Method, device, equipment and medium for relative accuracy quality inspection between orthophotos

By automatically matching corresponding points and removing error points from orthophoto data, and combining a block geometric polynomial model and an average error multiplier threshold algorithm, the problem of automating and quantifying the quality inspection of relative geometric accuracy between orthophotos is solved, and efficient quality inspection report generation is achieved.

CN118505709BActive Publication Date: 2024-11-15BEIJING AEROSPACE HONGTU INFORMATION TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410977496.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-22
Publication Date
2024-11-15
Estimated Expiration
2044-07-22

AI Technical Summary

Technical Problem

Existing technologies cannot automate or quantify the problem of relative geometric accuracy quality inspection between orthophotos.

Method used

By acquiring at least two orthophoto images with geographical overlap, automatic matching of corresponding points is performed. Then, multiple error point removals are carried out using a block geometric polynomial model and an average error multiplier threshold algorithm. The relative accuracy error is calculated, and a relative accuracy quality inspection report is generated.

Benefits of technology

It enables automated and quantitative quality inspection of the relative geometric accuracy between orthophotos, generates accurate quality inspection reports, and can quickly locate and quantitatively assess the relative accuracy error of images.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118505709B_ABST
    Figure CN118505709B_ABST
Patent Text Reader

Abstract

The present invention provides a relative accuracy quality inspection method, device, equipment and medium between orthophotos, including: obtaining orthophoto data of at least two scenes with geographical overlap; automatically matching the orthophoto data with the same name points to obtain the initial same name points with the same ground feature between the orthophoto data; using the block geometric polynomial model and the average error magnification threshold algorithm to perform multiple error point elimination processing on the initial same name points to obtain the target same name points between the orthophoto data; based on the target same name points contained in any orthophoto data, determining the relative accuracy error between the orthophoto data and the orthophoto data with geographical overlap; generating a relative accuracy quality inspection report corresponding to each orthophoto data according to the relative accuracy error. The present invention can automatically and quantitatively realize the quality inspection of the relative geometric accuracy between orthophotos.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of image accuracy quality inspection, and in particular to a relative accuracy quality inspection method, device, equipment and medium between orthophoto images. Background Art

[0002] At present, relevant technologies provide solutions for image accuracy quality inspection, such as optical satellite image geometric quality evaluation methods, remote sensing satellite image geometric accuracy automated inspection methods, and stereo satellite image accuracy quality inspection methods. However, the above solutions are not suitable for relative accuracy quality inspection between orthophotos. In other words, the above solutions cannot automatically and quantitatively solve the problem of relative geometric accuracy quality inspection between orthophotos. Summary of the invention

[0003] In view of this, the purpose of the present invention is to provide a method, device, equipment and medium for quality inspection of relative accuracy between orthophotos, which can automatically and quantitatively realize quality inspection of relative geometric accuracy between orthophotos.

[0004] In a first aspect, an embodiment of the present invention provides a relative accuracy quality inspection method between orthophotos, comprising:

[0005] Obtain orthophoto data with at least two scenes having geographical overlap;

[0006] Automatically match the orthophoto data with the same name points to obtain the initial same name points with the same ground features between the orthophoto data;

[0007] The block geometric polynomial model and average error ratio threshold algorithm are used to eliminate the initial homonymous points for multiple times, and the target homonymous points between orthophoto data are obtained.

[0008] Based on the target same-name points contained in any orthophoto data, determine the relative accuracy error between the orthophoto data and the orthophoto data with which it has a geographical overlap;

[0009] Based on the relative accuracy error, a relative accuracy quality inspection report corresponding to each orthophoto data is generated.

[0010] In one implementation, a block geometric polynomial model and an average error magnification threshold algorithm are used to perform multiple error point elimination processes on the initial homonymous points to obtain target homonymous points between orthophoto data, including:

[0011] The orthophoto data is divided into grids, and the target geometric polynomial model corresponding to each grid is determined, so as to use the target geometric polynomial model to perform the first error point elimination process on the initial homonymous points contained in each grid to obtain the intermediate homonymous points;

[0012] The average error ratio threshold algorithm is used to perform a second error point elimination process on the middle homonymous points between two orthophoto data with geographical overlap to obtain the target homonymous points.

[0013] In one implementation, determining a target geometric polynomial model corresponding to each grid includes:

[0014] If the number of initial points of the same name contained in the grid is less than the first point number threshold, the initial points of the same name contained in the grid are not subjected to the first error point elimination process;

[0015] Alternatively, if the number of initial points of the same name contained in the grid is greater than the first point number threshold and less than or equal to the second point number threshold, the linear geometric polynomial model is determined as the target geometric polynomial model;

[0016] Alternatively, if the number of initial points of the same name contained in the grid is greater than a second point number threshold, the quadratic geometric polynomial model is determined as the target geometric polynomial model.

[0017] In one implementation, the target geometric polynomial model is used to perform a first error point elimination process on the initial homonymous points contained in each grid to obtain intermediate homonymous points, including:

[0018] Using the target geometry polynomial model, based on the first pixel coordinates of the initial same-name points contained in the grid in the orthophoto data where they are located, determine the second pixel coordinates of the initial same-name points in another orthophoto data;

[0019] The difference between the first pixel coordinate and the second pixel coordinate is compared with a preset difference threshold, so as to perform a first error point elimination process on the initial same-name points contained in the grid based on the comparison result to obtain intermediate same-name points.

