CN110569745A - remote sensing image building area detection method - Google Patents
remote sensing image building area detection method Download PDFInfo
- Publication number
- CN110569745A CN110569745A CN201910768249.XA CN201910768249A CN110569745A CN 110569745 A CN110569745 A CN 110569745A CN 201910768249 A CN201910768249 A CN 201910768249A CN 110569745 A CN110569745 A CN 110569745A
- Authority
- CN
- China
- Prior art keywords
- building
- target
- candidate
- elements
- building elements
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 38
- 238000000034 method Methods 0.000 claims abstract description 19
- 239000000126 substance Substances 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims 5
- 238000004364 calculation method Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000000903 blocking effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/176—Urban or other man-made structures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/194—Terrestrial scenes using hyperspectral data, i.e. more or other wavelengths than RGB
Abstract
the embodiment of the invention provides a method for detecting a remote sensing image building area, which comprises the following steps: acquiring a building image shot with a building; extracting candidate building elements in the building image to obtain a plurality of candidate building elements; detecting whether part or all of the candidate building elements can form a preset building model or not based on a pre-constructed azimuth configuration template; and when detecting that part or all of the candidate building elements can form the preset building model, determining the area where the part or all of the candidate building elements forming the preset building model are located as a building area. The embodiment of the invention improves the efficiency of building area detection.
Description
Technical Field
The invention relates to the technical field of building detection, in particular to a method for detecting a remote sensing image building area.
Background
in recent years, the application of remote sensing image processing technology is more and more extensive, and the remote sensing image processing technology plays a vital role in a plurality of fields. The detection of suburban buildings in remote sensing images is one of the research hotspots in the field of remote sensing image processing at present. In the aspect of national economic development, due to the rapid increase of commercial land use and the rapid decrease of cultivated land resources, the monitoring of land utilization dynamic change is particularly important.
At present, the building detection technology based on remote sensing images is various, and mainly comprises a detection method based on a straight line segment, a detection method based on a local descriptor matching algorithm and a detection method based on a space strict topological structure. However, when the detection method based on the line segments is adopted, the line segment area in the large-view-field remote sensing image is more, and the line segment combination information in the building is complex and changeable, so that excessive interference line segments generated by complex ground objects such as vegetation areas and the like are difficult to distinguish by the algorithm, the false alarm rate of detection is greatly improved, and the accuracy rate of detection is reduced; in addition, only the local structure information of the building is considered based on the local descriptor matching algorithm, the consideration of the whole structure of the building is lacked, and the method needs to carry out full-image scanning on the input image to obtain key points of a full image, so that the calculation amount is large and the time is consumed; in addition, the detection method based on the space strict topological structure is strictly restricted by the azimuth topological relation of the building area, is not resistant to intra-class variation, and is not suitable for distributed target detection, so that the detection efficiency is influenced.
Disclosure of Invention
The embodiment of the invention provides a method for detecting a building area by using a remote sensing image, which aims to improve the detection efficiency during building area detection.
The embodiment of the invention provides a method for detecting a remote sensing image building area, which comprises the following steps:
Acquiring a building image shot with a building;
extracting candidate building elements in the building image to obtain a plurality of candidate building elements;
detecting whether part or all of the candidate building elements can form a preset building model or not based on a pre-constructed azimuth configuration template;
And when detecting that part or all of the candidate building elements can form the preset building model, determining the area where the part or all of the candidate building elements forming the preset building model are located as a building area.
optionally, the extracting candidate building elements from the building image to obtain a plurality of candidate building elements includes:
Extracting all straight line segments in the building image based on a straight line segment detection algorithm LSD;
acquiring target straight-line segments with parallel relation in all the straight-line segments;
detecting whether the distance between two target straight-line segments is smaller than a preset distance threshold value or not;
And when the distance between two target straight-line segments is smaller than the preset distance threshold value, connecting two ends of the two target straight-line segments to obtain a target rectangular area, and determining the target rectangular area as the candidate building element.
optionally, after extracting the candidate building elements in the building image to obtain a plurality of candidate building elements, the method further includes:
And acquiring a centroid point of each candidate building element, and identifying each candidate building element through the centroid point.
optionally, the orientation configuration template is obtained by projecting the building image, and the orientation configuration template is divided into a plurality of rectangular blocks with the same size;
The detecting whether part or all of the candidate building elements can form a preset building model based on the pre-constructed orientation configuration template comprises the following steps:
For each candidate building element, mapping the candidate building element into a rectangular block corresponding to the position in the orientation configuration template according to the position of the candidate building element in the building image;
acquiring all target rectangular blocks mapped with candidate building elements;
and when the arrangement mode formed by part or all of the target rectangular blocks is the same as the arrangement mode of the preset building model, determining that the candidate building elements mapped in the part or all of the target rectangular blocks can form the preset building model.
