CN114894164B - Oblique image matching screening method and system - Google Patents
Oblique image matching screening method and system Download PDFInfo
- Publication number
- CN114894164B CN114894164B CN202210366430.XA CN202210366430A CN114894164B CN 114894164 B CN114894164 B CN 114894164B CN 202210366430 A CN202210366430 A CN 202210366430A CN 114894164 B CN114894164 B CN 114894164B
- Authority
- CN
- China
- Prior art keywords
- horizontal distance
- track
- track points
- navigation
- angle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 62
- 238000012216 screening Methods 0.000 title claims abstract description 62
- 238000004364 calculation method Methods 0.000 claims abstract description 29
- 238000013507 mapping Methods 0.000 claims description 57
- 238000010276 construction Methods 0.000 claims description 20
- 238000005259 measurement Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 abstract description 9
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C11/00—Photogrammetry or videogrammetry, e.g. stereogrammetry; Photographic surveying
- G01C11/04—Interpretation of pictures
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
The invention provides a matching screening method for inclined images, which relates to the technical field of aerial survey, wherein the horizontal distance between adjacent track points of the same navigation belt and the horizontal distance between the track points of the adjacent navigation belt are determined through calculation, one track point is taken as an ellipse center, and an ellipse half axis is respectively determined through the horizontal distance between the adjacent track points of the same navigation belt and the horizontal distance between the track points of the adjacent navigation belt, so that an ellipse standard equation is constructed; and (3) performing image matching screening by judging the position relation between the rest track points and the ellipse, wherein if the track points are in the elliptical area, the images of the track points and the images of the track points in the center of the ellipse can form matching pairs. The invention also provides a matching and screening system for the inclined images, which utilizes the elliptical area to screen, and other images form matching pairs with the current image if the images are in the elliptical area, so that the matching and screening of the inclined images is realized, the number of the image matching pairs can be effectively reduced after screening, and the later image matching processing time is reduced.
Description
Technical Field
The invention relates to the technical field of aerial survey, in particular to an inclined image matching screening method and system.
Background
At present, unmanned aerial vehicles are used for aerial survey, and technical methods for producing three-dimensional models by utilizing shot images for three-dimensional reconstruction are becoming popular and widely used. Along with the better performance of computer hardware, the data size of the aerial survey photographed image is larger and larger, tens of thousands or even hundreds of thousands of images are photographed, according to the principle of aerial triangulation, the homonym points are obtained according to the matching relation between the images, and the image matching can be usually calculated through a SIFT algorithm, as the low-altitude multi-view remote sensing image matching method based on an improved SIFT operator disclosed by the prior art, but because the photographed images are too many, the images are matched one by one through the SIFT algorithm, the time is too long and the efficiency is low. According to the principle of aerial survey, a certain degree of overlapping between the heading and the sideways direction is required, so that only the image within a certain range of the position where the shot image is located has an overlapping part and can be matched. Therefore, the track point positions recorded by the images can be screened, then the images are accurately matched through sift and other algorithms, and the image matching processing time is shortened.
The conventional method is to screen images by manually arranging the navigation belts, which is not in line with full automation. In addition, the method also uses a method of calculating the projected quadrangle of each image on the ground to screen, but the method needs to know the average ground height of the current area, and the average ground height of the current area is not necessarily obtained.
Disclosure of Invention
The invention aims to solve at least one technical defect, provides a screening method and a screening system for matching inclined images, and aims to screen matching inclined images, so that the number of image matching pairs can be effectively reduced after screening, and the later image matching processing time is shortened.
In order to solve the technical problems, the technical scheme of the invention is as follows:
a tilt image matching screening method comprises the following steps:
s1: acquiring a navigation measurement track point file, and recording all track points and corresponding image information thereof as one piece of data into an original data list;
s2: calculating horizontal distance values of every two track points according to track point information of an original data list, and constructing a track mapping table;
s3: calculating the horizontal distance between adjacent track points of the same navigation band based on a track mapping table by adopting a method of calculating a histogram through statistics;
s4: constructing an angle list based on the track mapping table, and calculating the course angle of the navigation belt;
s5: judging according to the track mapping table and the horizontal distance between adjacent track points of the same navigation belt, dividing all track points into different navigation belts, and constructing a navigation belt code list;
s6: calculating the horizontal distance between one navigation belt code and the other navigation belt code in the navigation belt code list, and obtaining the horizontal distance of the track point of the adjacent navigation belt by adopting a method of calculating a histogram by statistics;
S7: reading a track point from an original data list, taking the track point as an ellipse center, and determining the sizes of two half shafts of the ellipse by the horizontal distance between adjacent track points of the same navigation belt and the horizontal distance between adjacent navigation belt track points respectively to obtain a standard equation of the ellipse;
s8: calculating the rest track points in the original data list according to the standard equation of the ellipse to obtain track points in the ellipse area, wherein the images of the track points in the ellipse area and the images of the track points in the center of the ellipse can form matching pairs;
s9: and repeating the steps S7-S8 until all the image matching pairs are calculated and screened out, outputting all the image matching pairs, and completing the inclined image matching screening process.
Because the image matching is usually calculated by adopting an algorithm such as sift and the like, if images are not screened and are matched pairwise, the number of image matching pairs formed is huge, so that the scheme aims at screening matching of inclined images, the number of image matching pairs is reduced after screening, and the later image matching processing time is shortened.
According to the scheme, the method for matching and screening the images shot by the unmanned aerial vehicle at different positions in the air does not need to divide the navigation belt manually or acquire the ground height of the current area, the related information of the navigation belt can be estimated automatically only by the position information of the track points recorded by the images, an elliptical area is formed at the track points of the current image, screening is carried out by utilizing the elliptical area, and other images form matching pairs with the current image in the elliptical area, so that matching and screening of inclined images is realized, the number of image matching pairs can be effectively reduced after screening, and the later image matching processing time is shortened.
