CN114166188B - Unmanned aerial vehicle inclined aerial survey image control point mark layout method and inclined aerial survey method - Google Patents
Unmanned aerial vehicle inclined aerial survey image control point mark layout method and inclined aerial survey method Download PDFInfo
- Publication number
- CN114166188B CN114166188B CN202111518445.5A CN202111518445A CN114166188B CN 114166188 B CN114166188 B CN 114166188B CN 202111518445 A CN202111518445 A CN 202111518445A CN 114166188 B CN114166188 B CN 114166188B
- Authority
- CN
- China
- Prior art keywords
- control point
- image
- hollowed
- image control
- out area
- 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 28
- 239000003973 paint Substances 0.000 claims abstract description 9
- 238000003384 imaging method Methods 0.000 claims abstract description 6
- 238000003709 image segmentation Methods 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 10
- 238000012360 testing method Methods 0.000 claims description 9
- 238000011176 pooling Methods 0.000 claims description 6
- 239000010426 asphalt Substances 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000002372 labelling Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000010422 painting Methods 0.000 claims 1
- 238000005507 spraying Methods 0.000 abstract description 3
- 238000013136 deep learning model Methods 0.000 abstract description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000013214 routine measurement 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
-
- 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
-
- 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
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an unmanned aerial vehicle inclined aerial survey control point mark layout method and an inclined aerial survey method, wherein the image control point mark layout method comprises the following steps: (S1) inserting a measuring nail on the ground surface of an image control point, and measuring the three-dimensional coordinates of the measuring nail; (S2) arranging a mask plate with a hollowed-out area on the image control point, wherein the hollowed-out area comprises a circular hollowed-out area matched with the measuring nail and two triangular hollowed-out areas symmetrically distributed on two sides of the circular hollowed-out area; the circular hollowed-out area is nested with the measuring nail; and (S3) spraying paint on the hollowed-out area so as to form an imaging control point mark around the measuring nail. The identification pattern has reasonable and simple structure, and the combination of the triangle and the circle is used for facilitating the recognition of the deep learning model.
Description
Technical Field
The invention belongs to the field of remote sensing mapping, and particularly relates to an unmanned aerial vehicle inclined aerial survey image control point mark layout method and an inclined aerial survey method.
Background
According to unmanned aerial vehicle oblique photogrammetry technology, need lay certain quantity of image control points in the district, this image control point is the basis of aerial survey interior industry sky three resolving encryption, directly influences the precision of output result. In the data processing process, the precision can be effectively improved through accurate puncturing work. Because unmanned aerial vehicle oblique photogrammetry increases the camera lens of the visual angle of looking at on the basis of conventional photogrammetry, lead to traditional ground to like the corner point difficult discernment under oblique visual angle, cause stab point position deviation, finally influence the precision of three-dimensional model achievement. Meanwhile, the pixel coordinates of the current image control points are largely manually extracted, and the degree of automation is low. The manual extraction in a large number of photographs is inefficient.
Disclosure of Invention
The invention aims to provide an unmanned aerial vehicle inclined aerial survey image control point mark layout method and an inclined aerial survey method.
The technical scheme of the invention is that the unmanned aerial vehicle inclined aerial survey control point mark layout method comprises the following steps:
(S1) inserting a measuring nail on the ground surface of an image control point, and measuring the three-dimensional coordinates of the measuring nail;
(S2) arranging a mask plate with a hollowed-out area on the image control point, wherein the hollowed-out area comprises a circular hollowed-out area matched with the measuring nail and two triangular hollowed-out areas symmetrically distributed on two sides of the circular hollowed-out area; the circular hollowed-out area is nested with the measuring nail;
and (S3) spraying paint on the hollowed-out area so as to form an imaging control point mark around the measuring nail.
The invention further improves that: the two triangular hollowed-out areas are isosceles triangles with congruent shapes.