Optionally, after detecting that part or all of the candidate building elements can form a preset building model, the method further includes:
determining a target building element from part or all of candidate building elements forming a preset building model;
acquiring the deformation quantity of each target building element and the accumulated deformation quantity of all the target building elements;
And judging whether the area where part or all of candidate building elements forming the preset building model are located is determined as a building area or not according to the deformation amount of each target building element and the accumulated deformation amount.
optionally, the determining a target building element from part or all of the candidate building elements constituting the preset building model includes:
Aiming at each target rectangular block, acquiring the number of candidate building elements in the target rectangular block; wherein the content of the first and second substances,
When the number of the candidate building elements in the target rectangular block is detected to be 1, directly determining the candidate building elements in the target rectangular block as target building elements;
and when the number of the candidate building elements in the target rectangular block is detected to be more than 1, clustering the candidate building elements in the target rectangular block, and determining a clustering center of mass obtained after clustering as the target building element.
Optionally, obtaining the deformation amount of each target building element comprises: aiming at each target building element, acquiring the central position of a target rectangular block where the target building element is located; calculating the position offset between the target building element and the central position, and determining the position offset as the deformation amount of the target building element;
Acquiring the accumulated deformation quantity of all target building elements, including: and acquiring the sum of the deformation quantities of all the target building elements, and determining the sum as the accumulated deformation quantity.
Optionally, the determining, according to the deformation amount of each target building element and the accumulated deformation amount, whether to determine an area where part or all of candidate building elements constituting a preset building model are located as a building area includes:
When the deformation amount of each target building element is smaller than a first threshold value and the accumulated deformation amount is smaller than a second threshold value, determining the area where the partial or all candidate building elements are located as a building area; and the second threshold is less than or equal to N times of the first threshold, wherein N is the number of the target building elements.
According to the remote sensing image building area detection method provided by the embodiment of the invention, the building image shot with the building is obtained, the candidate building elements in the building image are extracted to obtain a plurality of candidate building elements, then based on the pre-constructed azimuth configuration template, when the situation that part or all of the candidate building elements can form the preset building model is detected, the area where the part or all of the candidate building elements forming the preset building model are located is determined as the building area, the purpose that the candidate building elements in the building image are extracted firstly is achieved, and then whether the candidate building elements are the building area or not is accurately analyzed through the azimuth configuration template is achieved, so that the building area detection efficiency is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
FIG. 1 is a flow chart illustrating steps of a method for detecting a building area according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of an architectural image in an embodiment of the present invention;
FIG. 3 is a schematic illustration of an orientation configuration template in an embodiment of the present invention;
FIG. 4 is a flowchart illustrating steps after detecting that some or all of the candidate building elements can form a preset building model according to an embodiment of the present invention;
FIG. 5 is a diagram illustrating clustering of candidate building elements according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, 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 some, but not all, embodiments of the present invention. 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.
in particular, in a suburban remote sensing scene with a large field of view, buildings are sparsely distributed and generally appear in the form of a building area, namely, a plurality of buildings are gathered into one building area. At this time, due to the sparsity of the distribution of the building areas, most areas in the scene to be detected are non-target areas, namely non-building areas, so that the detection efficiency is low when the building areas in the suburban remote sensing scene with a large field of view are detected.
At this time, in order to improve the detection efficiency in the building area detection, the present invention provides the following embodiments:
As shown in fig. 1, which is a flow chart of steps of a method for detecting a building area of a remote sensing image in an embodiment of the present invention, the method includes the following steps:
step 101: a building image in which a building is photographed is acquired.
