioning, but when the signal is poor, accurate positioning is impossible. In this case, multiple reference objects with obvious features around the photographed object can be used for comparison to locate the shooting location. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a positioning analysis method based on multi-source images in view of the defects in the prior art.

[0004] The technical solution adopted by the present invention to solve the technical problem is: a positioning analysis method based on multi-source images, comprising the following steps:

[0005] 1) Collect images of surrounding objects with identifiers at the location to be located, including buildings, shops, factories, roads, and traffic signs;

[0006] 2) Calculate the contour features of the marker based on the collected image and compare them with the contour and feature library of the marker object;

[0007] Extraction of the feature library of the positioning points. This process can be done by calculating the region establishment through the regionprops function in MATLAB;

[0008] The contour features of the landmark are calculated by using methods such as resection;

[0009] 3) Real-time fusion calculation of intelligence multi-source measurement data;

[0010] The distance between the shooting point (x, y, z) and the source target (x1, y1, z1) can be expressed as:

[0011]

[0012] Three markers (i.e. three buildings: 1, 2, 3) are distributed in different locations. Each building can calculate the measurement data to the target object, including the distance between any two points, longitude and latitude, elevation H value, and time;

[0013] After time alignment, the automatic optimal intersection method is used according to the coordinates of the photographer's site (B longitude, L latitude, H elevation), and the target is automatically selected to perform trajectory coordinate intersection calculation with 1, 2, or 1 and 3 to obtain the coordinates (X, Y, Z) of the photographed target;

[0014] According to the intersection result, the position guidance data of the shooting target is set;

[0015] Set a time period T. At the end of each time period, calculate the fitted extrapolation function W(t) based on the (t, gx, gy, gz) obtained by the intersection, substitute the current moment into the function to obtain the trajectory extrapolation value at this moment, convert it into azimuth and pitch angle according to the photographer's station location (latitude, longitude and altitude), and send it to the server system as guidance data.

[0016] The calculation steps are as follows:

[0017] 3.1) Calculate the cosine vector (l, m, n) between the coordinate position of the two mobile phone shots and the position of the captured intelligence target. The calculation formula is as follows:

[0018]

[0019] Where: At is the azimuth of the photographer's position; L0, B0 are the latitude and longitude of the origin of the reference marker coordinate system; L, B are the latitude and longitude of the reference marker; A, E are the azimuth and elevation angles that the photographer may move to measure, and the subscript i represents the number of shots;

[0020] 3.2) Calculate the angle between two different shooting positions and the intelligence target The angle between the cosine vector of the target and the line connecting the target station and the line connecting the two station positions The cosine vector of is as follows:

[0021]

[0022] Where: x 0i ,y 0i , z 0i (i=1, 2) is the position coordinate of the i-th reference marker in the reference marker coordinate system, is the reference value for optimal intersection, the intersection angle between the two stations Small and medium ones are automatically selected for intersection;

[0023] 3.3) Calculate the slant distance of the target detected twice according to the sine theorem. The formula is as follows:

[0024]

[0025] 3.4) Calculate the coordinates of the target in the reference identification coordinate system. The formula is as follows:

[0026]

[0027] Where: x, y, z are the coordinates of the target in the reference identification coordinate system;

[0028] 3.5) Integration and analysis;

[0029] The guidance data is fused, and the fusion formula is:

[0030] P=(P0+2×P E ) / 3;

[0031] Among them, P0 is the location guidance data for setting the shooting target, P E The calculated position value of the target in the reference identification coordinate system;

[0032] The fusion result is used as the final location recognition result.

