CN113033701A - Optimization method for manufacturing remote sensing image deep learning sample based on GIS spatial data - Google Patents
Optimization method for manufacturing remote sensing image deep learning sample based on GIS spatial data Download PDFInfo
- Publication number
- CN113033701A CN113033701A CN202110421660.7A CN202110421660A CN113033701A CN 113033701 A CN113033701 A CN 113033701A CN 202110421660 A CN202110421660 A CN 202110421660A CN 113033701 A CN113033701 A CN 113033701A
- Authority
- CN
- China
- Prior art keywords
- polygon
- learning sample
- boundary rectangle
- minimum boundary
- data
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- 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
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The invention relates to the technical field of remote sensing, and particularly discloses an optimization method for manufacturing a remote sensing image deep learning sample based on GIS spatial data, which comprises the following steps: s1, obtaining polygon data; s2, calculating the minimum boundary rectangle of the polygon through the break point coordinates; s3, setting the length threshold of the learning sample as S0Calculating the length S of the long side of the minimum boundary rectangle1If S is1≤S0Then the minimum bounding rectangle is the clipping unit and proceeds to S5, if S1>S0Then proceed to S4; s4, dividing the minimum boundary rectangle into multiple cutting units along the long edge of the minimum boundary rectangle until the length S of the cutting units2≤S0Then, the process proceeds to S5; s5, clipping the image data or the polygon part in the clipping unitAnd cutting the image data to obtain learning samples, and numbering the learning samples. The optimization method for manufacturing the remote sensing image deep learning sample based on the GIS spatial data can improve the precision and speed of machine learning.
Description
Technical Field
The invention relates to the technical field of remote sensing, in particular to an optimization method for manufacturing a remote sensing image deep learning sample based on GIS spatial data.
Background
In the remote sensing image deep learning process, a large number of samples are needed, and the GIS is a computer system for collecting, storing, managing, displaying and analyzing data on the earth surface related to space and geographic distribution and used for inputting, storing, inquiring, analyzing and displaying geographic data, combines geography and cartography, is widely applied to different fields, and can integrate the unique visualization effect and the geographic analysis function of a map with general database operation. The GIS spatial database stores raster image data and corresponding vector data, the vector data is an accurate coordinate sequence obtained by carrying out visual interpretation and manual vectorization on the raster data, and is associated with detailed attribute fields, for example, the land utilization database has a land cover type; the forestry database has forest species, tree species, forest age and other forest stand structure information, and attribute fields of the forest species, the tree species, the forest age and other forest stand structure information represent attributes corresponding to the remote sensing image and can be used as samples of deep learning training, but the data cannot be directly applied to a deep learning framework and needs to be further processed. After the existing processing method processes the graphs in the data, the manufactured part of learning samples are too large or too fine, the main body information which can be identified by a machine is not uniformly distributed, and the proportion of the object main body information in the part of learning samples is too small, so that the machine cannot accurately identify, and the machine learning precision is not high.
Disclosure of Invention
The invention aims to solve at least one of the above technical problems, and provides an optimization method for manufacturing a remote sensing image deep learning sample based on GIS spatial data, so that the learning precision and speed of a machine are improved.
In order to achieve the purpose, the invention adopts the technical scheme that: an optimization method for manufacturing a remote sensing image deep learning sample based on GIS spatial data comprises the following steps:
s1, acquiring image data in a GIS spatial database, extracting polygon data in the image data to obtain the break point coordinates of the polygon, and numbering the polygon;
s2, calculating the minimum boundary rectangle of the polygon through the break point coordinates, and acquiring the end point coordinates of the long side and the short side of the minimum boundary rectangle;
s3, setting the length threshold of the learning sample as S0Calculating the length S of the long side of the minimum boundary rectangle1If S is1≤S0Then the minimum bounding rectangle is the clipping unit and proceeds to S5, if S1>S0Then proceed to S4;
s4, dividing the minimum boundary rectangle into multiple cutting units along the long edge of the minimum boundary rectangle until the length S of the cutting units2≤S0Then, the process proceeds to S5;
s5, clipping the image data by the clipping unit, or clipping the image data by the polygon part in the clipping unit to obtain the learning sample, numbering the learning sample so that the number of the learning sample corresponds to the number of the polygon.
Preferably, in step S2, the minimum bounding rectangle of the polygon data is calculated by a least square method.