Wherein, the step S2 specifically comprises the following substeps:
s21: reading information of every two track points one by one from an original data list, and calculating horizontal distance values of every two track points;
s22: setting a distance threshold, judging whether the horizontal distance value of each two track points is larger than the set distance threshold, if so, calculating the angle values of the two track points corresponding to the horizontal distance value;
s23: and (3) after all the horizontal distance values and the angle values are rounded downwards, taking every two track point information and the corresponding horizontal distance values and angle values as a record to construct a track mapping table.
The step S3 specifically includes: and reading the horizontal distance values recorded in the track mapping table one by one, accumulating the same number of the horizontal distance values to obtain a distance histogram, and taking the horizontal distance value with the largest number in the distance histogram as the horizontal distance between adjacent track points of the same navigation belt.
Wherein, the step S4 specifically comprises the following substeps:
s41: reading the horizontal distance values recorded in the track mapping table one by one, judging whether the absolute value of the difference value between each horizontal distance value and the horizontal distance value of the adjacent track point of the same navigation belt is smaller than a set threshold value, and if yes, recording the angle value corresponding to the horizontal distance value into an angle list;
S42: and reading the angle values recorded in the angle list one by one, accumulating the number of the same angle values to obtain an angle histogram, and taking the angle with the largest number in the angle histogram as the course angle of the navigation belt.
Wherein, the step S5 specifically comprises the following substeps:
s51: reading the horizontal distance value and the angle value recorded in the track mapping table one by one, and judging the two track points as the same navigation belt if two track points exist, wherein the absolute value of the difference value between the horizontal distance value and the horizontal distance value of the adjacent track points of the same navigation belt is smaller than a set threshold value, and the absolute value of the difference value between the corresponding angle value of the horizontal distance value and the heading angle of the navigation belt is smaller than the set threshold value;
s52: constructing a navigation belt code list, searching in the navigation belt code list by taking the track points as keywords, if the track points exist, recording the track points on sub-items of the same navigation belt code list, and if the track points do not exist, recording the track points in the navigation belt code list as new codes;
s53: and repeating the steps S51-S52 until all the track points are divided into different navigation bands, and obtaining a navigation band code list.
Wherein, the step S6 specifically comprises the following substeps:
s61: calculating the horizontal distance between each navigation belt code and the other navigation belt code in the navigation belt code list one by one, namely calculating the horizontal distance value between each track point of one navigation belt code and each track point of the other navigation belt code, and recording all the horizontal distance values into the distance list;
S62: and reading the horizontal distance values recorded in the distance list one by one, accumulating the number of the same horizontal distance values to obtain a ribbon horizontal distance histogram, and taking the horizontal distance value with the largest number in the ribbon horizontal distance histogram as the horizontal distance of the adjacent ribbon track point.
The step S8 specifically includes the following steps:
s81: the rest track points in the original data list are rotated anticlockwise, and the rotation angle is the course angle of the navigation belt;
s82: and (3) taking the rotated track point coordinates into a standard equation of an ellipse to calculate, wherein if the values of the track point coordinates are less than 1, the track point coordinates are represented to be in the elliptical region, and an image of the track point in the elliptical region and an image of the track point in the center of the ellipse can form a matching pair.
More specifically, the scheme also provides an inclined image matching screening system for realizing an inclined image matching screening method, which specifically comprises an original data acquisition module, a track mapping table construction module, an adjacent track point horizontal distance calculation module, a course angle calculation module, a navigation belt code list construction module, an adjacent navigation belt track point horizontal distance calculation module, an elliptic standard equation construction module, an image matching module, a circulation judgment module and a result output module; wherein:
The original data acquisition module is used for acquiring a navigation survey track point file, and recording all track points and corresponding image information thereof as one piece of data into an original data list;
the track mapping table construction module is used for calculating horizontal distance values of every two track points according to track point information of the original data list and constructing a track mapping table;
the adjacent track point horizontal distance calculation module is used for calculating the adjacent track point horizontal distance of the same navigation band based on a track mapping table by adopting a statistical histogram calculation method;
the course angle calculation module is used for constructing an angle list based on the track mapping table and calculating the course angle of the navigation belt;
the navigation belt code list construction module is used for judging the horizontal distance between adjacent track points of the same navigation belt according to the track mapping table, dividing all the track points into different navigation belts, and constructing a navigation belt code list;
the horizontal distance calculation module of the adjacent ribbon track points is used for calculating the horizontal distance between one ribbon code and the other ribbon code in the ribbon code list, and obtaining the horizontal distance of the adjacent ribbon track points by adopting a method of calculating a histogram through statistics;
the ellipse standard equation building module is used for reading a track point from the original data list, taking the track point as an ellipse center, and determining the sizes of two half shafts of the ellipse by the horizontal distance between adjacent track points of the same navigation belt and the horizontal distance between adjacent navigation belt track points respectively, so as to build a standard equation of the ellipse;
The image matching module is used for calculating the rest track points in the original data list according to an elliptic standard equation to obtain track points in an elliptic region, and the images of the track points in the elliptic region and the images of the track points in the center of the ellipse can form matching pairs;
the circulation judging module is used for judging whether the screening of all the image matching pairs is finished, if yes, the result output module outputs all the image matching pairs to finish the inclined image matching screening process; otherwise, the elliptic standard equation building module reselects the track points and continues to carry out image matching screening.