The invention further improves that: and the opposite vertexes of the two triangular hollowed-out areas are coincident with the circle centers of the circular hollowed-out areas.
The invention further improves that: the two triangular hollowed-out areas are centrally symmetrical relative to the circle center of the circular hollowed-out area.
The invention further improves that: in the step S3, the image control point is sprayed with red paint when positioned on an asphalt pavement, and is sprayed with white paint when positioned on other pavements.
The invention further improves that: in the step S1, the three-dimensional coordinates of the measuring nail are measured by adopting an RTK or a total station.
The invention also provides an unmanned aerial vehicle inclination aerial survey method, which comprises the following steps:
(S01) marking each image control point by adopting the unmanned aerial vehicle inclined navigation image control point marking layout method;
(S02) performing unmanned aerial vehicle oblique photography on the target area to obtain image data;
(S03) identifying an image control point identification in the image data using an image control point identification model.
The invention further improves that: the image control point identification model adopts an encoder-decoder-classifier structure based on a VGG11 network, and the improvement relative to the VGG11 network comprises: the parameters of each convolution layer are halved and one full connection layer is reduced.
The invention further improves that: the network structure of the image control point identification model is as follows: the encoder structure is input layer 224×224×3, the convolutional layers are 3×3×32, 3×3×64, 3×3×128, 3×3×256, wherein the pooling layer adopts 2×2 maximum pooling; the input of the classifier and the image segmentation decoder is the output of the encoder; the classifier adopts full-connection layering and softmax classification, and the image segmentation decoder structure adopts an up-sampling structure with symmetrical encoder structure.
The invention further improves that: the construction process of the image control point identification model comprises the following steps:
setting up a plurality of image control point marks, and shooting an original image of each image control point mark;
marking the original image marked by each image control point by using a marking program to form a marking file; the labeling piece corresponds to the original image to form a sample set; dividing the sample set into a training set and a test set, wherein the training set accounts for 70% of the total number of samples, and the test set accounts for 30% of the total number of samples;
the model training, testing and predicting adopts a two-stage method, firstly, the original image is subjected to image classification, whether the image is controlled by a point or not is judged, and if the image is controlled by a point, the image is subjected to image segmentation, so that the classification of each pixel is accurate.
The beneficial effects of the invention are as follows:
1. the invention has novel control point identification style, easy identification, lower cost and easy popularization.
2. The identification pattern has reasonable and simple structure, and the combination of the triangle and the circle is used for facilitating the recognition of the deep learning model.
3. The invention can be used for ground image control point layout under a no-measuring-nail mode, is not only convenient for the field routine measurement to acquire the three-dimensional coordinates of the image control point, but also is easy for the aerial survey of the manual stabbing point work and the automatic identification of the image control point position in the field industry
4. The automatic identification algorithm of the image control points can greatly improve the automation of identification and identification of the image control points, greatly improve the production efficiency and reduce the time of the industry.
Drawings
FIG. 1 is a schematic view of a mask;
FIG. 2 is a schematic diagram of an image control point identification model structure;
fig. 3 is a model structure diagram of an encoder.
Detailed Description
Examples: the embodiment provides a method for laying unmanned aerial vehicle inclined navigation imaging control point marks, which comprises the following steps:
(S1) inserting a measuring nail on the ground surface of an image control point, and measuring the three-dimensional coordinates of the measuring nail; in the step, the three-dimensional coordinates of the measuring nail are measured by adopting conventional means such as RTK or total station and the like.
(S2) arranging a mask plate with a hollowed-out area shown in the figure 1 on the image control point, wherein the hollowed-out area comprises a circular hollowed-out area 2 matched with the measuring nail and two triangular hollowed-out areas 1 symmetrically distributed on two sides of the circular hollowed-out area; and enabling the circular hollowed-out area to be nested with the measuring nail. In this embodiment, the shapes of the two triangular hollowed-out areas are isosceles triangles with congruence. And the opposite vertexes of the two triangular hollowed-out areas are coincident with the circle centers of the circular hollowed-out areas. The two triangular hollowed-out areas are centrally symmetrical relative to the circle center of the circular hollowed-out area.