Preferably, in step S4, the center line of the long side of the minimum boundary rectangle is acquired as a dividing line, the minimum boundary rectangle is divided equally into two equal-length clipping units by the dividing line, and the length S of the clipping unit after division is set2≥S0Continuously acquiring the center line of the sideline of the cutting unit along the long side direction of the minimum boundary rectangle as a dividing line, and continuously dividing the cutting unit through the dividing line until the divided cutting unit has the length S2≤S0。
Preferably, two intersection points of the dividing line and the polygon are obtained, the numbers are p0 and q0, the folding points of the polygon are sequentially searched by moving to the left and right sides along the side line of the polygon with p0 as a starting point, the folding points are recorded as pi, the folding points of the polygon are sequentially searched by moving to the left and right sides along the side line of the polygon with q0 as a starting point, the folding points are recorded as qi, pi and qi are connected, the shortest connecting line is selected as a fine dividing line, and the minimum boundary rectangle is divided through the fine dividing line.
Preferably, the pi and qi searches are the same in number, and the sum of the pi and qi does not exceed 1/4 of the total number of polygonal break points.
Preferably, the polygon data is rotated in a forward or reverse direction such that the long side of the minimum bounding rectangle is parallel to the X-axis or Y-axis of the coordinate axes.
The beneficial effects are that: compared with the prior art, the optimization method for manufacturing the remote sensing image deep learning sample based on the GIS spatial data extracts the polygonal data from the image data, calculates the minimum boundary rectangle of the polygonal data, equally divides the polygonal data along the long side direction of the minimum boundary rectangle, divides the polygonal data into the cutting units meeting the learning sample length threshold, and finally cuts and divides the image sample through the cutting units, so that the object main body information in the obtained learning sample is more uniformly distributed, and the precision and the speed of machine learning are improved.
Drawings
The following detailed description of embodiments of the invention is provided in conjunction with the appended drawings, in which:
FIG. 1 is a flow chart of an optimization method for making a remote sensing image deep learning sample based on GIS spatial data according to the present invention;
FIG. 2 is a schematic diagram of one embodiment of a conventional learning sample preparation method;
FIG. 3 is a schematic diagram of another embodiment of a conventional learning sample preparation method;
FIG. 4 is a schematic diagram of polygon data obtained from image data;
FIG. 5 is a schematic illustration of solving a minimum bounding rectangle for a polygon;
FIG. 6 is a diagram of a minimum bounding rectangle after segmentation;
FIG. 7 is a schematic view of a polygon after rotation;
FIG. 8 is a schematic diagram of searching nodes on both sides of a partition line and an intersection of a polygon;
FIG. 9 is a schematic diagram of nodes after they are connected;
fig. 10 is a schematic diagram of the optimized dividing line.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It will be understood that when an element is referred to as being "secured to" another element, it can be directly on the other element or intervening elements may also be present. When a component is referred to as being "connected" to another component, it can be directly connected to the other component or intervening components may also be present. When a component is referred to as being "disposed on" another component, it can be directly on the other component or there can be intervening components, and when a component is referred to as being "disposed in the middle," it is not just disposed in the middle, so long as it is not disposed at both ends, but rather is within the scope of the middle. The terms "vertical," "horizontal," "left," "right," and the like as used herein are for illustrative purposes only.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
As shown in fig. 2 and 3, the conventional learning sample preparation method generally adopts a cross segmentation method to segment image data, and in some segmented learning samples, the information of the subject to be recognized is less, while background information occupies most of the learning samples, such as portions a and D in fig. 2 and portions a and C in fig. 3, and the information of the subject to be recognized only occupies a very small part of the segmented learning samples, so that the machine is difficult to accurately recognize the samples, and the learning precision of the machine is low, and the learning speed is not high.