Wherein, in the track mapping table construction module, the following processes are specifically executed: reading information of every two track points one by one from an original data list, and calculating horizontal distance values of every two track points; setting a distance threshold, judging whether the horizontal distance value of each two track points is larger than the set distance threshold, if so, calculating the angle values of the two track points corresponding to the horizontal distance value; after all the horizontal distance values and angle values are rounded downwards, taking every two track point information and the corresponding horizontal distance values and angle values as a record to construct a track mapping table;
In the adjacent track point horizontal distance calculation module, the following process is specifically executed, horizontal distance values recorded in a track mapping table are read one by one, the number of the same horizontal distance values is accumulated, a distance histogram is obtained, and the horizontal distance value with the largest number in the distance histogram is used as the horizontal distance of the adjacent track points of the same navigation belt;
in the course angle calculation module, the following processes are specifically executed: reading the horizontal distance values recorded in the track mapping table one by one, judging whether the absolute value of the difference value between each horizontal distance value and the horizontal distance value of the adjacent track point of the same navigation belt is smaller than a set threshold value, and if yes, recording the angle value corresponding to the horizontal distance value into an angle list; and reading the angle values recorded in the angle list one by one, accumulating the number of the same angle values to obtain an angle histogram, and taking the angle with the largest number in the angle histogram as the course angle of the navigation belt.
In the navigation belt code list construction module, the following process is specifically executed: reading the horizontal distance value and the angle value recorded in the track mapping table one by one, and judging the two track points as the same navigation belt if two track points exist, wherein the absolute value of the difference value between the horizontal distance value and the horizontal distance value of the adjacent track points of the same navigation belt is smaller than a set threshold value, and the absolute value of the difference value between the corresponding angle value of the horizontal distance value and the heading angle of the navigation belt is smaller than the set threshold value; constructing a navigation belt code list, searching in the navigation belt code list by taking the track points as keywords, if the track points exist, recording the track points on sub-items of the same navigation belt code list, and if the track points do not exist, recording the track points in the navigation belt code list as new codes; repeating the operation until all the track points are divided into different navigation bands, and obtaining a navigation band code list;
In the adjacent navigation belt track point horizontal distance calculating module, the following process is specifically executed: calculating the horizontal distance between each navigation belt code and the other navigation belt code in the navigation belt code list one by one, namely calculating the horizontal distance value between each track point of one navigation belt code and each track point of the other navigation belt code, and recording all the horizontal distance values into the distance list; reading the horizontal distance values recorded in the distance list one by one, accumulating the number of the same horizontal distance values to obtain a ribbon horizontal distance histogram, and taking the horizontal distance value with the largest number in the ribbon horizontal distance histogram as the horizontal distance of the adjacent ribbon track point;
in the image matching module, the following process is specifically executed: the rest track points in the original data list are rotated anticlockwise, and the rotation angle is the course angle of the navigation belt; and (3) taking the rotated track point coordinates into a standard equation of an ellipse to calculate, wherein if the values of the track point coordinates are less than 1, the track point coordinates are represented to be in the elliptical region, and an image of the track point in the elliptical region and an image of the track point in the center of the ellipse can form a matching pair.
Compared with the prior art, the technical scheme of the invention has the beneficial effects that:
The invention provides a matching screening method and a matching screening system for oblique images, which do not need to divide a navigation belt manually or acquire the ground height of a current area, can automatically estimate the relevant information of the navigation belt only by the position information of track points recorded by images, form an elliptical area at the track points of the current image, screen by using the elliptical area, and form matching pairs with the current image if other images form matching pairs in the elliptical area, so that the matching screening of the oblique images is realized, the number of the image matching pairs can be effectively reduced after screening, and the later image matching processing time is reduced.
Drawings
FIG. 1 is a flow chart of an oblique image matching screening method according to the present invention;
FIG. 2 is a schematic diagram of a navigation track point in an embodiment of the present invention;
FIG. 3 is a schematic view of a portion of a navigation track point according to an embodiment of the present invention.
Detailed Description
The drawings are for illustrative purposes only and are not to be construed as limiting the present patent;
for the purpose of better illustrating the embodiments, certain elements of the drawings may be omitted, enlarged or reduced and do not represent the actual product dimensions;
it will be appreciated by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.
The technical scheme of the invention is further described below with reference to the accompanying drawings and examples.
Example 1
The embodiment provides a method for matching and screening oblique images, as shown in fig. 1, comprising the following steps:
s1: acquiring a navigation measurement track point file, and recording all track points and corresponding image information thereof as one piece of data into an original data list;
s2: calculating horizontal distance values of every two track points according to track point information of an original data list, and constructing a track mapping table;
s3: calculating the horizontal distance between adjacent track points of the same navigation band based on a track mapping table by adopting a method of calculating a histogram through statistics;
s4: constructing an angle list based on the track mapping table, and calculating the course angle of the navigation belt;
s5: judging according to the track mapping table and the horizontal distance between adjacent track points of the same navigation belt, dividing all track points into different navigation belts, and constructing a navigation belt code list;
s6: calculating the horizontal distance between one navigation belt code and the other navigation belt code in the navigation belt code list, and obtaining the horizontal distance of the track point of the adjacent navigation belt by adopting a method of calculating a histogram by statistics;
s7: reading a track point from an original data list, taking the track point as an ellipse center, and determining the sizes of two half shafts of the ellipse by the horizontal distance between adjacent track points of the same navigation belt and the horizontal distance between adjacent navigation belt track points respectively to obtain a standard equation of the ellipse;
S8: calculating the rest track points in the original data list according to the standard equation of the ellipse to obtain track points in the ellipse area, wherein the images of the track points in the ellipse area and the images of the track points in the center of the ellipse can form matching pairs;
s9: and repeating the steps S7-S8 until all the image matching pairs are calculated and screened out, outputting all the image matching pairs, and completing the inclined image matching screening process.
In the implementation process of the step S7, 2.5 times of the horizontal distance between adjacent track points of the same navigation belt and 1.5 times of the horizontal distance between the track points of the adjacent navigation belt are taken as two half shafts of the ellipse.
More specifically, step S2 specifically includes the following sub-steps:
s21: reading information of every two track points one by one from an original data list, and calculating horizontal distance values of every two track points;
s22: setting a distance threshold, such as 1 cm, wherein the threshold is used for filtering a plurality of groups of images with different angles shot by oblique shooting, the track points are similar in position, judging whether the horizontal distance value of each two track points is larger than the set distance threshold, if so, calculating the angle value of the two track points corresponding to the horizontal distance value, wherein the angle value is the included angle value between a straight line formed by the two track points and an X axis;
S23: and (3) after all the horizontal distance values and the angle values are rounded downwards, taking every two track point information and the corresponding horizontal distance values and angle values as a record to construct a track mapping table.