[0020] In one embodiment, an average error magnification threshold algorithm is used to perform a second error point elimination process on the middle homonymous points between two orthophoto data with geographical overlap to obtain target homonymous points, including:

[0021] Determine the geographic coordinate error of the middle points with the same name between two orthophoto data with geographic overlap;

[0022] Determine the average geographic coordinate error between two orthophoto data based on the geographic coordinate error of the middle same-name point;

[0023] The geographic coordinate error is compared with the average value of the geographic coordinate error of the specified magnification, and the intermediate homonymous points between the two orthophoto data are subjected to a second error point elimination process based on the comparison result to obtain the target homonymous points.

[0024] In one embodiment, based on the target same-name points contained in any orthophoto data, determining the relative accuracy error between the orthophoto data and the orthophoto data having a geographical overlap relationship therewith comprises:

[0025] Convert the geographic coordinates of the target homonymous point contained in any orthophoto data into the third pixel coordinates of the target homonymous point on another orthophoto data;

[0026] Determine the pixel error between the homonymous points based on the third pixel coordinate and the fourth pixel coordinate of the homonymous point matched with the target homonymous point on the orthophoto data of another scene;

[0027] The relative accuracy error between the orthophoto data and another orthophoto data is determined based on the pixel error between the same-name points.

[0028] In one embodiment, the relative accuracy quality inspection report includes a scene-by-scene relative accuracy quality inspection report and an overall relative accuracy quality inspection report;

[0029] The relative accuracy quality inspection report for each scene includes a quality inspection report between any scene orthophoto data and all orthophoto data that have geographical overlap with it, and / or a quality inspection report between any scene orthophoto data and another scene orthophoto data that has geographical overlap with it;

[0030] The overall relative accuracy quality inspection report includes the quality inspection report between each orthophoto data.

[0031] In a second aspect, an embodiment of the present invention further provides a relative accuracy quality inspection device between orthophotos, comprising:

[0032] An image acquisition module is used to acquire orthophoto data of at least two scenes having a geographically overlapping relationship;

[0033] The homonymous point matching module is used to automatically match homonymous points in orthophoto data to obtain initial homonymous points with the same ground features between orthophoto data;

[0034] The error point elimination module is used to use the block geometric polynomial model and the average error magnification threshold algorithm to perform multiple error point elimination processes on the initial homonymous points to obtain the target homonymous points between the orthophoto data;

[0035] The error determination module is used to determine the relative accuracy error between the orthophoto data and the orthophoto data with which the orthophoto data has a geographical overlap, based on the target same-name points contained in any orthophoto data;

[0036] The report generation module is used to generate a relative accuracy quality inspection report corresponding to each orthophoto data according to the relative accuracy error.

[0037] In a third aspect, an embodiment of the present invention further provides an electronic device, comprising a processor and a memory, wherein the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement any one of the methods provided in the first aspect.

[0038] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to implement any one of the methods provided in the first aspect.

[0039] The embodiment of the present invention provides a relative accuracy quality inspection method, device, equipment and medium between orthophotos. First, orthophoto data of at least two scenes with geographical overlap are obtained, and the orthophoto data are automatically matched with same-name points to obtain initial same-name points with the same ground feature between the orthophoto data; then, a block geometric polynomial model and an average error magnification threshold algorithm are used to perform multiple error point elimination processing on the initial same-name points to obtain target same-name points between the orthophoto data; then, based on the target same-name points contained in any orthophoto data, the relative accuracy error between the orthophoto data and the orthophoto data with geographical overlap is determined; finally, a relative accuracy quality inspection report corresponding to each orthophoto data is generated according to the relative accuracy error. The above method aims at the problem of automatic quality inspection of relative geometric accuracy between orthophoto data. After automatically matching the same-name points between orthophoto data, the method automatically eliminates the wrong same-name points by combining the block geometric polynomial model method and the average error magnification threshold method, calculates the relative accuracy error between one scene of orthophoto data and another scene of orthophoto data, and then statistically calculates the relative accuracy error between each scene of orthophoto data and all orthophoto data with overlapping relationships, and automatically generates a quality inspection report for each scene of orthophoto data. The embodiment of the present invention can automatically and quantitatively realize the quality inspection of the relative geometric accuracy between orthophotos.

[0040] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0041] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0043] Figure 1 A schematic diagram of a flow chart of a relative accuracy quality inspection method between orthophotos provided by an embodiment of the present invention;

[0044] Figure 2 A schematic diagram of a scene segmentation relative accuracy quality inspection report provided by an embodiment of the present invention;

[0045] Figure 3 A schematic diagram of another scene segmentation relative accuracy quality inspection report provided by an embodiment of the present invention;

[0046] Figure 4 A schematic diagram of an overall relative accuracy quality inspection report provided by an embodiment of the present invention;

[0047] Figure 5 An overall flow chart of a relative accuracy quality inspection method between orthophotos provided by an embodiment of the present invention;

[0048] Figure 6 A schematic diagram of the structure of a relative accuracy quality inspection device between orthophotos provided by an embodiment of the present invention;

[0049] Figure 7 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described in combination with the embodiments below. 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.