Specifically, when detecting a building area in an image, a building image in which a building is photographed may be acquired first.
the building image can be a satellite optical remote sensing image and can be obtained from a Google map.
specifically, a building is photographed in the building image. For example, referring to fig. 2, it can be seen from the architectural image in fig. 2 that a plurality of buildings, such as buildings in a dotted ellipse, are photographed in the architectural image.
step 102: and extracting candidate building elements in the building image to obtain a plurality of candidate building elements.
In this step, specifically, after the building image is acquired, candidate building elements in the building image may be extracted to obtain a plurality of candidate building elements.
specifically, the extracted candidate building elements are suspected building structures. Based on the sparsity of the distribution of the building area, the suspected building structure can be extracted firstly, and then the suspected building structure is accurately analyzed, so that the detection efficiency of the building area is improved.
step 103: and detecting whether part or all of the candidate building elements can form a preset building model or not based on the pre-constructed orientation configuration template.
in this step, specifically, after obtaining a plurality of candidate building elements, it may be detected whether some or all of the candidate building elements can form a preset building model based on the pre-constructed orientation configuration template, so as to determine whether the candidate building elements are aggregated into a building area.
the orientation configuration template is obtained by projecting the building image and is divided into a plurality of rectangular blocks with the same size. The size of the orientation configuration template may then be the same as the size of the building image, which enables the positions in the orientation configuration template to correspond to the positions in the building image, thereby enabling the candidate building elements in the building image to be mapped into the orientation configuration template.
of course, it should be noted here that, in the orientation configuration template, the number of the rectangular blocks obtained by dividing may be determined according to actual situations, and it is not specifically limited how many rectangular blocks the orientation configuration template is divided into. For example, as shown in the orientation configuration template of fig. 3, the orientation configuration template is divided into 9 rectangular blocks.
It should be further noted that the size of the orientation configuration template may be M times smaller than the building image, and the position in the orientation configuration template may also correspond to the position in the building image, so that the candidate building elements in the building image can be mapped into the orientation configuration template, and convenience is provided for the user to view the orientation configuration template.
At this time, when detecting whether part or all of a plurality of candidate building elements can form a preset building model, mapping the candidate building elements into rectangular blocks corresponding to the positions in the orientation configuration template according to the positions of the candidate building elements in the building image for each candidate building element; and then acquiring all target rectangular blocks mapped with the candidate building elements, and determining that the candidate building elements mapped in the partial or all rectangular blocks can form the preset building model when the arrangement mode formed by the detected partial or all target rectangular blocks is the same as the arrangement mode of the preset building model.
Specifically, the topological constraint of the suburban building area has a typical shape, for example, a typical letter type arrangement is a common layout of the building area, that is, the preset building model may be an L-type, I-type, rectangular model, and the like. Of course, the preset building models can be obtained by training the models through big data, so that the types of the preset building models are enriched.
In addition, specifically, the building image is projected based on the orientation configuration template, so that the same position corresponding to the position in the building image can be found in the orientation configuration template, that is, the candidate building elements can be mapped to the same position in the orientation configuration template according to the position in the building image, and at this time, a plurality of rectangular blocks are divided in the orientation configuration template, so that the candidate building elements can be mapped to the rectangular blocks at the same time. In addition, all candidate building elements can be mapped into the rectangular blocks of the azimuth configuration template one by one according to the mode, so that all target rectangular blocks mapped with the candidate building elements are obtained, and whether the mapped candidate building elements can form the preset building model or not can be determined through the arrangement mode of the target rectangular blocks.
Therefore, by forming the azimuth configuration template in advance, the problem of complexity of the changeable topological structure of the building area is solved, and more kinds of topological configurations can be covered; in addition, whether part or all of the candidate building elements can form a preset building model or not is detected through the orientation configuration template, and the multiple preset building models can be displayed based on the orientation configuration template, so that the efficiency of detecting whether part or all of the candidate building elements can form the preset building model or not is improved.
Step 104: and when detecting that part or all of the candidate building elements can form the preset building model, determining the area where the part or all of the candidate building elements forming the preset building model are located as a building area.
In this step, specifically, when it is detected that part or all of the candidate building elements can form the preset building model, it is determined that the candidate building elements are arranged according to the arrangement manner of the preset building model, and therefore, the area where the part or all of the candidate building elements forming the preset building model are located may be determined as the building area.