More specifically, step S3 specifically includes: and reading the horizontal distance values recorded in the track mapping table one by one, and accumulating the same number of the horizontal distance values to obtain a distance histogram with the horizontal distance values as horizontal axes and the number as vertical axes, and taking the horizontal distance value with the largest number in the distance histogram as the horizontal distance of the adjacent track points of the same navigation belt.
More specifically, step S4 specifically includes the following sub-steps:
s41: reading the horizontal distance values recorded in the track mapping table one by one, judging whether the absolute value of the difference value between each horizontal distance value and the horizontal distance value of the adjacent track point of the same navigation belt is smaller than a set threshold (for example, 1 meter), and if yes, recording the angle value corresponding to the horizontal distance value into an angle list;
s42: and reading the angle values recorded in the angle list one by one, accumulating the number of the same angle values to obtain an angle histogram, and taking the angle with the largest number in the angle histogram as the course angle of the navigation belt.
More specifically, step S5 specifically includes the following sub-steps:
s51: reading the horizontal distance value and the angle value recorded in the track mapping table one by one, and judging the two track points as the same navigation belt if two track points exist, wherein the absolute value of the difference between the horizontal distance value and the horizontal distance value of the adjacent track points of the same navigation belt is smaller than a set threshold value (for example, 1 meter), and the absolute value of the difference between the angle value corresponding to the horizontal distance value and the heading angle of the navigation belt is smaller than the set threshold value (for example, 3 degrees);
s52: constructing a navigation belt code list, searching in the navigation belt code list by taking the track points as keywords, if the track points exist, recording the track points on sub-items of the same navigation belt code list, and if the track points do not exist, recording the track points in the navigation belt code list as new codes;
s53: and repeating the steps S51-S52 until all the track points are divided into different navigation bands, and obtaining a navigation band code list.
More specifically, step S6 specifically includes the following sub-steps:
s61: calculating the horizontal distance between each navigation belt code and the other navigation belt code in the navigation belt code list one by one, namely calculating the horizontal distance value between each track point of one navigation belt code and each track point of the other navigation belt code, and recording all the horizontal distance values into the distance list;
S62: and reading the horizontal distance values recorded in the distance list one by one, accumulating the number of the same horizontal distance values to obtain a ribbon horizontal distance histogram, and taking the horizontal distance value with the largest number in the ribbon horizontal distance histogram as the horizontal distance of the adjacent ribbon track point.
More specifically, step S8 specifically includes the steps of:
s81: the rest track points in the original data list are rotated anticlockwise, and the rotation angle is the course angle of the navigation belt;
s82: and (3) taking the rotated track point coordinates into a standard equation of an ellipse to calculate, wherein if the values of the track point coordinates are less than 1, the track point coordinates are represented to be in the elliptical region, and an image of the track point in the elliptical region and an image of the track point in the center of the ellipse can form a matching pair.
Because image matching is usually calculated by adopting an algorithm such as sift, if images need to be matched two by two, the number of image matching pairs formed in this way is huge, so that the embodiment aims to perform matching screening on inclined images, reduce the number of image matching pairs after screening, and reduce the later image matching processing time.
In a specific implementation process, the embodiment is a method for matching and screening images shot by an unmanned aerial vehicle at different positions in the air, which does not need to divide a navigation belt manually or acquire the ground height of a current area, can automatically estimate relevant information of the navigation belt only by position information of track points recorded by the images, forms an elliptical area at the track points of the current image, screens other images if the images form matching pairs with the current image in the elliptical area, realizes the screening of matching the inclined images, can effectively reduce the number of image matching pairs after screening, and reduces the later image matching processing time.
Example 2
More specifically, on the basis of embodiment 1, the present embodiment provides an oblique image matching and screening system, which is configured to implement an oblique image matching and screening method, and specifically includes an original data acquisition module, a track mapping table construction module, an adjacent track point horizontal distance calculation module, a course angle calculation module, a navigation belt code list construction module, an adjacent navigation belt track point horizontal distance calculation module, an ellipse standard equation construction module, an image matching module, a circulation judgment module, and a result output module; wherein:
the original data acquisition module is used for acquiring a navigation survey track point file, and recording all track points and corresponding image information thereof as one piece of data into an original data list;
the track mapping table construction module is used for calculating horizontal distance values of every two track points according to track point information of the original data list and constructing a track mapping table;
the adjacent track point horizontal distance calculation module is used for calculating the adjacent track point horizontal distance of the same navigation band based on a track mapping table by adopting a statistical histogram calculation method;
the course angle calculation module is used for constructing an angle list based on the track mapping table and calculating the course angle of the navigation belt;
The navigation belt code list construction module is used for judging the horizontal distance between adjacent track points of the same navigation belt according to the track mapping table, dividing all the track points into different navigation belts, and constructing a navigation belt code list;
the horizontal distance calculation module of the adjacent ribbon track points is used for calculating the horizontal distance between one ribbon code and the other ribbon code in the ribbon code list, and obtaining the horizontal distance of the adjacent ribbon track points by adopting a method of calculating a histogram through statistics;
the ellipse standard equation building module is used for reading a track point from the original data list, taking the track point as an ellipse center, and determining the sizes of two half shafts of the ellipse by the horizontal distance between adjacent track points of the same navigation belt and the horizontal distance between adjacent navigation belt track points respectively, so as to build a standard equation of the ellipse;
the image matching module is used for calculating the rest track points in the original data list according to an elliptic standard equation to obtain track points in an elliptic region, and the images of the track points in the elliptic region and the images of the track points in the center of the ellipse can form matching pairs;
the circulation judging module is used for judging whether the screening of all the image matching pairs is finished, if yes, the result output module outputs all the image matching pairs to finish the inclined image matching screening process; otherwise, the elliptic standard equation building module reselects the track points and continues to carry out image matching screening.