[0051] At present, the existing technical solutions are not suitable for relative accuracy quality inspection between orthophotos. Based on this, the present invention provides a method, device, equipment and medium for relative accuracy quality inspection between orthophotos, which can automatically and quantitatively realize the quality inspection of the relative geometric accuracy between orthophotos.

[0052] To facilitate understanding of this embodiment, firstly, a relative accuracy quality inspection method between orthophotos disclosed in an embodiment of the present invention is described in detail, see Figure 1The flowchart of a relative accuracy quality inspection method between orthophotos shown in FIG. 1 mainly includes the following steps S102 to S110:

[0053] Step S102, obtaining orthophoto data of at least two scenes having a geographically overlapping relationship.

[0054] Orthophoto data refers to images that have been orthorectified and have their own geographic coordinate values. They are advanced image products and are not limited to satellite orthophotos, but also include orthophotos from aviation and drones. In one example, the data that needs to be prepared is orthophoto data of at least two scenes with geographical overlap.

[0055] Step S104, automatically matching the orthophoto data with the same-name points to obtain initial same-name points with the same ground feature between the orthophoto data.

[0056] Among them, the same-name points are the basis of quantitative geometric accuracy quality inspection. In one example, the same-name points between orthophoto data can be matched using fully automatic matching technology to obtain initial same-name points.

[0057] Step S106, using a block geometric polynomial model and an average error magnification threshold algorithm, multiple error point elimination processes are performed on the initial homonymous points to obtain target homonymous points between orthophoto data.

[0058] The purpose of error point elimination is to screen and eliminate the points with large errors in the initial homonymous points obtained by automatic matching, so that the target homonymous points obtained after elimination have a higher accuracy rate, thereby ensuring the accuracy of the final relative accuracy quality inspection report. In one example, the initial homonymous points are firstly eliminated by using the block geometric polynomial model to obtain the intermediate homonymous points, and then the average error multiplier threshold algorithm is used to perform the second error point elimination on the intermediate homonymous points to obtain the target homonymous points.

[0059] Step S108, based on the target same-name points contained in any orthophoto data, determine the relative accuracy error between the orthophoto data and the orthophoto data that has a geographical overlap relationship with the orthophoto data.

[0060] In one example, target homonymous points of each orthophoto data and all orthophoto data with geographical overlap are cyclically counted, and pixel coordinate errors between target homonymous points, referred to as pixel errors between homonymous points, are calculated. Then, the relative accuracy errors between orthophoto data are calculated based on the pixel errors between homonymous points.

[0061] Step S110: generating a relative accuracy quality inspection report corresponding to each orthophoto data according to the relative accuracy error.

[0062] Among them, the relative accuracy quality inspection report includes the scene-by-scene relative accuracy quality inspection report and the overall relative accuracy quality inspection report; the scene-by-scene relative accuracy quality inspection report includes the quality inspection report between the orthophoto data of any scene and all the orthophoto data that have a geographical overlap relationship with it, and / or, the quality inspection report between the orthophoto data of any scene and the orthophoto data of another scene that has a geographical overlap relationship with it; the overall relative accuracy quality inspection report includes the quality inspection report between the orthophoto data of each scene.

[0063] The above method aims at the problem of automatic quality inspection of relative geometric accuracy between orthophoto data. After automatically matching the same-name points between orthophoto data, the method automatically eliminates the wrong same-name points by combining the block geometric polynomial model method and the average error magnification threshold method, calculates the relative accuracy error between one scene of orthophoto data and another scene of orthophoto data, and then statistically calculates the relative accuracy error between each scene of orthophoto data and all orthophoto data with overlapping relationships, and automatically generates a quality inspection report for each scene of orthophoto data. The embodiment of the present invention can automatically and quantitatively realize the quality inspection of the relative geometric accuracy between orthophotos.

[0064] For ease of understanding, an embodiment of the present invention provides a specific implementation of a relative accuracy quality inspection method between orthophotos.

[0065] The embodiment of the present invention provides a specific implementation of step S104. The automation mainly relies on the fully automatic matching technology. The fully automatic matching algorithm adopted by the embodiment of the present invention is as follows: (i) the feature point extraction algorithm adopts the Harris operator; (ii) the automatic matching strategy adopts the pyramid level-by-level matching strategy; (iii) the main matching measurement adopts the correlation coefficient method.

[0066] After matching each orthophoto data, a homonymous point file is generated, which records the homonymous points between the orthophoto data of this scene and all adjacent orthophoto data with geographical overlap. The format of homonymous points is as follows:

[0067]

[0068] Where, (1) ID: initial same-name point number; (2) : Column coordinates of orthophoto data; (3) : row coordinates of orthophoto data; (4) : geographic X coordinate of orthophoto data; (5) : Geographic Y coordinate of orthophoto data; (6) : The image path of the nth scene with overlapping images; (7) : The column coordinates of the image with overlapping area in the nth scene; (8) : The row coordinates of the image with overlapping area in the nth scene; (9) : The geographic X coordinate of the image with overlapping area in the nth scene; (10) : The geographic Y coordinate of the nth image with overlapping area images.