And (S3) spraying paint on the hollowed-out area so as to form an imaging control point mark around the measuring nail. Red paint is used when the image control point is positioned on asphalt pavement, and white paint is used when the image control point is positioned on other pavement.
Compared with the conventional image control point mark, the image control point mark determines the point position through two congruent isosceles triangles, and has higher identification degree.
The embodiment of the invention also provides an unmanned aerial vehicle inclination aerial survey method, which is characterized by comprising the following steps:
(S01) marking each image control point by adopting the unmanned aerial vehicle inclined navigation image control point marking layout method;
(S02) performing unmanned aerial vehicle oblique photography on the target area to obtain image data;
and (S03) carrying out image recognition on the image data by adopting an image control point recognition model, distinguishing whether the image contains an image control point object through image classification, and then recognizing and acquiring a pixel region of the image control point in the image by adopting an image segmentation method, wherein the centroid coordinates of the region are used as the center pixel coordinates of the image control point.
In this embodiment, the image control point identification model adopts an encoder-decoder-classifier structure based on the VGG11 network, and as shown in fig. 2, the improvement thereof with respect to the VGG11 network includes: the parameters of each convolution layer are halved and one full connection layer is reduced. The image control point identification model is realized by adopting a *** deep learning framework TensorFlow.
Specifically, the network structure of the image control point identification model is as follows: as shown in fig. 3, the encoder structure is an input layer (224×224×3), and the convolution layer is (3×3×32), (3×3×64), (3×3×128), (3×3×256), and the convolution layer is in order, where the pooling layer adopts 2×2 max pooling; the input of the classifier and the image segmentation decoder is the output of the encoder; the classifier adopts full-connection layering and softmax classification, and the image segmentation decoder structure adopts an up-sampling structure with symmetrical encoder structure.
The construction process of the image control point identification model comprises the following steps:
setting up a plurality of image control point marks, and shooting an original image of each image control point mark;
marking the original image marked by each image control point by using a marking program to form a marking file; the labeling piece corresponds to the original image to form a sample set; dividing a sample set (an original image S and a mark file T) into a training set and a test set, wherein the training set accounts for 70% of the total number of samples, and the test set accounts for 30% of the total number of samples;
the model training, testing and predicting adopts a two-stage method, firstly, the original image is subjected to image classification, whether the image is controlled by a point or not is judged, and if the image is controlled by a point, the image is subjected to image segmentation, so that the classification of each pixel is accurate.
The improvement of the VGG11 network model is beneficial to improving the calculation speed of unmanned aerial vehicle image training, and improving the receptive field, so that the recognition accuracy of a small target (the image control point mark is smaller than the whole image) is improved.
And screening the identification result by combining the characteristics of the image control point marks, wherein the screening conditions mainly comprise graphic symmetry, arc marks, triangle marks and the like so as to exclude false positive identification results.
The foregoing describes in detail preferred embodiments of the present invention. It should be understood that numerous modifications and variations can be made in accordance with the concepts of the invention by one of ordinary skill in the art without undue burden. Therefore, all technical solutions which can be obtained by logic analysis, reasoning or limited experiments based on the prior art by the person skilled in the art according to the inventive concept shall be within the scope of protection defined by the claims.
