CN116434072B - Geological disaster early identification method and device based on multi-source data - Google Patents
Geological disaster early identification method and device based on multi-source data Download PDFInfo
- Publication number
- CN116434072B CN116434072B CN202310686349.4A CN202310686349A CN116434072B CN 116434072 B CN116434072 B CN 116434072B CN 202310686349 A CN202310686349 A CN 202310686349A CN 116434072 B CN116434072 B CN 116434072B
- Authority
- CN
- China
- Prior art keywords
- points
- point
- corner
- adjacent pixel
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 64
- 230000008859 change Effects 0.000 claims abstract description 26
- 238000001514 detection method Methods 0.000 claims abstract description 11
- 238000004364 calculation method Methods 0.000 claims description 17
- 230000008569 process Effects 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 8
- 230000009466 transformation Effects 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 7
- 238000005259 measurement Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 4
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 description 9
- 230000009471 action Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001427 coherent effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 229910052500 inorganic mineral Inorganic materials 0.000 description 1
- 238000005305 interferometry Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 239000010687 lubricating oil Substances 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000011707 mineral Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- 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
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
- G06V10/763—Non-hierarchical techniques, e.g. based on statistics of modelling distributions
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A50/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE in human health protection, e.g. against extreme weather
Abstract
The application belongs to the technical field of geological disaster identification, and provides a geological disaster early identification method and device based on multi-source data, comprising the following steps: acquiring a remote sensing image; extracting corner points in the remote sensing image by using a Fast corner point detection algorithm; analyzing local change information of the corner points and the adjacent pixel points to obtain characteristic points; analyzing local change information of the feature points and adjacent pixel points thereof, and selecting an optimal reference image; performing image registration between the optimal reference image and the remote sensing image according to the feature points to obtain a registration image; the registered images are analyzed to identify a geological disaster area. The device and the method improve the accuracy and the reliability of the image registration result, and further improve the accuracy and the efficiency of early identification of geological disasters in the InSAR algorithm.
Description
Technical Field
The application relates to the technical field of geological disaster identification, in particular to a geological disaster early-stage identification method and device based on multi-source data.
Background
Along with the continuous aggravation of global climate change, mineral resource exploitation and large-scale human engineering activities, various geological disasters such as glaciers, collapse, landslide, ground subsidence, ground cracks and the like are induced, and the trend of high frequency and chain nature is presented, so that disaster consequences are more serious. The early identification of the geological disasters with large range and high efficiency is an important premise for disaster prevention and reduction, and is also a technical guarantee for engineering safety.
Synthetic aperture radar interferometry (Interferometric Synthetic Aperture Radar, inSAR) technology has been a rapidly evolving geodetic technology based on satellite-borne sensor-to-earth observations over the last three decades. The basic principle is that the elevation information and the deformation information of the earth surface are obtained by carrying out phase interference on two radar images in the same area. The multi-baseline InSAR is a technology for acquiring surface deformation information by utilizing a plurality of remote sensing images, and is suitable for long-term monitoring by using a plurality of baselines. Compared with conventional measurement, the InSAR technology has the characteristics of wider measurement range, higher precision, all weather, full time of day and high efficiency, and has become one of the common geodetic technologies.
In the InSAR algorithm, image registration is a very important step, and the accuracy of the image registration has important influence on subsequent phase difference calculation and surface deformation monitoring. Therefore, when image registration is performed, a proper feature point extraction algorithm, a proper matching algorithm and a proper registration algorithm are required to be selected, so that the accuracy and reliability of registration results are ensured, and further the geological disaster identification efficiency and accuracy are improved.
Disclosure of Invention
The application provides a geological disaster early recognition method and device based on multi-source data, which are used for improving the recognition efficiency and accuracy of geological disasters.
According to a first aspect of an embodiment of the present application, there is provided a geological disaster early identification method based on multi-source data, the method including: a method for early identification of geologic hazards based on multi-source data, the method comprising:
acquiring a remote sensing image;
extracting corner points in the remote sensing image by a Fast corner point detection algorithm;
analyzing local change information of the corner points and the adjacent pixel points to obtain characteristic points;
analyzing the local change information of the characteristic points and the adjacent pixel points, and selecting an optimal reference image;
performing image registration between the optimal reference image and the remote sensing image according to the characteristic points to obtain a registration image;
and analyzing the registration image to identify a geological disaster area.
In some embodiments of the present application, analyzing local variation information of the corner point and its neighboring pixel points to obtain feature points includes:
sequencing M pixel points near the corner points according to the difference value of the reflection intensity with the corner points from small to large to obtain a sequence of adjacent pixel points of the corner points;
acquiring corresponding adjacent pixel points of all adjacent pixel points of the corner points;
obtaining the reflection intensity variation of the corner point in different local directions of adjacent pixel points according to the reflection intensity of the adjacent pixel points of the corner point and the corresponding adjacent pixel pointsThe calculation formula is as follows:
in the method, in the process of the application,for the reflection intensity variation in the local direction of the corner point adjacent to the pixel point,/->Arranging reflection intensity values of the ith adjacent pixel point in the adjacent pixel point sequence of the corner point,/for the adjacent pixel point sequence of the corner point>The reflection intensity value of the i-th adjacent pixel point corresponding to the adjacent pixel point i' is obtained, wherein the i-th adjacent pixel point represents a pixel point with the large i-th row of the reflection intensity difference value with the corner q point in the q-point adjacent pixel point;
based on the reflection intensity variation in the local direction of the adjacent pixel points of the corner points to be analyzed and the reflection intensity variation in the local direction of the adjacent pixel points of the rest arbitrary corner points, the correlation feature GL between each corner point to be analyzed and the rest arbitrary corner points is obtained, and the calculation formula is as follows:
in the formula, q represents the corner point to be analyzed, a represents any other corner points,、/>representing the reflection intensity values of corner a and corner q, respectively,/->For the number of adjacent pixels, < >>Representing the clockwise included angle between the straight line formed by the ith adjacent pixel point of the corner q point and the horizontal line, and +.>Is the clockwise included angle between the straight line formed by the ith adjacent pixel point of the angular point a and the horizontal line,representing the reflection intensity variation in the local direction of the ith adjacent pixel point of the corner q point, +.>Representing the reflection intensity change of the point a in the local direction of the ith adjacent pixel point;
according to the relevance between the corner points to be analyzed and any other corner points, a possible value KN taking the corner points to be analyzed as characteristic points is obtained, and the calculation formula is as follows:
wherein KN is the possible value of the characteristic point of the corner point to be analyzed,for the association between the ith corner and the corner to be analyzed, +.>For the frequency of occurrence of the association of the ith corner point with the corner point to be analyzed,/->Representing the number of corner points;
and carrying out normalization processing on the possible value KN of which the corner point to be analyzed is the characteristic point, setting a possibility judgment first threshold value, and when the KN corresponding to the corner point to be analyzed is larger than the first threshold value, obtaining the corner point to be analyzed as an actual characteristic point, otherwise, obtaining the corner point to be analyzed as a noise interference point.
In some embodiments of the present application, obtaining corresponding neighboring pixel points of all neighboring pixel points of the corner includes:
connecting the corner point q with the adjacent pixel point f to obtainAdjacent pixel point nearest to q point in direction +.>Adjacent pixel dot->And then sequentially acquiring the corresponding adjacent pixel points of all the adjacent pixel points of the corner point as the corresponding points of the adjacent pixel points f.
In some embodiments of the application, theThe obtaining method of (1) comprises the following steps: count the occurrence number of the correlation feature GL and record as + ->Then->。
In some embodiments of the application, theThe obtaining method of (2) further comprises: counting the number of occurrences of the relevance feature GL when +.>At 0.6, p->When the number of occurrences is counted, if the association index between a certain corner point and a q point is located in the interval [ -f ]>-0.05，/>+0.05]When in use, make->But when->And if so, normally counting.
In some embodiments of the present application, analyzing the local variation information of the feature point and the adjacent pixel point thereof, and selecting an optimal reference image includes:
analyzing the local change information of the characteristic points and the adjacent pixel points to obtain the optimal value Y of the optimal reference image of the remote sensing image, wherein the calculation formula is as follows:
in the formulaThe number of the feature points is M, the number of the adjacent pixel points of the feature points is +.>Reflection intensity variation in local direction of jth adjacent pixel point which is ith feature point, +.>The possibility that the ith corner point is a feature point;
and selecting the remote sensing image with the maximum optimal value Y as an optimal reference image.
In some embodiments of the present application, performing image registration between the optimal reference image and the remote sensing image according to the feature points to obtain a registered image, including:
analyzing the feature points obtained by the optimal reference image and the remote sensing image, and when the correlation feature GL between the feature points is larger than a second threshold value, the two feature points belong to the same feature point sequence; the steps are repeated continuously until the characteristic points in the remote sensing image are divided into different characteristic point sequences;
calculating the matching degree P of all the characteristic point sequences B in the remote sensing image W and the characteristic point sequences A in the optimal reference image Q, and obtaining the maximum matching degree Pmax of the characteristic point sequences in the remote sensing image W and the characteristic point sequences in the optimal reference image, wherein the calculation formula is as follows:
in the method, in the process of the application,for the average relevance between the ith feature point in the feature point sequence B and the rest of the feature points in the feature point sequence, +.>For the average relevance between the ith feature point in the feature point sequence A and the rest of the feature points in the feature point sequence,/the combination of the feature points is given by->For the distance between the centroid b of the remote sensing image W and the ith feature point,/and>for the distance between the optimal reference image Q centroid a and the ith feature point, +.>Is an extremely small positive number;
judging whether the maximum matching degree Pmax is larger than a third threshold value, if the maximum matching degree Pmax is larger than the third threshold value, the two feature point sequences can be considered to be successfully matched, otherwise, the remote sensing image can be considered to be changed, and the matching is unsuccessful;
based on the successfully matched characteristic point sequence pairs, calculating a transformation matrix from the remote sensing image to the optimal reference image;
and registering the remote sensing image according to the calculated transformation matrix.
In some embodiments of the application, analyzing the registered images, identifying a geological disaster area, comprises:
subtracting the phases of the pixels corresponding to the registration images to obtain a phase difference image;
processing pixel values of pixel points in the phase difference image through a phase unwrapping algorithm to obtain earth surface deformation information Z;
using k-means mean clustering, setting k=2, and dividing the image into two types of areas by using a distance measurement mode as deformation information difference;
obtaining deformation difference U=1-exp of two types of regions) Wherein->Representing the average deformation information of the pixel points in the first type of region,/and>the average deformation information of the pixel points in the second type area is represented;
when U is larger than the fourth threshold value, it is indicated that geological disaster occurs in the region corresponding to the remote sensing image, otherwise, it is indicated that geological disaster does not occur in the region, wherein when>/>When the first type of region is represented as a geological disaster region,and when the second type of region is represented as a geological disaster region.
According to a second aspect of embodiments of the present application, there is provided a multi-source data based geological disaster early identification device comprising a processor and a memory, wherein:
the memory is used for storing program codes;
the processor is configured to read the program code stored in the memory and execute the method according to the first aspect of the embodiment of the present application.
As can be seen from the above embodiments, the geological disaster early identification method and device based on multi-source data provided by the embodiments of the present application have the following beneficial effects:
according to the application, the corner point in the remote sensing image is extracted through the Fast corner point detection algorithm, the self-adaptive analysis of the corner point is completed based on the local information change of the corner point and the characteristics of the object area in the remote sensing image, the characteristic point is obtained, the extraction precision and efficiency of the characteristic point are improved, the extracted characteristic point information is more abundant, and the matching efficiency and precision are higher; the acquisition of the optimal reference image is completed based on the local change of the characteristic points, the matching is completed based on the difference of the characteristic point sequence between the optimal reference image and the remote sensing image, and the characteristic point pairs are acquired; the accurate registration of the images is completed based on the feature point pairs, so that the precision and the efficiency of registration are improved; and further improves the accuracy and efficiency of early identification of geological disasters in the InSAR algorithm.
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 as claimed.
Drawings
In order to more clearly illustrate the technical solution of the present application, the drawings that are needed in the embodiments will be briefly described below, and it will be obvious to those skilled in the art that other drawings can be obtained from these drawings without inventive effort.
Fig. 1 is a basic flow diagram of a geological disaster early recognition method based on multi-source data according to an embodiment of the present application;
fig. 2 is a basic flow chart of a method for obtaining feature points according to an embodiment of the present application;
FIG. 3 is a basic flow chart of a method for selecting an optimal reference image according to an embodiment of the present application;
fig. 4 is a basic flow chart of an image registration method according to an embodiment of the present application;
fig. 5 is a basic flow chart of a method for analyzing and registering images to identify geological disaster areas according to an embodiment of the present application.
Detailed Description
In order to make the technical solution of the present application better understood by those skilled in the art, the technical solution of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
The method for detecting the impurity pollution of the lubricating oil based on the artificial intelligence provided by the embodiment will be described in detail below with reference to the accompanying drawings.
Fig. 1 is a basic flow diagram of a geological disaster early recognition method based on multi-source data according to an embodiment of the present application. As shown in fig. 1, the method specifically includes the following steps:
step 100: and acquiring a remote sensing image.
The method comprises the steps of collecting an image of a region to be monitored through a radar satellite, wherein the collected image is a plurality of remote sensing images, and the collected image comprises a task region, namely a region for identifying geological disasters and a surrounding reference region so as to perform registration and phase difference. The information corresponding to the pixels of the image includes intensity information of the selected ground after radar backscatter of the resolution element, and also includes information related to the phase between the heel and the range.
Step 200: and extracting corner points in the remote sensing image by using a Fast corner point detection algorithm.
And performing corner detection on the pixel points in the remote sensing image by using Fast corner detection, and judging whether the pixel points are corner points or not based on the sum of intensity differences of the pixel points and surrounding M pixel points and the magnitude of a threshold value to obtain the corner points in the image, wherein in some embodiments of the application, M=16 is set according to an empirical value. The Fast corner detection process is a well-known technique and will not be described in detail here.
Step 300: and analyzing the local change information of the corner points and the adjacent pixel points to obtain the feature points.
The SAR image (remote sensing image) acquired by the radar is a coherent system, and the speckle noise is an inherent characteristic thereof. The uniform area shows obvious brightness random change, has no direct relation with resolution, polarization and incidence angle, and belongs to random noise. When Fast corner detection is used, since whether the pixel point is a corner is determined based on the sum of the intensity differences of the pixel point and the surrounding M pixel points and the threshold value, when noise points exist in the image, errors exist in the corner extracted by the Fast corner detection. Therefore, the diagonal points need to be further extracted, and the characteristic points are screened out for subsequent image selection and registration.
The pixel points in the acquired image have corresponding reflection intensity information and phase information. The reflected intensity information is usually the reflected intensity of the radar beam by the object, and the image can be adaptively segmented based on the intensity information of the pixels in the image, so as to be segmented into different object areas. Different areas in the remote sensing image often reflect different types of objects, such as various types of areas of houses, forests, lakes, etc. For the same type of object, the reflection intensity of the surface of the object should be relatively close, but the reflection intensity of the surface of different types of objects often has a certain difference due to different materials. According to analysis of the features of the remote sensing image, when artificially constructed reference objects such as houses exist in the image, the surfaces of the reference objects tend to be regular, for example, the houses tend to be rectangular on the remote sensing image, and the patterns such as the rectangular have certain correlation features among corner points obtained in the image. And noise points in SAR images corresponding to the InSAR algorithm are random noise, and correlation among corresponding angular points is poor, so that characteristic point screening is carried out on the angular points based on the correlation.
Fig. 2 is a basic flow chart of a method for obtaining feature points according to an embodiment of the present application, as shown in fig. 2, and analyzes local variation information of corner points and adjacent pixel points thereof to obtain feature points, including the following steps:
step 301: and sequencing M pixel points near the corner points from small to large according to the difference value of the reflection intensity with the corner points to obtain a sequence of adjacent pixel points of the corner points.
Assuming that N corner points are obtained in step 200, in this step, M pixel points near the N corner points are respectively ordered according to the difference value of reflection intensity from the corner points from small to large, so as to obtain a sequence of adjacent pixel points of the corner points. More specifically, taking a certain corner q as an example for analysis, according to step 200, it is known that the obtaining of the corner q is determined based on the sum of the intensity differences of m=16 pixels adjacent to the corner q, and for the corner q, the m=16 pixels adjacent to the corner q are ordered from small to large according to the intensity differences with the corner q, so as to obtain a sequence of adjacent pixels of the corner q.
Step 302: and acquiring corresponding adjacent pixel points of all adjacent pixel points of the corner point.
Each adjacent pixel point of the corner points is provided with a corresponding pixel point, and the obtained pixel points areAnd taking the corresponding adjacent pixel points of all the corner points. In some embodiments of the present application, the specific acquisition method corresponding to the neighboring pixel point is to connect the corner q with the neighboring pixel point f to acquireAdjacent pixel point nearest to q point in direction +.>Adjacent pixel dot->And then sequentially acquiring the corresponding adjacent pixel points of all the adjacent pixel points of the corner point as the corresponding points of the adjacent pixel points f.
Step 303: obtaining the reflection intensity changes of the corner points in different local directions of the adjacent pixel points according to the reflection intensities of the adjacent pixel points of the corner points and the corresponding adjacent pixel points。
Obtaining the reflection intensity changes of the corner points in different local directions of the adjacent pixel points according to the reflection intensities of the adjacent pixel points of the corner points and the corresponding adjacent pixel pointsThe calculation formula is as follows:
in the method, in the process of the application,for the reflection intensity variation in the local direction of the corner point adjacent to the pixel point,/->Arranging reflection intensity values of the ith adjacent pixel point in the adjacent pixel point sequence of the corner point,/for the adjacent pixel point sequence of the corner point>The reflection intensity value of the i-th adjacent pixel point corresponding to the adjacent pixel point i', wherein the i-th adjacent pixel point represents a pixel point with the i-th large row of the reflection intensity difference value with the corner q point in the q-point adjacent pixel point.
Step 304: and acquiring the relevance feature GL between each corner point to be analyzed and the rest arbitrary corner points based on the reflection intensity change in the local direction of the adjacent pixel points of the corner points to be analyzed and the reflection intensity change in the local direction of the adjacent pixel points of the rest arbitrary corner points.
This step requires the acquisition of the correlation features GL between each corner to be analyzed and the rest of any corner. Here, the point q of the corner to be analyzed and the point a of any other corner are selected as examples for specific display. Based on the reflection intensity variation in the local direction of the adjacent pixel point of the corner q to be analyzed and the reflection intensity variation in the local direction of the adjacent pixel point of the rest arbitrary corner a, the correlation feature GL between the corner q and the corner a is obtained, and the calculation formula is as follows:
in the formula, q represents a corner point to be analyzed, and a represents any other corner points;、/>the reflection intensity values of the corner points a and q are respectively represented, and the larger the difference is, the smaller the relevance among the corner points is; />For the number of adjacent pixel points, the setting can be 16 in the application;representing the clockwise included angle between the straight line formed by the ith adjacent pixel point of the corner q point and the horizontal line, and +.>Is the ith neighbor of the corner point aA clockwise included angle between a straight line formed by the near pixel point and the point a and a horizontal line is +.>And->The larger the difference is, the larger the difference of the reflection intensity distribution of the local pixel points at the corner points is, and the smaller the relevance is; />Representing the reflection intensity variation in the local direction of the ith adjacent pixel point of the corner q point, +.>Representing the reflection intensity variation in the local direction of the ith adjacent pixel point of the corner point a, +.>And->The larger the difference, the less similar the local variations of the corner points are, the less likely the corner points of the same region are, and the smaller the correlation is.
Step 305: and obtaining a possible value KN taking the corner point to be analyzed as a characteristic point according to the relevance between the corner point to be analyzed and any other corner points.
According to the relevance between the corner points to be analyzed and the rest corner points obtained in the step 304, screening and extraction of the feature points can be completed based on the relevance between the corner points, the possible value KN with the corner point q as the feature point is obtained, and the calculation formula is as follows:
wherein KN is a possible value of the corner point to be analyzed as a characteristic point, and the larger KN indicates the larger possibility that the corner point q is the characteristic point of a certain area;for the ith corner point and to be analyzedThe relevance among the corner points is that the larger the value is, the greater the possibility that the q point of the corner point is a characteristic point is; />For the frequency of occurrence of the association of the ith corner point with the corner point to be analyzed,/->Representing the number of corner points.
In some embodiments of the present application,the obtaining method of (1) comprises the following steps: count the occurrence number of the correlation feature GL and record as + ->Then->. Further, since the relevance between corner points is often not completely consistent, the number of occurrences of relevance feature GL is counted when +.>At 0.6, p->When the number of occurrences is counted, if the association index between a certain corner point and a q point is located in the interval [ -f ]>-0.05，/>+0.05]When makingBut when->And if so, normally counting.
Step 306: and carrying out normalization processing on the possible value KN of which the corner point to be analyzed is the characteristic point, setting a possibility judgment first threshold value, and when the KN corresponding to the corner point to be analyzed is larger than the first threshold value, obtaining the corner point to be analyzed as an actual characteristic point, otherwise, obtaining the corner point to be analyzed as a noise interference point.
And carrying out normalization processing on a possible value KN of which the angular point to be analyzed is a characteristic point, setting the possibility to judge that the first threshold value can be 0.75, setting the first threshold value according to an empirical value, adjusting by an implementer, and when the KN corresponding to the angular point to be analyzed is more than 0.75, the angular point to be analyzed is an actual characteristic point, otherwise, the angular point to be analyzed is a noise interference point.
Step 400: and analyzing the local change information of the feature points and the adjacent pixel points, and selecting an optimal reference image.
Fig. 3 is a basic flow chart of a method for selecting an optimal reference image according to an embodiment of the present application, as shown in fig. 3, the method includes the following steps:
step 401: and analyzing the local change information of the characteristic points and the adjacent pixel points to obtain the optimal value Y of the optimal reference image of the remote sensing image.
Analyzing local change information of the feature points and adjacent pixel points thereof, and obtaining a preferred value Y of an optimal reference image of the remote sensing image, wherein the calculation formula is as follows:
in the formulaThe larger the value of the feature points is, the richer the image information is, and the larger the preferred value of the feature points is used as a reference image; m is the number of adjacent pixel points of the feature points; />The change of the reflection intensity in the local direction of the jth adjacent pixel point which is the ith feature point; />For the ith corner pointThe larger the value of the feature point, the more obvious the difference between the image information, the higher the information contrast, the easier the matching is, and the higher the precision is.
Step 402: and selecting the remote sensing image with the maximum optimal value Y as an optimal reference image.
The purpose of acquiring the optimal reference image is to acquire images with more characteristic points, obvious differences among areas in the images and less possibility of geological disasters. When the remote sensing image and the optimal reference image are aligned, the characteristic points in the remote sensing image and the characteristic points in the optimal reference image are matched, and as the number of the characteristic points in the optimal reference image is large and the difference among the areas is obvious, the efficiency and the precision of the characteristic point matching are improved, and the efficiency and the precision of the image registration are further improved.
Step 500: and carrying out image registration between the optimal reference image and the remote sensing image according to the feature points to obtain a registration image.
Fig. 4 is a basic flow chart of an image registration method provided by the embodiment of the application, as shown in fig. 4, performing image registration between an optimal reference image and a remote sensing image according to feature points to obtain a registered image, and includes the following steps:
step 501: analyzing the feature points obtained by the optimal reference image and the remote sensing image, and when the correlation feature GL between the feature points is larger than a second threshold value, the two feature points belong to the same feature point sequence; the steps are repeated until the characteristic points in the remote sensing image are divided into different characteristic point sequences.
According to the optimal reference image obtained in the above step 400, the registration between the remote sensing image and the optimal reference image is required to be completed by dividing the N remote sensing images into the optimal reference image and N-1 remote sensing images. The feature points of the collected images are extracted according to the step 300, and the number of the feature points in the obtained remote sensing image is often smaller than that of the optimal reference image because the optimal reference image is selected in the application.
And analyzing the characteristic points obtained by the remote sensing image, wherein when the relevance between the characteristic points is larger than a second threshold value, the two characteristic points belong to the same characteristic point sequence, and in some embodiments of the application, the second threshold value is set to be 0.8, and the threshold value is set according to an empirical value, so that an implementer can adjust. The steps are repeated until the characteristic points in the remote sensing image are divided into different characteristic point sequences.
Step 502: and calculating the matching degree P of all the characteristic point sequences B in the remote sensing image W and the characteristic point sequences A in the optimal reference image Q, and obtaining the maximum matching degree Pmax of the characteristic point sequences in the remote sensing image W and the characteristic point sequences in the optimal reference image.
For the optimal reference image and the remote sensing image, different corresponding characteristic point sequences are respectively arranged, and the optimal reference image Q and the remote sensing image W are taken as examples. The characteristic point sequence in the optimal reference image Q is thatAnd the characteristic point sequence in the remote sensing image W is +.>And calculating the matching degree P of the characteristic point sequence in the remote sensing image W and the characteristic sequence in the optimal reference image. Taking the optimal reference image Q feature point sequence a and the remote sensing image W feature point sequence B as examples. For the feature point sequence B, since feature points of the same sequence are often edge points of the same object region, for feature points in the same feature point sequence, a fitted closed region can be obtained by a closed curve fitting algorithm such as a polynomial fitting algorithm, which is not described in detail in the prior art. And acquiring a centroid B of the closed region corresponding to the characteristic point sequence B, wherein the sum of distances from the centroid point to the edge of the fitted closed region is minimum, the shape characteristic of the region formed by the characteristic points can be described by acquiring the relative distance and angle between the centroid B and the characteristic points, and the acquisition of the matching degree of the characteristic point sequence is completed based on the shape characteristic difference between the characteristic point sequences and the relevance difference between the characteristic point sequences.
Calculating the matching degree P of all feature point sequences B in the remote sensing image W and feature point sequences A in the optimal reference image Q, firstly acquiring corresponding feature points of feature points in the feature point sequences B in the feature point sequences A, wherein the acquiring method is to acquire straight lines between the feature points in the feature point sequences B and centroid points, the feature points corresponding to the smallest angle difference corresponding to the straight lines are the corresponding feature points, the acquiring method is not described in detail in the known technology, the matching degree P can be acquired, and then the maximum matching degree Pmax of the feature point sequences in the remote sensing image W and the feature point sequences in the optimal reference image is acquired, and the calculating formula is as follows:
in the method, in the process of the application,the average relevance between the ith feature point in the feature point sequence B and the rest feature points in the feature point sequence; />The average relevance between the ith feature point (corresponding feature point in B) in the feature point sequence A and the rest feature points in the feature point sequence; />For the distance between the centroid b of the remote sensing image W and the ith feature point,/and>for the distance between the optimal reference image Q centroid a and the ith feature point, +.>And->The smaller the difference, the greater the degree of matching; />For a very small positive number, the prevent denominator is 0 and an example value may be 0.001.
It should be noted that early geological disasters may have occurred in the remote sensing image WResulting in a change of the target within the remote sensing image. When the characteristic points are completely matched under normal conditions, the number of the characteristic points in the characteristic point sequence A is consistent with that of the characteristic points in the characteristic point sequence B, and when early geological disasters occur in the remote sensing image W, deformation occurs in the image, and the number of the characteristic points in the characteristic point sequence B is possibly smaller than that of the corresponding matched characteristic point sequence A. Therefore, in the application, when the matching degree of the characteristic point sequence B and the characteristic point sequence A is calculated, the phenomenon occursWhen the matching degree is calculated, the matching degree calculation between the characteristic points and the characteristic point sequence B is required to be cut out from the characteristic point sequence A, and then the cutting out is carried out for a plurality of times to obtain the maximum matching degree as the final matching degree between the characteristic point sequence A and the characteristic point sequence B; when->In this case, the matching degree does not need to be calculated.
Step 503: judging whether the maximum matching degree Pmax is larger than a third threshold value, if the maximum matching degree Pmax is larger than the third threshold value, the two feature point sequences can be considered to be successfully matched, otherwise, the remote sensing image can be considered to be changed, and the matching is unsuccessful.
According to step 502, the maximum matching degree Pmax between the feature point sequence in the remote sensing image W and the feature point sequence in the optimal reference image is obtained, whether the maximum matching degree Pmax is greater than a third threshold value is judged, the third threshold value is an empirical value, the value can be 0.75 in some embodiments of the application, if Pmax is greater than 0.75, the two feature point sequences can be considered to be successfully matched, otherwise, the remote sensing image is considered to be changed, and the matching is unsuccessful.
Step 504: and calculating a transformation matrix from the remote sensing image to the optimal reference image based on the successfully matched characteristic point sequence pairs.
Step 505: and registering the remote sensing image according to the calculated transformation matrix.
According to step 503, feature point matching between the remote sensing image and the optimal reference image is completed, and based on the successfully matched feature point pairs, a transformation matrix from the remote sensing image to be registered to the optimal reference image is calculated. And registering the remote sensing images to be registered according to the calculated transformation matrix. The process is a well-known technology and will not be described herein.
Step 600: the registered images are analyzed to identify a geological disaster area.
Fig. 5 is a basic flow chart of a method for analyzing a registered image to identify a geological disaster area, which is provided by an embodiment of the present application, and as shown in fig. 5, the method includes the following steps:
step 601: and subtracting the phases of the corresponding pixels of the registration image to obtain a phase difference image.
Step 500 completes the registration between the plurality of remote sensing images acquired during the observation period in the InSAR algorithm. The phase difference image, namely the phase image of interference, can be obtained by subtracting the phases of the corresponding pixels from the registered image, namely the two radar images after registration.
Step 602: and processing pixel values of pixel points in the phase difference image through a phase unwrapping algorithm to obtain the earth surface deformation information Z.
The information of the phase difference obtained in step 601 is the topography relief on the earth's surface, and is typically represented by the deformation of the earth's surface. The phase difference image can be further analyzed, and the pixel values of the pixel points in the phase difference image (the phase difference of the corresponding pixel points in the registered image) are processed through a phase unwrapping algorithm to obtain the earth surface deformation information Z, wherein the specific process is a known technology and is not repeated.
Step 603: and (3) using k-means mean clustering, setting k=2, and dividing the image into two types of regions by using a distance measurement mode as deformation information difference.
Step 602 completes the acquisition of the surface deformation information of the area to be monitored, namely, each pixel point in the image has the corresponding surface deformation information Z, and the larger the deformation information is, the greater the possibility of geological disasters is. K-means mean clustering can be used, k=2 is set, the distance measurement mode is deformation information difference, the image is divided into two types of areas, and the clustering process is a known technology and is not repeated here.
Step 604: obtaining deformation difference U=1-exp of two types of regions) Wherein->Representing the average deformation information of the pixel points in the first type of region,/and>the average deformation information of the pixel points in the second class area is represented.
Step 603 completes the segmentation of the image into two classes of regions and then continues to acquire the deformation differences U for the two classes of regions. The calculation formula is as follows:
U=1-exp（-）
wherein the method comprises the steps ofRepresenting the average deformation information of the pixel points in the first type of region,/and>then mean deformation information of the pixels in the second class of regions is represented, < >>And->The greater the difference, the greater the likelihood of a geological disaster occurring within the image.
Step 605: when U is larger than the fourth threshold value, it is indicated that geological disaster occurs in the region corresponding to the remote sensing image, otherwise, it is indicated that geological disaster does not occur in the region, wherein when>/>When the first type of region is a geological disaster region, < ->And when the second type of region is represented as a geological disaster region.
In some embodiments of the present application, the fourth threshold value may be 0.6, which may be set according to an empirical value, and may be adjusted by an practitioner. And when U is more than 0.6, indicating that geological disasters occur in the region corresponding to the remote sensing image, otherwise, indicating that no geological disasters occur in the region. Wherein when>/>When the first type of region is a geological disaster region, < ->And when the second type of region is represented as a geological disaster region.
Based on the same inventive concept as the above method, the present embodiment further provides a geological disaster early recognition device based on multi-source data, the device including a processor and a memory, wherein:
a memory for storing program code; a processor for reading the program code stored in the memory and executing: acquiring a remote sensing image; extracting corner points in the remote sensing image by using a Fast corner point detection algorithm; analyzing local change information of the corner points and the adjacent pixel points to obtain characteristic points; analyzing local change information of the feature points and adjacent pixel points thereof, and selecting an optimal reference image; performing image registration between the optimal reference image and the remote sensing image according to the feature points to obtain a registration image; the registered images are analyzed to identify a geological disaster area.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
It is noted that unless specified and limited otherwise, relational terms such as "first" and "second", and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a circuit structure, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such article or apparatus. Without further limitation, the statement "comprises one … …" does not exclude that an additional identical element is present in an article or device that comprises the element. In addition, the term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.
Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. This application is intended to cover any variations, uses, or adaptations of the application following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice within the art to which the application pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the application being indicated by the following claims.
It is to be understood that the application is not limited to the precise arrangements and instrumentalities shown in the drawings, which have been described above, and that various modifications and changes may be effected without departing from the scope thereof. The scope of the application is limited only by the appended claims.
Claims (5)
1. A method for early identification of geologic hazards based on multi-source data, the method comprising:
acquiring a remote sensing image;
extracting corner points in the remote sensing image by a Fast corner point detection algorithm;
analyzing local change information of the corner points and the adjacent pixel points to obtain characteristic points;
analyzing the local change information of the characteristic points and the adjacent pixel points, and selecting an optimal reference image;
performing image registration between the optimal reference image and the remote sensing image according to the characteristic points to obtain a registration image;
analyzing the registration image and identifying a geological disaster area;
analyzing the local change information of the corner points and the adjacent pixel points to obtain characteristic points, wherein the method comprises the following steps:
sequencing M pixel points near the corner points according to the difference value of the reflection intensity with the corner points from small to large to obtain a sequence of adjacent pixel points of the corner points;
acquiring corresponding adjacent pixel points of all adjacent pixel points of the corner points;
obtaining the reflection intensity variation of the corner point in different local directions of adjacent pixel points according to the reflection intensity of the adjacent pixel points of the corner point and the corresponding adjacent pixel pointsThe calculation formula is as follows:
in the method, in the process of the application,for the reflection intensity variation in the local direction of the corner point adjacent to the pixel point,/->Arranging reflection intensity values of the ith adjacent pixel point in the adjacent pixel point sequence of the corner point,/for the adjacent pixel point sequence of the corner point>The reflection intensity value of the i-th adjacent pixel point corresponding to the i' adjacent pixel point is the i-th adjacent imageThe pixel points represent the ith row of the i-th large pixel points in the adjacent pixel points of the q points, wherein the reflection intensity difference value between the pixel points and the q points of the corner points is equal to the i-th large pixel point;
based on the reflection intensity variation in the local direction of the adjacent pixel points of the corner points to be analyzed and the reflection intensity variation in the local direction of the adjacent pixel points of the rest arbitrary corner points, the correlation feature GL between each corner point to be analyzed and the rest arbitrary corner points is obtained, and the calculation formula is as follows:
in the formula, q represents the corner point to be analyzed, a represents any other corner points,、/>representing the reflection intensity values of corner a and corner q, respectively,/->For the number of adjacent pixels, < >>Representing the clockwise included angle between the straight line formed by the ith adjacent pixel point of the corner q point and the horizontal line, and +.>Is the clockwise included angle between the straight line formed by the ith adjacent pixel point of the angular point a and the horizontal line, and is +.>Representing the reflection intensity variation in the local direction of the ith adjacent pixel point of the corner q point, +.>Representing the reflection intensity change of the point a in the local direction of the ith adjacent pixel point;
according to the relevance between the corner points to be analyzed and any other corner points, a possible value KN taking the corner points to be analyzed as characteristic points is obtained, and the calculation formula is as follows:
wherein KN is the possible value of the characteristic point of the corner point to be analyzed,for the association between the ith corner and the corner to be analyzed, +.>For the frequency of occurrence of the association of the ith corner point with the corner point to be analyzed,/->Representing the number of corner points;
carrying out normalization processing on a possible value KN of which the corner point to be analyzed is a characteristic point, setting a possibility judging first threshold value, and when the possible value KN corresponding to the corner point to be analyzed is larger than the first threshold value, the corner point to be analyzed is an actual characteristic point, otherwise, the corner point to be analyzed is a noise interference point;
analyzing the local change information of the feature points and the adjacent pixel points, and selecting an optimal reference image, wherein the method comprises the following steps:
analyzing the local change information of the characteristic points and the adjacent pixel points to obtain the optimal value Y of the optimal reference image of the remote sensing image, wherein the calculation formula is as follows:
in the formulaThe number of the feature points is M, the number of the adjacent pixel points of the feature points is +.>Reflection intensity variation in local direction of jth adjacent pixel point which is ith feature point, +.>The possibility that the ith corner point is a feature point;
selecting a remote sensing image with the maximum optimal value Y as an optimal reference image;
performing image registration between the optimal reference image and the remote sensing image according to the feature points to obtain a registration image, including:
analyzing the feature points obtained by the optimal reference image and the remote sensing image, and when the correlation feature GL between the feature points is larger than a second threshold value, the two feature points belong to the same feature point sequence; the steps are repeated continuously until the characteristic points in the remote sensing image are divided into different characteristic point sequences;
calculating the matching degree P of all the characteristic point sequences B in the remote sensing image W and the characteristic point sequences A in the optimal reference image Q, and obtaining the maximum matching degree Pmax of the characteristic point sequences in the remote sensing image W and the characteristic point sequences in the optimal reference image, wherein the calculation formula is as follows:
in the method, in the process of the application,for the average relevance between the ith feature point in the feature point sequence B and the rest of the feature points in the feature point sequence, +.>For the average relevance between the ith feature point in the feature point sequence A and the rest of the feature points in the feature point sequence,/the combination of the feature points is given by->For the distance between the centroid b of the remote sensing image W and the ith feature pointLeave, go up>For the distance between the optimal reference image Q centroid a and the ith feature point, +.>Is an extremely small positive number;
judging whether the maximum matching degree Pmax is larger than a third threshold value, if the maximum matching degree Pmax is larger than the third threshold value, the two feature point sequences can be considered to be successfully matched, otherwise, the remote sensing image can be considered to be changed, and the matching is unsuccessful;
based on the successfully matched characteristic point sequence pairs, calculating a transformation matrix from the remote sensing image to the optimal reference image;
registering the remote sensing image according to the calculated transformation matrix;
analyzing the registered image, identifying a geological disaster area, comprising:
subtracting the phases of the pixels corresponding to the registration images to obtain a phase difference image;
processing pixel values of pixel points in the phase difference image through a phase unwrapping algorithm to obtain earth surface deformation information Z;
using k-means mean clustering, setting k=2, and dividing the image into two types of areas by using a distance measurement mode as deformation information difference;
obtaining deformation difference U=1-exp of two types of regions) Wherein->Representing the average deformation information of the pixel points in the first type of region,/and>the average deformation information of the pixel points in the second type area is represented;
when U is larger than the fourth threshold value, the remote sensing image is indicated to generate the ground in the region corresponding to the remote sensing imageA physical disaster, otherwise, indicating that no physical disaster occurs in the region, wherein, when>/>When the first type of region is a geological disaster region, < ->And when the second type of region is represented as a geological disaster region.
2. The method for early recognition of a geological disaster based on multi-source data according to claim 1, wherein obtaining corresponding neighboring pixels of all neighboring pixels of the corner point comprises:
connecting the corner point q with the adjacent pixel point f to obtainAdjacent pixel point nearest to q point in direction +.>Adjacent pixel dot->And then sequentially acquiring the corresponding adjacent pixel points of all the adjacent pixel points of the corner point as the corresponding points of the adjacent pixel points f.
3. The method for early identification of a geological disaster based on multi-source data according to claim 1, wherein said method comprises the steps ofThe obtaining method of (1) comprises the following steps: count the occurrence number of the correlation feature GL and record as + ->Then->。
4. The method for early identification of a geological disaster based on multi-source data according to claim 3, wherein said method comprises the steps ofThe obtaining method of (2) further comprises: counting the number of occurrences of the relevance feature GL when +.>At 0.6, p->When the number of occurrences is counted, if the association index between a certain corner point and a q point is located in the interval [ -f ]>-0.05，/>+0.05]When in use, make->But when->And if so, normally counting.
5. A multi-source data based geological disaster early identification device, the device comprising a processor and a memory, wherein:
the memory is used for storing program codes;
the processor being configured to read the program code stored in the memory and to perform the method of any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310686349.4A CN116434072B (en) | 2023-06-12 | 2023-06-12 | Geological disaster early identification method and device based on multi-source data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310686349.4A CN116434072B (en) | 2023-06-12 | 2023-06-12 | Geological disaster early identification method and device based on multi-source data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116434072A CN116434072A (en) | 2023-07-14 |
| CN116434072B true CN116434072B (en) | 2023-08-18 |
Family
ID=87083555
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310686349.4A Active CN116434072B (en) | 2023-06-12 | 2023-06-12 | Geological disaster early identification method and device based on multi-source data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116434072B (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104167003A (en) * | 2014-08-29 | 2014-11-26 | 福州大学 | Method for fast registering remote-sensing image |
| CN107132539A (en) * | 2017-05-03 | 2017-09-05 | 中国地质科学院探矿工艺研究所 | Landslide early-stage identification method of time sequence InSAR (interferometric synthetic Aperture Radar) based on small baseline set |
| CN108957456A (en) * | 2018-08-13 | 2018-12-07 | 伟志股份公司 | Landslide monitoring and EARLY RECOGNITION method based on multi-data source SBAS technology |
| KR20190063670A (en) * | 2017-11-30 | 2019-06-10 | (주)코어센스 | Intelligent system for ignition point surveillance using composite image of thermal camera and color camera |
| CN110427857A (en) * | 2019-07-26 | 2019-11-08 | 国网湖北省电力有限公司检修公司 | A kind of transmission line of electricity geological disasters analysis method based on Remote Sensing Data Fusion Algorithm |
| CN110765934A (en) * | 2019-10-22 | 2020-02-07 | 成都理工大学 | Geological disaster identification method based on multi-source data fusion |
| CN111142119A (en) * | 2020-01-10 | 2020-05-12 | 中国地质大学（北京） | Mine geological disaster dynamic identification and monitoring method based on multi-source remote sensing data |
| WO2020166972A1 (en) * | 2019-02-13 | 2020-08-20 | 제주대학교 산학협력단 | Method for classifying and analyzing disasters by utilizing sar-based big data, and computer-readable recording medium in which program for executing same method is recorded |
| CN113807272A (en) * | 2021-09-22 | 2021-12-17 | 应急管理部国家减灾中心 | Method for rapidly extracting disaster situations of disaster damaged roads on remote sensing platform based on vector guidance |
| US11521379B1 (en) * | 2021-09-16 | 2022-12-06 | Nanjing University Of Information Sci. & Tech. | Method for flood disaster monitoring and disaster analysis based on vision transformer |
| US11521377B1 (en) * | 2021-10-26 | 2022-12-06 | Nanjing University Of Information Sci. & Tech. | Landslide recognition method based on laplacian pyramid remote sensing image fusion |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101984463A (en) * | 2010-11-02 | 2011-03-09 | 中兴通讯股份有限公司 | Method and device for synthesizing panoramic image |
| CN110765941B (en) * | 2019-10-23 | 2022-04-26 | 北京建筑大学 | Seawater pollution area identification method and equipment based on high-resolution remote sensing image |
-
2023
- 2023-06-12 CN CN202310686349.4A patent/CN116434072B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104167003A (en) * | 2014-08-29 | 2014-11-26 | 福州大学 | Method for fast registering remote-sensing image |
| CN107132539A (en) * | 2017-05-03 | 2017-09-05 | 中国地质科学院探矿工艺研究所 | Landslide early-stage identification method of time sequence InSAR (interferometric synthetic Aperture Radar) based on small baseline set |
| KR20190063670A (en) * | 2017-11-30 | 2019-06-10 | (주)코어센스 | Intelligent system for ignition point surveillance using composite image of thermal camera and color camera |
| CN108957456A (en) * | 2018-08-13 | 2018-12-07 | 伟志股份公司 | Landslide monitoring and EARLY RECOGNITION method based on multi-data source SBAS technology |
| WO2020166972A1 (en) * | 2019-02-13 | 2020-08-20 | 제주대학교 산학협력단 | Method for classifying and analyzing disasters by utilizing sar-based big data, and computer-readable recording medium in which program for executing same method is recorded |
| CN110427857A (en) * | 2019-07-26 | 2019-11-08 | 国网湖北省电力有限公司检修公司 | A kind of transmission line of electricity geological disasters analysis method based on Remote Sensing Data Fusion Algorithm |
| CN110765934A (en) * | 2019-10-22 | 2020-02-07 | 成都理工大学 | Geological disaster identification method based on multi-source data fusion |
| CN111142119A (en) * | 2020-01-10 | 2020-05-12 | 中国地质大学（北京） | Mine geological disaster dynamic identification and monitoring method based on multi-source remote sensing data |
| US11521379B1 (en) * | 2021-09-16 | 2022-12-06 | Nanjing University Of Information Sci. & Tech. | Method for flood disaster monitoring and disaster analysis based on vision transformer |
| CN113807272A (en) * | 2021-09-22 | 2021-12-17 | 应急管理部国家减灾中心 | Method for rapidly extracting disaster situations of disaster damaged roads on remote sensing platform based on vector guidance |
| US11521377B1 (en) * | 2021-10-26 | 2022-12-06 | Nanjing University Of Information Sci. & Tech. | Landslide recognition method based on laplacian pyramid remote sensing image fusion |
Non-Patent Citations (1)
| Title |
|---|
| 吴伟 ; 丁香乾 ; 闫明 ; .基于异常区域感知的多时相高分辨率遥感图像配准.计算机应用.2016,(10),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116434072A (en) | 2023-07-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gamba et al. | Improving urban road extraction in high-resolution images exploiting directional filtering, perceptual grouping, and simple topological concepts | |
| Tian et al. | Building change detection based on satellite stereo imagery and digital surface models | |
| Pirotti et al. | Ground filtering and vegetation mapping using multi-return terrestrial laser scanning | |
| Awrangjeb et al. | Building detection in complex scenes thorough effective separation of buildings from trees | |
| Prieto et al. | A similarity metric for edge images | |
| US20060013455A1 (en) | Image analysis | |
| CN108629760B (en) | Method and device for detecting change area of remote sensing image | |
| CN103400137A (en) | Method for extracting geometrical building parameters of synthetic aperture radar (SAR) image | |
| Princen et al. | A comparison of Hough transform methods | |
| CN116051822A (en) | Concave obstacle recognition method and device, processor and electronic equipment | |
| Shi et al. | A method to detect earthquake-collapsed buildings from high-resolution satellite images | |
| CN116434072B (en) | Geological disaster early identification method and device based on multi-source data | |
| CN112906659A (en) | Remote sensing image change detection method based on virtual sample | |
| CN112348750A (en) | SAR image change detection method based on threshold fusion and neighborhood voting | |
| CN114742849B (en) | Leveling instrument distance measuring method based on image enhancement | |
| Sghaier et al. | Combination of texture and shape analysis for a rapid rivers extraction from high resolution SAR images | |
| Zhao et al. | An automatic SAR-based change detection method for generating large-scale flood data records: the UK as a test case | |
| Li et al. | Landslide detection based on GLCM using SAR images | |
| CN113822361B (en) | SAR image similarity measurement method and system based on Hamming distance | |
| Tian et al. | Region based forest change detection from Cartosat-1 stereo imagery | |
| Santi et al. | Measuring the forest cover change in burned area using change vector analysis approach | |
| CN112270675B (en) | Urban waterlogging area detection method based on polarized radar remote sensing image | |
| CN116542526B (en) | Disaster risk assessment method, device and equipment in homeland space planning | |
| CN116758086B (en) | Bulldozer part quality detection method based on image data | |
| Spadavecchia et al. | Preliminary Test on Structural Elements Health Monitoring with a LiDAR-Based Approach |
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 |