[0069] The embodiment of the present invention provides a specific implementation method of step S106. The orthophoto data itself does not have a strict imaging model, so it cannot perform free network adjustment to delete error points through the rational polynomial model (RPC) like satellite L1 remote sensing images. The embodiment of the present invention intends to use a block geometric polynomial model and an average error multiple threshold method to detect and eliminate error points.

[0070] For details, please refer to steps 1 to 2 below:

[0071] Step 1: Divide the orthophoto data into grids, and determine the target geometric polynomial model corresponding to each grid, so as to use the target geometric polynomial model to perform the first error point elimination process on the initial homonymous points contained in each grid to obtain intermediate homonymous points;

[0072] Orthophoto data usually does not conform to the geometric polynomial model, except for small area images without internal distortion. Therefore, the embodiment of the present invention uses a block geometric polynomial model to more approximately characterize the model of orthophoto data, avoiding the problem that the entire area is too large for the geometric polynomial model to fit. Specifically, it includes:

[0073] Step 1.1, divide each orthophoto data into grids.

[0074] Step 1.2, count all the IDs of the points with the same name in each grid to determine the target geometric polynomial model corresponding to each grid, including:

[0075] (1.2.1) If the number of initial same-name points contained in the grid is less than the first point number threshold, the first error point elimination process is not performed on the initial same-name points contained in the grid. (1.2.2) If the number of initial same-name points contained in the grid is greater than the first point number threshold and less than or equal to the second point number threshold, the first-order geometric polynomial model is determined as the target geometric polynomial model. (1.2.3) If the number of initial same-name points contained in the grid is greater than the second point number threshold, the second-order geometric polynomial model is determined as the target geometric polynomial model.

[0076] For example, the first point number threshold can be set to 4, and the second point number threshold can be set to 6. In one example, when the number of initial points of the same name contained in the grid is less than 4 points, no elimination processing is performed, when the number of initial points of the same name contained in the grid is greater than 4 and less than or equal to 6, a first-order geometric polynomial model is used, and when the number of initial points of the same name contained in the grid is greater than 6, a second-order geometric polynomial model is used.

[0077] Specifically, the linear geometric polynomial is as follows: , , , , , is the coefficient of a first-order geometric polynomial, calculated using at least three initial homonymous points, where x and y represent the pixel coordinates of the initial homonymous points on a scene orthophoto data. and Represents the pixel coordinates of another orthophoto data calculated using the linear geometric polynomial coefficients.

[0078] ;

[0079] Specifically, the quadratic geometric polynomial is as follows, , , , , , , , , are the coefficients of a quadratic geometric polynomial, calculated using at least 6 initial homonymous points, where x and y represent the pixel coordinates of the initial homonymous points on a scene orthophoto data. and Represents the pixel coordinates of another orthophoto calculated using the coefficients of a linear geometric polynomial.

[0080] .

[0081] Step 1.3, loop through each grid to solve the geometric polynomial error, including:

[0082] (1.3.1) Using the target geometry polynomial model, based on the first pixel coordinates of the initial homonymous points contained in the grid in the orthophoto data where they are located, determine the second pixel coordinates of the initial homonymous points in another orthophoto data; (1.3.2) Compare the difference between the first pixel coordinate and the second pixel coordinate with a preset difference threshold, and perform a first error point elimination process on the initial homonymous points contained in the grid based on the comparison results to obtain intermediate homonymous points.

[0083] Specifically, , represents the polynomial error for each initial homonymous point, , A polynomial error threshold is set for the pixel coordinates of another orthophoto data. In the embodiment of the present invention, the polynomial error threshold is set to 3, and initial same-name points exceeding the polynomial error threshold are considered to be error points and deleted.

[0084] .

[0085] Step 2: Using the average error ratio threshold algorithm, the middle homonymous points between two orthophoto data with geographical overlap are subjected to a second error point elimination process to obtain the target homonymous points.

[0086] In addition to using block geometric polynomials to delete homonymous points, in order to further ensure the accuracy of homonymous points, the embodiment of the present invention introduces the use of an average error multiplier threshold to remove error points. The basic core idea is that the geographic coordinate error values ​​of homonymous points between two orthophoto data are relatively uniform and fixed. This idea can be used to automatically count the middle homonymous points with particularly large geographic coordinate error values ​​as error points for removal. Specifically, it includes:

[0087] Step 2.1, determine the geographic coordinate error of the middle points with the same name between two orthophoto data with geographic overlap.

[0088] Specifically, the geographic coordinate errors of the common points of each orthophoto data with which it has geographical overlap are cyclically counted. , , rms, if there are N orthophoto data in total, numbered 1, 2, 3...N, and loop to orthophoto data 1, orthophoto data 1 has geographical overlap with orthophoto data 2, orthophoto data 3, and orthophoto data 5, then it is necessary to loop three times to perform error statistics on orthophoto data 1 and orthophoto data 2, orthophoto data 1 and orthophoto data 3, and orthophoto data 1 and orthophoto data 5 respectively. Take the calculation of the same-name point error between orthophoto data 1 and orthophoto data 2 as an example:

[0089] ;

[0090] , Represents the geographic coordinate value of orthophoto data 1, , is the geographic coordinate value of the orthophoto data 2, It is the final statistical geographic coordinate error value of the same-name point.

