CN113822247B - Method and system for identifying illegal building based on aerial image - Google Patents

Method and system for identifying illegal building based on aerial image Download PDF

Info

Publication number
CN113822247B
CN113822247B CN202111386033.0A CN202111386033A CN113822247B CN 113822247 B CN113822247 B CN 113822247B CN 202111386033 A CN202111386033 A CN 202111386033A CN 113822247 B CN113822247 B CN 113822247B
Authority
CN
China
Prior art keywords
violation
building
image
area
aerial
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
Application number
CN202111386033.0A
Other languages
Chinese (zh)
Other versions
CN113822247A (en
Inventor
黄山
王宇翔
郑少杰
林俊杰
舒世嘉
张澳腾
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangdong Airace Technology Development Co ltd
Original Assignee
Guangdong Airace Technology Development Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Guangdong Airace Technology Development Co ltd filed Critical Guangdong Airace Technology Development Co ltd
Priority to CN202111386033.0A priority Critical patent/CN113822247B/en
Publication of CN113822247A publication Critical patent/CN113822247A/en
Application granted granted Critical
Publication of CN113822247B publication Critical patent/CN113822247B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods

Abstract

The application relates to an identification method and system for illegal buildings based on aerial images. The method comprises the following steps: acquiring a first aerial image; performing region segmentation according to the first aerial image to obtain a region segmentation image; obtaining difference region information based on a comparison result of the second aerial image and the region segmentation image; the difference region information includes: a difference region coordinate; the first aerial image and the second aerial image are images of the same area, which have the same icon size and shooting angle and are shot at different time; acquiring a first building height of the difference area according to the coordinates of the difference area; obtaining violation assessment information of the difference area based on the comparison result of the first building height and the second building height; the violation assessment information includes: violation evaluation results and violation building coordinates; and initiating a evidence obtaining instruction to the violation management terminal according to the violation evaluation information. The scheme provided by the application can improve the accuracy of illegal building identification.

Description

Method and system for identifying illegal building based on aerial image
Technical Field
The application relates to the technical field of image recognition, in particular to a method and a system for recognizing illegal buildings based on aerial images.
Background
Along with the rapid development of national economy, the urbanization speed of China is higher and higher, the number of various buildings in cities and towns is increased sharply, and meanwhile, the maturity of the unmanned aerial vehicle aerial photography technology enables cities and towns to be widely applied to city planning tasks, images and videos. In the urban planning process, various illegal buildings not only occupy social public resources and urban space, but also bring harm to economic order of market and social management, and are important problems to be solved urgently in the urban management process. Method for producing a composite material
Once the violation buildings can not be stopped and dismantled in time at the initial construction stage, the difficulty of subsequent dismantling work is increased and higher dismantling cost is brought, so that the violation buildings need to be warned in time. How to early warn the existing illegal buildings in time becomes a problem to be solved urgently at present.
At present, aiming at monitoring of buildings against regulations, a scheme of monitoring by mounting a camera and combining manual identification is generally adopted, so that the efficiency is relatively low, and human resources are wasted.
In the related art, patent document CN112200769A discloses a fixed-point monitoring new and old time phase image change detection method for violation detection, which performs data preprocessing operations such as registration and histogram matching on a new time phase image and an old time phase image of a monitored point, then constructs a difference image by fusing a logarithm ratio method and a mean ratio method, and segments the difference image by using a level set model, then extracts a contour of a change region and calculates a minimum rectangle of a vertical boundary of the contour, and finally merges overlapping rectangles and draws the overlapping rectangles on the new time phase image to obtain a change detection effect graph, thereby determining whether the monitored point has a violation phenomenon.
However, in the above-mentioned scheme, only whether the building image changes or not is taken as a criterion for illegal building judgment, and in actual illegal building management, situations such as building floor removal, building planning, building land restoration and the like exist, which all generate large differences on aerial photography images but do not belong to the category of illegal buildings, and since the above-mentioned scheme judges only the degree of change of the building image, the situation that the degree of change is large but does not belong to illegal buildings is mistakenly judged as illegal buildings, and the monitoring accuracy of illegal buildings is low.
Disclosure of Invention
In order to solve the problems in the related art, the application provides the illegal building identification method based on the aerial image, which can avoid the situation that the situation with large change degree but not illegal building is mistaken to be the illegal building, thereby improving the accuracy of illegal building identification.
The application provides a violation building identification method based on aerial images in a first aspect, and the method comprises the following steps:
acquiring a first aerial image;
performing region segmentation according to the first aerial image to obtain a region segmentation image;
obtaining difference region information based on a comparison result of the second aerial image and the region segmentation image; the difference region information includes: a difference region coordinate; the first aerial image and the second aerial image are images of the same area, which have the same icon size and shooting angle and are shot at different time;
acquiring a first building height of the difference area according to the coordinates of the difference area;
obtaining violation assessment information of the difference area based on the comparison result of the first building height and the second building height; the violation assessment information includes: violation evaluation results and violation building coordinates;
and initiating a evidence obtaining instruction to the violation management terminal according to the violation evaluation information.
In one embodiment, obtaining violation assessment information based on a comparison of the first building height and the second building height comprises:
subtracting the second building height from the first building height to obtain a building height difference;
judging whether the building height difference is larger than zero, if so, judging that the violation evaluation result of the difference area is a violation building, and taking the coordinates of the difference area as the coordinates of the violation building;
if not, judging that the violation evaluation result of the difference area is a construction without violations.
In one embodiment, obtaining the difference region information based on the comparison result of the second aerial image and the region segmentation image includes:
judging whether the area segmentation image meets a preset inspection-free condition, if not, judging whether the area segmentation image changes according to the second aerial image, if so, judging that the area corresponding to the area segmentation image is a difference area, and acquiring a difference area coordinate of the difference area;
the preset inspection-free conditions comprise: the area of the area segmentation image is smaller than the detection area threshold, the area segmentation image meets the soil accumulation area characteristic and the area is smaller than the soil accumulation area threshold, the area segmentation image meets the water area characteristic or the area segmentation image meets the green land characteristic.
In one embodiment, the determining whether the area segmentation image changes according to the second aerial image includes:
utilizing a histogram matching algorithm to perform illumination equalization processing on the region segmentation image to obtain an equalized region segmentation image;
graying the balanced region segmentation image and the second aerial image to obtain a region grayscale image and a second grayscale image;
calculating to obtain the Hamming distance between the regional gray image and the second gray image by using a difference Hash algorithm;
and judging whether the Hamming distance is larger than a Hamming distance threshold value or not.
In one embodiment, acquiring a first building height for a difference region from difference region coordinates includes:
controlling the aerial photographing device to detect the echo duration through an electromagnetic wave distance measuring technology under the position of the coordinate of the difference region and the preset distance measuring height;
and calculating to obtain a first building height based on the echo time length and the preset ranging height.
In one embodiment, the difference region coordinates include: region center coordinates and N region edge coordinates; n is an integer greater than or equal to zero;
acquiring a first building height of the difference area according to the difference area coordinates, wherein the first building height comprises the following steps:
acquiring N +1 first building heights of the difference region according to the region center coordinate and the N region edge coordinates;
obtaining violation assessment information of the difference area based on the comparison result of the first building height and the second building height, wherein the violation assessment information comprises:
calculating the number of the first building heights with the height greater than the second building height in the N +1 first building heights to obtain the height increase number;
judging whether the increase number is greater than or equal to ⌊ (N +1)/2 ⌋, if so, judging that the violation assessment result of the difference area is a violation building, and taking the area center coordinate as a violation building coordinate;
if not, judging that the violation evaluation result of the difference area is a construction without violations.
In one embodiment, performing region segmentation from the first aerial image comprises:
and carrying out region segmentation on the first aerial image by utilizing the aerial building segmentation model.
In one embodiment, prior to performing region segmentation on the first aerial image using the aerial building segmentation model, the method comprises:
acquiring a building aerial photo album; building aerial photograph album includes: building aerial images and building outline labeling;
carrying out data enhancement processing on the building aerial photo album to obtain an extended album;
dividing an expansion diagram set into a training set, a verification set and a test set;
performing iterative training on the Mask ordering R-CNN model by using the training set as the input of the Mask ordering R-CNN model to obtain a primary segmentation model;
calling a verification set to evaluate the primary segmentation model to obtain network parameters of the primary segmentation model;
calling a test set to test the primary segmentation model, judging whether the deviation between the building segmentation result output by the primary segmentation model and the building outline label in the test set is smaller than a deviation threshold value, if so, outputting the primary segmentation model as an aerial photography building segmentation model;
if not, reconstructing the network parameters of the primary segmentation model until the deviation between the building segmentation result output by the primary segmentation model and the building outline label in the test set is smaller than a deviation threshold value.
In one embodiment, after initiating evidence obtaining indication to the violation management terminal according to the violation identification information, the method comprises the following steps:
classifying according to the violation evidence obtaining information obtained by evidence obtaining according to administrative divisions to obtain M district-level violation data sets; m is a positive integer; violation forensics information, including: the violation evidence obtaining date, the evidence obtaining person information, the violation case number, the violation assessment information and the violation type;
respectively carrying out data statistics of violation evidence obtaining information on the M district-level violation data sets to obtain violation building analysis reports;
in the data statistics of the violation evidence obtaining information of M district-level violation data sets, the data statistics of the violation evidence obtaining information of one district-level violation data set comprises the following steps:
calculating according to the violation building coordinates and the area segmentation image to obtain the violation building area, and associating the violation building area with the violation case number;
calculating to obtain the zone-level violation number according to the violation case number;
calculating to obtain the area of the district-level violation type corresponding to each violation type according to the violation type and the area of the violation building;
and calculating to obtain the zone-level violation type number corresponding to each violation type according to the violation type and the violation case number.
In one embodiment, the violation assessment information further comprises: the type of the place for the illegal building and the height of the illegal building are increased; the illegal building land type is determined by searching in a building land type mapping table based on the illegal building coordinate; the illegal building heightening height is a height difference obtained by subtracting the second building height from the first building height;
initiating a evidence obtaining instruction to the violation management terminal according to the violation evaluation information, comprising the following steps:
judging whether the height of the illegal building is greater than a second height increasing threshold value or not, if so, initiating an emergency evidence obtaining instruction to the illegal management terminal;
if not, judging whether the heightening height of the illegal building is greater than a first heightening threshold value, if so, initiating a preferential evidence obtaining instruction to the illegal management terminal;
if not, initiating an instruction to be subjected to evidence obtaining to the violation management terminal;
the first boost threshold is less than the second boost threshold.
The application second aspect provides a building identification system violating regulations based on image of taking photo by plane, includes:
the processing device is used for executing the illegal building identification method based on the aerial image;
the aerial photographing device is used for acquiring aerial photographing images according to the instruction of the processing device;
and the distance measuring device is used for measuring the height above the ground of the building according to the indication of the processing device.
The technical scheme provided by the application can comprise the following beneficial effects:
according to the method and the device, the first aerial image is firstly segmented according to the area to obtain the area segmentation image, and the areas with differences are screened out by utilizing an image comparison technology, so that the workload of subsequent violation judgment is reduced, and the violation building identification efficiency is improved; and after the screening of the difference area is completed, data acquisition is carried out again, the first building height under the coordinate of the difference area is acquired, and the first building height and the second building height are compared, so that the height change condition of the building in the difference area is judged.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application.
Drawings
The foregoing and other objects, features and advantages of the application will be apparent from the following more particular descriptions of exemplary embodiments of the application, as illustrated in the accompanying drawings wherein like reference numbers generally represent like parts throughout the exemplary embodiments of the application.
Fig. 1 is a first process schematic diagram of a violation building identification method based on aerial images according to an embodiment of the application;
fig. 2 is a second process schematic diagram of the illegal building identification method based on aerial images according to the embodiment of the application;
fig. 3 is a third process schematic diagram of the illegal building identification method based on aerial images according to the embodiment of the application;
FIG. 4 is a flow chart of a method for constructing an aerial photography building segmentation model according to an embodiment of the present application;
fig. 5 is a flowchart illustrating a method for acquiring difference region information according to an embodiment of the present application;
fig. 6 is a flowchart illustrating a forensic instruction generation method according to an embodiment of the present application;
fig. 7 is a schematic flow chart of a method for obtaining a violation building analysis report according to an embodiment of the present application;
fig. 8 is a schematic structural diagram of a violation building identification system based on aerial images according to an embodiment of the application;
FIG. 9-a is a schematic diagram of a second region image in a second aerial image shown in an embodiment of the present application;
FIG. 9-b is a schematic alignment chart of FIG. 9-a shown in an embodiment of the present application;
FIG. 10-a is another schematic diagram of a second region image in a second aerial image shown in an embodiment of the present application;
FIG. 10-b is a schematic alignment diagram of FIG. 10-a shown in an embodiment of the present application.
Detailed Description
Preferred embodiments of the present application will be described in more detail below with reference to the accompanying drawings. While the preferred embodiments of the present application are shown in the drawings, it should be understood that the present application may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in this application and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
It should be understood that although the terms "first," "second," "third," etc. may be used herein to describe various information, these information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present application. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the present application, "a plurality" means two or more unless specifically limited otherwise.
Example one
The patent document CN112200769A discloses that only whether a building image has changed is used as a criterion for illegal building determination, but in actual illegal building management, there are cases such as building floor removal, building planning, and building land greening, which all have great difference in aerial photography images but do not belong to the category of illegal buildings.
In view of the above problems, the embodiment of the application provides an identification method for a violation building based on an aerial image, which can avoid the situation that the situation with a large change degree but not illegal building is mistakenly judged as the violation building, so that the accuracy of the violation building identification is improved.
The technical solutions of the embodiments of the present application are described in detail below with reference to the accompanying drawings.
Fig. 1 is a first process schematic diagram of a violation building identification method based on aerial images according to an embodiment of the application.
Referring to fig. 1, the illegal building identification method based on aerial images comprises the following steps:
101. acquiring a first aerial image;
in the embodiment of the application, the first aerial image is obtained by shooting at the selected position and height by the aerial photographing device at the set angle, and in practical application, fixed-point aerial photographing at multiple positions can be performed on the violation assessment area, so that the obtained multiple first aerial images can cover the whole violation assessment area.
In the practical application process, the longitude/latitude coordinate where the aerial photographing device executes the photographing action, the photographing angle of the aerial photographing device and the photographing parameters of the aerial photographing device can be set in real time according to the practical situation, or photographing can be directly carried out according to the preset aerial photographing strategy.
It is to be understood that the above description of the acquisition process of the first aerial image is only an example in the embodiments of the present application and should not be taken as the only limitation of the present application.
102. Performing region segmentation according to the first aerial image to obtain a region segmentation image;
since the first aerial image covers a large range and includes a plurality of building areas, it is necessary to perform area division processing on the first aerial image in advance for facilitating image comparison by a computer and also for accuracy of image comparison, and after obtaining the area division image, image comparison is performed in units of building areas.
103. Obtaining difference region information based on a comparison result of the second aerial image and the region segmentation image;
wherein the difference region information includes: a difference region coordinate; in this embodiment of the present application, the difference region information may further include: the area of the difference region and a comparison map of the difference region; the difference region comparison map comprises: the local images corresponding to the difference regions in the first aerial image and the local images corresponding to the difference regions in the second aerial image.
In the embodiment of the present application, the coordinates of the difference region may be embodied in the form of longitude/latitude coordinates. The disparity region coordinates can be coordinates of the geometric center of the disparity region.
In the embodiment of the application, the first aerial image and the second aerial image are images of the same area, which have the same icon size and shooting angle and are shot at different time; it can be understood that the first aerial image and the second aerial image are aerial images of the same aerial device at the same position and height and at different times of the same area and at the same shooting angle under the same shooting parameters, the light conditions of the first aerial image and the second aerial image are approximately the same, and the shooting time of the first aerial image is later than that of the second aerial image.
Referring to fig. 9-a to 10-b, in the embodiment of the present application, for the purpose of comparing the accuracy of the result, the second aerial image needs to be preprocessed before the image comparison, that is, the region-based segmented image needs to be located in the corresponding region on the second aerial image, the region image on the second aerial image needs to be extracted, and then the region image is compared with the segmented region image. Step 103 is to compare the area segmentation image with the second area image at the corresponding position on the second aerial image, and determine the difference area information according to the comparison result. In the embodiment of the application, if the obtained comparison result is that the difference is large, it is indicated that the building of the area is changed greatly, and the area is considered to have a risk of violation of the badge, the coordinate of the geometric center of the area is taken as the coordinate of the difference area and is added to the information of the difference area; if the obtained comparison result shows that the difference is small, the building of the area is unchanged, the area is considered to have no violation risk, and subsequent operation is not required to be performed on the area.
It should be noted that, in the practical application process, the second aerial image may be subjected to region segmentation processing in advance, after the second region image is obtained, the corresponding second region image is found according to the region coordinate information of the region segmentation image, and then image comparison is performed; or taking the area segmentation image as a positioning template, performing positioning identification on the second aerial image to obtain a second area image, and then performing image comparison.
It is to be understood that the above description of the pre-processing applied to the second aerial image is only an example given in the embodiments of the present application, and may be adjusted according to actual situations, and should not be taken as the only limitation of the present application.
In the embodiment of the application, the algorithm adopted in the image comparison process is not strictly required, and in the practical application process, the hamming distance between the region segmentation image and the image at the corresponding position on the second aerial image can be calculated, and whether the region is a difference region or not can be judged according to the hamming distance, or a histogram comparison method is adopted for image comparison, or a BOW model is adopted to be combined with a K-means clustering algorithm for image comparison.
It will be appreciated that the algorithms employed for image comparison enumerated above do not constitute the only limitations of the present application.
104. Acquiring a first building height of a difference area according to the difference area coordinates;
the following are exemplary:
controlling the aerial photographing device to obtain echo duration through detection of an electromagnetic wave ranging technology at the position of the difference region coordinate and a preset ranging height;
and calculating to obtain the first building height based on the echo duration and the preset ranging height.
In the embodiments of the present application, the electromagnetic wave ranging technology includes, but is not limited to: laser ranging technology, ultrasonic ranging technology or microwave radar ranging technology; the method comprises the steps that echo time length can be obtained through the electromagnetic wave distance measuring technology, the height difference between an aerial photographing device and a building can be calculated according to the echo time length, and then the building height of the building, namely the first building height, is calculated according to the preset distance measuring height and the height difference; specifically, the first building height can be obtained by subtracting the height difference from the preset ranging height.
The preset distance measurement height is lower than the shooting height of the first aerial image, and can be preset in the actual application process and adjusted according to the actual situation.
105. Obtaining violation assessment information of the difference area based on the comparison result of the first building height and the second building height;
the violation assessment information includes: violation evaluation results and violation building coordinates;
the following are exemplary:
subtracting the second building height from the first building height to obtain a building height difference;
judging whether the building height difference is larger than zero, if so, judging that the violation assessment result of the difference area is a violation building, and taking the coordinates of the difference area as the coordinates of the violation building;
if not, judging that the violation evaluation result of the difference area is a construction without violations.
In the embodiment of the application, when the building height difference is greater than zero, compared with the shooting of the second aerial image, the detected height of the building is increased by an increment in the shooting time period of the first aerial image, namely the building is increased by the building height and belongs to a violation building; when the building height difference is equal to zero, the detected building height is unchanged and does not belong to a violation building when the building height difference is compared with that shot by the second aerial image and the detected building height is not changed in the first aerial image shooting time period; when the building height difference is smaller than zero, the detected height of the building is reduced in the first aerial image shooting time period compared with the second aerial image shooting time period, namely the building is dismantled and does not belong to a violation building.
In practical applications, step 105 may also be performed as follows:
determining whether the first building height is greater than the second building height,
if so, judging that the violation evaluation result of the difference area is a violation building, and taking the coordinates of the difference area as the coordinates of the violation building;
if not, judging that the violation evaluation result of the difference area is a construction without violations.
It should be noted that the method for comparing the first building height with the second building height is not exclusive, and the above description of the process for comparing the first building height with the second building height is only an example given in the embodiment of the present application, and does not constitute the only limitation to the present application.
It should be noted that the violation assessment result is a preliminary identification result of the violation building identification obtained by the computer processing, and is used as a basis for initiating a evidence obtaining indication to the violation management terminal, rather than being a final judgment result of the violation building. Specifically, the violation evaluation result is that the violation building explanation computer evaluates whether a violation building exists in the building area and whether the building needs to be dismantled or not based on the first aerial image and the first building height, and a final conclusion is obtained after a forensics person obtains evidence on the spot; and if the violation evaluation result is that the violation building explanation computer evaluates that no violation building exists in the building area based on the first aerial image and the first building height, the evidence obtaining personnel is not needed to obtain evidence on the spot.
106. And initiating a evidence obtaining instruction to the violation management terminal according to the violation evaluation information.
In the embodiment of the application, after the computer finishes the discrimination of the illegal building, the computer sends the coordinates of the illegal building, the area segmentation image of the corresponding area of the illegal building and the second area image to the illegal management terminal according to the violation evaluation result so as to remind related evidence obtaining personnel to carry out evidence obtaining work.
In the embodiment of the application, the violation management terminal can be a fixed violation management system operation desk or a handheld violation management system operator, and can also be an intelligent terminal such as a mobile phone or a computer with a violation management system.
In the embodiment of the application, the first aerial image is segmented according to the area to obtain the area segmentation image, and the areas with differences are screened out by utilizing an image comparison technology, so that the workload of subsequent violation judgment is reduced, and the violation building identification efficiency is improved; and after the screening of the difference area is completed, data acquisition is carried out again, the first building height under the coordinate of the difference area is acquired, and the first building height and the second building height are compared, so that the height change condition of the building in the difference area is judged.
Example two
In practical application, after the aerial image is subjected to region segmentation, a large number of green areas and water areas can be obtained, the areas do not need to be subjected to illegal building determination substantially, and if the area segmentation image is not screened, and image comparison operation is performed on each area, the workload is huge, and a large amount of comparison time is consumed.
The technical solutions of the embodiments of the present application are described in detail below with reference to the accompanying drawings.
Fig. 2 is a second process schematic diagram of the illegal building identification method based on the aerial image according to the embodiment of the application.
The illegal building identification method based on the aerial image comprises the following steps:
201. acquiring a first aerial image;
in the embodiment of the present application, the content of step 201 is the same as that of step 101 in the first embodiment, and details are not described here.
202. Performing region segmentation according to the first aerial image to obtain a region segmentation image;
in the embodiment of the application, the first aerial image can be subjected to region segmentation by utilizing an aerial building segmentation model, the aerial building segmentation model belongs to an example segmentation model, and the boundary of each building region can be marked on the first aerial image, so that a plurality of region segmentation images are obtained.
In the practical application process, high-precision depth and dense three-dimensional point cloud information of a building can be obtained through the multi-angle aerial image, three-dimensional point cloud feature statistical information is obtained through extraction of the dense three-dimensional point cloud information, semantic classification is carried out on a three-dimensional point cloud scene through global energy optimization, and then fine segmentation of the building in the aerial image is achieved.
It should be noted that different schemes can be selected according to actual conditions and actual conditions in the specific method for obtaining the region segmentation image, in the embodiment of the present application, an aerial photography building segmentation model is adopted, and in actual application, other schemes capable of realizing image segmentation can be adopted, that is, the above description on the aerial photography building segmentation model and the semantic classification of the three-dimensional point cloud scene does not form the only limitation on the present application.
203. Judging whether the area segmentation image meets a preset inspection-free condition, if not, judging whether the area segmentation image changes according to the second aerial image, if so, judging that an area corresponding to the area segmentation image is the difference area, and acquiring the difference area coordinates of the difference area;
if the area segmentation image meets the preset inspection-free condition, updating the area segmentation image and judging whether the updated area segmentation image meets the preset inspection-free condition again until the updated area segmentation image meets the preset inspection-free condition; in practical application, the above process can be understood as follows: and the computer traverses the plurality of region segmentation images to judge whether the preset inspection-free condition is met.
And if the result of judging whether the area segmentation image changes is negative according to the second aerial image, the fact that the area corresponding to the area segmentation image does not have the violation risk is shown, and the area segmentation image does not need to be processed.
Wherein, the preset non-inspection condition comprises: the area of the region segmentation image is smaller than a detection area threshold value, the region segmentation image meets the soil accumulation region characteristic and the area is smaller than the soil accumulation area threshold value, the region segmentation image meets the water area characteristic or the region segmentation image meets the green land characteristic.
In the embodiment of the present application, the range of the detection area threshold is 1m2To 40m2(ii) a Preferably, the value range of the detection area threshold is 30m2And in practical application, the specific value of the detection area threshold can be adjusted according to the identification requirement of the illegal building.
In the embodiment of the present application, the region segmentation image satisfies the feature of the mound region, that is, the region is a loess coverage region, that is, an empty region without a building, but since a large amount of loess is required for construction engineering, there is a situation that too much loess is accumulated in the mound region in practice, and an empty region with an excessively large area is occupied illegally, so that the area of the mound region needs to be limited, that is, the image comparison needs to be performed on the mound region greater than or equal to the mound area threshold, in the embodiment of the present application, the value range of the mound area threshold is 100m2To 150m2。
In practical applications, the method for determining whether the corresponding region of the region segmentation image is a mound region is not unique, for example: the standard soil-piled region image can be used as a template, after the feature extraction is carried out on the region segmentation image, the feature comparison is carried out by using a feature comparison method, so that whether the region corresponding to the region segmentation image is the soil-piled region or not is judged, and specifically, the feature comparison can be carried out by using an SIFT algorithm or an ORB algorithm; or classifying the region segmentation images by using a BP neural network, thereby screening the region segmentation images corresponding to the soil accumulation region from a large number of region segmentation images.
It is to be understood that the above description of the method for determining whether the region segmentation image satisfies the feature of the mound region is only an example given in the embodiment of the present application, and does not constitute a unique limitation to the present application.
In the embodiment of the present application, it may also be determined whether the region segmentation image satisfies the water area feature or the green space feature by using a feature comparison method or a BP neural network classification method.
204. Acquiring a first building height of a difference area according to the difference area coordinates;
in the embodiment of the present application, the content of step 204 is the same as that of step 104 in the first embodiment, and is not described herein again.
205. Obtaining violation assessment information of the difference area based on the comparison result of the first building height and the second building height;
the violation assessment information includes: violation assessment results and violation building coordinates.
The following are exemplary:
subtracting the second building height from the first building height to obtain a building height difference;
judging whether the building height difference is larger than zero, if so, judging that the violation assessment result of the difference area is a violation building, and taking the coordinates of the difference area as the coordinates of the violation building;
if not, judging that the violation evaluation result of the difference area is a construction without violations.
Or
Judging whether the first building height is greater than the second building height, if so, judging that the violation evaluation result of the difference area is a violation building, and taking the coordinates of the difference area as the coordinates of the violation building;
if not, judging that the violation evaluation result of the difference area is a construction without violations.
It should be noted that the method for comparing the first building height with the second building height is not exclusive, and the above description of the process for comparing the first building height with the second building height is only an example given in the embodiment of the present application, and does not constitute the only limitation to the present application.
206. And initiating a evidence obtaining instruction to the violation management terminal according to the violation evaluation information.
In the embodiment of the present application, the content of step 206 is the same as that of step 106 in the first embodiment, and is not described herein again.
In the embodiment of the application, after the first aerial image is subjected to region segmentation to obtain the region segmentation image, the region segmentation image is screened by using the preset inspection-free condition, and the region with the too small area, the mound region smaller than the mound area threshold, the water region and the greenbelt region are screened out, so that the image comparison task of determining the region without the violation risk is omitted, the workload of computer image comparison is greatly simplified, the processing speed of the image comparison is improved, and the efficiency of illegal building identification is improved.
EXAMPLE III
In the process of identifying the illegal buildings, the areas of the different areas are large, and in actual conditions, the illegal behaviors of part of illegal buildings cannot cover the whole area, for example, a roof is covered on a house, and the area of the covered roof is only half of the area of the house; or the loess accumulated in the mound area only occupies one corner in the mound area and does not occupy the whole mound area, so that misjudgment is caused when whether the difference area has the violation risk according to the height change condition, and therefore, in order to improve the accuracy of the violation building judging process, the embodiment of the application provides the violation building identifying method based on the aerial photography image.
The technical solutions of the embodiments of the present application are described in detail below with reference to the accompanying drawings.
Fig. 3 is a third process schematic diagram of the illegal building identification method based on the aerial image according to the embodiment of the application.
The illegal building identification method based on the aerial image comprises the following steps:
301. acquiring a first aerial image;
in the embodiment of the present application, the content of step 301 is the same as that of step 101 in the first embodiment, and details are not described here.
302. Performing region segmentation according to the first aerial image to obtain a region segmentation image;
in the embodiment of the present application, the content of step 302 is the same as that of step 102 in the first embodiment, and is not described herein again.
303. Obtaining difference region information based on a comparison result of the second aerial image and the region segmentation image;
wherein the difference region information includes: a difference region coordinate; the difference region coordinates include: region center coordinates and N region edge coordinates; n is an integer greater than or equal to zero;
in this embodiment of the present application, the area center coordinate is a longitude/latitude coordinate of a geometric center of the area, the area edge coordinates are circumferentially distributed in the difference area around the geometric center as a circle center, and distances between the N area edge coordinates and the geometric center may be adjusted according to an actual situation, which is not limited herein.
304. Acquiring N +1 first building heights of the difference region according to the region center coordinate and the N region edge coordinates respectively;
in the embodiment of the application, the aerial photographing device respectively measures the distance by using an electromagnetic wave distance measurement technology in the area center coordinate and the N area edge coordinates to obtain N +1 first building heights.
305. Calculating the number of the first building heights which are greater than the second building height in the N +1 first building heights to obtain an increase number;
in this application embodiment, with N +1 first building height is corresponding, has in advance among the image based on taking photo by plane's the building identification system violating regulations the regional center coordinate with N +1 second building height of a regional edge coordinate will N +1 first building height with N +1 second building height compares according to the coordinate one by one to calculate in N +1 first building height, is greater than the quantity of its first building height of corresponding second building height, obtains to increase the number.
306. Obtaining violation evaluation information according to the increase number;
the following are exemplary:
judging whether the increased number is greater than or equal to ⌊ (N +1)/2 ⌋, if so, judging that the violation assessment result of the difference area is a violation building, and taking the area center coordinate as a violation building coordinate;
if not, judging that the violation evaluation result of the difference area is a construction without violations.
In the embodiment of the application, ⌊ (N +1)/2 ⌋ represents that (N +1)/2 is rounded downwards, if the condition that the first building height is greater than the second building height is greater than or equal to ⌊ (N +1)/2 ⌋, the condition that the building height of the difference area reaches the violation requirement can be judged, the violation evaluation result of the difference area is judged to be a violation building, and the area center coordinate of the difference area is used as the violation building coordinate and added to the violation evaluation information;
if the first building height is greater than the second building height and is less than ⌊ (N +1)/2 ⌋, the situation that the building is increased in the difference area is an accidental situation, in the practical application process, sundries or loess may be stacked in a local area, the situation does not belong to a violation building, and therefore the violation assessment result of the difference area is judged to be the violation-free building.
307. And initiating a evidence obtaining instruction to the violation management terminal according to the violation evaluation information.
In the embodiment of the present application, the content of step 307 is the same as that of step 106 in the first embodiment, and details are not described here.
The method for identifying the illegal buildings based on the aerial images comprises the steps of collecting the height of a first building at a plurality of coordinate positions of a difference area, wherein the coordinate positions comprise an area center coordinate and N area edge coordinates, so that the heightening detection of the multi-position buildings in the difference area is realized, and when the heightening detection of the unit built buildings is avoided, the selected unit is a special position, for example, a sundry stacking point or a vacant area with a roof is added, so that illegal building identification errors are caused, the illegal sundry stacking point is identified as the illegal building or the illegal roof with the roof is identified as the illegal building, and the accuracy and the reliability of illegal building identification are improved.
Example four
For step 102 in the first embodiment, the first aerial image is subjected to region segmentation by using an aerial building segmentation model to obtain a region segmentation image. Before the first aerial image is subjected to region segmentation by utilizing the aerial building segmentation model, the aerial building segmentation model needs to be constructed. The construction method of the segmentation model of the aerial photography building is designed in the embodiment of the application.
The technical solutions of the embodiments of the present application are described in detail below with reference to the accompanying drawings.
Fig. 4 is a flowchart illustrating a method for constructing an aerial photography building segmentation model according to an embodiment of the present application.
Referring to fig. 4, the method for constructing the segmentation model of the aerial photography building includes:
401. acquiring a building aerial photo album;
the building aerial photograph album comprises: building aerial images and building outline labeling;
in the embodiment of the application, the building outline is marked on the building aerial image through marking software to form a building aerial image set.
402. Carrying out data enhancement processing on the building aerial photo album to obtain an extended album;
in the embodiment of the present application, the data enhancement process includes, but is not limited to: horizontal flipping, brightness adjustment, contrast adjustment, and saturation adjustment. It is to be understood that the above description of data enhancement processing is not to be taken as the only limitation on the present application.
403. Dividing the expansion diagram set into a training set, a verification set and a test set;
in the embodiment of the application, the expansion atlas can be divided into a training set, a verification set and a test set according to a certain proportion; the training set is used for training a Mask ordering R-CNN model, the verification set is used for evaluating the Mask ordering R-CNN model to obtain network parameters of the model, and the test set is used for testing whether the obtained model meets the segmentation requirements.
404. Performing iterative training on the Mask ordering R-CNN model by using the training set as the input of the Mask ordering R-CNN model to obtain a primary segmentation model;
in the embodiment of the application, the Mask screening R-CNN model comprises: a basic backbone network, a regional suggestion network, a head network and a segmentation quality scoring network; the basic backbone network is used for processing aerial images of the building and inputting the obtained characteristic diagram into the regional suggestion network; the area suggestion network is used for processing the characteristic diagram and inputting the obtained interesting area into the head network; the head network carries out classification and frame regression based on the region of interest to obtain a coordinate set, a segmentation mask and classification information of each building, and outputs the predicted segmentation mask to a segmentation quality scoring network after maximum pooling processing; the segmentation quality scoring network quality scores the predictive segmentation mask in the head network.
405. Calling a verification set to evaluate the primary segmentation model to obtain network parameters of the primary segmentation model;
406. and calling a test set to test the primary segmentation model to obtain an aerial photography building segmentation model.
Specifically, the method comprises the following steps:
calling a test set to test the primary segmentation model, and judging whether the deviation between the building segmentation result output by the primary segmentation model and the building outline label in the test set is smaller than a deviation threshold value or not;
if so, outputting the primary segmentation model as the aerial photography building segmentation model;
if not, reconstructing the network parameters of the primary segmentation model until the deviation between the building segmentation result output by the primary segmentation model and the building outline label in the test set is smaller than a deviation threshold value.
In the embodiment of the present application, the deviation threshold may be set according to the requirement of the segmentation accuracy in practical situations, and is not limited herein.
The embodiment of the application selects a Mask Scoring R-CNN model as an example segmentation model, compared with the Mask R-CNN model, the example segmentation model is additionally provided with a segmentation quality Scoring network, and high-quality segmentation results are selected by calculating scores of quality confidence degrees of segmentation areas, so that the example segmentation performance of the example segmentation model is improved; moreover, the example segmentation model is evaluated through the verification set, so that network parameters influencing the segmentation result of the example segmentation model can be obtained, and then the network parameters are optimized and iterated through the test set, so that the aerial photography building segmentation model meeting the requirements is obtained; accurate segmentation of aerial images can be achieved based on the aerial building segmentation model, and then the accuracy of subsequent image comparison is guaranteed.
EXAMPLE five
The embodiment of the present application designs step 203 in the second embodiment.
The technical solutions of the embodiments of the present application are described in detail below with reference to the accompanying drawings.
Fig. 5 is a flowchart illustrating a method for acquiring difference region information according to an embodiment of the present application.
Referring to fig. 5, the method for acquiring the difference region information includes:
501. utilizing a histogram matching algorithm to perform illumination equalization processing on the region segmentation image to obtain an equalized region segmentation image;
the histogram matching algorithm is an image processing method which takes an image histogram as a template and enables the image histogram to be processed to be approximate to the template histogram; in the embodiment of the application, the histogram of the second aerial image is used as the template histogram, and the histogram matching algorithm is used for carrying out light and color evening processing on the region segmentation image, so that the light condition of the obtained balanced region segmentation image is closer to that of the second aerial image, the error caused by the light condition is reduced, and the accuracy of image contrast is ensured.
It should be noted that the embodiment of the present application does not have strict requirements on the employed dodging algorithm, and in practical applications, other algorithms may also be employed to process the region segmentation image, for example: mask self-adaptive dodging algorithm or Gamma correction algorithm.
502. Graying the balanced region segmentation image and the second aerial image to obtain a region grayscale image and a second grayscale image;
in the embodiment of the application, a weighted average method is adopted to perform weighted average on three color components in a color image by different weights to perform graying processing, so that an area grayscale image and a second grayscale image are obtained.
It should be noted that, in the practical application process, the graying algorithm that can be adopted is not unique, and a component method or an average value method can be selected and applied to the present application according to the practical situation. The above description of the weighted average method is only an example given in the embodiments of the present application and is not intended as the only limitation of the present application.
503. Calculating to obtain the Hamming distance between the regional gray image and the second gray image by using a difference hash algorithm;
the following are exemplary:
respectively comparing the intensity of each row of adjacent pixels in the regional gray image and the second gray image to respectively obtain a difference value array of the regional gray image and the second gray image;
respectively calculating to obtain hash values of the regional gray image and the second gray image according to the difference value arrays of the regional gray image and the second gray image;
and carrying out XOR on the hash values of the regional gray image and the second gray image to obtain the Hamming distance between the regional gray image and the second gray image.
It should be noted that, in the embodiment of the present application, a difference hash algorithm is used to calculate the hamming distance between the area grayscale image and the second grayscale image, and in an actual application process, other algorithms may also be used to calculate the hamming distance, for example: a mean hash algorithm or a perceptual hash algorithm.
It is to be understood that the above description of the difference hash algorithm is not to be taken as the only limitation of the present application.
504. And obtaining difference region information according to the comparison result of the Hamming distance and the Hamming distance threshold.
Specifically, the method comprises the following steps:
judging whether the Hamming distance is larger than a Hamming distance threshold value, if so, judging that the corresponding area of the area segmentation image is the difference area, and acquiring the difference area coordinates of the difference area and adding the difference area coordinates into the difference area information; if not, the fact that the area corresponding to the area segmentation image has no violation risk is shown, and the area segmentation image does not need to be processed.
In the embodiment of the present application, the hamming distance threshold may be adjusted according to actual situations, and is not limited herein.
The embodiment of the application provides a method for acquiring difference region information, wherein the influence of light conditions on similarity judgment of a region segmentation image and a second aerial image is reduced by performing illumination equalization processing on the region segmentation image, hash values of the region grayscale image and the second grayscale image are obtained by calculating through a difference hash algorithm, the calculation is equivalent to obtaining respective identity codes of the region segmentation image and the second aerial image, the identity codes of the region segmentation image and the second aerial image are compared to obtain a hamming distance capable of expressing the difference degree of the identity codes of the region segmentation image and the second aerial image, corresponding regions of the region segmentation image with the hamming distance larger than a hamming distance threshold value are screened out, a difference region can be obtained, and difference region coordinates are obtained, so that the difference region information is obtained.
EXAMPLE six
In the practical application process, the evidence obtaining personnel need to obtain evidence on the spot according to the coordinates of the violation buildings, and also need to know the corresponding application type of the violation buildings, and further can judge the construction condition of the violation buildings according to the height increasing height of the violation buildings, so that the established violation buildings, the violation buildings under construction and the initial violation buildings are distinguished, and the evidence obtaining task is managed in a grading manner, thereby providing the evidence obtaining efficiency.
Fig. 6 is a flowchart illustrating a forensic instruction generation method according to an embodiment of the present application.
In an embodiment of the present application, the violation assessment information further includes: the type of the place for the illegal building and the height of the illegal building are increased; the type of the illegal building land is determined by searching in a building land type mapping table based on the coordinates of the illegal building; and the illegal building heightening height is a height difference obtained by subtracting the second building height from the first building height.
In practical applications, types of places for construction violations include, but are not limited to: construction land and greening land.
Referring to fig. 6, the forensic indication generating method includes:
601. judging whether the height of the illegal building is greater than a second height increasing threshold value or not, if so, executing a step 602; if not, go to step 603;
602. initiating an emergency evidence obtaining instruction to the violation management terminal;
603. judging whether the height of the illegal building is greater than a first height increasing threshold value or not, if so, executing a step 604; if not, go to step 605;
604. initiating a preferential evidence obtaining instruction to the violation management terminal;
605. and initiating an instruction to be subjected to evidence obtaining to the violation management terminal.
In an embodiment of the application, the first increase threshold is smaller than the second increase threshold.
When the height of the violation building is greater than the second height increasing threshold, the fact that the height of the violation building is too high is indicated, namely the violation building is built, and as the height of the violation building is increased, the dismantling difficulty and the dismantling cost of the violation building in the later stage of construction are higher, on-site evidence obtaining and dismantling guidance are urgently needed, so that an emergency evidence obtaining indication is initiated to the violation management terminal; in the practical application process, the emergency evidence obtaining indication can be realized by combining text prompt with sound and light alarm.
When the height of the illegal building is less than or equal to the second height increasing threshold and greater than the first height increasing threshold, the illegal building is in the state of construction, and the illegal building should be processed preferentially, so that a preferential evidence obtaining instruction should be sent to the illegal management terminal; in the practical application process, the prior evidence obtaining indication can be realized by combining text prompt with voice broadcast or combining text prompt with an alarm lamp.
When the height of the violation building is smaller than the first height increasing threshold, the violation building is in the initial stage of the building, the emergency degree is low, and an emergency evidence obtaining instruction and a priority evidence obtaining instruction can be processed preferentially, so that an instruction to be subjected to evidence obtaining is sent to the violation management terminal; in the practical application process, the indication to be forensically indicated can be realized by adopting a text prompting mode.
In the embodiment of the application, the illegal buildings are classified by utilizing the height increased by the illegal buildings in the illegal assessment information, the illegal buildings with the height increased by more than a second height increased threshold value are marked as illegal buildings to be processed urgently, and the evidence obtaining personnel is reminded to obtain evidence on site through the urgent evidence obtaining indication; the illegal buildings with the height higher than the first height-increasing threshold value and not higher than the second height-increasing threshold value are marked as illegal buildings to be processed preferentially, and forensics personnel are reminded to carry out on-site forensics through the preferential forensics instruction; the violation building is increased and is highly be less than the first violation building that is marked as the pending violation building that increases of threshold value, indicates through waiting to collect evidence and reminds the personnel of collecting evidence, can handle again after handling urgent instruction of collecting evidence and the instruction of preferentially collecting evidence to the work of collecting evidence of the personnel of collecting evidence has been planned according to the emergency degree of collecting evidence, makes the work of collecting evidence orderly, high-efficient.
EXAMPLE seven
After the evidence obtaining personnel complete the on-site evidence obtaining work, a large amount of violation evidence obtaining information is obtained, therefore, according to the violation identification information, after the violation evidence obtaining instruction is sent to the violation management terminal, the violation evidence obtaining information returned by the evidence obtaining personnel is counted, and a violation building analysis report is formed.
Fig. 7 is a schematic flow chart of a method for acquiring a violation building analysis report according to an embodiment of the present application.
Referring to fig. 7, the method for obtaining the analysis report of the illegal building includes:
701. classifying according to the violation evidence obtaining information obtained by evidence obtaining according to administrative divisions to obtain M district-level violation data sets;
wherein M is a positive integer; the violation forensics information comprises: the violation evidence obtaining date, the evidence obtaining person information, the violation case number, the violation assessment information and the violation type.
In the embodiment of the application, since the violation assessment information comprises the violation building coordinate, the administrative division to which the violation building belongs can be judged according to the violation building coordinate and the local administrative division file, and further, the street and the community to which the violation building belongs can also be judged, so that the number of the violation buildings or the density of the violation buildings on a certain street can be obtained through statistics, and the responsibility of the street in charge of the violation building management can be asked for duty or guided for work.
In the embodiment of the application, the unique violation case number is formed for each identified violation building and is associated with the violation evidence obtaining date, the evidence obtaining person information, the violation evaluation information and the violation type, so that a violation data network is generated and stored according to the violation evidence obtaining information of the violation building, and the follow-up evidence obtaining personnel can conveniently carry out data backtrack, for example, all evidence obtaining works executed on the date are obtained according to the violation evidence obtaining date.
In practical applications, the types of violations include, but are not limited to: building, covering floors, covering roofs and stacking occupation.
702. And respectively carrying out data statistics of violation evidence obtaining information on the M district-level violation data sets to obtain violation building analysis reports.
Wherein, the data statistics of the violation forensic information of a district-level violation data set includes but is not limited to:
calculating the area of the violation buildings according to the coordinates of the violation buildings and the area segmentation image, and associating the area of the violation buildings with the number of the violation cases;
calculating to obtain the district-level violation number according to the violation case number;
calculating to obtain the area of the district-level violation type corresponding to each violation type according to the violation type and the area of the violation building;
and calculating to obtain the zone-level violation type number corresponding to each violation type according to the violation type and the violation case number.
In the embodiment of the application, the illegal building analysis report can be displayed in multiple data dimensions, for example, an administrative division is used as a first level, a violation type is used as a second level, and the illegal building area is used as a numerical value for displaying; or the administrative division is used as a first level, and the district-level violation number is used as a numerical value for displaying.
For example, the following table takes the street or the village and the town where the illegal building belongs as a father level, the type of the violation as a child level, and the number of the violation and the area of the illegal building as numerical values to generate an analysis report of the illegal building; in the report, the district-level violation type number and the district-level violation type area of each street are obtained through statistics.
Figure 677426DEST_PATH_IMAGE002
According to the embodiment of the application, the violation evidence obtaining information returned by the evidence obtaining personnel is counted to form a violation building analysis report, data backtracking can be achieved based on the violation building analysis report, and other related data are inquired by taking certain data as nodes; the data dimension of difference can also be used for displaying, and a network topological structure with different levels and display values is formed, so that different statistical demands are met, ordered storage and efficient access of various data are realized, and the efficiency of related data management of the illegal building is improved.
Example eight
Corresponding to the embodiment of the application function implementation method, the application also provides an aerial image-based illegal building identification system and a corresponding embodiment.
Referring to fig. 8, an identification system for illegal buildings based on aerial images includes:
a processing device for executing the illegal building identification method based on aerial images as described in any one of the first to seventh embodiments;
the aerial photographing device is used for acquiring aerial photographing images according to the instruction of the processing device;
and the distance measuring device is used for measuring the height above the ground of the building according to the indication of the processing device.
In practical applications, the Processing device may be a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC); in an embodiment of the application, the processing means is a microprocessor.
In the embodiment of the application, the aerial photographing device can be an unmanned aerial photographing device, and the specific model of the aerial photographing device can be selected according to the actual situation, which is not limited herein.
In the embodiment of the present application, the distance measuring device may be any one of a laser distance meter, an ultrasonic distance meter and a radar distance meter, and specifically may be an infrared distance meter; in practical application, the distance measuring device can be arranged on the aerial photographing device.
The aspects of the present application have been described in detail hereinabove with reference to the accompanying drawings. In the above embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments. Those skilled in the art should also appreciate that the acts and modules referred to in the specification are not necessarily required in the present application. In addition, it can be understood that the steps in the method of the embodiment of the present application may be sequentially adjusted, combined, and deleted according to actual needs, and the modules in the device of the embodiment of the present application may be combined, divided, and deleted according to actual needs.
Furthermore, the method according to the present application may also be implemented as a computer program or computer program product comprising computer program code instructions for performing some or all of the steps of the above-described method of the present application.
Alternatively, the present application may also be embodied as a non-transitory machine-readable storage medium (or computer-readable storage medium, or machine-readable storage medium) having stored thereon executable code (or a computer program, or computer instruction code) which, when executed by a processor of an electronic device (or electronic device, server, etc.), causes the processor to perform part or all of the various steps of the above-described method according to the present application.
Those of skill would further appreciate that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the applications disclosed herein may be implemented as electronic hardware, computer software, or combinations of both.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems and methods according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
Having described embodiments of the present application, the foregoing description is intended to be exemplary, not exhaustive, and not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen in order to best explain the principles of the embodiments, the practical application, or improvements made to the technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.

Claims (9)

1. An identification method for illegal buildings based on aerial images is characterized by comprising the following steps:
acquiring a first aerial image;
performing region segmentation according to the first aerial image to obtain a region segmentation image;
obtaining difference region information based on a comparison result of the second aerial image and the region segmentation image; the difference region information includes: a difference region coordinate; the first aerial image and the second aerial image are images of the same area, which have the same icon size and shooting angle and are shot at different time; the difference region coordinates include: region center coordinates and N region edge coordinates; n is an integer greater than or equal to zero;
acquiring a first building height of a difference area according to the difference area coordinates; the method specifically comprises the following steps:
acquiring N +1 first building heights of the difference region according to the region center coordinate and the N region edge coordinates respectively;
obtaining violation assessment information of the difference area based on the comparison result of the first building height and the second building height, wherein the violation assessment information specifically comprises the following steps: calculating the number of the first building heights which are greater than the second building height in the N +1 first building heights to obtain an increase number; determining whether the increase number is greater than or equal to
Figure FDA0003458039430000011
If so, judging that the violation evaluation result of the difference area is a violation building, and taking the area center coordinate as the violation building coordinate; if not, judging that the violation evaluation result of the difference area is a construction without violations; the violation assessment information includes: violation evaluation results and violation building coordinates;
and initiating a evidence obtaining instruction to the violation management terminal according to the violation evaluation information.
2. The method for identifying illegal buildings based on aerial images as claimed in claim 1, wherein the step of obtaining the difference region information based on the comparison result of the second aerial image and the region segmentation image comprises the steps of:
judging whether the area segmentation image meets a preset inspection-free condition, if not, judging whether the area segmentation image changes according to the second aerial image, if so, judging that an area corresponding to the area segmentation image is the difference area, and acquiring the difference area coordinates of the difference area;
the preset inspection-free condition comprises the following steps: the area of the region segmentation image is smaller than a detection area threshold value, the region segmentation image meets the soil accumulation region characteristic and the area is smaller than the soil accumulation area threshold value, the region segmentation image meets the water area characteristic or the region segmentation image meets the green land characteristic.
3. The method for identifying illegal buildings based on aerial images as claimed in claim 2, wherein the step of judging whether the area segmentation image changes or not according to the second aerial image comprises the following steps:
utilizing a histogram matching algorithm to perform illumination equalization processing on the region segmentation image to obtain an equalized region segmentation image;
graying the balanced region segmentation image and the second aerial image to obtain a region grayscale image and a second grayscale image;
calculating to obtain the Hamming distance between the regional gray image and the second gray image by using a difference hash algorithm;
and judging whether the Hamming distance is larger than a Hamming distance threshold value or not.
4. The method for identifying illegal buildings based on aerial images as claimed in claim 1, wherein the step of collecting the first building height of the difference area according to the coordinates of the difference area comprises the following steps:
controlling the aerial photographing device to obtain echo duration through detection of an electromagnetic wave ranging technology at the position of the difference region coordinate and a preset ranging height;
and calculating to obtain the first building height based on the echo duration and the preset ranging height.
5. The method for identifying illegal buildings based on aerial images as claimed in claim 1, wherein the area segmentation according to the first aerial image comprises:
and carrying out region segmentation on the first aerial image by utilizing an aerial building segmentation model.
6. The method of claim 5 for identifying violating buildings based on aerial images, wherein before the area segmentation of the first aerial image by the aerial building segmentation model, the method comprises:
acquiring a building aerial photo album; the building aerial photograph album comprises: building aerial images and building outline labeling;
carrying out data enhancement processing on the building aerial photo album to obtain an extended album;
dividing the expansion diagram set into a training set, a verification set and a test set;
performing iterative training on the Mask ordering R-CNN model by using the training set as the input of the Mask ordering R-CNN model to obtain a primary segmentation model;
calling the verification set to evaluate the primary segmentation model to obtain network parameters of the primary segmentation model;
calling the test set to test the primary segmentation model, judging whether the deviation between the building segmentation result output by the primary segmentation model and the building outline label in the test set is smaller than a deviation threshold value, if so, outputting the primary segmentation model as the aerial photography building segmentation model;
if not, the network parameters of the primary segmentation model are reconstructed until the deviation between the building segmentation result output by the primary segmentation model and the building outline label in the test set is smaller than the deviation threshold value.
7. The method for identifying the illegal building based on the aerial image as claimed in claim 1, wherein after the evidence obtaining instruction is sent to the illegal management terminal according to the illegal identification information, the method comprises the following steps:
classifying according to the violation evidence obtaining information obtained by evidence obtaining according to administrative divisions to obtain M district-level violation data sets; m is a positive integer; the violation forensics information comprises: the violation evidence obtaining date, the evidence obtaining person information, the violation case number, the violation assessment information and the violation type;
carrying out data statistics of violation evidence obtaining information on the M district-level violation data sets respectively to obtain violation building analysis reports;
in the data statistics of the violation evidence obtaining information of the M district-level violation data sets, the data statistics of the violation evidence obtaining information of one district-level violation data set comprises the following steps:
calculating the area of the violation buildings according to the coordinates of the violation buildings and the area segmentation image, and associating the area of the violation buildings with the number of the violation cases;
calculating to obtain the district-level violation number according to the violation case number;
calculating to obtain the area of the district-level violation type corresponding to each violation type according to the violation type and the area of the violation building;
and calculating to obtain the zone-level violation type number corresponding to each violation type according to the violation type and the violation case number.
8. The method of claim 1 wherein the identification of violation buildings based on aerial images,
the violation assessment information further includes: the type of the place for the illegal building and the height of the illegal building are increased; the type of the illegal building land is determined by searching in a building land type mapping table based on the coordinates of the illegal building; the illegal building heightening height is a height difference obtained by subtracting the second building height from the first building height;
the step of initiating evidence obtaining indication to the violation management terminal according to the violation evaluation information comprises the following steps:
judging whether the heightening height of the illegal building is greater than a second heightening threshold value or not, if so, initiating an emergency evidence obtaining instruction to the illegal management terminal;
if not, judging whether the heightening height of the violation building is greater than a first heightening threshold value, if so, initiating a preferential evidence obtaining instruction to the violation management terminal;
if not, initiating an instruction to be subjected to evidence obtaining to the violation management terminal;
the first boost threshold is less than the second boost threshold.
9. The utility model provides a building identification system violating regulations based on image of taking photo by plane which characterized in that includes:
processing means for executing the aerial image-based violation building identification method according to any one of claims 1 to 8;
the aerial photographing device is used for acquiring aerial photographing images according to the instruction of the processing device;
and the distance measuring device is used for measuring the height above the ground of the building according to the indication of the processing device.
CN202111386033.0A 2021-11-22 2021-11-22 Method and system for identifying illegal building based on aerial image Active CN113822247B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111386033.0A CN113822247B (en) 2021-11-22 2021-11-22 Method and system for identifying illegal building based on aerial image

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111386033.0A CN113822247B (en) 2021-11-22 2021-11-22 Method and system for identifying illegal building based on aerial image

Publications (2)

Publication Number Publication Date
CN113822247A CN113822247A (en) 2021-12-21
CN113822247B true CN113822247B (en) 2022-02-18

Family

ID=78918051

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111386033.0A Active CN113822247B (en) 2021-11-22 2021-11-22 Method and system for identifying illegal building based on aerial image

Country Status (1)

Country Link
CN (1) CN113822247B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114820748B (en) * 2022-06-30 2022-09-27 广东泰一高新技术发展有限公司 Full-view-angle characteristic intelligent evidence obtaining method based on panoramic data
CN115049935B (en) * 2022-08-12 2022-11-11 松立控股集团股份有限公司 Urban illegal building division detection method
CN116050842B (en) * 2023-02-01 2023-08-04 东华理工大学 Dynamic control method and system for urban underground engineering construction safety risk
CN116863099B (en) * 2023-06-29 2023-12-26 广州城市职业学院 Building automatic modeling method and system based on point cloud data

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111368615A (en) * 2019-06-26 2020-07-03 杭州海康威视系统技术有限公司 Violation building early warning method and device and electronic equipment
CN111640125A (en) * 2020-05-29 2020-09-08 广西大学 Mask R-CNN-based aerial photograph building detection and segmentation method and device
CN111692978A (en) * 2020-06-04 2020-09-22 福建睿思特科技股份有限公司 Multi-source image data-based illegal building identification method
CN112200769A (en) * 2020-09-08 2021-01-08 东南大学 Fixed-point monitoring new and old time phase image change detection method for illegal construction detection

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9064150B2 (en) * 2013-05-08 2015-06-23 Honeywell International Inc. Aerial image segmentation for refineries
CN106683112B (en) * 2016-10-10 2019-09-27 国交空间信息技术（北京）有限公司 A kind of Road domain building change detection method based on high-definition picture
CN112241659A (en) * 2019-07-17 2021-01-19 杭州海康威视数字技术股份有限公司 Detection device and detection method for illegal building and terminal equipment

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111368615A (en) * 2019-06-26 2020-07-03 杭州海康威视系统技术有限公司 Violation building early warning method and device and electronic equipment
CN111640125A (en) * 2020-05-29 2020-09-08 广西大学 Mask R-CNN-based aerial photograph building detection and segmentation method and device
CN111692978A (en) * 2020-06-04 2020-09-22 福建睿思特科技股份有限公司 Multi-source image data-based illegal building identification method
CN112200769A (en) * 2020-09-08 2021-01-08 东南大学 Fixed-point monitoring new and old time phase image change detection method for illegal construction detection

Also Published As

Publication number Publication date
CN113822247A (en) 2021-12-21

Similar Documents

Publication Publication Date Title
CN113822247B (en) Method and system for identifying illegal building based on aerial image
US11455805B2 (en) Method and apparatus for detecting parking space usage condition, electronic device, and storage medium
CN110148196B (en) Image processing method and device and related equipment
CN109816024B (en) Real-time vehicle logo detection method based on multi-scale feature fusion and DCNN
CN104700099B (en) The method and apparatus for recognizing traffic sign
CN108171136B (en) System and method for searching images by images for vehicles at multi-task gate
CN110610166B (en) Text region detection model training method and device, electronic equipment and storage medium
CN105574550A (en) Vehicle identification method and device
CN111784685A (en) Power transmission line defect image identification method based on cloud edge cooperative detection
CN104166841A (en) Rapid detection identification method for specified pedestrian or vehicle in video monitoring network
GB2502187A (en) Determining parking lot occupancy from digital camera images
EP3028464A1 (en) System and method for detecting features in aerial images using disparity mapping and segmentation techniques
CN107563265A (en) A kind of high beam detection method and device
CN112329881B (en) License plate recognition model training method, license plate recognition method and device
CN110751012B (en) Target detection evaluation method and device, electronic equipment and storage medium
CN109389105B (en) Multitask-based iris detection and visual angle classification method
CN110852164A (en) YOLOv 3-based method and system for automatically detecting illegal building
CN111126393A (en) Vehicle appearance refitting judgment method and device, computer equipment and storage medium
CN111144372A (en) Vehicle detection method, device, computer equipment and storage medium
CN110636248B (en) Target tracking method and device
CN106548195A (en) A kind of object detection method based on modified model HOG ULBP feature operators
CN111695373A (en) Zebra crossing positioning method, system, medium and device
Wu et al. Block-based hough transform for recognition of zebra crossing in natural scene images
CN110765900A (en) DSSD-based automatic illegal building detection method and system
CN106920398A (en) A kind of intelligent vehicle license plate recognition system

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
PE01 Entry into force of the registration of the contract for pledge of patent right

Denomination of invention: A Method and System for Identifying Illegal Buildings Based on Aerial Images

Effective date of registration: 20230814

Granted publication date: 20220218

Pledgee: Guangdong Huaxing Bank Co.,Ltd. Huizhou Branch

Pledgor: GUANGDONG AIRACE TECHNOLOGY DEVELOPMENT CO.,LTD.

Registration number: Y2023980052084

PE01 Entry into force of the registration of the contract for pledge of patent right