More specifically, in the track mapping table construction module, the following process is specifically executed: reading information of every two track points one by one from an original data list, and calculating horizontal distance values of every two track points; setting a distance threshold, judging whether the horizontal distance value of each two track points is larger than the set distance threshold, if so, calculating the angle values of the two track points corresponding to the horizontal distance value; after all the horizontal distance values and angle values are rounded downwards, taking every two track point information and the corresponding horizontal distance values and angle values as a record to construct a track mapping table;
in the adjacent track point horizontal distance calculation module, the following process is specifically executed, horizontal distance values recorded in a track mapping table are read one by one, the number of the same horizontal distance values is accumulated, a distance histogram is obtained, and the horizontal distance value with the largest number in the distance histogram is used as the horizontal distance of the adjacent track points of the same navigation belt;
in the course angle calculation module, the following processes are specifically executed: reading the horizontal distance values recorded in the track mapping table one by one, judging whether the absolute value of the difference value between each horizontal distance value and the horizontal distance value of the adjacent track point of the same navigation belt is smaller than a set threshold value, and if yes, recording the angle value corresponding to the horizontal distance value into an angle list; and reading the angle values recorded in the angle list one by one, accumulating the number of the same angle values to obtain an angle histogram, and taking the angle with the largest number in the angle histogram as the course angle of the navigation belt.
More specifically, in the navigation belt code list construction module, the following process is specifically executed: reading the horizontal distance value and the angle value recorded in the track mapping table one by one, and judging the two track points as the same navigation belt if two track points exist, wherein the absolute value of the difference value between the horizontal distance value and the horizontal distance value of the adjacent track points of the same navigation belt is smaller than a set threshold value, and the absolute value of the difference value between the corresponding angle value of the horizontal distance value and the heading angle of the navigation belt is smaller than the set threshold value; constructing a navigation belt code list, searching in the navigation belt code list by taking the track points as keywords, if the track points exist, recording the track points on sub-items of the same navigation belt code list, and if the track points do not exist, recording the track points in the navigation belt code list as new codes; repeating the operation until all the track points are divided into different navigation bands, and obtaining a navigation band code list;
in the adjacent navigation belt track point horizontal distance calculating module, the following process is specifically executed: calculating the horizontal distance between each navigation belt code and the other navigation belt code in the navigation belt code list one by one, namely calculating the horizontal distance value between each track point of one navigation belt code and each track point of the other navigation belt code, and recording all the horizontal distance values into the distance list; reading the horizontal distance values recorded in the distance list one by one, accumulating the number of the same horizontal distance values to obtain a ribbon horizontal distance histogram, and taking the horizontal distance value with the largest number in the ribbon horizontal distance histogram as the horizontal distance of the adjacent ribbon track point;
In the image matching module, the following process is specifically executed: the rest track points in the original data list are rotated anticlockwise, and the rotation angle is the course angle of the navigation belt; and (3) taking the rotated track point coordinates into a standard equation of an ellipse to calculate, wherein if the values of the track point coordinates are less than 1, the track point coordinates are represented to be in the elliptical region, and an image of the track point in the elliptical region and an image of the track point in the center of the ellipse can form a matching pair.
The embodiment provides an oblique image matching and screening system, which does not need to divide a navigation belt manually or acquire the ground height of a current area, can automatically estimate relevant information of the navigation belt only by the position information of track points recorded by images, forms an elliptical area at the track points of the current image, screens other images by using the elliptical area, and forms matching pairs with the current image if the other images are in the elliptical area, so that the screening of matching of oblique images is realized, the number of image matching pairs can be effectively reduced after screening, and the later image matching processing time is reduced.
Example 3
The technical principle of the scheme is further described in the embodiment, and the technical principle is specifically described as follows:
First, unmanned aerial vehicle aerial survey flies by aerial belt, as shown in fig. 2, but aerial belt is not necessarily regular due to factors such as flight environment. For a navigation belt, the effect of actual flight is likely to be not a straight line, that is, the track points do not meet the straight line equation, and only a straight line can be fitted, so that the core of automatic navigation belt division can realize automatic division by calculating the course angle of the track points, and the course angle of the track points is within a set error range and can be considered to be in the same navigation belt. Therefore, the scheme realizes automatic division of the navigation belt by calculating the course angle of the track point.
Secondly, according to the principle requirement of aerial survey, a certain overlapping degree is required between the heading and the sideways, the overlapping degree of the heading (generally more than 60%) is larger than the overlapping degree of the sideways (generally more than 30%), namely the horizontal distance between two adjacent track points of the same navigation belt is smaller than the horizontal distance between the adjacent navigation belts (as shown in fig. 3, two track points A and B are on the same navigation belt, and A and C are on different navigation belts). And according to the requirement of the overlapping degree, the method of adopting an elliptical area is determined to screen, and the areas with other shapes cannot simultaneously satisfy the two conditions.
Thirdly, because unmanned aerial vehicle aerial survey is shot every set time, track points which are positioned between two zones and turn around and track points which are positioned at the back of the unmanned aerial vehicle appear, the track points have influence on the calculation of the course angle of the zone, the horizontal distance of adjacent track points of the same zone and the horizontal distance of the adjacent zone track points, but the number of the track points with influence is small, so that the values can be determined by a method of calculating the histogram of the track points, and the influence is eliminated.