[0091] Step 2.2, based on the geographic coordinate errors of the middle points with the same name, determine the average geographic coordinate error between the two orthophoto data. In one example, the average geographic coordinate error of all the points with the same name between the two images is calculated. , the formula is as follows:

[0092] ;

[0093] Where M is the number of points with the same name between the two images.

[0094] Step 2.3, compare the geographic coordinate error with the average value of the geographic coordinate error of the specified magnification, and perform a second error point elimination process on the intermediate homonymous points between the two orthophoto data based on the comparison result to obtain the target homonymous points.

[0095] In one example, the average error multiplier threshold T is set. The embodiment of the present invention recommends setting the interval to be between 1.5 and 2.0. Greater than T times , then the point is considered to be a point where the geographic coordinate error is discontinuous, that is, an error point and is eliminated.

[0096] The embodiment of the present invention further provides a specific implementation of step S108, including the following steps a to c:

[0097] Step a: converting the geographical coordinates of the target homonymous point contained in any orthophoto data into the third pixel coordinates of the target homonymous point on another orthophoto data.

[0098] Circularly count the target same-name points of each orthophoto data and all the orthophoto data that overlap with it. For example, there are N target same-name points between orthophoto data 1 and orthophoto data 2. The geographic coordinates of these N target same-name points on orthophoto data 1 ( , ) is converted to the third pixel coordinate of orthophoto data 2 ( , ), the conversion relationship is as follows: , is the resolution of the orthophoto data 2 in the x and y directions, is the minimum x geographic coordinate of orthophoto data 2, The maximum y geographic coordinate of image 2.

[0099] .

[0100] Step b: determining the pixel error between the homonymous points based on the third pixel coordinate and the fourth pixel coordinate of the homonymous point matching the target homonymous point on another orthophoto data.

[0101] The target homonymous point is based on orthophoto data 1. The pixel coordinate error between orthophoto data 1 and orthophoto data 2 is As follows:

[0102]

[0103] in( , ) is the fourth pixel coordinate of the point in the orthophoto data 2 obtained by matching the target point with the same name.

[0104] Step c: determining the relative accuracy error between the orthophoto data and another orthophoto data based on the pixel errors between the same-name points.

[0105] Pixel error of all target same-name points between orthophoto data 1 and orthophoto data 2 After the calculation is completed, a mean error is calculated , represents the relative accuracy value between orthophoto data 1 and orthophoto data 2.

[0106] ;

[0107] The same method is used to calculate the relative accuracy mean error of orthophoto data 1 and other overlapping orthophoto data. … The relative accuracy mean error of orthophoto data 1 can also be calculated using the above formula If you want to calculate , N represents the number of all the same-name points in the orthophoto data 1.

[0108] Calculate the relative accuracy mean error of orthophoto data 1 Then, the mean error of orthophoto data 2…orthophoto data N can be calculated by the same method. … .

[0109] The embodiment of the present invention further provides a specific implementation of step S110, which requires exporting two relative accuracy quality inspection reports, one is a scene-by-scene relative accuracy quality inspection report, and the other is an overall relative accuracy quality inspection report.

[0110] In one example, a mean error threshold M0 and a relative error threshold ratio (rms_ratio) are set.

[0111] Case 1, such as Figure 2 The following are some indicators that need to be counted when exporting the relative accuracy quality inspection report of the scene segmentation: (a1) The mean error of the target homonymous points ; (a2) Total number of target corresponding points (N); (a3) Number of target corresponding points less than the medium error threshold M0 (N1); (a4) Number of target corresponding points less than twice the medium error threshold (N2) -- (M0 < RMS < 2*M0); (a5) Number of target corresponding points greater than twice the medium error threshold (C) -- (RMS > 2*M0); (a6) Maximum error (Max(Di)) (Explanation: Medium error of the target corresponding point with the largest error); (a7) Gross error rate (Cp) (Explanation: Number of error points divided by the total number of points (C / N * 100%)); (a8) Quality inspection result (Explanation: The qualified condition is that the gross error rate (Cp) < 5% and the medium error ( ) <= medium error threshold (M0).