As shown in fig. 1, 4, 5 and 6, the present application discloses an optimization method for making a remote sensing image deep learning sample based on GIS spatial data, which includes the following steps:
s1, acquiring image data in a GIS space database, and extracting polygon data in the image data, wherein the extracted polygon data can not only comprise the outline shape of the polygon data, but also comprise attribute information recorded in the GIS space database, and can also directly acquire the breakpoint coordinates of the polygon in the GIS space database, and uniquely number the polygon, for example, different polygons are numbered 1, 2, 3.. once;
s2, calculating the minimum boundary rectangle of the polygon through the break point coordinates, and acquiring the end point coordinates of the long side and the short side of the minimum boundary rectangle, specifically, in the calculation process, the break point coordinate sequence can be loaded into a memory, the logic structure of the memory is a circular list, and can be simulated by arrays, lists, queues and the like, and the coordinate sequence in the memory is directly extracted during calculation, so that the minimum boundary rectangle of the polygon can be quickly acquired;
s3, setting the length threshold of the learning sample as S0Calculating the length S of the long side of the minimum boundary rectangle1If S is1≤S0Then the minimum bounding rectangle is the clipping unit and proceeds to S5, if S1>S0Then proceed to S4;
s4, dividing the minimum boundary rectangle into multiple cutting units along the long edge of the minimum boundary rectangle until the length S of the cutting units2≤S0Then, the process proceeds to S5;
s5, clipping the image data by the clipping unit, or clipping the image data by the polygon included in the clipping unit to obtain a learning sample, specifically, clipping the obtained learning sample by the clipping unit, which includes not only the subject body information to be recognized but also partial background information, clipping the image data by the polygon included in the clipping unit, similar to the cutout in the prior art, where the clipped learning sample only includes the subject body information to be recognized, and after obtaining the learning sample, numbering the learning sample so that the number of the learning sample corresponds to the number of the polygon, for example, the clipping unit is obtained by dividing the number 1, and the corresponding number is 1-1, 1-2, 1-3.
In a preferred embodiment, in step S2, the minimum bounding rectangle of the polygon data may be calculated by using a least square method, so as to obtain the minimum bounding rectangle quickly, where the least square method is a conventional calculation method in the art and is not described in detail.
In another preferred embodiment, in step S4, the center line of the long side of the minimum boundary rectangle is acquired as a dividing line, the minimum boundary rectangle is divided equally into two equal-length clipping units by the dividing line, and the length S of the clipping unit after division is equal2≥S0Continuously acquiring the center line of the sideline of the cutting unit along the long side direction of the minimum boundary rectangle as a dividing line, and continuously dividing the cutting unit through the dividing line until the divided cutting unit has the length S2≤S0When the minimum boundary rectangle is calculated, the coordinates of the long side and the short side of the minimum boundary rectangle are obtained, and the coordinates are stored in the memory, so that when the dividing line is obtained, the coordinate data of the long side and the short side can be directly obtained from the memory, the dividing line is rapidly obtained, and the rapid division of the minimum boundary rectangle is further realized.
As shown in fig. 8 to 10, in a more preferred embodiment, two intersection points of the dividing line and the polygon are obtained, which are numbered p0 and q0, the break points of the polygon are sequentially searched for moving to the left and right along the edge of the polygon with p0 as the starting point, which is denoted as pi, specifically, as shown in fig. 8, the break points of the polygon are sequentially searched for with p2, p4 and p6 on the left side of p0, the break points on the right side of p0 are p1, p3 and p4, and the break points of the polygon are sequentially searched for moving to the left and right along the edge of the polygon with q0 as the starting point, which is denoted as qi, specifically, the break points on the left side of q0 are q2, q4 and q6, the break points on the right side of q0 are q1, q3 and q4, pi and qi are connected, that each break point pi is connected to each other and the shortest connecting line is obtained, and the coordinate of the dividing line is calculated as the shortest connecting line, the minimum boundary rectangle is divided by the fine dividing line, so that the shape of the divided polygon is more round, and the generation of the protruded corners is reduced, therefore, in the learning sample divided by the cutting unit, the occupation ratio of the main body information to be identified is larger, and the identification precision is further improved.
In a preferred embodiment, the number of pi and qi searches may be the same, and the sum of the numbers of pi and qi does not exceed a preset threshold, and in general, the preset threshold may be 1/4 of the total number of polygonal break points, so as to ensure that the subject information to be identified contained in each sample after segmentation does not differ too much.
As shown in fig. 7, in a more preferred embodiment, the polygon data may be rotated in the forward direction or the reverse direction so that the long side of the minimum boundary rectangle is parallel to the X axis or the Y axis of the coordinate axes, and since the present application uses a computer to automatically divide the minimum boundary rectangle, the learning sample obtained after the division is stored in a square picture output by the computer, and the minimum boundary rectangle is placed and then divided, so that the learning sample data in the learning sample picture generated by the computer can occupy a larger proportion, and the proportion of blank unidentifiable parts in the learning sample picture generated by the computer is reduced, thereby further improving the identification accuracy of the machine.
