CN113793355A - Automatic matching method for central line of top surface of unmanned aerial vehicle image railway steel rail - Google Patents
Automatic matching method for central line of top surface of unmanned aerial vehicle image railway steel rail Download PDFInfo
- Publication number
- CN113793355A CN113793355A CN202111068004.XA CN202111068004A CN113793355A CN 113793355 A CN113793355 A CN 113793355A CN 202111068004 A CN202111068004 A CN 202111068004A CN 113793355 A CN113793355 A CN 113793355A
- Authority
- CN
- China
- Prior art keywords
- image
- line
- steel rail
- top surface
- straight line
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
Abstract
The invention belongs to the technical field of railway existing line surveying and mapping, and relates to an automatic matching method for a central line of an unmanned aerial vehicle image railway steel rail top surface.
Description
Technical Field
The invention relates to the technical field of unmanned aerial vehicle surveying and mapping, in particular to an automatic matching method for a central line of a top surface of a railway rail by using an unmanned aerial vehicle image.
Background
One of the core tasks of existing railway mapping is to acquire high-precision three-dimensional coordinates of the center line of the track. With the development of computer vision and photogrammetry technologies, a non-contact ground positioning method based on a multi-view geometric principle has been widely applied in the fields of basic mapping, engineering investigation and design and the like. Meanwhile, thanks to the rapid development of software and hardware equipment, the unmanned aerial vehicle becomes an indispensable surveying and mapping geographic information data acquisition platform, and has the advantages of low cost, flexibility and high image ground resolution (reaching a sub centimeter level). Through unmanned aerial vehicle oblique photography mode, can effectively increase image base height ratio and intersection angle, promote the positioning accuracy of multi-view geometry on the elevation direction to the problem that existing line elevation precision is not enough is gathered for solving traditional aerial photography three-dimensional provides new thinking. However, the premise of realizing the high-precision mapping work of the existing line based on the unmanned aerial vehicle image is to obtain the homonymy mapping relation of the image steel rail line, namely, the matching of the central line of the top surface of the unmanned aerial vehicle image steel rail is realized. The traditional method needs to manually interpret the same-name straight line segment of the image steel rail, and has high working strength and low working efficiency. In order to further improve the production efficiency, the automatic matching problem of the image steel trajectory needs to be studied. However, since the steel rail is a kind of weak texture low line-shaped ground object, the characteristics of the steel rail on the image are not significant, and how to realize the automatic matching of the image steel trajectory is still a difficulty.
Disclosure of Invention
Therefore, the invention aims to provide an automatic matching method for the central line of the top surface of the unmanned aerial vehicle image railway steel rail, which is used for obtaining the matching relation of the central line of the top surface of the image steel rail by adopting an image space measurement combination method, can achieve higher matching precision and can meet the subsequent requirement of surveying and mapping precision of the existing line based on the unmanned aerial vehicle image.
In order to achieve the purpose, the invention discloses an automatic matching method for the central line of the top surface of an unmanned aerial vehicle image railway steel rail, which comprises the following steps:
s1, acquiring an unmanned aerial vehicle image and a three-dimensional coordinate initial value of a central line of the top surface of the steel rail;
s2, extracting straight line segments from the unmanned aerial vehicle image, and screening the extracted straight line segments;
s3, segmenting the acquired central line of the top surface of the steel rail according to a preset segment length value, back-projecting the segmented straight-line segments onto an unmanned aerial vehicle image, and screening based on the Euclidean distance of the image to obtain a group of candidate image lines; re-screening the candidate image lines by utilizing the angle constraint of the back projection line and the image extraction line;
s4, projecting the candidate image lines onto an object space elevation surface to obtain a group of object space projection straight line segments, performing least square straight line fitting by using end points of the object space projection straight line segments, taking an average value of errors in two fitting residuals as a target function, and taking elevation values of two end points of the straight line segments as variables to be optimized; and solving an extreme value of the objective function to obtain an accurate matching relation of the straight-line segments of the steel rail images.
And S5, calculating the top surface center line of the steel rail image according to the accurate matching relation of the straight line segments of the steel rail image, and performing corresponding name on the calculated top surface center line of the steel rail image and the acquired top surface center line of the steel rail to finish matching.
Further preferably, in S2, the method for extracting the straight line segments from the drone image and screening the extracted straight line segments includes the following steps:
s201, extracting a straight line segment from an image by adopting an LSD algorithm;
s202, screening the extracted straight line segments according to a preset threshold value, and screening the straight line segments larger than the preset threshold value.