[0112] Case 2, another schematic diagram of the relative accuracy quality inspection report for sub-scenes such as Figure 3 shown, circularly statistics the relative accuracy indicators of Image 1 and other overlapping images: (b1) Medium error of target corresponding points , representing the medium error of relative accuracy between the orthoimage data of the Mth scene and the orthoimage data of the Nth scene; (b2) Total number of target corresponding points (N); representing the number of target corresponding points between the orthoimage data of the Mth scene and the orthoimage data of the Nth scene; (b3) Number of target corresponding points less than the medium error threshold M0 (N1); (b4) Number of target corresponding points less than twice the medium error threshold (N2) -- (M0 < RMS < 2*M0); (b5) Number of target corresponding points greater than twice the medium error threshold (C) -- (RMS > 2*M0); (b6) Maximum error (Max(Di)) (Explanation: Medium error of the target corresponding point with the largest error); (b7) Gross error rate (Cp) (Explanation: Number of error points divided by the total number of points (C / N * 100%)); (b8) Quality inspection result (Explanation: The qualified condition is that the gross error rate (Cp) < 5% and the medium error ( ) <= medium error threshold (M0); (b9) Thumbnail of target corresponding points, the thumbnail of target corresponding points between every two scenes of images, the target corresponding points less than the medium error threshold M0 are shown in green, the target corresponding points less than twice the medium error threshold are shown in yellow, and the target corresponding points greater than twice the medium error threshold are shown in red. This can not only quantitatively evaluate the quality inspection result, but also visually evaluate which positions of the image have larger errors. In the embodiment of the present invention, the thumbnail represents the relative error degree of each corresponding point with different color levels according to the set relative geometric accuracy error threshold, and introduces the parameter of the percentage exceeding the threshold to assist in determining whether the relative geometric accuracy of the image to be inspected meets the standard.

[0113] The automatically generated internal accuracy quality inspection report of the image can clearly and quickly locate visually which places in the image have larger relative accuracy errors, and can also accurately give quantitatively how much the internal geometric distortion medium error of this scene of the image is and whether it is qualified.

[0114] Case 3, such as Figure 4 The figure shows a schematic diagram of an overall relative accuracy quality inspection report. The overall relative accuracy quality inspection report is exported. All orthophoto data will output an overall accuracy quality inspection report. The main statistical indicators are as follows: (c1) the number of orthophoto data (ALL_IMG_NUM); (c2) the total number of target homonymous points (ALL_N); (c3) the mean error of all target homonymous points (ALL_M); (c4) the overall quality inspection result (mean error (ALL_M) <= mean error threshold (M0); (c5) the number of scenes that passed the quality inspection (Qualified_NUM); (c6) the number of scenes that failed the quality inspection (Qualified_NUM); (c7) the quality inspection pass rate (Qualified_Ratio): Qualified_NUM / ALL_IMG_NUM; (c8) lists the relative mean error value of each orthophoto data, adjacent orthophoto data, whether it is qualified, and the single-scene quality inspection report path.

[0115] In summary, the overall flow chart of the relative accuracy quality inspection method between orthophotos provided by the embodiment of the present invention is as follows: Figure 5 As shown, the process includes orthophoto data preparation, automatic matching of same-name points, error point elimination, relative accuracy error calculation, and relative accuracy quality inspection report generation. Through the method proposed in the embodiment of the present invention, not only can the relative geometric accuracy of the image to be inspected be accurately captured visually, but also the relative accuracy value of each image to be inspected and all adjacent images can be quantitatively described to determine whether it has passed the quality inspection.

[0116] Based on the above-mentioned embodiment, the present invention provides a relative accuracy quality inspection device between orthophotos. Figure 6 The schematic diagram of the structure of a relative accuracy quality inspection device between orthophotos shown in FIG. 1 mainly includes the following parts:

[0117] An image acquisition module 602 is used to acquire orthophoto data of at least two scenes having a geographically overlapping relationship;

[0118] The same-name point matching module 604 is used to automatically match the same-name points on the orthophoto data to obtain initial same-name points with the same ground feature between the orthophoto data;

[0119] The error point elimination module 606 is used to use a block geometric polynomial model and an average error magnification threshold algorithm to perform multiple error point elimination processes on the initial homonymous points to obtain target homonymous points between orthophoto data;

[0120] The error determination module 608 is used to determine the relative accuracy error between the orthophoto data and the orthophoto data with which the orthophoto data has a geographical overlap, based on the target same-name points contained in any orthophoto data;

[0121] The report generation module 610 is used to generate a relative accuracy quality inspection report corresponding to each orthophoto data according to the relative accuracy error.

[0122] The above-mentioned device aims at the problem of automatic quality inspection of relative geometric accuracy between orthophoto data. After automatically matching the same-name points between the orthophoto data, the device automatically eliminates the wrong same-name points by combining the block geometric polynomial model method and the average error magnification threshold method, calculates the relative accuracy error between one scene of orthophoto data and another scene of orthophoto data, and then statistically calculates the relative accuracy error between each scene of orthophoto data and all orthophoto data with overlapping relationships, and automatically generates a quality inspection report for each scene of orthophoto data. The embodiment of the present invention can automatically and quantitatively realize the quality inspection of the relative geometric accuracy between orthophotos.

[0123] In one implementation, the error point elimination module 606 is specifically used to:

[0124] The orthophoto data is divided into grids, and the target geometric polynomial model corresponding to each grid is determined, so as to use the target geometric polynomial model to perform the first error point elimination process on the initial homonymous points contained in each grid to obtain the intermediate homonymous points;

[0125] The average error ratio threshold algorithm is used to perform a second error point elimination process on the middle homonymous points between two orthophoto data with geographical overlap to obtain the target homonymous points.