Fourth, as can be seen from the overlapping degree of aerial survey, the captured image is actually overlapped with the image around the track point only, and there is a possibility of successful matching, so the track point (e.g. point a in fig. 3) is taken as the ellipse circle center, the size of the long half axis can be set to be 2.5 times of the horizontal distance (e.g. the horizontal distance of two points AB in fig. 3) between two adjacent track points of the same navigation belt, the size of the short half axis can be set to be 1.5 times of the horizontal distance (e.g. the horizontal distance of two points AC in fig. 3) between two adjacent navigation belt, so that the matching screening principle is that other track points are located in the ellipse area of the ellipse circle center track point (e.g. point a in fig. 3), and the image of the image and the image of the ellipse circle center track point are possible to be successfully matched, so as to form an image matching pair.
Finally, because of oblique photography, a plurality of groups of images (generally 5 groups) with different angles are photographed at one track point (shown as a point A in fig. 3), and track points recorded by the images with different angles are equal or have small phase difference, the scheme calculates the course angle of the navigation belt, and filters the equal or small phase difference (a threshold value can be set by software) before the horizontal distance between the adjacent track points of the same navigation belt and the horizontal distance between the adjacent track points of the navigation belt.
According to the scheme, the navigation belt is not required to be divided manually, the ground height of the current area is not required to be acquired, the relevant information of the navigation belt can be estimated automatically only by the position information of the track points recorded by the images, an elliptical area is formed at the track points of the current image, screening is carried out by utilizing the elliptical area, and if other images form matching pairs with the current image in the elliptical area, screening for matching the inclined images is realized.
It is to be understood that the above examples of the present invention are provided by way of illustration only and not by way of limitation of the embodiments of the present invention. Other variations or modifications of the above teachings will be apparent to those of ordinary skill in the art. It is not necessary here nor is it exhaustive of all embodiments. Any modification, equivalent replacement, improvement, etc. which come within the spirit and principles of the invention are desired to be protected by the following claims.
Claims (10)
1. The inclined image matching screening method is characterized by comprising the following steps of:
s1: acquiring a navigation measurement track point file, and recording all track points and corresponding image information thereof as one piece of data into an original data list;
s2: calculating horizontal distance values of every two track points according to track point information of an original data list, and constructing a track mapping table;
s3: calculating the horizontal distance between adjacent track points of the same navigation band based on a track mapping table by adopting a method of calculating a histogram through statistics;
s4: constructing an angle list based on the track mapping table, and calculating the course angle of the navigation belt;
s5: judging according to the track mapping table and the horizontal distance between adjacent track points of the same navigation belt, dividing all track points into different navigation belts, and constructing a navigation belt code list;
s6: calculating the horizontal distance between one navigation belt code and the other navigation belt code in the navigation belt code list, and obtaining the horizontal distance of the track point of the adjacent navigation belt by adopting a method of calculating a histogram by statistics;
s7: reading a track point from an original data list, taking the track point as an ellipse center, and determining the sizes of two half shafts of the ellipse by the horizontal distance between adjacent track points of the same navigation belt and the horizontal distance between adjacent navigation belt track points respectively to obtain a standard equation of the ellipse;
S8: calculating the rest track points in the original data list according to the standard equation of the ellipse to obtain track points in the ellipse area, wherein the images of the track points in the ellipse area and the images of the track points in the center of the ellipse can form matching pairs;
s9: and repeating the steps S7-S8 until all the image matching pairs are calculated and screened out, outputting all the image matching pairs, and completing the inclined image matching screening process.
2. The oblique image matching screening method as claimed in claim 1, wherein the step S2 specifically includes the following sub-steps:
s21: reading information of every two track points one by one from an original data list, and calculating horizontal distance values of every two track points;
s22: setting a distance threshold, judging whether the horizontal distance value of each two track points is larger than the set distance threshold, if so, calculating the angle values of the two track points corresponding to the horizontal distance value;
s23: and (3) after all the horizontal distance values and the angle values are rounded downwards, taking every two track point information and the corresponding horizontal distance values and angle values as a record to construct a track mapping table.
3. The oblique image matching screening method as claimed in claim 2, wherein step S3 specifically comprises: and reading the horizontal distance values recorded in the track mapping table one by one, accumulating the same number of the horizontal distance values to obtain a distance histogram, and taking the horizontal distance value with the largest number in the distance histogram as the horizontal distance between adjacent track points of the same navigation belt.
4. A method of screening for matching oblique images according to claim 3, wherein step S4 comprises the following steps:
s41: reading the horizontal distance values recorded in the track mapping table one by one, judging whether the absolute value of the difference value between each horizontal distance value and the horizontal distance value of the adjacent track point of the same navigation belt is smaller than a set threshold value, and if yes, recording the angle value corresponding to the horizontal distance value into an angle list;
s42: and reading the angle values recorded in the angle list one by one, accumulating the number of the same angle values to obtain an angle histogram, and taking the angle with the largest number in the angle histogram as the course angle of the navigation belt.
5. The oblique image matching screening method as claimed in claim 4, wherein the step S5 specifically includes the following sub-steps:
s51: reading the horizontal distance value and the angle value recorded in the track mapping table one by one, and judging the two track points as the same navigation belt if two track points exist, wherein the absolute value of the difference value between the horizontal distance value and the horizontal distance value of the adjacent track points of the same navigation belt is smaller than a set threshold value, and the absolute value of the difference value between the corresponding angle value of the horizontal distance value and the heading angle of the navigation belt is smaller than the set threshold value;
S52: constructing a navigation belt code list, searching in the navigation belt code list by taking the track points as keywords, if the track points exist, recording the track points on sub-items of the same navigation belt code list, and if the track points do not exist, recording the track points in the navigation belt code list as new codes;
s53: and repeating the steps S51-S52 until all the track points are divided into different navigation bands, and obtaining a navigation band code list.
6. The oblique image matching screening method as claimed in claim 5, wherein the step S6 specifically includes the following sub-steps:
s61: calculating the horizontal distance between each navigation belt code and the other navigation belt code in the navigation belt code list one by one, namely calculating the horizontal distance value between each track point of one navigation belt code and each track point of the other navigation belt code, and recording all the horizontal distance values into the distance list;
s62: and reading the horizontal distance values recorded in the distance list one by one, accumulating the number of the same horizontal distance values to obtain a ribbon horizontal distance histogram, and taking the horizontal distance value with the largest number in the ribbon horizontal distance histogram as the horizontal distance of the adjacent ribbon track point.