[0033] The beneficial effects produced by the present invention are:

[0034] The method of the present invention can be used to more accurately determine the location of the intelligence target object, and this high-precision positioning helps to improve the accuracy of the intelligence location. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0036] Figure 1 is a method flow chart of an embodiment of the present invention;

[0037] Figure 2 Schematic diagram of the fusion result of the embodiment of the present invention. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0039] like Figure 1 As shown, a positioning analysis method based on multi-source images includes the following steps:

[0040] 1) Collect images of surrounding objects with identifiers at the location to be located, including buildings, shops, factories, roads, and traffic signs;

[0041] 2) Calculate the contour features of the marker based on the collected image, compare it with the contour and feature library of the marker object established previously, and obtain the position data of the marker object;

[0042] The contour features of the landmarks are calculated using the resection method;

[0043] Extraction of the feature library of the positioning points. This process can be done by calculating the region establishment through the regionprops function in MATLAB; its prototype is:

[0044] D = regionprops(L, properties), where: L is a label matrix that can be obtained by the connected region labeling function bwlabel. properties can be a comma-separated string list.

[0045] During the back intersection, the camera's intrinsic parameters (x0, y0, f) and the spatial coordinates of the feature point, Y, Z are known, the feature point image coordinates (X, Y) are the observed values, and the image's exterior orientation element T (X U , Y U , Z U , φ, ω, κ) are the quantities to be determined. The observation equation can be listed for each photographic light, as shown below:

[0046]

[0047] The above formula is a nonlinear equation. To solve it, it is linearized and the initial value of the external orientation element is Perform Taylor expansion at and take the first-order term to obtain the error equation as shown below.

[0048]

[0049] Where V x and V y is the correction factor of the image point coordinates (x, y), ΔX s , ΔY s , ΔZ s , and Δκ is the increment of the exterior orientation element, and is the coordinate of the image point obtained in the previous iteration, that is,

[0050] 3) Real-time fusion calculation of intelligence multi-source measurement data;

[0051] The distance between the shooting point (x, y, z) and the source target (x1, y1, z1) can be expressed as:

[0052]

[0053] Three markers (i.e. three buildings: 1, 2, 3) are distributed in different locations. Each building can calculate the measurement data to the target object, including the distance between any two points, longitude and latitude, elevation H value, and time;

[0054] After time alignment, the automatic optimal intersection method is used according to the coordinates of the photographer's site (B longitude, L latitude, H elevation), and the target is automatically selected to perform trajectory coordinate intersection calculation with 1, 2, or 1 and 3 to obtain the coordinates (X, Y, Z) of the photographed target;

[0055] According to the intersection result, the position guidance data of the shooting target is set;

[0056] Set a time period T. At the end of each time period, calculate the fitted extrapolation function W(t) based on the (t, gx, gy, gz) obtained by the intersection, substitute the current moment into the function to obtain the trajectory extrapolation value at this moment, convert it into azimuth and pitch angle according to the photographer's station location (latitude, longitude and altitude), and send it to the server system as guidance data.

[0057] The calculation steps are as follows:

[0058] 3.1) Calculate the cosine vector (l, m, n) between the coordinate position of the two mobile phone shots and the position of the captured intelligence target. The calculation formula is as follows:

[0059]

[0060] Where: At is the azimuth of the photographer's position; L0, B0 are the latitude and longitude of the origin of the reference marker coordinate system; L, B are the latitude and longitude of the reference marker; A, E are the azimuth and elevation angles that the photographer may move to measure, and the subscript i represents the number of shots;

[0061] 3.2) Calculate the angle between two different shooting positions and the intelligence target The cosine vector of the target and the angle between the line connecting the target to the measuring station and the line connecting the two measuring stations The cosine vector of is as follows:

[0062]

[0063] Where: x 0i ,y 0i , z 0i (i=1, 2) is the position coordinate of the i-th reference marker in the reference marker coordinate system, is the reference value for optimal intersection, the intersection angle between the two stations Small and medium ones are automatically selected for intersection;

[0064] 3.3) Calculate the slant distance of the target detected twice according to the sine theorem. The formula is as follows:

[0065]

[0066] 3.4) Calculate the coordinates of the target in the reference identification coordinate system. The formula is as follows:

[0067]

[0068] Where: x, y, z are the coordinates of the target in the reference identification coordinate system;

[0069] Therefore, it can be concluded that the more reference marker objects (1…n) are associated based on the current station coordinate position, the more accurate the measurement result will be;

[0070] 3.5) Integration and analysis;

[0071] The guidance data is fused, and the fusion formula is:

[0072] P=(P0+2×P E ) / 3;

[0073] Among them, P0 is the location guidance data for setting the shooting target, P E The calculated position value of the target in the reference identification coordinate system;

[0074] The fusion result is used as the final location recognition result.

[0075] Record both the unfused guidance data and the fused guidance data, and compare the pitch and azimuth values ​​obtained after interpretation. The analysis results are as follows: Figure 2 As shown, A0 is the true value, A1 is the unfused guided value, and A2 is the fused guided value.

[0076] After the calculation is completed, the target image acquisition result is stored in the contour and feature history library of the acquired identification object. This process will increase the number of reference identification objects, and therefore, will continuously enhance the multi-source detection capability of the system of the present invention.

[0077] It should be understood that those skilled in the art can make improvements or changes based on the above description, and all these improvements and changes should fall within the scope of protection of the appended claims of the present invention.

Claims

1. A positioning analysis method based on multi-source images, characterized in that: The following steps are involved: 1) Collect images of surrounding objects with identifiers at the location to be located, including buildings, shops, factories, roads, and traffic signs; 2) Calculate the contour features of the marker based on the collected image and compare them with the contour and feature library of the marker object; 3) Real-time fusion and position calculation of intelligence multi-source measurement data; The distance between the shooting point (x, y, z) and the source target (x1, y1, z1) is expressed as: Take three markers 1, 2, and 3 distributed in different locations, and each building can calculate the measurement data to the target object, including the distance between any two points, longitude and latitude, elevation H value, and time; After time alignment, according to the coordinates of the photographer's site (B, L, H), where B represents longitude, L represents latitude, and H represents elevation; use the automatic optimal intersection method to automatically select the target and 1, 2 or 1, 3 for trajectory coordinate intersection calculation to obtain the coordinates (X, Y, Z) of the photographed target; According to the intersection result, the position guidance data of the shooting target is set; The steps for real-time fusion position calculation are as follows: 3.1) Calculate the cosine vector (l, m, n) between the coordinate position of the two mobile phone shots and the position of the captured intelligence target. The calculation formula is as follows: Where: At is the azimuth of the photographer's position; L0, B0 are the latitude and longitude of the origin of the reference marker coordinate system; L, B are the latitude and longitude of the reference marker; A, E are the azimuth and elevation angles that the photographer may move to measure, and the subscript i represents the number of shots; 3.2) Calculate the angle between two different shooting positions and the intelligence target The cosine vector of the target and the angle between the line connecting the target and the measuring station and the line connecting the two measuring stations The cosine vector of is as follows: Where: x 0i ,y 0i , z 0i (i=1, 2) is the position coordinate of the i-th reference marker in the reference marker coordinate system, is the reference value for optimal intersection, the intersection angle between the two stations Small and medium ones are automatically selected for intersection; 3.3) Calculate the slant distance of the target detected twice according to the sine theorem. The formula is as follows: 3.4) Calculate the coordinates of the target in the reference identification coordinate system. The formula is as follows: Where: x, y, z are the coordinates of the target in the reference identification coordinate system; 3.5) Integration and analysis; The guidance data is fused, and the fusion formula is: P=(P0+2×P E ) / 3; Among them, P0 is the location guidance data for setting the shooting target, P E The calculated position value of the target in the reference identification coordinate system; The fusion result is used as the final location recognition result.

2. The positioning analysis method based on multi-source images according to claim 1 is characterized in that: In the step 2), the contour features of the marker are calculated using the resection method.

Images