CN107133979B - Polarized radar building damage information extraction method - Google Patents

Polarized radar building damage information extraction method Download PDF

Info

Publication number
CN107133979B
CN107133979B CN201710283555.5A CN201710283555A CN107133979B CN 107133979 B CN107133979 B CN 107133979B CN 201710283555 A CN201710283555 A CN 201710283555A CN 107133979 B CN107133979 B CN 107133979B
Authority
CN
China
Prior art keywords
building
texture
statistical model
damage information
information extraction
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.)
Expired - Fee Related
Application number
CN201710283555.5A
Other languages
Chinese (zh)
Other versions
CN107133979A (en
Inventor
陈启浩
刘修国
李林林
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.)
China University of Geosciences
Original Assignee
China University of Geosciences
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 China University of Geosciences filed Critical China University of Geosciences
Priority to CN201710283555.5A priority Critical patent/CN107133979B/en
Publication of CN107133979A publication Critical patent/CN107133979A/en
Application granted granted Critical
Publication of CN107133979B publication Critical patent/CN107133979B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/40—Analysis of texture
    • G06T7/41—Analysis of texture based on statistical description of texture
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10032—Satellite or aerial image; Remote sensing
    • G06T2207/10044—Radar image

Abstract

The invention provides a polarized radar building damage information extraction method, which comprises the following steps: preprocessing an original PolSAR image, including selecting a research area and extracting a coherent matrix; carrying out eigenvalue-eigenvector decomposition on the coherent matrix, identifying pixels which are smaller than the characteristic threshold value in the research area as non-building areas, and rejecting the non-building areas; modeling a coherent matrix of a building area by using a statistical model, extracting texture features based on texture parameters of the statistical model, classifying pixels with the texture features larger than a preset texture threshold value as collapsed buildings, and classifying pixels with the texture features smaller than or equal to the preset texture threshold value as intact buildings; and partitioning the building area, and then evaluating the precision of the building damage information extraction result by combining the obtained building damage information extraction result. The invention only uses the texture parameters of the statistical model to extract the collapsed building, fully utilizes the information of the PolSAR image, and simultaneously improves the precision and the efficiency of extracting the damage information of the building.