7. The oblique image matching screening method as claimed in claim 6, wherein the step S8 specifically includes the steps of:
S81: the rest track points in the original data list are rotated anticlockwise, and the rotation angle is the course angle of the navigation belt;
s82: and (3) taking the rotated track point coordinates into a standard equation of an ellipse to calculate, wherein if the values of the track point coordinates are less than 1, the track point coordinates are represented to be in the elliptical region, and an image of the track point in the elliptical region and an image of the track point in the center of the ellipse can form a matching pair.
8. An oblique image matching screening system is characterized by being used for realizing the oblique image matching screening method according to any one of claims 1-7, and specifically comprises an original data acquisition module, a track mapping table construction module, an adjacent track point horizontal distance calculation module, a course angle calculation module, a navigation belt code list construction module, an adjacent navigation belt track point horizontal distance calculation module, an elliptic standard equation construction module, an image matching module, a circulation judgment module and a result output module; wherein:
the original data acquisition module is used for acquiring a navigation survey track point file, and recording all track points and corresponding image information thereof as one piece of data into an original data list;
the track mapping table construction module is used for calculating horizontal distance values of every two track points according to track point information of the original data list and constructing a track mapping table;
The adjacent track point horizontal distance calculation module is used for calculating the adjacent track point horizontal distance of the same navigation band based on a track mapping table by adopting a statistical histogram calculation method;
the course angle calculation module is used for constructing an angle list based on the track mapping table and calculating the course angle of the navigation belt;
the navigation belt code list construction module is used for judging the horizontal distance between adjacent track points of the same navigation belt according to the track mapping table, dividing all the track points into different navigation belts, and constructing a navigation belt code list;
the horizontal distance calculation module of the adjacent ribbon track points is used for calculating the horizontal distance between one ribbon code and the other ribbon code in the ribbon code list, and obtaining the horizontal distance of the adjacent ribbon track points by adopting a method of calculating a histogram through statistics;
the ellipse standard equation building module is used for reading a track point from the original data list, taking the track point as an ellipse center, and determining the sizes of two half shafts of the ellipse by the horizontal distance between adjacent track points of the same navigation belt and the horizontal distance between adjacent navigation belt track points respectively, so as to build a standard equation of the ellipse;
the image matching module is used for calculating the rest track points in the original data list according to an elliptic standard equation to obtain track points in an elliptic region, and the images of the track points in the elliptic region and the images of the track points in the center of the ellipse can form matching pairs;
The circulation judging module is used for judging whether the screening of all the image matching pairs is finished, if yes, the result output module outputs all the image matching pairs to finish the inclined image matching screening process; otherwise, the elliptic standard equation building module reselects the track points and continues to carry out image matching screening.
9. The oblique image matching screening system of claim 8, wherein in the trajectory mapping table construction module, the following process is specifically performed: reading information of every two track points one by one from an original data list, and calculating horizontal distance values of every two track points; setting a distance threshold, judging whether the horizontal distance value of each two track points is larger than the set distance threshold, if so, calculating the angle values of the two track points corresponding to the horizontal distance value; after all the horizontal distance values and angle values are rounded downwards, taking every two track point information and the corresponding horizontal distance values and angle values as a record to construct a track mapping table;
in the adjacent track point horizontal distance calculation module, the following process is specifically executed, horizontal distance values recorded in a track mapping table are read one by one, the number of the same horizontal distance values is accumulated, a distance histogram is obtained, and the horizontal distance value with the largest number in the distance histogram is used as the horizontal distance of the adjacent track points of the same navigation belt;
In the course angle calculation module, the following processes are specifically executed: reading the horizontal distance values recorded in the track mapping table one by one, judging whether the absolute value of the difference value between each horizontal distance value and the horizontal distance value of the adjacent track point of the same navigation belt is smaller than a set threshold value, and if yes, recording the angle value corresponding to the horizontal distance value into an angle list; and reading the angle values recorded in the angle list one by one, accumulating the number of the same angle values to obtain an angle histogram, and taking the angle with the largest number in the angle histogram as the course angle of the navigation belt.