In summary, the optimization method for making the remote sensing image deep learning sample based on the GIS spatial data has the following advantages:
1. by utilizing the graphs and attribute achievements in the existing GIS spatial data, the work of going-out investigation and large-scale manual marking is reduced, the workload is greatly reduced, and the working efficiency is improved;
2. the sample processing is directly carried out in the GIS spatial data, the speed is high, and the existing GIS spatial data platform has basic functions of polygon segmentation, combination, area and perimeter calculation, grid cutting and the like, so the GIS spatial data platform is convenient to integrate in a GIS system;
3. the divided polygons are regular, so that the main body information of the object to be recognized is prominent, and background information which cannot be recognized is less, so that a good learning sample is obtained, and the training precision and speed are obviously improved.
The above embodiments are only for illustrating the technical solutions of the present invention and are not limited thereto, and any modification or equivalent replacement without departing from the spirit and scope of the present invention should be covered within the technical solutions of the present invention.
Claims (6)
1. The optimization method for manufacturing the remote sensing image deep learning sample based on GIS spatial data is characterized by comprising the following steps of:
s1, acquiring image data in a GIS spatial database, extracting polygon data in the image data to obtain the break point coordinates of the polygon, and numbering the polygon;
s2, calculating the minimum boundary rectangle of the polygon through the break point coordinates, and acquiring the end point coordinates of the long side and the short side of the minimum boundary rectangle;
s3, setting the length threshold of the learning sample as S0Calculate the minimumLength S of the long side of the bounding rectangle1If S is1≤S0Then the minimum bounding rectangle is the clipping unit and proceeds to S5, if S1>S0Then proceed to S4;
s4, dividing the minimum boundary rectangle into multiple cutting units along the long edge of the minimum boundary rectangle until the length S of the cutting units2≤S0Then, the process proceeds to S5;
s5, clipping the image data by the clipping unit, or clipping the image data by the polygon part in the clipping unit to obtain the learning sample, numbering the learning sample so that the number of the learning sample corresponds to the number of the polygon.
2. The optimization method for making the remote sensing image deep learning sample based on the GIS spatial data as claimed in claim 1, wherein in the step S2, the minimum boundary rectangle of the polygon data is calculated by using a least square method.
3. The optimizing method for making the remote sensing image deep learning sample based on the GIS spatial data as claimed in claim 1, wherein in step S4, the central line of the long side of the minimum boundary rectangle is obtained as a dividing line, the minimum boundary rectangle is divided equally into two clipping units with equal length by the dividing line, and if the length of the clipping unit after division is S, the length of the clipping unit after division is equal to that of the minimum boundary rectangle2≥S0Continuously acquiring the center line of the sideline of the cutting unit along the long side direction of the minimum boundary rectangle as a dividing line, and continuously dividing the cutting unit through the dividing line until the divided cutting unit has the length S2≤S0。
4. The optimizing method for making the remote sensing image deep learning sample based on the GIS spatial data as claimed in claim 3, characterized in that two intersection points of the dividing line and the polygon are obtained, the numbers are p0 and q0, the folding points of the polygon are sequentially searched by moving to the left and right sides along the side line of the polygon with p0 as the starting point, the folding points are recorded as pi, the folding points of the polygon are sequentially searched by moving to the left and right sides along the side line of the polygon with q0 as the starting point, the folding points are recorded as qi, the pi and qi are connected, the shortest connecting line is selected as a fine dividing line, and the minimum boundary rectangle is divided by the fine dividing line.
5. The optimization method for making the remote sensing image deep learning sample based on the GIS spatial data as claimed in claim 4, wherein the pi and qi search numbers are the same, and the sum of the pi and qi does not exceed 1/4 of the total number of polygonal break points.