Description

polarized radar building damage information extraction method
Technical Field
the invention relates to the technical field of Polarimetric Synthetic Aperture Radar (Polarimetric SAR) image processing, in particular to a method for extracting building damage information of a Polarimetric Radar.
Background
The natural disasters such as earthquake, typhoon and the like bring great loss to the lives and properties of people. Buildings are the main places where people live, and most of the loss caused by disasters is caused by the collapse of the buildings. Therefore, accurate extraction of damage information of the post-disaster building can provide decision support for disaster rescue and post-disaster reconstruction. After the disaster happens, traffic and communication equipment are destroyed, the weather condition is severe, and meanwhile, the danger of secondary disaster happens, so that the disaster condition of a disaster area is very difficult to timely acquire. The SAR is a main way for acquiring disaster information due to the wide coverage range, strong penetrability and all-weather working capability all day long.
Aiming at different data used for extracting the building damage information, the existing extraction method comprises the following steps:
And extracting the building damage information based on the single-channel or multi-channel SAR image. The method mainly utilizes the changes of the SAR images before and after disasters, such as correlation, coherence and the like, or texture information of the SAR images after the disasters to extract collapsed buildings. However, the ground feature information acquired by the single-channel or multi-channel SAR is not comprehensive enough, and the detection effect on collapsed buildings is poor.
And extracting the building damage information based on the PolSAR image. PolSAR can obtain echoes of the ground objects under 4 polarization channels, is more sensitive to the ground objects, and provides rich information for extracting building damage information. The collapsed building is extracted based on the multi-temporal PolSAR image, and mainly changes of scattering mechanisms of the building before and after a disaster are utilized. However, in some cases, the matched pre-disaster and post-disaster data are difficult to obtain, and the registration of the SAR images is difficult, so that the extraction of collapsed buildings using the multi-temporal PolSAR images is limited to a certain extent. The method for extracting collapsed buildings by utilizing the post-disaster single PolSAR image mainly comprises the steps of polarization-based feature, texture-based feature, polarization-based feature and texture feature. The existing method for extracting collapsed buildings based on single PolSAR images after disasters has more use parameters and more complex process. And the texture information used for the PolSAR image to extract the collapsed building is usually extracted based on the total power image, only diagonal elements in a coherent matrix are utilized, and the information of the PolSAR image is not comprehensively utilized.
Disclosure of Invention
In view of the above, the present invention provides a method for extracting damaged information of a polarized radar building, so as to solve the defects of incomplete extracted information, many parameters, complex process, etc.
In order to achieve the above object, the present invention provides a polarized radar building damage information extraction method, which is based on a PolSAR image processing technology, and comprises the following steps:
step 101: preprocessing an original PolSAR image, including selecting a research area and extracting a coherent matrix;
Step 102: carrying out eigenvalue-eigenvector decomposition on the coherent matrix, setting a characteristic threshold value based on the decomposed eigenvalue, identifying pixels smaller than the characteristic threshold value in the research area as non-building areas, identifying pixels larger than the characteristic threshold value as building areas, and removing the non-building areas;
Step 103: modeling a coherent matrix of a building area by using a statistical model, extracting texture features based on texture parameters of the statistical model, classifying pixels with the texture features larger than a preset texture threshold value as collapsed buildings, classifying pixels with the texture features smaller than or equal to the preset texture threshold value as intact buildings, and obtaining a building damage information extraction result;
Step 104: and (4) partitioning the building area, and then evaluating the precision of the building damage information extraction result by combining the building damage information extraction result obtained in the step 103.
Further, in step 102, performing eigenvalue-eigenvector decomposition on the coherence matrix specifically includes:
Where T is the coherence matrix, λnand munRespectively representing eigenvalues and eigenvectors, lambda1≥λ2≥λ3H denotes the transpose of the conjugate, the feature vector munexpressed as:
Wherein alpha isnIndicating the scattering mechanism of the target to which the scattering vector corresponds, betanIs the target azimuth angle phin、δn、γnFor 3 phase angles, T stands for transpose operation, let λ2+λ3Pixels less than a characteristic threshold are identified as non-building areas, λ1＞0。
further, the extracting process of the texture feature in step 103 includes estimating the texture parameter of the statistical model and taking the logarithm of the estimated texture parameter of the statistical model as the texture feature.
further, the statistical model is G0The statistical model has the texture parameters as follows:
M＝tr(Σ-1T)
where T is a coherence matrix, Σ ═ E [ T ], λ is a texture parameter of the statistical model, L represents a view, Var { · } represents a variance, and d represents a dimension of the coherence matrix.
Further, the specific process of step 104 is as follows:
Step 401: setting the construction of the jth block in the obtained blocksThe damage index of buildings is DLIjAnd then:
Wherein, CjNumber of pixels, I, representing a collapsed building in the jth blockjthe number of pixels representing the good buildings in the jth block, and the building damage index DLIjRepresents the damage degree of the building;
Step 402: according to DLIjDividing the corresponding blocks into different grades by a preset grading threshold value to obtain a classification result of the damage degree of the building;
step 403: and evaluating the accuracy of the building information extraction by using the detection rate, the false alarm rate or the total accuracy by using the confusion matrix.
in order to achieve the above object, the present invention further provides another method for extracting the building damage information of the polarization radar, which is based on the polarisar image processing technology, and comprises the following steps:
Step 101: preprocessing an original PolSAR image, including selecting a research area and extracting a coherent matrix;
Step 102: identifying a building area and a non-building area in the research area based on the coherent matrix, and removing the non-building area;
Step 103: modeling a coherent matrix of a building area by using a statistical model, estimating texture parameters of the statistical model, taking logarithm of the estimated texture parameters as texture features, classifying pixels with the texture features larger than a preset texture threshold value as collapsed buildings, classifying pixels with the texture features smaller than or equal to the preset texture threshold value as intact buildings, and obtaining a building damage information extraction result;
step 104: and (4) partitioning the building area, and then evaluating the precision of the building damage information extraction result by combining the building damage information extraction result obtained in the step 103.
Further, in step 103, the statistical model is G0Statistical models whose estimation of texture parameters is based on the coherence matrix:
M＝tr(Σ-1T)
Where T is a coherence matrix, Σ ═ E [ T ], λ is a texture parameter of the statistical model, L represents a view, Var { · } represents a variance, and d represents a dimension of the coherence matrix.
Compared with the prior art, the invention has the beneficial effects that: and reflecting the uniformity degree of the building area by using the texture parameters of the statistical model, and extracting collapsed buildings and intact buildings by using texture features extracted based on the texture parameters of the statistical model according to the higher uniformity degree of the collapsed building area than the intact building area. The method has the advantages that non-buildings in a research area are eliminated by using the characteristic values of characteristic value-characteristic vector decomposition aiming at the fact that non-building areas such as collapsed building areas and roads and water bodies have high uniformity, confusion of the non-buildings and collapsed buildings is reduced, and accordingly building damage information extraction accuracy is improved. The polarized radar building damage information extraction method based on the PolSAR image statistical model texture parameters fully utilizes the information of the PolSAR image, reduces the dependence on data, and improves the efficiency and the evaluation precision.
drawings
FIG. 1 is a schematic diagram illustrating steps of a polarized radar building damage information extraction method according to a first embodiment of the present invention;
FIG. 2 is a schematic diagram illustrating the principle of the method for extracting the damage information of the polarized radar building.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be further described with reference to the accompanying drawings.
the technical scheme of the invention can realize automatic operation by adopting computer technology, and please refer to fig. 1 and fig. 2. The invention provides a polarized radar building damage information extraction method based on a PolSAR image processing technology, which comprises the following steps:
Step 101: the original PolSAR image is preprocessed, including selecting a research area and extracting a coherence matrix.
Before extracting the building damage information, the original PolSAR image needs to be preprocessed, including selecting a research area and extracting a coherence matrix. Because the polarisar image has a large amount of speckle, in order to reduce the influence of noise, the original polarisar image generally needs to be filtered, such as fine Lee filtering, Sigma Lee filtering, etc. When the resolution of the PolSAR is low, filtering may not be performed to retain the image information.
Step 102: and (3) carrying out eigenvalue-eigenvector decomposition on the coherent matrix, setting a characteristic threshold value based on the decomposed eigenvalue, identifying pixels smaller than the characteristic threshold value in the research area as a non-building area, identifying pixels larger than the characteristic threshold value as a building area, and rejecting the non-building area.
the eigenvalue-eigenvector decomposition of the coherence matrix is specifically as follows:
where T is the coherence matrix, λnAnd munrespectively representing eigenvalues and eigenvectors, lambda1≥λ2≥λ3h denotes the transpose of the conjugate, the feature vector munExpressed as:
Wherein alpha isnindicating the scattering mechanism of the target to which the scattering vector corresponds, betanIs the target azimuth angle phin、δn、γnfor 3 phase angles, T stands for transpose operation, let λ2+λ3and identifying the pixels smaller than the characteristic threshold value as non-building areas, and rejecting the non-building areas.
Because the collapsed buildings and the sound buildings are distinguished according to the uniformity of the collapsed buildings and the sound buildings, the uniformity of non-building areas such as roads, water bodies and the like and the collapsed buildings is higher, in order to reduce the interference of the non-buildings on the collapsed buildings and improve the extraction accuracy of the collapsed buildings, the general non-building areas such as the roads, the water bodies and the like are typical single scattering targets,Only one eigenvalue is large and satisfies lambda1＞0，λ2and λ3Are all smaller, approaching 0. So will2+λ3Pixels less than the characteristic threshold are classified as non-building regions, so the present invention utilizes λ2+λ3And removing non-building areas in the research area. And before the subsequent collapsed buildings are extracted, removing the non-building areas.
Step 103: and modeling a coherent matrix of the building area by using a statistical model, estimating texture parameters of the statistical model, taking logarithms of the estimated texture parameters as texture features, classifying pixels with the texture features larger than a preset texture threshold value as collapsed buildings, classifying pixels with the texture features smaller than or equal to the preset texture threshold value as intact buildings, and obtaining a building damage information extraction result.
Statistical model G established by coherent matrix based on building area0For example, statistical model G0Comprises the following steps:
estimated statistical model G0The texture parameters of (a) are:
M＝tr(Σ-1T)
Wherein, E [ T ═ E [ T ]]λ is G0and (3) texture parameters of the statistical model, wherein L represents a view, Var {. cndot } represents a variance, d represents the dimension of a coherence matrix, and d is 3 when the reciprocity theorem is satisfied.
And inputting a coherent matrix T, selecting a sliding window with the size of k x k (k is an odd number) to estimate a texture parameter lambda and assigning the texture parameter lambda to a central pixel in the window, thereby obtaining a texture parameter map of the whole research area. However, in general, the values of the texture parameters in the entire region of interest are very different and inconvenient to display and count. The invention can clearly reflect the uniformity degree of the ground features by taking the logarithmic texture characteristics of the estimated texture parameters. Based on G0Statistical model texturethe texture features of the parameter extraction are as follows:
TF_G0＝lg(λ)
wherein, TF _ G0Is based on G0And (5) counting the texture features extracted by the texture parameters of the model.
the types of ground objects in the perfect building area are complex and comprise buildings, roads, vegetation around the buildings and the like, so the uniformity degree of the perfect building area is low. The building is completely collapsed and covers the surrounding ground objects, so that the uniformity degree is increased. The textural feature values are therefore greater for collapsed areas of the building than for sound areas of the building.
and identifying the pixels with the extracted texture features larger than the texture threshold value as collapsed buildings, and classifying the pixels with the texture features smaller than or equal to the texture threshold value as intact buildings. The texture threshold value in the concrete implementation can determine the center of the texture threshold value by using the existing texture threshold value determination method, and then select the proper texture threshold value through manual fine adjustment.
Step 104: and (4) partitioning the building area, and then evaluating the precision of the building damage information extraction result by combining the building damage information extraction result obtained in the step 103.
Generally, urban areas are selected as research areas, blocks are divided based on roads, and buildings in the same block have similar structures. The damage degree of the blocks is divided according to the building damage information extraction result extracted in the step 103, and a basis is provided for verification of the building damage information extraction result, namely: calculating the building damage index (i.e. the proportion of collapsed buildings to total building pixels within a block) for a block divides all blocks into different classes: severe, moderate, mild. The classification threshold is set according to the criteria of the number of pixels of collapsed buildings and sound buildings in step 103. And calculating a confusion matrix, and evaluating the accuracy of building damage information extraction by using the detection rate, the false alarm rate and the total accuracy. The method specifically comprises the following steps:
step 401: setting the building damage index of the jth block in the obtained blocks as DLIjand then:
Wherein, CjNumber of pixels, I, representing a collapsed building in the jth blockjThe number of pixels representing the good buildings in the jth block, and the building damage index DLIjRepresents the damage degree of the building;
step 402: according to DLIjDividing the corresponding blocks into different grades by a preset grading threshold value to obtain a classification result of the damage degree of the building;
Step 403: and evaluating the accuracy of the building information extraction by using the detection rate or the false alarm rate or the total accuracy by using the confusion matrix.
Compared with the prior art, the invention has the advantages that: only a single PolSAR image after the disaster is utilized, so that the dependence on data is reduced; the collapsed building is extracted only by using the texture parameters of the statistical model, the information of the PolSAR image is fully utilized, and meanwhile, the precision and the efficiency of extracting the damage information of the building are improved.
It should be understood that the above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, so that the equivalent changes and modifications made in the light of the above disclosure and the appended claims are all within the scope of the present invention.

Claims (4)

1. a polarized radar building damage information extraction method is based on a polarized PolSAR image processing technology and is characterized by comprising the following steps:
Step 101: preprocessing an original PolSAR image, including selecting a research area and extracting a coherent matrix;
Step 102: carrying out eigenvalue-eigenvector decomposition on the coherent matrix, setting a characteristic threshold value based on the decomposed eigenvalue, identifying pixels smaller than the characteristic threshold value in the research area as non-building areas, identifying pixels larger than the characteristic threshold value as building areas, and removing the non-building areas;
step 103: modeling a coherent matrix of a building area by using a statistical model, extracting texture features based on texture parameters of the statistical model, classifying pixels with the texture features larger than a preset texture threshold value as collapsed buildings, and classifying pixels with the texture features smaller than or equal to the preset texture threshold valueclassifying the pixels as good buildings to obtain the building damage information extraction result; the extraction process comprises estimating the texture parameters of the statistical model, and taking the logarithm of the estimated texture parameters of the statistical model as the texture characteristics, wherein the statistical model is G0The statistical model has the texture parameters as follows:
M＝tr(Σ-1T)
wherein, T is a coherence matrix, Σ ═ E [ T ], λ is a texture parameter of the statistical model, L represents a view, Var { · } represents a variance, and d represents a dimension of the coherence matrix; step 104: and (4) partitioning the building area, and then evaluating the precision of the building damage information extraction result by combining the building damage information extraction result obtained in the step 103.
2. the method for extracting the building damage information of the polarized radar as claimed in claim 1, wherein in step 102, the eigenvalue-eigenvector decomposition of the coherent matrix is specifically as follows:
Where T is the coherence matrix, λnand munrespectively representing eigenvalues and eigenvectors, lambda1≥λ2≥λ3H denotes the transposition of the conjugate, λ1> 0, feature vector munExpressed as:
Wherein alpha isnindicating the scattering mechanism of the target to which the scattering vector corresponds, betanIs the target azimuth angle phin、δn、γnFor 3 phase angles, T stands for transpose operation, let λ2+λ3pixels that are less than the feature threshold are identified as non-building regions.
3. The method for extracting the polarized radar building damage information according to claim 1, wherein the specific process of the step 104 is as follows:
Step 401: setting the building damage index of the jth block in the obtained blocks as DLIjAnd then:
Wherein, CjNumber of pixels, I, representing a collapsed building in the jth blockjThe number of pixels representing the good buildings in the jth block, and the building damage index DLIjrepresents the damage degree of the building;
Step 402: according to DLIjDividing the corresponding blocks into different grades by a preset grading threshold value to obtain a classification result of the damage degree of the building;
step 403: and evaluating the accuracy of the building information extraction by using the detection rate, the false alarm rate or the total accuracy by using the confusion matrix.
4. A polarized radar building damage information extraction method is based on a PolSAR image processing technology and is characterized by comprising the following steps:
Step 101: preprocessing an original PolSAR image, including selecting a research area and extracting a coherent matrix;
step 102: identifying a building area and a non-building area in the research area based on the coherent matrix, and removing the non-building area;
Step 103: modeling a coherent matrix of a building area by using a statistical model, estimating texture parameters of the statistical model, taking logarithm of the estimated texture parameters as texture features, classifying pixels with the texture features larger than a preset texture threshold value as collapsed buildings, classifying pixels with the texture features smaller than or equal to the preset texture threshold value as intact buildings, and obtaining a building damage information extraction result; the statistical model is G0statistical models whose estimation of texture parameters is based on the coherence matrix:
M＝tr(Σ-1T)
Wherein, T is a coherence matrix, Σ ═ E [ T ], λ is a texture parameter of the statistical model, L represents a view, Var { · } represents a variance, and d represents a dimension of the coherence matrix;
Step 104: and (4) partitioning the building area, and then evaluating the precision of the building damage information extraction result by combining the building damage information extraction result obtained in the step 103.
CN201710283555.5A 2017-04-26 2017-04-26 Polarized radar building damage information extraction method Expired - Fee Related CN107133979B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710283555.5A CN107133979B (en) 2017-04-26 2017-04-26 Polarized radar building damage information extraction method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710283555.5A CN107133979B (en) 2017-04-26 2017-04-26 Polarized radar building damage information extraction method

Publications (2)

Publication Number Publication Date
CN107133979A CN107133979A (en) 2017-09-05
CN107133979B true CN107133979B (en) 2019-12-17

Family

ID=59716065

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710283555.5A Expired - Fee Related CN107133979B (en) 2017-04-26 2017-04-26 Polarized radar building damage information extraction method

Country Status (1)

Country Link
CN (1) CN107133979B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103761740A (en) * 2014-01-23 2014-04-30 武汉大学 Construction damage assessment method based on single post-earthquake POLSAR image
CN105321163A (en) * 2014-07-31 2016-02-10 中国科学院遥感与数字地球研究所 Method and apparatus for detecting variation region of fully polarimetric SAR (Synthetic Aperture Radar) image
CN105975986A (en) * 2016-05-03 2016-09-28 河海大学 Fully-polarimetric SAR image supervised classification method based on improved genetic algorithm

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103761740A (en) * 2014-01-23 2014-04-30 武汉大学 Construction damage assessment method based on single post-earthquake POLSAR image
CN105321163A (en) * 2014-07-31 2016-02-10 中国科学院遥感与数字地球研究所 Method and apparatus for detecting variation region of fully polarimetric SAR (Synthetic Aperture Radar) image
CN105975986A (en) * 2016-05-03 2016-09-28 河海大学 Fully-polarimetric SAR image supervised classification method based on improved genetic algorithm

Also Published As

Publication number Publication date
CN107133979A (en) 2017-09-05

Similar Documents

Publication Publication Date Title
CN108596108B (en) Aerial remote sensing image change detection method based on triple semantic relation learning
CN105701481B (en) A kind of collapsed building extracting method
CN105389799B (en) SAR image object detection method based on sketch map and low-rank decomposition
CN109977968B (en) SAR change detection method based on deep learning classification comparison
Li et al. An improved approach of information extraction for earthquake-damaged buildings using high-resolution imagery
CN103400137A (en) Method for extracting geometrical building parameters of synthetic aperture radar (SAR) image
Chen et al. A novel statistical texture feature for SAR building damage assessment in different polarization modes
Kwak et al. A new approach for rapid urban flood mapping using ALOS-2/PALSAR-2 in 2015 Kinu River Flood, Japan
Koppel et al. Sentinel-1 for urban area monitoring—Analysing local-area statistics and interferometric coherence methods for buildings' detection
CN107527035A (en) Earthquake damage to building information extracting method and device
CN105184804A (en) Sea surface small target detection method based on airborne infrared camera aerially-photographed image
CN110516552B (en) Multi-polarization radar image classification method and system based on time sequence curve
Gokon et al. Verification of a method for estimating building damage in extensive tsunami affected areas using L-band SAR data
CN111275680B (en) SAR image change detection method based on Gabor convolution network
CN107133979B (en) Polarized radar building damage information extraction method
CN112989940A (en) Raft culture area extraction method based on high-resolution three-satellite SAR image
Guo et al. Study of detecting method with advanced airborne and spaceborne synthetic aperture radar data for collapsed urban buildings from the Wenchuan earthquake
Liu et al. Development of building height data in Peru from high-resolution SAR imagery
Dell'Acqua et al. Experiences in optical and SAR imagery analysis for damage assessment in the Wuhan, May 2008 earthquake
Chen et al. Building collapse extraction using modified freeman decomposition from post-disaster polarimetric SAR image
CN107748875A (en) A kind of earthquake building recognition method based on multidate radar image texture feature
Duan et al. Collapsed houses automatic identification based on texture changes of post-earthquake aerial remote sensing image
CN105551021B (en) The building method of estimation of falling loss rate based on multidate full-polarization SAR
Cao et al. Detecting the number of buildings in a single high-resolution SAR image
Dell'Acqua et al. Mapping earthquake damage in VHR radar images of human settlements: Preliminary results on the 6 th April 2009, Italy case

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
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20191217

CF01 Termination of patent right due to non-payment of annual fee