In this way, in the embodiment, the building image in which the building is shot is obtained, the candidate building elements in the building image are extracted to obtain the plurality of candidate building elements, and then, based on the pre-constructed orientation configuration template, when it is detected that part or all of the candidate building elements can form the preset building model, the area in which the part or all of the candidate building elements forming the preset building model are located is determined as the building area, so that the candidate building elements in the building image are extracted first, and then whether the candidate building elements are the building areas or not is accurately analyzed, and the efficiency of building area detection is improved.
Further, when extracting candidate building elements in the building image to obtain a plurality of candidate building elements, the embodiment may include the following steps:
extracting all straight line segments in the building image based on a straight line segment detection algorithm LSD; then acquiring target straight-line segments with parallel relation in all the straight-line segments; detecting whether the distance between two target straight-line segments is smaller than a preset distance threshold value or not; and when the distance between two target straight-line segments is smaller than the preset distance threshold value, connecting two ends of the two target straight-line segments to obtain a target rectangular area, and determining the target rectangular area as the candidate building element.
specifically, when all the straight line segments in the building image are extracted based on a straight line segment detection algorithm (LSD for short), the building image may be sampled first to reduce the calculation amount of screening, and then the straight line segment detection is performed. Specifically, 8-time gaussian sampling can be performed on the building image, gradient calculation is performed on the building image, a region growing algorithm is adopted to generate a support domain, and rectangle estimation is performed based on the support domain to obtain a straight-line segment.
In addition, particularly, due to the fact that a large number of complicated ground object areas such as vegetation areas exist in the remote sensing image, straight line segments at the edge of the roof of a building are prone to being blocked by trees under vegetation coverage, and due to the influences of shooting angles, light rays, blocking and other factors, most roofs of key structural parts of the building are not in a standard rectangular shape, but mostly have parallel line structures. Therefore, when a plurality of candidate building elements are extracted, the embodiment may acquire target straight-line segments having a parallel relationship among all the straight-line segments, and then, when it is detected that a distance between two target straight-line segments is smaller than a preset distance threshold, connect two ends of the two target straight-line segments to form a completed target rectangular area, thereby obtaining the candidate building elements.
therefore, candidate building elements are obtained through screening based on the parallel line principle, the problem of low detection efficiency caused by incomplete extraction of the building edges is greatly reduced, the candidate building elements, namely the extraction efficiency of the key building structure, is improved, and the overall detection efficiency of the building area is further improved.
in addition, specifically, after obtaining a plurality of candidate building elements, the present embodiment may further obtain a centroid point of each candidate building element, and identify each candidate building element through the centroid point.
In this way, because buildings are sparsely distributed in the remote sensing image and the shape difference is large, the candidate building elements can be replaced and identified through the centroid points of the candidate building elements, so that the problem of the intra-class deformation of the building area caused by the appearance of the buildings is reduced.
Furthermore, as shown in fig. 4, when detecting that part or all of the candidate building elements can constitute the preset building model, the embodiment may further include the following steps:
step 401: and determining a target building element from part or all of the candidate building elements forming the preset building model.
in this step, specifically, when the target building element is determined from the part or all of the candidate building elements constituting the preset building model, since the part or all of the candidate building elements constituting the preset building model are already mapped into the template rectangular blocks of the azimuth configuration template, the number of the candidate building elements in each template rectangular block may be obtained at this time.
At this time, when the number of candidate building elements in the target rectangular block is detected to be 1, the candidate building elements in the target rectangular block can be directly determined as the target building elements; and when the number of the candidate building elements in the target rectangular block is detected to be more than 1, clustering the candidate building elements in the target rectangular block, and determining a clustering center of mass obtained after clustering as the target building element.
for example, as shown in fig. 5, when a plurality of candidate building elements (shown by black solid dots in the figure) are respectively mapped in the four target rectangular blocks of the azimuth configuration template, the plurality of candidate building elements in each target rectangular block may be clustered, so that only one cluster centroid is finally reserved in each target rectangular block, and at this time, the cluster centroid may be determined as the target building element (shown by white circle center dots in the figure).
therefore, based on the fact that under vegetation coverage, straight-line segments at the edge of a roof are not completely displayed due to tree shielding, or a building key structure is in a non-standard rectangular shape due to factors such as shooting angles, light rays and shielding, a plurality of candidate building elements in one rectangular block may be the same building structure in a real scene, and at the moment, the clustering centroid obtained by clustering in each target rectangular block is determined as a target building element, so that the accuracy of the determined target building element is guaranteed.
step 402: and acquiring the deformation quantity of each target building element and the accumulated deformation quantity of all the target building elements.