Claims (1)
1. An unmanned aerial vehicle tilt aerial survey method, characterized by comprising:
s01: marking each image control point by adopting an unmanned aerial vehicle inclined aerial survey control point marking layout method;
s02: performing unmanned aerial vehicle oblique photography on a target area to obtain image data;
s03: identifying an image control point identification in the image data by adopting an image control point identification model;
the image control point identification model adopts an encoder-decoder-classifier structure based on a VGG11 network, and the improvement relative to the VGG11 network comprises: halving the parameters of each convolution layer and reducing one full connection layer;
the network structure of the image control point identification model is as follows: the encoder structure is 224×224×3 in the form of input layer, the convolution layer is 3×3×32, 3×3×64, 3×3×128, 3×3×256 in turn, wherein the pooling layer adopts 2×2 maximum pooling; the input of the classifier and the image segmentation decoder is the output of the encoder; the classifier adopts full connection layering and softmax classification, and the image segmentation decoder structure adopts an up-sampling structure with symmetrical encoder structure;
the construction process of the image control point identification model comprises the following steps:
setting up a plurality of image control point marks, and shooting an original image of each image control point mark;
marking the original image marked by each image control point by using a marking program to form a marking file; the labeling piece corresponds to the original image to form a sample set; dividing the sample set into a training set and a test set, wherein the training set accounts for 70% of the total number of samples, and the test set accounts for 30% of the total number of samples;
the model training, testing and predicting adopts a two-stage method, firstly, the original image is subjected to image classification, whether the image is controlled by a point or not is judged, and if the image is controlled by a point, the image is subjected to image segmentation, so that the classification of each pixel is accurate;
the unmanned aerial vehicle inclined navigation imaging control point mark layout method in the step S01 comprises the following steps:
s1: inserting a measuring nail on the ground surface of the image control point, and measuring the three-dimensional coordinates of the measuring nail; in the step S1, measuring the three-dimensional coordinates of the measuring nail by adopting an RTK or a total station;
s2: setting a mask plate with a hollowed-out area on the image control point, wherein the hollowed-out area comprises a circular hollowed-out area matched with the measuring nail and two triangular hollowed-out areas symmetrically distributed on two sides of the circular hollowed-out area; the circular hollowed-out area is nested with the measuring nail; the shapes of the two triangular hollowed-out areas are isosceles triangles which are congruent; the opposite vertexes of the two triangular hollowed-out areas are overlapped with the circle center of the circular hollowed-out area; the two triangular hollowed-out areas are centrally symmetrical relative to the circle center of the circular hollowed-out area;
s3: painting the hollowed-out area so as to form an imaging control point mark around the measuring nail; the image control point is sprayed with red paint when being positioned on asphalt pavement, and is sprayed with white paint when being positioned on other pavement.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111518445.5A CN114166188B (en) | 2021-12-13 | 2021-12-13 | Unmanned aerial vehicle inclined aerial survey image control point mark layout method and inclined aerial survey method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111518445.5A CN114166188B (en) | 2021-12-13 | 2021-12-13 | Unmanned aerial vehicle inclined aerial survey image control point mark layout method and inclined aerial survey method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114166188A CN114166188A (en) | 2022-03-11 |
| CN114166188B true CN114166188B (en) | 2024-04-09 |
Family
ID=80486299
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111518445.5A Active CN114166188B (en) | 2021-12-13 | 2021-12-13 | Unmanned aerial vehicle inclined aerial survey image control point mark layout method and inclined aerial survey method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114166188B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN209524925U (en) * | 2019-04-13 | 2019-10-22 | 浙江泰和土地勘测规划有限公司 | A kind of reusable aerial survey of unmanned aerial vehicle photo control point identity device convenient for plug |