[0126] In one implementation, the error point elimination module 606 is specifically used to:

[0127] If the number of initial points of the same name contained in the grid is less than the first point number threshold, the initial points of the same name contained in the grid are not subjected to the first error point elimination process;

[0128] Alternatively, if the number of initial points of the same name contained in the grid is greater than the first point number threshold and less than or equal to the second point number threshold, the linear geometric polynomial model is determined as the target geometric polynomial model;

[0129] Alternatively, if the number of initial points of the same name contained in the grid is greater than a second point number threshold, the quadratic geometric polynomial model is determined as the target geometric polynomial model.

[0130] In one implementation, the error point elimination module 606 is specifically used to:

[0131] Using the target geometry polynomial model, based on the first pixel coordinates of the initial same-name points contained in the grid in the orthophoto data where they are located, determine the second pixel coordinates of the initial same-name points in another orthophoto data;

[0132] The difference between the first pixel coordinate and the second pixel coordinate is compared with a preset difference threshold, so as to perform a first error point elimination process on the initial same-name points contained in the grid based on the comparison result to obtain intermediate same-name points.

[0133] In one implementation, the error point elimination module 606 is specifically used to:

[0134] Determine the geographic coordinate error of the middle points with the same name between two orthophoto data with geographic overlap;

[0135] Determine the average geographic coordinate error between two orthophoto data based on the geographic coordinate error of the middle same-name point;

[0136] The geographic coordinate error is compared with the average value of the geographic coordinate error of the specified magnification, and the intermediate homonymous points between the two orthophoto data are subjected to a second error point elimination process based on the comparison result to obtain the target homonymous points.

[0137] In one implementation, the error determination module 608 is specifically configured to:

[0138] Convert the geographic coordinates of the target homonymous point contained in any orthophoto data into the third pixel coordinates of the target homonymous point on another orthophoto data;

[0139] Determine the pixel error between the homonymous points based on the third pixel coordinate and the fourth pixel coordinate of the homonymous point matched with the target homonymous point on the orthophoto data of another scene;

[0140] The relative accuracy error between the orthophoto data and another orthophoto data is determined based on the pixel error between the same-name points.

[0141] In one embodiment, the relative accuracy quality inspection report includes a scene-by-scene relative accuracy quality inspection report and an overall relative accuracy quality inspection report;

[0142] The relative accuracy quality inspection report for each scene includes a quality inspection report between any scene orthophoto data and all orthophoto data that have geographical overlap with it, and / or a quality inspection report between any scene orthophoto data and another scene orthophoto data that has geographical overlap with it;

[0143] The overall relative accuracy quality inspection report includes the quality inspection report between each orthophoto data.

[0144] The device provided in the embodiment of the present invention has the same implementation principle and technical effects as those of the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the device embodiment, reference may be made to the corresponding contents in the aforementioned method embodiment.

[0145] An embodiment of the present invention provides an electronic device. Specifically, the electronic device includes a processor and a storage device. The storage device stores a computer program, and when the computer program is executed by the processor, it executes the method described in any one of the above-mentioned implementation methods.

[0146] Figure 7 A structural diagram of an electronic device provided in an embodiment of the present invention, the electronic device 100 includes: a processor 70, a memory 71, a bus 72 and a communication interface 73, wherein the processor 70, the communication interface 73 and the memory 71 are connected via the bus 72; the processor 70 is used to execute an executable module stored in the memory 71, such as a computer program.

[0147] The memory 71 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 73 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0148] The bus 72 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 7 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0149] Among them, the memory 71 is used to store programs, and the processor 70 executes the program after receiving the execution instruction. The method executed by the device for flow process definition disclosed in any embodiment of the above-mentioned embodiment of the present invention can be applied to the processor 70 or implemented by the processor 70.

[0150] The processor 70 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 70. The above processor 70 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in the embodiment of the present invention can be directly embodied as a hardware decoding processor to execute, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 71, and the processor 70 reads the information in the memory 71 and completes the steps of the above method in combination with its hardware.

[0151] The computer program product of the readable storage medium provided in the embodiment of the present invention includes a computer-readable storage medium storing program code, and the instructions included in the program code can be used to execute the methods described in the previous method embodiments. The specific implementation can be referred to the previous method embodiments, which will not be repeated here.

[0152] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0153] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A relative accuracy quality inspection method between orthophotos, characterized in that: include: Obtain orthophoto data with at least two scenes having geographical overlap; Automatically matching the orthophoto data with the same points to obtain initial same points with the same ground features between the orthophoto data; Using a block geometric polynomial model and an average error magnification threshold algorithm, multiple error point elimination processes are performed on the initial homonymous points to obtain target homonymous points between the orthophoto data; Based on the target homonymous points contained in the orthophoto data of any scene, determining the relative accuracy error between the orthophoto data and the orthophoto data having a geographical overlap relationship therewith; Generate a relative accuracy quality inspection report corresponding to each scene of the orthophoto data according to the relative accuracy error; The block geometric polynomial model and the average error magnification threshold algorithm are used to perform multiple error point elimination processes on the initial homonymous points to obtain the target homonymous points between the orthophoto data, including: Dividing the orthophoto data into grids, and determining a target geometric polynomial model corresponding to each grid, so as to use the target geometric polynomial model to perform a first error point elimination process on the initial homonymous points contained in each grid to obtain intermediate homonymous points; The average error magnification threshold algorithm is used to perform a second error point elimination process on the middle homonymous points between two orthophoto data with geographical overlap to obtain the target homonymous points. The average error magnification threshold algorithm is used to perform a second error point elimination process on the middle homonymous points between the orthophoto data of the two scenes with geographical overlap to obtain the target homonymous points, including: Determine the geographic coordinate error of the middle same-name point between the orthophoto data of two views having a geographic overlap relationship; Determine the average geographic coordinate error between the orthophoto data of two scenes based on the geographic coordinate error of the middle same-name point; The geographic coordinate error is compared with the average value of the geographic coordinate error at a specified magnification, so as to perform a second error point elimination process on the intermediate same-name points between the two orthophoto image data based on the comparison result to obtain target same-name points.