In this step, specifically, after the target building elements are determined, the deformation amount of each target building element and the accumulated deformation amount of all target building elements can be obtained to perform deformation resistance discrimination, so that the orientation configuration template has a certain deformation adaptability, the capability of resisting intra-class variation is improved, and the distributed target detection is more suitable.
when the deformation amount of each target building element is obtained, the central position of the target rectangular block where the target building element is located can be obtained for each target building element, then the position offset between the target building element and the central position is calculated, and the position offset is determined as the deformation amount of the target building element.
Specifically, the position offset between the target building element and the center position can be calculated by the following formula:
Wherein the content of the first and second substances,
Representing the amount of positional offset between the target building element and the center position, x representing the position of the target building element in the target rectangular block, μ representing the mathematical expectation, 2 σ2the standard deviation is indicated.
in addition, when the accumulated deformation amounts of all the target building elements are acquired, the sum of the deformation amounts of all the target building elements may be acquired and determined as the accumulated deformation amount.
specifically, the formula of the accumulated deformation amount can be expressed by:
Wherein the content of the first and second substances,
Representing said accumulated deformation, N representing the number of target building elements, f1(x) Representing the amount of deformation of each target building element.
Step 403: and judging whether the area where part or all of candidate building elements forming the preset building model are located is determined as a building area or not according to the deformation amount of each target building element and the accumulated deformation amount.
in this step, specifically, after the deformation amount of each target building element and the accumulated deformation amount are calculated, it may be determined whether to determine an area where a part or all of candidate building elements constituting the preset building model are located as a building area according to the deformation amount of each target building element and the accumulated deformation amount.
specifically, when it is detected that the deformation amount of each target building element is smaller than a first threshold value, and the accumulated deformation amount is smaller than a second threshold value, it may be determined that the area where the part or all of the candidate building elements are located is determined as the building area. And the second threshold is less than or equal to N times of the first threshold, wherein N is the number of the target building elements.
Thus, through the steps, whether the area where the part or all of the candidate building elements forming the preset building model are located is determined as the building area is judged based on the deformation amount and the accumulated deformation amount of each target building element, so that the identification of the part or all of the candidate building elements forming the preset building model is realized, and the accuracy of building area detection is ensured.
In this way, according to the method for detecting the building area of the remote sensing image provided by the embodiment, the building image is primarily screened, the candidate building elements are extracted, and then whether part or all of the candidate building elements can form a preset building model is detected based on the orientation structural model, so that whether the area where the part or all of the candidate building elements are located is the building area is detected, the orientation topological relation among the comprehensive buildings is realized, the candidate building area is accurately judged again, and the high efficiency of building area detection and the accuracy of the detection result are further realized.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (8)
1. A method for detecting a building area by using a remote sensing image is characterized by comprising the following steps:
acquiring a building image shot with a building;
Extracting candidate building elements in the building image to obtain a plurality of candidate building elements;
Detecting whether part or all of the candidate building elements can form a preset building model or not based on a pre-constructed azimuth configuration template;
And when detecting that part or all of the candidate building elements can form the preset building model, determining the area where the part or all of the candidate building elements forming the preset building model are located as a building area.
2. The method for detecting the remote sensing image building area according to claim 1, wherein the step of extracting candidate building elements in the building image to obtain a plurality of candidate building elements comprises the steps of:
Extracting all straight line segments in the building image based on a straight line segment detection algorithm LSD;
acquiring target straight-line segments with parallel relation in all the straight-line segments;
Detecting whether the distance between two target straight-line segments is smaller than a preset distance threshold value or not;
and when the distance between two target straight-line segments is smaller than the preset distance threshold value, connecting two ends of the two target straight-line segments to obtain a target rectangular area, and determining the target rectangular area as the candidate building element.
3. the method for detecting the remote sensing image building area according to claim 1, wherein after extracting the candidate building elements in the building image to obtain a plurality of candidate building elements, the method further comprises:
And acquiring a centroid point of each candidate building element, and identifying each candidate building element through the centroid point.
4. the remote sensing image building area detection method according to claim 1, wherein the orientation configuration template is obtained by projection of the building image, and the orientation configuration template is divided into a plurality of rectangular blocks with the same size;
The detecting whether part or all of the candidate building elements can form a preset building model based on the pre-constructed orientation configuration template comprises the following steps:
For each candidate building element, mapping the candidate building element into a rectangular block corresponding to the position in the orientation configuration template according to the position of the candidate building element in the building image;
acquiring all target rectangular blocks mapped with candidate building elements;
and when the arrangement mode formed by part or all of the target rectangular blocks is the same as the arrangement mode of the preset building model, determining that the candidate building elements mapped in the part or all of the target rectangular blocks can form the preset building model.
5. the method for detecting the remote sensing image building area according to claim 4, wherein after detecting that part or all of the candidate building elements can form a preset building model, the method further comprises the following steps:
Determining a target building element from part or all of candidate building elements forming a preset building model;
Acquiring the deformation quantity of each target building element and the accumulated deformation quantity of all the target building elements;
and judging whether the area where part or all of candidate building elements forming the preset building model are located is determined as a building area or not according to the deformation amount of each target building element and the accumulated deformation amount.
6. The remote sensing image construction zone detection method according to claim 5, wherein the determining a target construction element from among part or all of candidate construction elements constituting a preset construction model comprises:
aiming at each target rectangular block, acquiring the number of candidate building elements in the target rectangular block; wherein the content of the first and second substances,
when the number of the candidate building elements in the target rectangular block is detected to be 1, directly determining the candidate building elements in the target rectangular block as target building elements;
And when the number of the candidate building elements in the target rectangular block is detected to be more than 1, clustering the candidate building elements in the target rectangular block, and determining a clustering center of mass obtained after clustering as the target building element.
7. The remote sensing image construction zone detection method according to claim 5,
acquiring the deformation quantity of each target building element, wherein the deformation quantity comprises the following steps:
aiming at each target building element, acquiring the central position of a target rectangular block where the target building element is located;
calculating the position offset between the target building element and the central position, and determining the position offset as the deformation amount of the target building element;
Acquiring the accumulated deformation quantity of all target building elements, including:
And acquiring the sum of the deformation quantities of all the target building elements, and determining the sum as the accumulated deformation quantity.
8. the method for detecting the remote sensing image building area according to claim 5, wherein the step of judging whether the area where part or all of candidate building elements forming a preset building model are located is determined as the building area according to the deformation amount of each target building element and the accumulated deformation amount comprises the steps of:
When the deformation amount of each target building element is smaller than a first threshold value and the accumulated deformation amount is smaller than a second threshold value, determining the area where the partial or all candidate building elements are located as a building area; and the second threshold is less than or equal to N times of the first threshold, wherein N is the number of the target building elements.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910768249.XA CN110569745B (en) | 2019-08-20 | 2019-08-20 | Remote sensing image building area detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910768249.XA CN110569745B (en) | 2019-08-20 | 2019-08-20 | Remote sensing image building area detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110569745A true CN110569745A (en) | 2019-12-13 |
| CN110569745B CN110569745B (en) | 2022-04-12 |
Family
ID=68773965
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910768249.XA Active CN110569745B (en) | 2019-08-20 | 2019-08-20 | Remote sensing image building area detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110569745B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111487643A (en) * | 2020-04-13 | 2020-08-04 | 中国科学院空天信息创新研究院 | Building detection method based on laser radar point cloud and near-infrared image |
| CN115049028A (en) * | 2022-08-17 | 2022-09-13 | 中建五局第三建设有限公司 | Construction area partitioning method, system, terminal and medium based on unsupervised learning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103971377A (en) * | 2014-05-27 | 2014-08-06 | 中国科学院遥感与数字地球研究所 | Building extraction method based on prior shape level set segmentation |
| CN105608691A (en) * | 2015-12-17 | 2016-05-25 | 武汉大学 | High-resolution SAR image individual building extraction method |
| CN105719306A (en) * | 2016-01-26 | 2016-06-29 | 郑州恒正电子科技有限公司 | Rapid building extraction method from high-resolution remote sensing image |
| CN107092877A (en) * | 2017-04-12 | 2017-08-25 | 武汉大学 | Remote sensing image roof contour extracting method based on basement bottom of the building vector |
-
2019
- 2019-08-20 CN CN201910768249.XA patent/CN110569745B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103971377A (en) * | 2014-05-27 | 2014-08-06 | 中国科学院遥感与数字地球研究所 | Building extraction method based on prior shape level set segmentation |
| CN105608691A (en) * | 2015-12-17 | 2016-05-25 | 武汉大学 | High-resolution SAR image individual building extraction method |
| CN105719306A (en) * | 2016-01-26 | 2016-06-29 | 郑州恒正电子科技有限公司 | Rapid building extraction method from high-resolution remote sensing image |
| CN107092877A (en) * | 2017-04-12 | 2017-08-25 | 武汉大学 | Remote sensing image roof contour extracting method based on basement bottom of the building vector |
Non-Patent Citations (2)
| Title |
|---|
| 毕福昆 等: "机载复杂遥感场景下特定建筑区检测跟踪算法", 《电子学报》 * |
| 赵赛男: "基于机载LiDAR的规则建筑物自动重建研究", 《中国优秀博硕士学位论文全文数据库(硕士)工程科技Ⅱ辑》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111487643A (en) * | 2020-04-13 | 2020-08-04 | 中国科学院空天信息创新研究院 | Building detection method based on laser radar point cloud and near-infrared image |
| CN111487643B (en) * | 2020-04-13 | 2021-06-08 | 中国科学院空天信息创新研究院 | Building detection method based on laser radar point cloud and near-infrared image |
| CN115049028A (en) * | 2022-08-17 | 2022-09-13 | 中建五局第三建设有限公司 | Construction area partitioning method, system, terminal and medium based on unsupervised learning |
| CN115049028B (en) * | 2022-08-17 | 2022-12-13 | 中建五局第三建设有限公司 | Construction area partitioning method, system, terminal and medium based on unsupervised learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110569745B (en) | 2022-04-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020258899A1 (en) | Mountain landscape architecture extraction method and equipment based on high-resolution remote sensing image | |
| Guo et al. | Mining parameter information for building extraction and change detection with very high-resolution imagery and GIS data | |
| Rottensteiner et al. | The ISPRS benchmark on urban object classification and 3D building reconstruction | |
| Culvenor | TIDA: an algorithm for the delineation of tree crowns in high spatial resolution remotely sensed imagery | |
| WO2017020528A1 (en) | Lane line recognition modeling method, apparatus, storage medium, and device, recognition method and apparatus, storage medium, and device | |
| US9064151B2 (en) | Device and method for detecting plantation rows | |
| Chen et al. | Building detection in an urban area using lidar data and QuickBird imagery | |
| Cheng et al. | Robust affine invariant feature extraction for image matching | |
| Nyaruhuma et al. | Verification of 2D building outlines using oblique airborne images | |
| CN105389799B (en) | SAR image object detection method based on sketch map and low-rank decomposition | |
| CN110569745B (en) | Remote sensing image building area detection method | |
| CN107330875A (en) | Based on the forward and reverse heterogeneous water body surrounding enviroment change detecting method of remote sensing images | |
| CN108898159B (en) | False forest fire hot spot filtering method based on DBSCAN algorithm | |
| CN107301649B (en) | Regional merged SAR image coastline detection algorithm based on superpixels | |
| CN107992856A (en) | High score remote sensing building effects detection method under City scenarios | |
| Hui et al. | Multi-level self-adaptive individual tree detection for coniferous forest using airborne LiDAR | |
| CN112668461A (en) | Intelligent supervision system with wild animal identification function | |
| Hao et al. | A hierarchical region-merging algorithm for 3-D segmentation of individual trees using UAV-LiDAR point clouds | |
| CN112085778A (en) | Oblique photography illegal building detection method and system based on superpixels and morphology | |
| Recky et al. | Window detection in complex facades | |
| CN113869133A (en) | Method, device and equipment for detecting change of remote sensing image and storage medium | |
| Michelin et al. | Quality evaluation of 3D city building models with automatic error diagnosis | |
| CN106570506B (en) | Solar activity recognition method based on scale transformation model | |
| Zhang et al. | Automatic extraction of building geometries based on centroid clustering and contour analysis on oblique images taken by unmanned aerial vehicles | |
| WO2023116359A1 (en) | Method, apparatus and system for classifying green, blue and gray infrastructures, and medium |
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 |