| CN112212835A (en) * | 2020-09-15 | 2021-01-12 | 广州全成多维信息技术有限公司 | Oblique photography and control method based on single-lens unmanned aerial vehicle |
| CN112669280A (en) * | 2020-12-28 | 2021-04-16 | 莆田市山海测绘技术有限公司 | Unmanned aerial vehicle oblique aerial photography right-angle image control point target detection method based on LSD algorithm |
| KR102263560B1 (en) * | 2020-10-29 | 2021-06-14 | 한국건설기술연구원 | System for setting ground control points using cluster RTK drones |
| CN113340277A (en) * | 2021-06-18 | 2021-09-03 | 深圳市武测空间信息有限公司 | High-precision positioning method based on unmanned aerial vehicle oblique photography |
-
2021
- 2021-12-13 CN CN202111518445.5A patent/CN114166188B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN209524925U (en) * | 2019-04-13 | 2019-10-22 | 浙江泰和土地勘测规划有限公司 | A kind of reusable aerial survey of unmanned aerial vehicle photo control point identity device convenient for plug |
| CN112212835A (en) * | 2020-09-15 | 2021-01-12 | 广州全成多维信息技术有限公司 | Oblique photography and control method based on single-lens unmanned aerial vehicle |
| KR102263560B1 (en) * | 2020-10-29 | 2021-06-14 | 한국건설기술연구원 | System for setting ground control points using cluster RTK drones |
| CN112669280A (en) * | 2020-12-28 | 2021-04-16 | 莆田市山海测绘技术有限公司 | Unmanned aerial vehicle oblique aerial photography right-angle image control point target detection method based on LSD algorithm |
| CN113340277A (en) * | 2021-06-18 | 2021-09-03 | 深圳市武测空间信息有限公司 | High-precision positioning method based on unmanned aerial vehicle oblique photography |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114166188A (en) | 2022-03-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108305264B (en) | A kind of unmanned plane precision landing method based on image procossing | |
| CN107830846B (en) | Method for measuring angle of communication tower antenna by using unmanned aerial vehicle and convolutional neural network | |
| CN106651953B (en) | A kind of vehicle position and orientation estimation method based on traffic sign | |
| CN104200086B (en) | Wide-baseline visible light camera pose estimation method | |
| CN106468553B (en) | A kind of localization method of the mobile object based on road sign | |
| CN109446973B (en) | Vehicle positioning method based on deep neural network image recognition | |
| CN109685855B (en) | Camera calibration optimization method under road cloud monitoring platform | |
| CN105335973A (en) | Visual processing method for strip steel processing production line | |
| CN107392929B (en) | Intelligent target detection and size measurement method based on human eye vision model | |
| CN109032174B (en) | Unmanned aerial vehicle operation route planning method and operation execution method | |
| CN102183250B (en) | Automatic navigation and positioning device and method for field road of agricultural machinery | |
| CN103729837A (en) | Rapid calibration method of single road condition video camera | |
| CN112308916A (en) | Target pose identification method based on image target | |
| CN108596165A (en) | Road traffic marking detection method based on unmanned plane low latitude Aerial Images and system | |
| CN109491384B (en) | Method and device for acquiring road marking data | |
| CN106529452B (en) | Mobile intelligent terminal building method for quickly identifying based on building threedimensional model | |
| CN113377118A (en) | Multi-stage accurate landing method for unmanned aerial vehicle hangar based on vision | |
| CN106446785A (en) | Passable road detection method based on binocular vision | |
| CN107564046A (en) | It is a kind of based on a cloud and the secondary accurate extracting method of registering contour of building of UAV images | |
| CN106908064A (en) | A kind of indoor night vision navigation method based on Kinect2 sensors | |
| CN110083177A (en) | A kind of quadrotor and control method of view-based access control model landing | |
| CN115578315A (en) | Bridge strain close-range photogrammetry method based on unmanned aerial vehicle image | |
| CN101950409B (en) | Quadrant-based encoding label point design method | |
| CN114166188B (en) | Unmanned aerial vehicle inclined aerial survey image control point mark layout method and inclined aerial survey method | |
| CN111380503B (en) | Monocular camera ranging method adopting laser-assisted calibration |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Country or region after: China Address after: 200093 No. 38 Shui Feng Road, Yangpu District, Shanghai. Applicant after: Shanghai Survey, Design and Research Institute (Group) Co.,Ltd. Address before: 200093 No. 38 Shui Feng Road, Yangpu District, Shanghai. Applicant before: SGIDI ENGINEERING CONSULTING (Group) Co.,Ltd. Country or region before: China |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |