CN112183434A - Building change detection method and device - Google Patents

Building change detection method and device Download PDF

Info

Publication number
CN112183434A
CN112183434A CN202011085627.3A CN202011085627A CN112183434A CN 112183434 A CN112183434 A CN 112183434A CN 202011085627 A CN202011085627 A CN 202011085627A CN 112183434 A CN112183434 A CN 112183434A
Authority
CN
China
Prior art keywords
building
stage image
change
later
image
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202011085627.3A
Other languages
Chinese (zh)
Other versions
CN112183434B (en
Inventor
赖余斌
洪巧章
彭飞
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.)
Energy Development Research Institute of China Southern Power Grid Co Ltd
Original Assignee
Energy Development Research Institute of China Southern Power Grid 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 Energy Development Research Institute of China Southern Power Grid Co Ltd filed Critical Energy Development Research Institute of China Southern Power Grid Co Ltd
Priority to CN202011085627.3A priority Critical patent/CN112183434B/en
Publication of CN112183434A publication Critical patent/CN112183434A/en
Application granted granted Critical
Publication of CN112183434B publication Critical patent/CN112183434B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • G06V20/176—Urban or other man-made structures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
    • G06V10/507—Summing image-intensity values; Histogram projection analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
    • G06V10/467—Encoded features or binary features, e.g. local binary patterns [LBP]

Abstract

The invention discloses a building change detection method and a building change detection device, wherein the method comprises the following steps: acquiring front and rear two-stage images; respectively extracting building graphs of front and back images by using the object-oriented morphological building index; acquiring laser radar point cloud data consistent with the later-stage image, and filtering the laser radar point cloud data to obtain a digital elevation model of the terrain surface; performing rasterization processing, difference operation and threshold screening on the digital surface model and the digital elevation model respectively to obtain alternative buildings; respectively projecting the top coordinates of the alternative buildings onto the front and rear images by adopting a collinear equation; excluding non-building areas, and acquiring actual building areas of front and back two-stage images; extracting the building characteristics in the actual building areas of the front-stage and back-stage images for comparison to obtain a change confidence map; and carrying out binary segmentation on the change confidence coefficient map to obtain a building change area. The invention can improve the accuracy of building change detection.

Description

Building change detection method and device
Technical Field
The invention relates to the technical field of building change detection, in particular to a building change detection method and device.
Background
The land feature change detection is to find the change process on the earth surface by using two or more remote sensing images acquired in the same geographical area at different time points. Building change detection is one type of terrain change detection. For some fields, the detection of the change of buildings is extremely important, for example, when buildings are newly built along the transmission line, the safety problems such as the collapse of wire piles and the short circuit of electric wires can be caused, and therefore, the detection of the change of the buildings along the transmission line is an important content of safety management work. The current building change detection method comprises the following steps: the method comprises a building change detection method based on satellite images and a building change detection method based on lidar point cloud. The building change detection method based on the satellite image needs to provide a registered orthoimage, and the problem of projection difference cannot be solved when the change detection of the building is carried out, so that the detection result is not accurate enough. The building change detection method based on the lidar point cloud usually needs two stages of lidar laser point clouds, the acquisition cost is high, the lidar laser point clouds in the old stage are probably not available, and the point clouds obtained through dense matching are usually wrong, so the detection result is not accurate enough.
Disclosure of Invention
The invention aims to provide a building change detection method and a building change detection device, which are used for solving the technical problem that the existing building change detection method is not accurate enough, and can improve the accuracy of building change detection.
In a first aspect, an embodiment of the present invention provides a building change detection method, including:
acquiring detection range data, and determining a detection range according to the detection range data;
acquiring an early-stage image and a later-stage image of the detection range; wherein the early-stage image and the later-stage image are aerial images;
respectively extracting the building graph of the early-stage image and the building graph of the later-stage image by using an object-oriented morphological building index;
acquiring laser radar point cloud data consistent with the later-stage image, and filtering the laser radar point cloud data to obtain a digital elevation model of the terrain surface;
rasterizing the digital surface model and the digital elevation model respectively, and performing difference operation to obtain an elevation difference;
taking the ground object with the elevation difference larger than a preset threshold value as an alternative building;
projecting the top coordinates of the alternative building onto the early-stage image and the later-stage image respectively by using an collinear equation and utilizing the information of internal and external orientation elements of a navigation piece to obtain the area of the alternative building in the early-stage image and the area of the alternative building in the later-stage image;
according to the building graph of the early-stage image and the area of the alternative building in the early-stage image, excluding a non-building area by adopting a polygon intersection method to obtain an actual building area in the early-stage image; according to the building graph of the later-stage image and the area of the alternative building in the later-stage image, excluding a non-building area by adopting a polygon intersection method to obtain an actual building area in the later-stage image;
respectively extracting the building features in the actual building region in the early-stage image and the building features in the actual building in the later-stage image, and comparing the building features in the actual building region in the early-stage image with the building features in the actual building region in the later-stage image to obtain a change position degree information map;
and carrying out binary segmentation on the change confidence coefficient map to obtain a building change area.
Further, the filtering processing is performed on the laser radar point cloud data to obtain a digital elevation model of the terrain surface, and specifically the method comprises the following steps:
and filtering the laser radar point cloud data by using a triangulation network filtering method to obtain a digital elevation model of the terrain surface.
Further, extracting building features by extracting an LBP operator and an HOG operator;
comparing the building features in the actual building region in the early-stage image with the building features in the actual building region in the later-stage image to obtain a change confidence map, specifically:
and comparing the building features in the actual building region in the early-stage image with the building features in the actual building region in the later-stage image, and calculating the change probability of the building by using a correlation coefficient method to obtain a change confidence map.
Further, x and y are defined as the coordinates of the plane of the aerial photo where the image point is located, and x0,y0F, inner orientation elements of the navigation plate; xs,YsZsObject space coordinates, X, for aerial photography sitesA、YA、ZAObject space coordinates of object space points, ai,bi,cii is 1,2,3 is the cosine of 9 directions formed by 3 azimuth elements of the image, then the collinearity equation is:
Figure BDA0002720283650000021
Figure BDA0002720283650000022
further, the change confidence map is subjected to binary segmentation to obtain a building change region, specifically:
based on the change confidence map, all buildings are divided into change type patches wcAnd unchanged class of patch wn；
Wherein the set of patches defining the change confidence map is H ═ { H ═ H1,h2,...hN)，hiThe ith pattern spot is shown, c and n respectively show the number of the variation pattern spotsAnd the number of unchanged type patches;
according to hiA posteriori probability formula
Figure BDA0002720283650000031
Calculate hiPosterior probability p (w) of a pattern spot belonging to a variation classc|hi) And hiPosterior probability p (w) of patches belonging to unchanged classn|hi) (ii) a Wherein, p (h)i) A joint probability density function, p (h), for each patch confidence distributioni)＝p(wc)×p(hi|wc)+p(wn)×p(hi|wn)，p(wc) And p (w)n) The ratio of the changed pattern spots and the ratio of the unchanged pattern spots are obtained; p (h)iI k) is a conditional probability density function,
Figure BDA0002720283650000032
if p (w)c|hi)>p(wn|hi) Then h isiBelongs to the variation class of pattern spots, otherwise hiA pattern spot belonging to an unchanged class;
and traversing the whole change confidence map to obtain a binary change map, and extracting a foreground value to obtain a change area of the building.
Further, iterative computation is carried out by adopting an EM algorithm to obtain the mean value u of the change-like pattern spotscAnd mean u of unchanged class patchesnVariance σ of variation-like patchesc 2And variance σ of unchanged class of patchesn 2。
Further, the preset threshold is two meters.
Further, the detection range data comprises power transmission line data information and safety distance information between a building area and the power transmission line, and the power transmission line data information comprises trend information of the power transmission line;
the determining the detection range according to the detection range data specifically includes:
and analyzing a buffer zone according to the trend information of the power transmission line and the safety distance information, and determining a detection range.
In a second aspect, an embodiment of the present invention provides a building change detection apparatus, including:
the detection range determining unit is used for acquiring detection range data and determining a detection range according to the detection range data;
the image acquisition unit is used for acquiring an early-stage image and a later-stage image of the detection range; wherein the early-stage image and the later-stage image are aerial images;
a building graph obtaining unit, configured to extract a building graph of the early-stage image and a building graph of the later-stage image respectively by using an object-oriented morphological building index;
the digital elevation model acquisition unit is used for acquiring laser radar point cloud data consistent with the later-stage image, and filtering the laser radar point cloud data to obtain a digital elevation model of the terrain surface;
the elevation difference acquisition unit is used for respectively carrying out rasterization processing on the digital surface model and the digital elevation model and carrying out difference value operation to obtain elevation difference;
the alternative building acquisition unit is used for taking the ground object with the elevation difference larger than a preset threshold value as an alternative building;
the projection unit is used for projecting the top coordinates of the alternative buildings to the early-stage image and the later-stage image respectively by using an collinear equation and utilizing the information of internal and external orientation elements of a navigation film to obtain the areas of the alternative buildings in the early-stage image and the areas of the alternative buildings in the later-stage image;
the actual building area acquisition unit is used for eliminating a non-building area by adopting a polygon intersection method according to the building graph of the early-stage image and the area of the candidate building in the early-stage image to obtain an actual building area in the early-stage image; according to the building graph of the later-stage image and the area of the alternative building in the later-stage image, excluding a non-building area by adopting a polygon intersection method to obtain an actual building area in the later-stage image;
a change confidence map obtaining unit, configured to extract building features in the actual building region in the previous image and building features in the actual building in the later image, respectively, and compare the building features in the actual building region in the previous image and the building features in the actual building region in the later image to obtain a change confidence map;
and the building change region acquisition unit is used for carrying out binary segmentation on the change confidence coefficient map to obtain a building change region.
Further, the detection range data comprises power transmission line data information and safety distance information between a building area and the power transmission line, and the power transmission line data information comprises trend information of the power transmission line;
the determining the detection range according to the detection range data specifically includes:
and analyzing a buffer zone according to the trend information of the power transmission line and the safety distance information, and determining a detection range.
In summary, the embodiment of the invention has the following beneficial effects:
according to the embodiment of the invention, the change condition of the building is detected based on the front and back two-stage images and the radar point cloud data consistent with the later-stage images, so that the technical problems that the cost is high and the detection result is not accurate enough due to intensive matching because the change detection of the building can be realized only by depending on the front and back two-stage laser radar point cloud data in the prior art are solved. The embodiment of the invention also projects the building roof obtained from the digital surface model onto the image by adopting a method of a collinearity equation, thereby eliminating the change interference caused by the projection difference of the building and improving the accuracy of the detection of the change of the building. Finally, the embodiment of the invention realizes the improvement of the accuracy of the building change detection by adopting a series of technologies such as front and back two-stage images, morphological building indexes, laser radar point cloud data consistent with the later-stage images, a digital elevation model of a terrain surface, a digital surface model, projection of a building roof onto the images, a polygon intersection method and the like.
Further, when the newly added buildings along the power transmission line are detected, the method and the device can improve the accuracy of detecting the newly added buildings along the power transmission line, further improve the accuracy of predicting whether safety problems such as pile collapse and short circuit of the power transmission line exist in advance, and further reduce the occurrence of safety accidents along the power transmission line.
Drawings
In order to more clearly illustrate the technical solution of the present invention, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a simplified flow chart of a building change detection method according to an embodiment of the present invention.
Fig. 2 is a structural diagram of a building change detection device according to the present embodiment.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be understood that the step numbers used herein are for convenience of description only and are not intended as limitations on the order in which the steps are performed.
It is to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in the specification of the present invention 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.
The terms "comprises" and "comprising" indicate the presence of the described features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
The term "and/or" refers to and includes any and all possible combinations of one or more of the associated listed items.
Example 1:
referring to fig. 1, a building change detection method provided in embodiment 1 of the present invention may be executed by a device for detecting a building change, where the device may be implemented in a software and/or hardware manner, and the device may be formed by a plurality of physical entities or may be formed by one physical entity, and specifically, the device may be a server, a computer, a mobile phone, a tablet or an intelligent interactive tablet. In the embodiment of the present invention, a server is described as an example of a device for detecting a change in a building. The building change detection method provided by the embodiment of the invention comprises the following steps of S1-S10 (step numbers are not shown in the figure):
and S1, acquiring detection range data, and determining the detection range according to the detection range data.
In the embodiment of the invention, in order to detect the change condition of the building within a certain range, firstly, detection range data needs to be acquired, for example, when the change condition of the building along the power transmission line is detected, power transmission line data information needs to be acquired, wherein the power transmission line data information comprises the trend information of the power transmission line; in addition, safety distance information between the building and the power transmission line needs to be acquired, so that buffer area analysis can be performed according to the trend information of the power transmission line and the safety distance information between the building and the power transmission line to obtain a detection range, and the detection range can be further determined.
S2, acquiring an early-stage image and a later-stage image of the detection range; and the early-stage image and the later-stage image are aerial images.
In the embodiment of the invention, the early-stage image and the later-stage image are images shot in different time periods in the same geographic area, the shooting time of the early-stage image is earlier than that of the later-stage image, and the image in the embodiment of the invention is an aerial image. The early-stage images of the detection range and the later-stage images of the detection range are obtained from the early-stage aerial photographs and the later-stage aerial photographs which are obtained through original shooting respectively.
And S3, respectively extracting the building graph of the early-stage image and the building graph of the later-stage image by using the object-oriented morphological building index.
In the embodiment of the present invention, the object-oriented morphological index is used to obtain the building polygon, and preferably, the object-oriented morphological building index is obtained by using a huangxin research method, such as Huangxin, Zhang liang, and zhu ting. building change detection from multitudinal high-resolution Sensing sensed images d on an amorphic building index, ieee Journal of selected properties in applied Earth orientations and movement Sensing,2014,7(1): 105-115.
And S4, acquiring laser radar point cloud data consistent with the later-stage image, and filtering the laser radar point cloud data to obtain a digital elevation model of the terrain surface.
In the embodiment of the present invention, it is preferable to perform filtering processing on the laser radar point cloud data by using a triangulation network filtering method to obtain a Digital Elevation Model (DEM) of a terrain surface.
And S5, rasterizing the digital surface model and the digital elevation model respectively, and performing difference operation to obtain an elevation difference.
In the embodiment of the invention, the digital surface model can be abbreviated as DSM, the purpose of the step is to obtain the elevation difference, and ground objects which are possibly buildings can be preliminarily screened out through the elevation difference, and the ground objects are abbreviated as alternative buildings hereinafter.
And S6, taking the ground object with the elevation difference larger than the preset threshold value as a candidate building.
In the embodiment of the invention, in order to ensure that the screened ground object is a building as much as possible, the embodiment of the invention preferably adopts a ground object more than two meters as the alternative building.
And S7, projecting the top coordinates of the alternative building onto the early-stage image and the later-stage image respectively by using an collinear equation and utilizing the information of the internal and external orientation elements of the navigation film to obtain the area of the alternative building in the early-stage image and the area of the alternative building in the later-stage image.
In the embodiment of the invention, x and y are defined as the coordinates of the plane of the aerial photo where the image point is located, and x0,y0F, inner orientation elements of the navigation plate; xs,Ys ZsObject space coordinates, X, for aerial photography sitesA、YA、ZAObject space coordinates of object space points, ai,bi,cii is 1,2,3 is the cosine of 9 directions formed by 3 azimuth elements of the image, then the collinearity equation is:
Figure BDA0002720283650000071
Figure BDA0002720283650000072
s8, according to the building graph of the early-stage image and the area of the alternative building in the early-stage image, excluding a non-building area by adopting a polygon intersection method to obtain an actual building area in the early-stage image; and according to the building graph of the later-stage image and the area of the candidate building in the later-stage image, eliminating a non-building area by adopting a polygon intersection method to obtain an actual building area in the later-stage image.
In the embodiment of the present invention, the polygon intersection method is used to exclude non-buildings in the candidate buildings, and specifically, the polygon intersection method is used to determine whether the building graph of the previous image and the area of the candidate building in the previous image are buildings at the same time, and if so, the non-building area may be excluded, so that the non-buildings can be excluded.
And S9, respectively extracting the building features in the actual building region in the early-stage image and the building features in the actual building in the later-stage image, and comparing the building features in the actual building region in the early-stage image with the building features in the actual building region in the later-stage image to obtain a change position confidence map.
In the embodiment of the invention, the building characteristics are obtained by extracting an LBP operator and an HOG operator, wherein the LBP operator is a characteristic operator for describing local textures of the image and has the advantages of rotation invariance and gray invariance. The HOG operator is a feature descriptor for object detection in computer vision and image processing, and is used for constructing features by calculating and counting a gradient direction histogram of a local region of an image.
In an embodiment of the present invention, the comparing the building features in the actual building area in the previous image with the building features in the actual building area in the later image to obtain a change location degree map specifically includes:
and comparing the building features in the actual building region in the early-stage image with the building features in the actual building region in the later-stage image, and calculating the change probability of the building by using a correlation coefficient method to obtain a change confidence map.
And S10, carrying out binary segmentation on the change confidence coefficient map to obtain a building change area.
In the embodiment of the present invention, the change area of the change confidence map is set as the foreground, and the other areas are set as the background, so as to convert the building change detection into the problem of binary segmentation, and obtain the changed building in the detection range based on the binary segmentation method, which is specifically implemented as follows:
defining the image spot set of the change confidence map as H ═ H1,h2,...hN)，hiRepresenting the ith pattern spot, the pattern spots in H are classified into variation type pattern spots wcAnd unchanged class of patch wnC and n respectively represent the number of the variation type patches and the number of the unchanged type patchesThe joint probability density function of each patch confidence distribution can be expressed as follows:
p(hi)＝p(wc)×p(hi|wc)+p(wn)×p(hi|wn) (3)
wherein, p (w)c) And p (w)n(the ratio of the changed pattern spots and the ratio of the unchanged pattern spots.
The conditional probability density adopts a Gaussian mixture function form and is realized by the following steps:
Figure BDA0002720283650000081
where k ∈ { w ∈ [ ]c，wn}。
To solve the mean value u of the variation class patches in equation (4)cAnd mean u of unchanged class patchesnVariance σ of variation-like patchesc 2And variance σ of unchanged class of patchesn 2Iterative computation can be performed by adopting an EM algorithm, and the computation process is as follows:
(1) determining an initial value: dividing all buildings into variation-like pattern spots w by using a K-means algorithm based on a variation confidence mapcAnd unchanged class of patch wnObtaining the initial value of the probability function p (k), the initial value of the mean function u (k) and the initial value of the variance function sigma (k) of the variation type image spots;
(2) calculation of expectation (step E): calculate hiThe posterior probabilities of the image spots belonging to the variation class and the unchanged class are normalized, and the formula is as follows:
Figure BDA0002720283650000082
(3) maximization (M step): the values of the parameters p (k), u (k), σ (k) are continuously updated in iterations with the following formulas;
Figure BDA0002720283650000083
Figure BDA0002720283650000084
Figure BDA0002720283650000085
wherein t is the number of iterations;
and E, iteratively calculating the step E and the step M until p (k), u (k) and sigma (k) converge to obtain the mean value u of the variation type image spotscAnd mean u of unchanged class patchesnVariance σ of variation-like patchesc 2And variance σ of unchanged class of patchesn 2. H can be calculated by using the parameters according to the following formulaiPosterior probability of the pattern spots belonging to the variation class and the pattern spots belonging to the non-variation class:
Figure BDA0002720283650000091
if p (w)c|hi)>p(wn|hi) Then h isiBelongs to the variation class of pattern spots, otherwise hiA pattern spot belonging to an unchanged class;
and traversing the whole change confidence map to obtain a binary change map, and extracting a foreground value to obtain a change area of the building.
According to the embodiment of the invention, the change condition of the building is detected based on the front and back two-stage images and the radar point cloud data consistent with the later-stage images, so that the technical problems that the cost is high and the detection result is not accurate enough due to intensive matching because the change detection of the building can be realized only by depending on the front and back two-stage laser radar point cloud data in the prior art are solved. The embodiment of the invention also projects the building roof obtained from the digital surface model onto the image by adopting a method of a collinearity equation, thereby eliminating the change interference caused by the projection difference of the building and improving the accuracy of the detection of the change of the building. Finally, the embodiment of the invention realizes the improvement of the accuracy of the building change detection by adopting a series of technologies such as front and back two-stage images, morphological building indexes, laser radar point cloud data consistent with the later-stage images, a digital elevation model of a terrain surface, a digital surface model, projection of a building roof onto the images, a polygon intersection method and the like.
Further, when the newly added buildings along the power transmission line are detected, the method and the device can improve the accuracy of detecting the newly added buildings along the power transmission line, further improve the accuracy of predicting whether safety problems such as pile collapse and short circuit of the power transmission line exist in advance, and further reduce the occurrence of safety accidents along the power transmission line.
Example 2:
referring to fig. 2, an embodiment of the present invention further provides a building change detection apparatus, including:
the detection range determining unit 1 is configured to acquire detection range data and determine a detection range according to the detection range data.
In the embodiment of the invention, in order to detect the change condition of the building within a certain range, firstly, detection range data needs to be acquired, for example, when the change condition of the building along the power transmission line is detected, power transmission line data information needs to be acquired, wherein the power transmission line data information comprises the trend information of the power transmission line; in addition, safety distance information between the building and the power transmission line needs to be acquired, so that buffer area analysis can be performed according to the trend information of the power transmission line and the safety distance information between the building and the power transmission line to obtain a detection range, and the detection range can be further determined.
The image acquisition unit 2 is used for acquiring an early-stage image and a later-stage image of the detection range; and the early-stage image and the later-stage image are aerial images.
In the embodiment of the invention, the early-stage image and the later-stage image are images shot in different time periods in the same geographic area, the shooting time of the early-stage image is earlier than that of the later-stage image, and the image in the embodiment of the invention is an aerial image. The early-stage images of the detection range and the later-stage images of the detection range are obtained from the early-stage aerial photographs and the later-stage aerial photographs which are obtained through original shooting respectively.
A building graph obtaining unit 3, configured to extract the building graph of the early-stage image and the building graph of the later-stage image respectively by using an object-oriented morphological building index.
In the embodiment of the present invention, the object-oriented morphological index is used to obtain the building polygon, and preferably, the object-oriented morphological building index is obtained by using a huangxin research method, such as Huangxin, Zhang liang, and zhu ting. building change detection from multitudinal high-resolution Sensing sensed images d on an amorphic building index, ieee Journal of selected properties in applied Earth orientations and movement Sensing,2014,7(1): 105-115.
And the digital elevation model acquisition unit 4 is used for acquiring the laser radar point cloud data consistent with the later-stage image, and filtering the laser radar point cloud data to obtain a digital elevation model of the terrain surface.
In the embodiment of the present invention, it is preferable to perform filtering processing on the laser radar point cloud data by using a triangulation network filtering method to obtain a Digital Elevation Model (DEM) of a terrain surface.
And the elevation difference acquisition unit 5 is used for respectively carrying out rasterization processing on the digital surface model and the digital elevation model and carrying out difference value operation to obtain the elevation difference.
In the embodiment of the invention, the digital surface model can be abbreviated as DSM, the purpose of the step is to obtain the elevation difference, and ground objects which are possibly buildings can be preliminarily screened out through the elevation difference, and the ground objects are abbreviated as alternative buildings hereinafter.
And the alternative building acquisition unit 6 is used for taking the ground object with the elevation difference larger than a preset threshold value as the alternative building.
In the embodiment of the invention, in order to ensure that the screened ground object is a building as much as possible, the embodiment of the invention preferably adopts a ground object more than two meters as the alternative building.
And the projection unit 7 is used for projecting the top coordinates of the alternative buildings to the early-stage image and the later-stage image respectively by using an collinear equation and utilizing the information of internal and external orientation elements of the aerial photo to obtain the areas of the alternative buildings in the early-stage image and the areas of the alternative buildings in the later-stage image.
In the embodiment of the invention, x and y are defined as the coordinates of the plane of the aerial photo where the image point is located, and x0,y0F, inner orientation elements of the navigation plate; xs,Ys ZsObject space coordinates, X, for aerial photography sitesA、YA、ZAObject space coordinates of object space points, ai,bi,cii is 1,2,3 is the cosine of 9 directions formed by 3 azimuth elements of the image, then the collinearity equation is:
Figure BDA0002720283650000101
Figure BDA0002720283650000102
the actual building area obtaining unit 8 is configured to eliminate a non-building area by using a polygon intersection method according to the building graph of the previous image and the area of the candidate building in the previous image, and obtain an actual building area in the previous image; and according to the building graph of the later-stage image and the area of the candidate building in the later-stage image, eliminating a non-building area by adopting a polygon intersection method to obtain an actual building area in the later-stage image.
The change confidence map obtaining unit 9 is configured to extract building features in the actual building area in the previous image and building features in the actual building in the later image, and compare the building features in the actual building area in the previous image and the building features in the actual building area in the later image to obtain a change confidence map.
In the embodiment of the invention, the building characteristics are obtained by extracting an LBP operator and an HOG operator, wherein the LBP operator is a characteristic operator for describing local textures of the image and has the advantages of rotation invariance and gray invariance. The HOG operator is a feature descriptor for object detection in computer vision and image processing, and is used for constructing features by calculating and counting a gradient direction histogram of a local region of an image.
In an embodiment of the present invention, the comparing the building features in the actual building area in the previous image with the building features in the actual building area in the later image to obtain a change location degree map specifically includes:
and comparing the building features in the actual building region in the early-stage image with the building features in the actual building region in the later-stage image, and calculating the change probability of the building by using a correlation coefficient method to obtain a change confidence map.
And the building change region acquisition unit 10 is configured to perform binary segmentation on the change confidence map to obtain a building change region.
In the embodiment of the present invention, the change area of the change confidence map is set as the foreground, and the other areas are set as the background, so as to convert the building change detection into the problem of binary segmentation, and obtain the changed building in the detection range based on the binary segmentation method, which is specifically implemented as follows:
defining the image spot set of the change confidence map as H ═ H1,h2,...hN)，hiRepresenting the ith pattern spot, the pattern spots in H are classified into variation type pattern spots wcAnd unchanged class of patch wnC and n respectively represent the number of the changed type patches and the number of the unchanged type patches, and then the joint probability density function of the confidence distribution of each patch can be represented as follows:
p(hi)＝p(wc)×p(hi|wc)+p(wn)×p(hi|wn) (3)
wherein, p (w)c) And p (w)n) The ratio of the changed pattern spots to the unchanged pattern spotsFor example.
The conditional probability density adopts a Gaussian mixture function form and is realized by the following steps:
Figure BDA0002720283650000111
where k ∈ { w ∈ [ ]c，wn}。
To solve the mean value u of the variation class patches in equation (4)cAnd mean u of unchanged class patchesnVariance σ of variation-like patchesc 2And variance σ of unchanged class of patchesn 2Iterative computation can be performed by adopting an EM algorithm, and the computation process is as follows:
(1) determining an initial value: dividing all buildings into variation-like pattern spots w by using a K-means algorithm based on a variation confidence mapcAnd unchanged class of patch wnObtaining the initial value of the probability function p (k), the initial value of the mean function u (k) and the initial value of the variance function sigma (k) of the variation type image spots;
(2) calculation of expectation (step E): calculate hiThe posterior probabilities of the image spots belonging to the variation class and the unchanged class are normalized, and the formula is as follows:
Figure BDA0002720283650000121
(3) maximization (M step): the values of the parameters p (k), u (k), σ (k) are continuously updated in iterations with the following formulas;
Figure BDA0002720283650000122
Figure BDA0002720283650000123
Figure BDA0002720283650000124
wherein t is the number of iterations;
and E, iteratively calculating the step E and the step M until p (k), u (k) and sigma (k) converge to obtain the mean value u of the variation type image spotscAnd mean u of unchanged class patchesnVariance σ of variation-like patchesc 2And variance σ of unchanged class of patchesn 2. H can be calculated by using the parameters according to the following formulaiPosterior probability of the pattern spots belonging to the variation class and the pattern spots belonging to the non-variation class:
Figure BDA0002720283650000125
if p (w)c|hi)>p(wn|hi) Then h isiBelongs to the variation class of pattern spots, otherwise hiA pattern spot belonging to an unchanged class;
and traversing the whole change confidence map to obtain a binary change map, and extracting a foreground value to obtain a change area of the building.
Example 3:
the embodiment of the invention also provides a computer-readable storage medium, which comprises a stored computer program, wherein when the computer program runs, the device where the storage medium is located is controlled to execute the building change method, and the technical effect consistent with the building change detection method is achieved.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and can include the processes of the embodiments of the methods described above when the computer program is executed. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention.

Claims (10)

1. A building change detection method, comprising:
acquiring detection range data, and determining a detection range according to the detection range data;
acquiring an early-stage image and a later-stage image of the detection range; wherein the early-stage image and the later-stage image are aerial images;
respectively extracting the building graph of the early-stage image and the building graph of the later-stage image by using an object-oriented morphological building index;
acquiring laser radar point cloud data consistent with the later-stage image, and filtering the laser radar point cloud data to obtain a digital elevation model of the terrain surface;
rasterizing the digital surface model and the digital elevation model respectively, and performing difference operation to obtain an elevation difference;
taking the ground object with the elevation difference larger than a preset threshold value as an alternative building;
projecting the top coordinates of the alternative building onto the early-stage image and the later-stage image respectively by using an collinear equation and utilizing the information of internal and external orientation elements of a navigation piece to obtain the area of the alternative building in the early-stage image and the area of the alternative building in the later-stage image;
according to the building graph of the early-stage image and the area of the alternative building in the early-stage image, excluding a non-building area by adopting a polygon intersection method to obtain an actual building area in the early-stage image; according to the building graph of the later-stage image and the area of the alternative building in the later-stage image, excluding a non-building area by adopting a polygon intersection method to obtain an actual building area in the later-stage image;
respectively extracting the building features in the actual building region in the early-stage image and the building features in the actual building in the later-stage image, and comparing the building features in the actual building region in the early-stage image with the building features in the actual building region in the later-stage image to obtain a change position degree information map;
and carrying out binary segmentation on the change confidence coefficient map to obtain a building change area.
2. The building change detection method according to claim 1, wherein the filtering processing is performed on the lidar point cloud data to obtain a digital elevation model of the terrain surface, specifically:
and filtering the laser radar point cloud data by using a triangulation network filtering method to obtain a digital elevation model of the terrain surface.
3. The building change detection method according to claim 1, wherein the building features are extracted by extracting an LBP operator and an HOG operator;
comparing the building features in the actual building region in the early-stage image with the building features in the actual building region in the later-stage image to obtain a change confidence map, specifically:
and comparing the building features in the actual building region in the early-stage image with the building features in the actual building region in the later-stage image, and calculating the change probability of the building by using a correlation coefficient method to obtain a change confidence map.
4. The building change detection method of claim 1, wherein x and y are defined as coordinates of a photo plane where the image point is located, and x is0,y0F, inner orientation elements of the navigation plate; xs,Ys ZsObject space coordinates, X, for aerial photography sitesA、YA、ZAObject space coordinates of object space points, ai,bi,cii is 1,2,3 is the cosine of 9 directions formed by 3 azimuth elements of the image, then the collinearity equationComprises the following steps:
Figure FDA0002720283640000021
Figure FDA0002720283640000022
5. the building change detection method according to claim 1, wherein the change confidence map is binary-segmented to obtain a building change region, specifically:
based on the change confidence map, all buildings are divided into change type patches wcAnd unchanged class of patch wn；
Wherein the set of patches defining the change confidence map is H ═ { H ═ H1,h2,...hN)，hiThe ith pattern spot is shown, c and n respectively show the number of the variation pattern spots and the number of the unchanged pattern spots;
according to hiA posteriori probability formula
Figure FDA0002720283640000023
Calculate hiPosterior probability p (w) of a pattern spot belonging to a variation classc|hi) And hiPosterior probability p (w) of patches belonging to unchanged classn|hi) (ii) a Wherein, p (h)i) A joint probability density function, p (h), for each patch confidence distributioni)＝p(wc)×p(hi|wc)+p(wn)×p(hi|wn)，p(wc) And p (w)n) The ratio of the changed pattern spots and the ratio of the unchanged pattern spots are obtained; p (h)iI k) is a conditional probability density function,
Figure FDA0002720283640000024
if p (w)c|hi)>p(wn|hi) Then h isiBelongs to the variation class of pattern spots, otherwise hiA pattern spot belonging to an unchanged class;
and traversing the whole change confidence map to obtain a binary change map, and extracting a foreground value to obtain a change area of the building.
6. The building change detection method according to claim 5, further comprising: iterative calculation is carried out by adopting an EM algorithm to obtain the mean value u of the change type image spotscAnd mean u of unchanged class patchesnVariance σ of variation-like patchesc 2And variance σ of unchanged class of patchesn 2。
7. The building change detection method according to claim 1, wherein the preset threshold value is two meters.
8. The building change detection method according to any one of claims 1 to 6, wherein the detection range data includes transmission line data information and safety distance information between a building area and the transmission line, and the transmission line data information includes strike information of the transmission line;
the determining the detection range according to the detection range data specifically includes:
and analyzing a buffer zone according to the trend information of the power transmission line and the safety distance information, and determining a detection range.
9. Building change detection device, characterized by, includes:
the detection range determining unit is used for acquiring detection range data and determining a detection range according to the detection range data;
the image acquisition unit is used for acquiring an early-stage image and a later-stage image of the detection range; wherein the early-stage image and the later-stage image are aerial images;
a building graph obtaining unit, configured to extract a building graph of the early-stage image and a building graph of the later-stage image respectively by using an object-oriented morphological building index;
the digital elevation model acquisition unit is used for acquiring laser radar point cloud data consistent with the later-stage image, and filtering the laser radar point cloud data to obtain a digital elevation model of the terrain surface;
the elevation difference acquisition unit is used for respectively carrying out rasterization processing on the digital surface model and the digital elevation model and carrying out difference value operation to obtain elevation difference;
the alternative building acquisition unit is used for taking the ground object with the elevation difference larger than a preset threshold value as an alternative building;
the projection unit is used for projecting the top coordinates of the alternative buildings to the early-stage image and the later-stage image respectively by using an collinear equation and utilizing the information of internal and external orientation elements of a navigation film to obtain the areas of the alternative buildings in the early-stage image and the areas of the alternative buildings in the later-stage image;
the actual building area acquisition unit is used for eliminating a non-building area by adopting a polygon intersection method according to the building graph of the early-stage image and the area of the candidate building in the early-stage image to obtain an actual building area in the early-stage image; according to the building graph of the later-stage image and the area of the alternative building in the later-stage image, excluding a non-building area by adopting a polygon intersection method to obtain an actual building area in the later-stage image;
a change confidence map obtaining unit, configured to extract building features in the actual building region in the previous image and building features in the actual building in the later image, respectively, and compare the building features in the actual building region in the previous image and the building features in the actual building region in the later image to obtain a change confidence map;
and the building change region acquisition unit is used for carrying out binary segmentation on the change confidence coefficient map to obtain a building change region.
10. The building change detection device according to claim 9, wherein the detection range data includes transmission line data information and safety distance information between a building area and the transmission line, and the transmission line data information includes information on a trend of the transmission line;
the determining the detection range according to the detection range data specifically includes:
and analyzing a buffer zone according to the trend information of the power transmission line and the safety distance information, and determining a detection range.
CN202011085627.3A 2020-10-12 2020-10-12 Building change detection method and device Active CN112183434B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011085627.3A CN112183434B (en) 2020-10-12 2020-10-12 Building change detection method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011085627.3A CN112183434B (en) 2020-10-12 2020-10-12 Building change detection method and device

Publications (2)

Publication Number Publication Date
CN112183434A true CN112183434A (en) 2021-01-05
CN112183434B CN112183434B (en) 2022-06-24

Family

ID=73949391

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011085627.3A Active CN112183434B (en) 2020-10-12 2020-10-12 Building change detection method and device

Country Status (1)

Country Link
CN (1) CN112183434B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111323788A (en) * 2020-01-19 2020-06-23 北京建筑大学 Building change monitoring method and device and computer equipment
CN112819753A (en) * 2021-01-12 2021-05-18 香港理工大学深圳研究院 Building change detection method and device, intelligent terminal and storage medium
CN113516777A (en) * 2021-05-13 2021-10-19 天讯方舟(北京)信息科技有限公司 Three-dimensional automatic modeling and visualization method for urban building

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101405566A (en) * 2006-03-17 2009-04-08 株式会社博思科 Building shape change detecting method, and building shape change detecting system
US20100150431A1 (en) * 2008-12-11 2010-06-17 National Central University Method of Change Detection for Building Models
CN103093466A (en) * 2013-01-21 2013-05-08 武汉大学 Building three-dimensional change detection method based on LiDAR point cloud and image
CN108197583A (en) * 2018-01-10 2018-06-22 武汉大学 The building change detecting method of optimization and image structure feature is cut based on figure
CN109919944A (en) * 2018-12-29 2019-06-21 武汉大学 A kind of joint super-pixel figure of complex scene building variation detection cuts optimization method

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101405566A (en) * 2006-03-17 2009-04-08 株式会社博思科 Building shape change detecting method, and building shape change detecting system
US20100150431A1 (en) * 2008-12-11 2010-06-17 National Central University Method of Change Detection for Building Models
CN103093466A (en) * 2013-01-21 2013-05-08 武汉大学 Building three-dimensional change detection method based on LiDAR point cloud and image
CN108197583A (en) * 2018-01-10 2018-06-22 武汉大学 The building change detecting method of optimization and image structure feature is cut based on figure
CN109919944A (en) * 2018-12-29 2019-06-21 武汉大学 A kind of joint super-pixel figure of complex scene building variation detection cuts optimization method

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
J.C.TRINDER ET AL: ""Disaster Change Detection Using Airborne LiDAR"", 《ISPRS ANNALS OF THE PHOTOGRAMMETRY,REMOTE SENSING AND SPATIAL INFORMATION SCIENCES》 *
SHOUJI DU ET AL: ""Building Change Detection Using Old Aerial Images and New LiDAR Data"", 《REMOTE SENSING》 *
彭代锋等: ""结合LiDAR点云和航空影像的建筑物三维变化检测"", 《武汉大学学报 信息科学版》 *
洪巧章等: ""高分辨率遥感影像在电网工程建设中的应用研究"", 《绿色科技》 *
赖余斌等: ""基于卫星遥感技术的电网建设全过程动态监控技术研究"", 《绿色科技》 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111323788A (en) * 2020-01-19 2020-06-23 北京建筑大学 Building change monitoring method and device and computer equipment
CN111323788B (en) * 2020-01-19 2022-04-15 北京建筑大学 Building change monitoring method and device and computer equipment
CN112819753A (en) * 2021-01-12 2021-05-18 香港理工大学深圳研究院 Building change detection method and device, intelligent terminal and storage medium
CN112819753B (en) * 2021-01-12 2021-11-30 香港理工大学深圳研究院 Building change detection method and device, intelligent terminal and storage medium
CN113516777A (en) * 2021-05-13 2021-10-19 天讯方舟(北京)信息科技有限公司 Three-dimensional automatic modeling and visualization method for urban building

Also Published As

Publication number Publication date
CN112183434B (en) 2022-06-24

Similar Documents

Publication Publication Date Title
CN112183434B (en) Building change detection method and device
CN111028277B (en) SAR and optical remote sensing image registration method based on pseudo-twin convolution neural network
Lu et al. Object-oriented change detection for landslide rapid mapping
CN110119438B (en) Airborne LiDAR point cloud filtering method based on active learning
Zhou et al. A fast and accurate segmentation method for ordered LiDAR point cloud of large-scale scenes
CN109829423B (en) Infrared imaging detection method for frozen lake
US11841421B2 (en) Synthetic aperture radar image analysis system, synthetic aperture radar image analysis method, and synthetic aperture radar image analysis program
Moussa et al. A new object based method for automated extraction of urban objects from airborne sensors data
CN108230375A (en) Visible images and SAR image registration method based on structural similarity fast robust
Li et al. An improved approach of information extraction for earthquake-damaged buildings using high-resolution imagery
Chehata et al. LIDAR data classification using hierarchical K-means clustering
Kalantar et al. Smart counting–oil palm tree inventory with UAV
Guo et al. Extraction of dense urban buildings from photogrammetric and LiDAR point clouds
CN110636248B (en) Target tracking method and device
CN114694030A (en) Landslide detection method, device, equipment and storage medium
Guo et al. KD-tree-based euclidean clustering for tomographic SAR point cloud extraction and segmentation
Parmehr et al. Automatic registration of optical imagery with 3d lidar data using local combined mutual information
Dos Santos et al. Automatic building change detection using multi-temporal airborne LiDAR data
dos Santos et al. The use of Otsu algorithm and multi-temporal airborne LiDAR data to detect building changes in urban space
Kusetogullari et al. Unsupervised change detection in landsat images with atmospheric artifacts: a fuzzy multiobjective approach
Wang et al. Adaptive superpixel segmentation with Fisher vectors for ship detection in SAR images
Wen et al. Mixture multiscale autoregressive modelling of SAR imagery for segmentation
CN116091998A (en) Image processing method, device, computer equipment and storage medium
Zhang et al. Robust DEM co-registration method for terrain changes assessment using least trimmed squares estimator
Xie et al. SMRD: A Local Feature Descriptor for Multi-modal Image Registration

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