6. The optimization method for making the remote sensing image deep learning sample based on the GIS spatial data as claimed in claim 1, wherein the polygon data is rotated forward or backward so that the long side of the minimum boundary rectangle is parallel to the X-axis or Y-axis of the coordinate axes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110421660.7A CN113033701B (en) | 2021-04-20 | 2021-04-20 | Optimization method for manufacturing remote sensing image deep learning sample based on GIS spatial data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110421660.7A CN113033701B (en) | 2021-04-20 | 2021-04-20 | Optimization method for manufacturing remote sensing image deep learning sample based on GIS spatial data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113033701A true CN113033701A (en) | 2021-06-25 |
| CN113033701B CN113033701B (en) | 2022-06-03 |
Family
ID=76457013
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110421660.7A Active CN113033701B (en) | 2021-04-20 | 2021-04-20 | Optimization method for manufacturing remote sensing image deep learning sample based on GIS spatial data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113033701B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113438484A (en) * | 2021-06-28 | 2021-09-24 | 遥聚信息服务(上海)有限公司 | Data transmission method and system for remote sensing image |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093484A (en) * | 2013-01-05 | 2013-05-08 | 武汉中地数码科技有限公司 | Integrated batch drawing method of remote-sensing image and vector data |
| CN110322556A (en) * | 2019-04-29 | 2019-10-11 | 武汉大学 | A kind of high-speed, high precision arrow grid Overlap Analysis method based on border cuts |
| CN111582218A (en) * | 2020-05-18 | 2020-08-25 | 长光卫星技术有限公司 | Remote sensing image island extraction method based on deep learning |
| US20200327725A1 (en) * | 2015-01-27 | 2020-10-15 | Splunk Inc. | Clipping polygons based on a scan of a storage grid |
| CN112396128A (en) * | 2020-12-08 | 2021-02-23 | 中国铁路设计集团有限公司 | Automatic labeling method for railway external environment risk source sample |
-
2021
- 2021-04-20 CN CN202110421660.7A patent/CN113033701B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103093484A (en) * | 2013-01-05 | 2013-05-08 | 武汉中地数码科技有限公司 | Integrated batch drawing method of remote-sensing image and vector data |
| US20200327725A1 (en) * | 2015-01-27 | 2020-10-15 | Splunk Inc. | Clipping polygons based on a scan of a storage grid |
| CN110322556A (en) * | 2019-04-29 | 2019-10-11 | 武汉大学 | A kind of high-speed, high precision arrow grid Overlap Analysis method based on border cuts |
| CN111582218A (en) * | 2020-05-18 | 2020-08-25 | 长光卫星技术有限公司 | Remote sensing image island extraction method based on deep learning |
| CN112396128A (en) * | 2020-12-08 | 2021-02-23 | 中国铁路设计集团有限公司 | Automatic labeling method for railway external environment risk source sample |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113438484A (en) * | 2021-06-28 | 2021-09-24 | 遥聚信息服务(上海)有限公司 | Data transmission method and system for remote sensing image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113033701B (en) | 2022-06-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103258345A (en) | Method for extracting parameters of tree branches based on ground laser radar three-dimensional scanning | |
| Najdataei et al. | Continuous and parallel lidar point-cloud clustering | |
| CN111339692B (en) | Laser point cloud data-based power line sag determination method and device | |
| WO2020225889A1 (en) | Point cloud annotation device, method and program | |
| CN110322497A (en) | A kind of interactive point cloud object extraction method based on three-dimensional visualization | |
| CN114241217B (en) | Trunk point cloud efficient extraction method based on cylindrical features | |
| US5144679A (en) | Graphic data searching and storage method | |
| CN113033701B (en) | Optimization method for manufacturing remote sensing image deep learning sample based on GIS spatial data | |
| CN114359226A (en) | Three-dimensional model set visual area extraction method based on hierarchical superposition and region growth | |
| CN106204719B (en) | Magnanimity model real-time scheduling method in three-dimensional scenic based on two-dimensional neighbourhood retrieval | |
| CN111159451B (en) | Power line point cloud dynamic monomer method based on spatial database | |
| CN105068918A (en) | Page test method and device | |
| CN112527916A (en) | Grid visual definition and intelligent feature analysis method based on remote sensing image | |
| CN108629315B (en) | Multi-plane identification method for three-dimensional point cloud | |
| CN107993242B (en) | Method for extracting boundary of missing area based on airborne LiDAR point cloud data | |
| WO2022022186A1 (en) | Data point identification method, identification device, and computer readable storage medium | |
| CN116958485A (en) | Visual field analysis method and device | |
| CN112116612A (en) | Pavement tree image example segmentation method based on Mask R-CNN | |
| Wibowo et al. | Mapping and grouping of farm land with Graham scan algorithm on convex hull method | |
| CN107220481B (en) | Quantitative evaluation method for comprehensive quality of planar element map | |
| CN113269202B (en) | Method for extracting point cloud of gate-type electric tower | |
| CN113408895B (en) | Ecological quality index construction method and system based on pixel scale | |
| CN112991300B (en) | Single wood skeleton extraction and visualization method based on neighborhood characteristics | |
| CN115731560A (en) | Slot line identification method and device based on deep learning, storage medium and terminal | |
| CN112082475B (en) | Living stumpage species identification method and volume 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 |