Further preferably, in S3, when segmenting the acquired top surface center line of the steel rail according to the preset segment length value, the segment length value is 10 to 15m when segmenting a straight line portion of the acquired top surface center line of the steel rail, and the segment length value is 5 to 10m when segmenting a curved portion of the top surface center line of the steel rail.
Further preferably, in S3, the screening based on the euclidean distance of the image to obtain a set of candidate image lines includes the following steps:
for a certain section of steel rail line R, back projecting the image external orientation element subjected to POS auxiliary leveling and the camera internal parameters onto the unmanned aerial vehicle image to obtain a back projection straight line section R;
taking the back projection straight-line segment r as a central line, searching surrounding straight-line segments extracted from the image, and taking all image lines intersected with the straight-line segments as candidate straight-line segments to form a group of candidate image lines.
Further preferably, in S3, the re-screening the candidate image lines by using the angle constraint between the back projection line and the image extraction line includes the following steps:
setting an angle threshold, calculating an included angle between the image extraction line and the back projection line, reserving the image lines with the included angle less than or equal to the angle threshold, and deleting the image lines exceeding the angle threshold.
Further preferably, the angle threshold is set to 10 ° to 15 °.
Further preferably, in S4, a least-squares straight line fitting is performed using the end points of the object projection straight line segment; the method comprises the following steps:
setting a search range according to the elevation value and the elevation precision of the end point of the projection straight-line segment; searching all projection line end points;
selecting two end points and adopting RANSAC least square method to perform straight line fitting to obtain error Res in fitting residual error1(ii) a Fitting again by using the residual points to obtain error Res in fitting residual error2The average of the errors in the two residuals, RES, (RES) is calculated1+Res2) And/2 as an objective function.
More preferably, in the RANSAC least squares method, the residual threshold is set to 0.02m, and the inner point rate threshold is set to 0.4.
Further preferably, when the extremum of the objective function is obtained, the extremum of the objective function is obtained by iteration through a Powell algorithm.
Further preferably, in S5, calculating the rail image top surface center line according to the exact matching relationship of the rail image straight line segments includes the following steps:
taking two end points with similar distances between two straight line segments of the image as corresponding points, and calculating the middle point of the two end points as the end point of the central line of the top surface of the steel rail;
and connecting the end points of the calculated center lines pairwise to obtain the center line of the top surface of the steel rail.
The application discloses unmanned aerial vehicle image railway steel rail top surface central line automatic matching method compares and has following advantage at least in prior art:
1. according to the automatic matching method for the central line of the top surface of the unmanned aerial vehicle image railway steel rail, the unmanned aerial vehicle image multi-view geometric relation is utilized, firstly, a straight line segment is extracted from an image, then, the matching relation of the central line of the top surface of the image steel rail is obtained by adopting an image object measure combination method, higher matching precision can be achieved, and the following requirement of surveying and mapping precision of an existing line based on the unmanned aerial vehicle image can be met; the automatic matching of the central line of the top surface of the image steel rail is realized, manual interpretation is not needed, the operation efficiency is effectively improved, and the automatic matching device has strong practical and popularization values.
2. The automatic matching method for the central line of the top surface of the unmanned aerial vehicle image railway steel rail can be applied to image line feature matching optimization processing. Compared with a matching method based on an image space alone, the method disclosed by the invention fully utilizes the geometric constraint condition of the object space, applies the mathematical local optimization algorithm to image matching, can successfully match the situation that only the edge of one side of the steel rail can be extracted, greatly improves the matching accuracy and robustness, and has obvious advantages in the practical application process.
Drawings
FIG. 1 is a block flow diagram of the present invention.
FIG. 2 is a schematic diagram of coarse image line screening based on image distance.
FIG. 3 is a schematic diagram of coarse screening of image lines based on image space angle.
FIG. 4 is a schematic view of image line projection.
FIG. 5 is a schematic diagram illustrating the calculation of the center line of the top surface of the rail.
Detailed Description
The invention is described in further detail below with reference to the figures and the detailed description.
As shown in figure 1 of the drawings, in which,
the invention discloses an automatic matching method for the center line of the top surface of an unmanned aerial vehicle image railway steel rail, which comprises the following steps:
s1, acquiring an unmanned aerial vehicle image and a three-dimensional coordinate initial value of a central line of the top surface of the steel rail; acquiring required data, including acquiring unmanned aerial vehicle images, accurate camera internal parameters and image external orientation elements (POS auxiliary image area network adjustment results); and a three-dimensional coordinate initial value of a central line of the top surface of the steel rail (DOM/DSM steel rail line measurement result);
s2, extracting straight line segments from the unmanned aerial vehicle image, and screening the extracted straight line segments; in S2, the method for extracting the straight line segments from the drone image and screening the extracted straight line segments includes the following steps:
s201, extracting a straight line segment from an image by adopting an LSD algorithm;
s202, screening the extracted straight line segments according to a preset threshold value, and screening the straight line segments larger than the preset threshold value.
The LSD is a linear detection segmentation algorithm, the gradient of gray scale is obtained mainly by derivation, because the direction perpendicular to the direction of the gray scale gradient is the direction of a line, vectors with the same direction are circled by a rectangle, and the image straight line segment can be obtained by refining the rectangle. The algorithm can obtain a detection result with sub-pixel level precision in linear time, and does not need parameter adjustment on any digital image.
S3, segmenting the acquired central line of the top surface of the steel rail according to a preset segment length value, back-projecting the segmented straight-line segments onto an unmanned aerial vehicle image, and screening based on the Euclidean distance of the image to obtain a group of candidate image lines; re-screening the candidate image lines by utilizing the angle constraint of the back projection line and the image extraction line;
s4, projecting the candidate image lines onto an object space elevation surface to obtain a group of object space projection straight line segments, performing least square straight line fitting by using end points of the object space projection straight line segments, taking an average value of errors in two fitting residuals as a target function, and taking elevation values of two end points of the straight line segments as variables to be optimized; and solving an extreme value of the objective function to obtain an accurate matching relation of the straight-line segments of the steel rail images.
And S5, calculating the top surface center line of the steel rail image according to the accurate matching relation of the straight line segments of the steel rail image, and performing corresponding name on the calculated top surface center line of the steel rail image and the acquired top surface center line of the steel rail to finish matching.
In S3, when segmenting the acquired center line of the top surface of the steel rail according to the preset segment length value, the segment length value is 10 to 15m when segmenting the straight line portion of the acquired center line of the top surface of the steel rail, and the segment length value is 5 to 10m when segmenting the curved portion of the center line of the top surface of the steel rail.
Further, the screening based on the Euclidean distance of the image to obtain a group of candidate image lines includes the following steps: for a certain section of steel rail line R, back projecting the image external orientation element subjected to POS auxiliary leveling and the camera internal parameters onto the unmanned aerial vehicle image to obtain a back projection straight line section R;
taking the back projection straight-line segment r as a central line, searching surrounding straight-line segments extracted from the image, and taking all image lines intersected with the straight-line segments as candidate straight-line segments to form a group of candidate image lines.
As shown in fig. 2, R is a straight line segment of a certain rail, R is a back projection straight line segment of the certain rail on the image, R is a central line to search for surrounding straight line segments extracted from the image, all image lines intersected with R are taken as candidate straight line segments and are marked as candidate straight line segments
In the figure, the black line is the back projection line of the orbit line, and the gray line (steel trajectory of the same name) and the white line (non-steel trajectory of the same name) are the straight line segments extracted from the image, and finally the image line within the range of the dotted line is reserved. The step can eliminate most of the non-homonymous straight line segments (if another steel rail line of the track), and the matching range of the image line is greatly reduced.
Further, the re-screening of the candidate image lines by using the angle constraint of the back projection line and the image extraction line includes the following steps: setting an angle threshold, calculating an included angle between the image extraction line and the back projection line, reserving the image lines with the included angle less than or equal to the angle threshold, and deleting the image lines exceeding the angle threshold. The angle threshold is set at 10-15 deg. As shown in fig. 3, the black line is the back projection line of the straight line segment of the initial trajectory, and the white line is the error image line remaining after distance screening. Since the difference between the white line and the track line direction is large, the white line and the track line direction can be eliminated through the step.
Further preferably, in S4, a least-squares straight line fitting is performed using the end points of the object projection straight line segment; the method comprises the following steps:
s401, setting a search range according to the elevation value and the elevation precision of the end point of the projection straight-line segment; searching all projection line end points;
if the initial values of the coordinates of two end points of a certain steel rail straight-line segment R are respectively [ X ]1，Y1，Z1]TAnd [ X ]2，Y2，Z2]TThe candidate image line is
Projecting two end points of the candidate image line to a sum elevation plane Z1And Z2Thus, a set of object projection lines, denoted as L, can be obtainedi(i ═ 1, 2, 3.. n). If the accurate height values of the end points of the straight line sections of the steel rails are Z respectivelyR1And ZR2The elevation combination is denoted as PR(ZR1，ZR2). As shown in FIG. 4, the black line and the gray line are the two side edges of the rail, respectively, and when the projected elevation error of the two end points is large, the projected line L isi(i ═ 1, 2, 3.. n) appears as scattered straight line segments on the projection plane, such as PH(ZH1，ZH2) And PL(ZL1，ZL2) (ii) a When the projected elevations of the two end points are close to the real elevation value, the projected lines approach to form two clusters of lines (P)R(ZR1，ZR2) And then an accurate image steel rail line homonymous mapping relation can be obtained.
S402, selecting two end points and adopting RANSAC least square method to perform straight line fitting to obtain error Res in fitting residual error1(ii) a Fitting again by using the residual points to obtain error Res in fitting residual error2The average of the errors in the two residuals, RES, (RES) is calculated1+Res2) And/2 as an objective function. In the RANSAC least square method, a residual error threshold value is set to be 0.02m, and an interior point rate threshold value is 0.4.
If the elevation values of the two end points of the initial steel rail line are Z respectively1And Z2If the elevation precision is sigma, the search ranges of the elevation values of the two end points are respectively [ Z ]1-σ，Z1+σ]And [ Z2-σ，Z2+σ]。
And S403, when the extreme value of the objective function is obtained, iterative obtaining of the extreme value of the objective function is carried out by adopting a Powell algorithm.
It should be noted that the candidate image lines include a small number of non-rail-top edge lines (e.g., image lines extracted from the bottom of the rail), which can be eliminated in this step.
Further preferably, in S5, calculating the rail image top surface center line according to the exact matching relationship of the rail image straight line segments includes the following steps:
taking two end points with similar distances between two straight line segments of the image as corresponding points, and calculating the middle point of the two end points as the end point of the central line of the top surface of the steel rail;
and connecting the end points of the calculated center lines pairwise to obtain the center line of the top surface of the steel rail.
And 5, obtaining the center line of the top surface of the steel rail and the matching relation thereof through the homonymous mapping relation of the straight-line segments at the edge of the steel rail.
The method comprises the following specific steps:
step 501, calculating the center line of straight line segments on the left side and the right side of the top surface of the image steel rail. And taking two end points of the image with the closer distance between the two straight line segments as corresponding points, calculating the middle point of the corresponding point as the end point of the central line of the top surface of the steel rail, and connecting the two calculated points to obtain the central line of the top surface of the steel rail. As shown in fig. 5, AB and CD are respectively straight line segments of the edge extracted from both sides of the rail, E is the center line of a and C, F is the midpoint of B and D, and then the straight line segment of the top center line of the rail is EF.
And 502, obtaining the homonymous corresponding relation of the center line of the top surface of the steel rail according to the homonymous corresponding relation of the straight line segments of the edge of the steel rail obtained in the step 4.
It should be understood that the above examples are only for clarity of illustration and are not intended to limit the embodiments. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. And obvious variations or modifications therefrom are within the scope of the invention.
Claims (10)
1. An automatic matching method for central lines of top surfaces of unmanned aerial vehicle image railway steel rails is characterized by comprising the following steps:
s1, acquiring an unmanned aerial vehicle image and a three-dimensional coordinate initial value of a central line of the top surface of the steel rail;
s2, extracting straight line segments from the unmanned aerial vehicle image, and screening the extracted straight line segments;
s3, segmenting the acquired central line of the top surface of the steel rail according to a preset segment length value, back-projecting the segmented straight-line segments onto an unmanned aerial vehicle image, and screening based on the Euclidean distance of the image to obtain a group of candidate image lines; re-screening the candidate image lines by utilizing the angle constraint of the back projection line and the image extraction line;
s4, projecting the candidate image lines onto an object space elevation surface to obtain a group of object space projection straight line segments, performing least square straight line fitting by using end points of the object space projection straight line segments, taking an average value of errors in two fitting residuals as a target function, and taking elevation values of two end points of the straight line segments as variables to be optimized; solving an extreme value of the objective function to obtain an accurate matching relation of the straight-line segments of the steel rail images;
and S5, calculating the top surface center line of the steel rail image according to the accurate matching relation of the straight line segments of the steel rail image, and performing corresponding name on the calculated top surface center line of the steel rail image and the acquired top surface center line of the steel rail to finish matching.
2. The method of claim 1, wherein in step S2, the method comprises extracting straight line segments from the drone image and screening the extracted straight line segments, comprising the steps of:
s201, extracting a straight line segment from an image by adopting an LSD algorithm;
s202, screening the extracted straight line segments according to a preset threshold value, and screening the straight line segments larger than the preset threshold value.
3. The method of claim 1, wherein the step of segmenting the centerline of the top surface of the steel rail obtained according to the preset segment length values in step S3 includes
When the obtained straight line part of the central line of the top surface of the steel rail is segmented, the segment length value is 10-15 m, and when the curve part of the central line of the top surface of the steel rail is segmented, the segment length value is 5-10 m.
4. The method of claim 1, wherein in step S3, the step of screening based on Euclidean image distances to obtain a set of candidate image lines includes:
for a certain section of steel rail line R, back projecting the image external orientation element subjected to POS auxiliary leveling and the camera internal parameters onto the unmanned aerial vehicle image to obtain a back projection straight line section R;
taking the back projection straight-line segment r as a central line, searching surrounding straight-line segments extracted from the image, and taking all image lines intersected with the straight-line segments as candidate straight-line segments to form a group of candidate image lines.
5. The method of claim 1, wherein in step S3, the re-screening of the candidate image lines is performed by using the angle constraint between the back projection line and the image extraction line, and the method comprises the following steps:
setting an angle threshold, calculating an included angle between the image extraction line and the back projection line, reserving the image lines with the included angle less than or equal to the angle threshold, and deleting the image lines exceeding the angle threshold.
6. The method of claim 5, wherein the angle threshold is set to be 10 ° -15 °.
7. The automatic matching method for the top surface center line of the unmanned aerial vehicle image railway steel rail of claim 1, wherein in S4, least square straight line fitting is performed by using the end points of the object projection straight line segment; the method comprises the following steps:
setting a search range according to the elevation value and the elevation precision of the end point of the projection straight-line segment; searching all projection line end points;
selecting two end points and performing linear fitting by using RANSAC least square method to obtain error in fitting residual error
(ii) a Fitting again by using the residual points to obtain errors in fitting residual errors
Calculating the average of the errors in the two residuals
As an objective function.
8. The automatic matching method for the central line of the top surface of the unmanned aerial vehicle image railway steel rail as claimed in claim 1, wherein in the RANSAC least square method, a residual threshold value is set to be 0.02m, and an interior point rate threshold value is 0.4.
9. The method of claim 1, wherein the extreme value of the objective function is obtained by iterative Powell algorithm.
10. The method of claim 1, wherein in step S5, the top centerline of the rail image is calculated according to the exact matching relationship of the straight line segments of the rail image, comprising the steps of:
taking two end points with similar distances between two straight line segments of the image as corresponding points, and calculating the middle point of the two end points as the end point of the central line of the top surface of the steel rail;
and connecting the end points of the calculated center lines pairwise to obtain the center line of the top surface of the steel rail.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111068004.XA CN113793355B (en) | 2021-09-13 | 2021-09-13 | Automatic matching method for top surface center line of unmanned aerial vehicle image railway steel rail |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111068004.XA CN113793355B (en) | 2021-09-13 | 2021-09-13 | Automatic matching method for top surface center line of unmanned aerial vehicle image railway steel rail |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113793355A true CN113793355A (en) | 2021-12-14 |
| CN113793355B CN113793355B (en) | 2022-12-16 |
Family
ID=78880029
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111068004.XA Active CN113793355B (en) | 2021-09-13 | 2021-09-13 | Automatic matching method for top surface center line of unmanned aerial vehicle image railway steel rail |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113793355B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116756808A (en) * | 2023-05-23 | 2023-09-15 | 济青高速铁路有限公司 | Railway line shape optimization method and device, electronic equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106845493A (en) * | 2016-12-06 | 2017-06-13 | 西南交通大学 | The identification at railroad track close-range image rail edge and matching process |
| JP6852936B1 (en) * | 2019-11-15 | 2021-03-31 | 広東工業大学Ｇｕａｎｇｄｏｎｇ Ｕｎｉｖｅｒｓｉｔｙ Ｏｆ Ｔｅｃｈｎｏｌｏｇｙ | Drone visual odometer method based on depth dotted line features |
| CN113155098A (en) * | 2021-04-20 | 2021-07-23 | 中国铁路设计集团有限公司 | Existing railway track line high-precision three-dimensional reconstruction method based on unmanned aerial vehicle multi-view images |
-
2021
- 2021-09-13 CN CN202111068004.XA patent/CN113793355B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106845493A (en) * | 2016-12-06 | 2017-06-13 | 西南交通大学 | The identification at railroad track close-range image rail edge and matching process |
| JP6852936B1 (en) * | 2019-11-15 | 2021-03-31 | 広東工業大学Ｇｕａｎｇｄｏｎｇ Ｕｎｉｖｅｒｓｉｔｙ Ｏｆ Ｔｅｃｈｎｏｌｏｇｙ | Drone visual odometer method based on depth dotted line features |
| CN113155098A (en) * | 2021-04-20 | 2021-07-23 | 中国铁路设计集团有限公司 | Existing railway track line high-precision three-dimensional reconstruction method based on unmanned aerial vehicle multi-view images |
Non-Patent Citations (2)
| Title |
|---|
| 于英等: "无人机影像物方多视匹配算法", 《中国图象图形学报》 * |
| 董莉等: "VLL影像匹配的线路地形断面测绘方法", 《测绘科学》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116756808A (en) * | 2023-05-23 | 2023-09-15 | 济青高速铁路有限公司 | Railway line shape optimization method and device, electronic equipment and storage medium |
| CN116756808B (en) * | 2023-05-23 | 2024-04-23 | 济青高速铁路有限公司 | Railway line shape optimization method and device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113793355B (en) | 2022-12-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Holgado‐Barco et al. | Semiautomatic extraction of road horizontal alignment from a mobile LiDAR system | |
| CN112767490B (en) | Outdoor three-dimensional synchronous positioning and mapping method based on laser radar | |
| CN110006408B (en) | LiDAR data cloud control aerial image photogrammetry method | |
| CN113155098B (en) | Existing railway track line high-precision three-dimensional reconstruction method based on unmanned aerial vehicle multi-view images | |
| CN104574421A (en) | Large-breadth small-overlapping-area high-precision multispectral image registration method and device | |
| US6442292B1 (en) | Measuring apparatus | |
| CN114859374B (en) | Newly-built railway cross measurement method based on unmanned aerial vehicle laser point cloud and image fusion | |
| CN113971768A (en) | Unmanned aerial vehicle-based three-dimensional dynamic detection method for power transmission line illegal building | |
| CN111915517A (en) | Global positioning method for RGB-D camera in indoor illumination adverse environment | |
| Zhang et al. | Automatic registration of urban aerial imagery with airborne LiDAR data | |
| CN117274499B (en) | Unmanned aerial vehicle oblique photography-based steel structure processing and mounting method | |
| CN113793355B (en) | Automatic matching method for top surface center line of unmanned aerial vehicle image railway steel rail | |
| CN116563377A (en) | Mars rock measurement method based on hemispherical projection model | |
| CN113221883A (en) | Real-time correction method for flight navigation route of unmanned aerial vehicle | |
| CN116518864A (en) | Engineering structure full-field deformation detection method based on three-dimensional point cloud comparison analysis | |
| Lu et al. | A lightweight real-time 3D LiDAR SLAM for autonomous vehicles in large-scale urban environment | |
| CN114612412A (en) | Processing method of three-dimensional point cloud data, application of processing method, electronic device and storage medium | |
| Habib et al. | Linear features in photogrammetric activities | |
| CN111754556B (en) | Incremental unmanned aerial vehicle aerial photography overlapping degree detection method and system | |
| CN113673011A (en) | Method for intelligently identifying tunnel invasion boundary in operation period based on point cloud data | |
| CN115790539B (en) | Cooperative target underwater photogrammetry method | |
| Hasheminasab et al. | Linear Feature-based image/LiDAR integration for a stockpile monitoring and reporting technology | |
| Zhang et al. | Automatic Extrinsic Parameter Calibration for Camera-LiDAR Fusion using Spherical Target | |
| Lu et al. | Estimation of Transformation Parameters Between Centre‐Line Vector Road Maps and High Resolution Satellite Images | |
| Chen et al. | A comparison of stereo-matching cost between convolutional neural network and census for satellite images |
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 |