2. The relative accuracy quality inspection method between orthophotos according to claim 1 is characterized in that: Determining a target geometric polynomial model corresponding to each of the grids comprises: If the number of the initial points of the same name contained in the square grid is less than the first point number threshold, the first error point elimination process is not performed on the initial points of the same name contained in the square grid; Alternatively, if the number of the initial points of the same name contained in the square grid is greater than the first point number threshold and less than or equal to the second point number threshold, the linear geometric polynomial model is determined as the target geometric polynomial model; Alternatively, if the number of the initial points of the same name contained in the grid is greater than the second point number threshold, the quadratic geometric polynomial model is determined as the target geometric polynomial model.

3. The relative accuracy quality inspection method between orthophotos according to claim 1 is characterized in that: Using the target geometric polynomial model, the initial same-name points contained in each of the square grids are subjected to a first error point elimination process to obtain intermediate same-name points, including: Determine the second pixel coordinates of the initial same-name points in the orthophoto data of another scene by using the target geometric polynomial model based on the first pixel coordinates of the initial same-name points contained in the grid in the orthophoto data where the initial same-name points are located; The difference between the first pixel coordinate and the second pixel coordinate is compared with a preset difference threshold, so as to perform a first error point elimination process on the initial same-name points contained in the grid based on the comparison result to obtain intermediate same-name points.

4. The relative accuracy quality inspection method between orthophotos according to claim 1 is characterized in that: Based on the target homonymous points contained in the orthophoto data of any scene, determining the relative accuracy error between the orthophoto data and the orthophoto data having a geographical overlap relationship therewith, comprises: Convert the geographic coordinates of the target homonymous point contained in the orthophoto data of any scene into the third pixel coordinates of the target homonymous point on the orthophoto data of another scene; Determine the pixel error between homonymous points based on the third pixel coordinate and the fourth pixel coordinate of the homonymous point matching the target homonymous point on the orthophoto data of another scene; The relative accuracy error between the orthophoto data and the orthophoto data of another scene is determined according to the pixel error between the same-name points.

5. The relative accuracy quality inspection method between orthophotos according to claim 1 is characterized in that: The relative accuracy quality inspection report includes a scene-by-scene relative accuracy quality inspection report and an overall relative accuracy quality inspection report; The relative accuracy quality inspection report for each scene includes a quality inspection report between the orthophoto data of any scene and all the orthophoto data having a geographical overlap relationship with it, and / or a quality inspection report between the orthophoto data of any scene and the orthophoto data of another scene having a geographical overlap relationship with it; The overall relative accuracy quality inspection report includes a quality inspection report between the orthophoto data of each scene.

6. A relative accuracy quality inspection device between orthophotos, characterized in that: include: An image acquisition module is used to acquire orthophoto data of at least two scenes having a geographically overlapping relationship; A homonymous point matching module is used to automatically match homonymous points on the orthophoto data to obtain initial homonymous points with the same ground feature between the orthophoto data; An error point elimination module is used to perform multiple error point elimination processes on the initial homonymous points by using a block geometric polynomial model and an average error magnification threshold algorithm to obtain target homonymous points between the orthophoto data; An error determination module, for determining a relative accuracy error between the orthophoto data and the orthophoto data having a geographical overlap relationship therewith based on the target homonymous points contained in the orthophoto data of any scene; A report generation module, used for generating a relative accuracy quality inspection report corresponding to each scene of the orthophoto data according to the relative accuracy error; The error point elimination module is specifically used for: Dividing the orthophoto data into grids, and determining a target geometric polynomial model corresponding to each grid, so as to use the target geometric polynomial model to perform a first error point elimination process on the initial homonymous points contained in each grid to obtain intermediate homonymous points; The average error magnification threshold algorithm is used to perform a second error point elimination process on the middle homonymous points between two orthophoto data with geographical overlap to obtain the target homonymous points. The error point elimination module is specifically used for: Determine the geographic coordinate error of the middle same-name point between the orthophoto data of two views having a geographic overlap relationship; Determine the average value of the geographic coordinate error between the orthophoto data of two scenes based on the geographic coordinate error of the middle same-name point; The geographic coordinate error is compared with the average value of the geographic coordinate error at a specified magnification, so as to perform a second error point elimination process on the intermediate same-name points between the two orthophoto image data based on the comparison result to obtain target same-name points.

7. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement the method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to implement the method according to any one of claims 1 to 5.