10. The oblique image matching screening system of claim 9, wherein in the navigation ribbon code list building module, the following is specifically performed: reading the horizontal distance value and the angle value recorded in the track mapping table one by one, and judging the two track points as the same navigation belt if two track points exist, wherein the absolute value of the difference value between the horizontal distance value and the horizontal distance value of the adjacent track points of the same navigation belt is smaller than a set threshold value, and the absolute value of the difference value between the corresponding angle value of the horizontal distance value and the heading angle of the navigation belt is smaller than the set threshold value; constructing a navigation belt code list, searching in the navigation belt code list by taking the track points as keywords, if the track points exist, recording the track points on sub-items of the same navigation belt code list, and if the track points do not exist, recording the track points in the navigation belt code list as new codes; repeating the operation until all the track points are divided into different navigation bands, and obtaining a navigation band code list;
In the adjacent navigation belt track point horizontal distance calculating module, the following process is specifically executed: calculating the horizontal distance between each navigation belt code and the other navigation belt code in the navigation belt code list one by one, namely calculating the horizontal distance value between each track point of one navigation belt code and each track point of the other navigation belt code, and recording all the horizontal distance values into the distance list; reading the horizontal distance values recorded in the distance list one by one, accumulating the number of the same horizontal distance values to obtain a ribbon horizontal distance histogram, and taking the horizontal distance value with the largest number in the ribbon horizontal distance histogram as the horizontal distance of the adjacent ribbon track point;
in the image matching module, the following process is specifically executed: the rest track points in the original data list are rotated anticlockwise, and the rotation angle is the course angle of the navigation belt; and (3) taking the rotated track point coordinates into a standard equation of an ellipse to calculate, wherein if the values of the track point coordinates are less than 1, the track point coordinates are represented to be in the elliptical region, and an image of the track point in the elliptical region and an image of the track point in the center of the ellipse can form a matching pair.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210366430.XA CN114894164B (en) | 2022-04-08 | 2022-04-08 | Oblique image matching screening method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210366430.XA CN114894164B (en) | 2022-04-08 | 2022-04-08 | Oblique image matching screening method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114894164A CN114894164A (en) | 2022-08-12 |
| CN114894164B true CN114894164B (en) | 2023-08-29 |
Family
ID=82715376
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210366430.XA Active CN114894164B (en) | 2022-04-08 | 2022-04-08 | Oblique image matching screening method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114894164B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100591608B1 (en) * | 2005-03-18 | 2006-06-20 | 전남대학교산학협력단 | Method for searching matching point in image matching |
| CN105716605A (en) * | 2016-03-30 | 2016-06-29 | 北京理工大学 | Matching method of gravity-aided inertial navigation system |
| CN106778588A (en) * | 2016-12-09 | 2017-05-31 | 国家测绘地理信息局四川测绘产品质量监督检验站 | State of flight detection method and device based on same place |
| CN107480727A (en) * | 2017-08-28 | 2017-12-15 | 荆门程远电子科技有限公司 | The unmanned plane image fast matching method that a kind of SIFT and ORB are combined |
| CN109978997A (en) * | 2019-04-03 | 2019-07-05 | 广东电网有限责任公司 | A kind of transmission line of electricity three-dimensional modeling method and system based on inclination image |
| CN111754556A (en) * | 2020-06-23 | 2020-10-09 | 自然资源部四川测绘产品质量监督检验站(四川省测绘产品质量监督检验站) | Incremental unmanned aerial vehicle aerial photography overlapping degree detection method and system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103167395B (en) * | 2011-12-08 | 2015-08-12 | 腾讯科技（深圳）有限公司 | Based on photo localization method and the system of mobile terminal navigation feature |
-
2022
- 2022-04-08 CN CN202210366430.XA patent/CN114894164B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100591608B1 (en) * | 2005-03-18 | 2006-06-20 | 전남대학교산학협력단 | Method for searching matching point in image matching |
| CN105716605A (en) * | 2016-03-30 | 2016-06-29 | 北京理工大学 | Matching method of gravity-aided inertial navigation system |
| CN106778588A (en) * | 2016-12-09 | 2017-05-31 | 国家测绘地理信息局四川测绘产品质量监督检验站 | State of flight detection method and device based on same place |
| CN107480727A (en) * | 2017-08-28 | 2017-12-15 | 荆门程远电子科技有限公司 | The unmanned plane image fast matching method that a kind of SIFT and ORB are combined |
| CN109978997A (en) * | 2019-04-03 | 2019-07-05 | 广东电网有限责任公司 | A kind of transmission line of electricity three-dimensional modeling method and system based on inclination image |
| CN111754556A (en) * | 2020-06-23 | 2020-10-09 | 自然资源部四川测绘产品质量监督检验站(四川省测绘产品质量监督检验站) | Incremental unmanned aerial vehicle aerial photography overlapping degree detection method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114894164A (en) | 2022-08-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210141378A1 (en) | Imaging method and device, and unmanned aerial vehicle | |
| CN108519094B (en) | Local path planning method and cloud processing terminal | |
| CN110223380B (en) | Scene modeling method, system and device fusing aerial photography and ground visual angle images | |
| CN110895819B (en) | Target tracking method, target tracking device, computer-readable storage medium and computer equipment | |
| CN105069843A (en) | Rapid extraction method for dense point cloud oriented toward city three-dimensional modeling | |
| CN107131883A (en) | The full-automatic mobile terminal indoor locating system of view-based access control model | |
| KR101558467B1 (en) | System for revising coordinate in the numerical map according to gps receiver | |
| CN109214254B (en) | Method and device for determining displacement of robot | |
| CN106096497B (en) | A kind of house vectorization method for polynary remotely-sensed data | |
| CN115049700A (en) | Target detection method and device | |
| CN110555377B (en) | Pedestrian detection and tracking method based on fish eye camera overlooking shooting | |
| CN112711034B (en) | Object detection method, device and equipment | |
| CN110264397B (en) | Method and device for extracting effective region of fisheye image | |
| CN113066112B (en) | Indoor and outdoor fusion method and device based on three-dimensional model data | |
| WO2020006378A1 (en) | Sensor calibration | |
| US11948329B2 (en) | System using image connectivity to reduce bundle size for bundle adjustment | |
| JP2016039390A (en) | Image generation method and device | |
| CN112084853B (en) | Footprint prediction method, footprint prediction device and humanoid robot | |
| CN114332291A (en) | Oblique photography model building outer contour rule extraction method | |
| CN112132853A (en) | Method and device for constructing ground guide arrow, electronic equipment and storage medium | |
| CN110197489B (en) | Method and device for selecting landing area of unmanned aerial vehicle | |
| CN114894164B (en) | Oblique image matching screening method and system | |
| CN111982076B (en) | Single-lens unmanned aerial vehicle flight parameter setting method | |
| CN116958218A (en) | Point cloud and image registration method and equipment based on calibration plate corner alignment | |
| CN114742876B (en) | Land vision stereo measurement method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20231213 Address after: 510000 Si Cheng Road, Tianhe District, Guangzhou, Guangdong Province, No. 39 Patentee after: SOUTH SURVEYING & MAPPING TECHNOLOGY CO.,LTD. Address before: 510000 Tianhe District District, Guangzhou, Guangdong, 39 Tianhe District 4 Building A area, 5 floor, A area, 6 floor A area Patentee before: GUANGZHOU SOUTH SATELLITE NAVIGATION INSTRUMENT Co.,Ltd. |