CN108090898B - Dictionary representation-based satellite remote sensing image typical landmark detection method - Google Patents

Dictionary representation-based satellite remote sensing image typical landmark detection method Download PDF

Info

Publication number
CN108090898B
CN108090898B CN201711380350.5A CN201711380350A CN108090898B CN 108090898 B CN108090898 B CN 108090898B CN 201711380350 A CN201711380350 A CN 201711380350A CN 108090898 B CN108090898 B CN 108090898B
Authority
CN
China
Prior art keywords
landmark
image
typical
representation
alpha
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201711380350.5A
Other languages
Chinese (zh)
Other versions
CN108090898A (en
Inventor
梅少辉
张易凡
孙俊
田晋
彭杨
魏江
陈文�
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.)
Northwestern Polytechnical University
Original Assignee
Northwestern Polytechnical University
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 Northwestern Polytechnical University filed Critical Northwestern Polytechnical University
Priority to CN201711380350.5A priority Critical patent/CN108090898B/en
Publication of CN108090898A publication Critical patent/CN108090898A/en
Application granted granted Critical
Publication of CN108090898B publication Critical patent/CN108090898B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/29—Geographical information databases
    • 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
    • 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
    • G06V10/513—Sparse representations

Abstract

The invention discloses a satellite remote sensing image typical landmark detection method based on dictionary representation, which comprises the following steps of S1, selecting a landmark area, and generating a typical landmark database of the landmark area; and S2, acquiring images of the landmark areas, and matching the acquired images with the typical landmark database acquired in S1 by using a dictionary representation method to acquire detection results. The typical landmark database contains typical landmark image information under complex time-of-day environments of different weather conditions, different time, different illumination, different backgrounds and the like, so the method has good adaptability and can obtain good detection effect on the typical landmark detection under the complex time-of-day environments. The dictionary representation method comprises sparse representation and collaborative representation, the typical landmark library and the dictionary representation method are designed and integrated to carry out the typical landmark detection of the satellite remote sensing image, and the detection effect and the stability are good.

Description

Dictionary representation-based satellite remote sensing image typical landmark detection method
Technical Field
The invention relates to the technical field of remote sensing, in particular to a satellite remote sensing image typical landmark detection method based on dictionary representation.
Background
Since the first artificial earth satellite is successfully launched, the satellite image data is widely applied to a plurality of fields such as space detection, weather, resource survey, navigation, communication, investigation and the like due to the advantages of rapid acquisition, low cost and no region limitation. The satellite acquires images of the ground, the atmosphere and the space through the satellite-borne sensor, and corresponding functions are completed through analyzing and processing the satellite images. In the application of satellite image analysis processing, the technologies of geometric correction of satellite images, satellite orbit determination based on satellite remote sensing images, image navigation and the like are important applications in the satellite image data processing process, and are also the technical basis of satellite image data application research and product development. How to realize the detection of the typical landmark of the satellite remote sensing image is one of the key research contents of the geometric correction of the satellite image, the satellite orbit determination based on the image, the image navigation and the like.
In practical application, due to imaging errors in a satellite image, ground feature characteristics contained in the image are deformed, and noise influence in the image, a point diffusion phenomenon of an optical system during imaging of an optical sensor and shielding of cloud-containing areas in the satellite image on the ground feature characteristics all affect the accuracy of landmark detection in the satellite image. The average cloud amount in the satellite image is large, and the landmark is greatly influenced by weather, sunlight conditions and the like, so that the image characteristics can be changed under different illumination conditions. In addition, the terrestrial object features in the satellite image have distortion, offset and other phenomena, and the sensor also has noise, which all cause the loss or invalid acquisition of the terrestrial object feature information in the satellite image, and further cause wrong matching or reduce the matching precision of the landmark, so that the traditional landmark detection method cannot achieve an ideal detection effect.
Disclosure of Invention
The invention aims to provide a dictionary representation-based satellite remote sensing image typical landmark detection method to overcome the defect that the traditional satellite image typical landmark detection is not suitable for efficient detection in a complex environment.
In order to achieve the purpose, the invention adopts the following technical scheme:
the satellite remote sensing image typical landmark detection method based on dictionary representation comprises the following steps:
s1, selecting a landmark area and generating a typical landmark database of the landmark area;
and S2, acquiring images of the landmark areas, and matching the acquired images with the typical landmark database acquired in S1 by using a dictionary representation method to acquire detection results.
Further, step S1 specifically includes:
s11, carrying out image acquisition on the selected landmark region to obtain a prior image of the landmark region and geographic coordinate information of the landmark region;
s12, extracting the feature of the prior image of the landmark region;
and S13, generating a typical landmark database of the landmark area, wherein the typical landmark database comprises prior image information, characteristic information and geographical position information.
Further, the prior image in step S11 includes image information under different weather conditions, different time, different illumination or different background.
Further, the step S2 specifically includes:
s21, acquiring a real-time image of the landmark region, and performing image preprocessing;
s22, extracting the characteristics of the preprocessed image;
and S23, matching the obtained image with the typical landmark database obtained in the S1 by utilizing sparse representation or collaborative representation, and obtaining a detection result.
Further, the sparse representation in step S23 is specifically:
s231, a typical landmark database is set as a dictionary D, the real-time image acquired in the step S2 is set as a test sample y, and the sparsely represented objective function is represented as
Figure GDA0002651982200000021
In the formula, λ1Is a regular parameter, alpha is a sparse coefficient to be estimated, the superscript SR represents sparse representation, alpha(SR)Representing coefficients for the sparsity to be estimated;
s232, solving a sparse coefficient alpha;
s233, calculating the product of each atom in the dictionary D and the sparse coefficient alpha to obtain an estimated value of the real-time image y;
and S234, calculating a two-norm of a residual error between the test sample y and the estimated value D & alpha of the dictionary, namely Res | | | y-D & alpha | |, selecting a threshold value, and if | | | y-D & alpha | | <, successfully matching the real-time image with the landmark library.
Further, in step S232, the sparse coefficient α is solved through an orthogonal matching pursuit algorithm, a synchronous orthogonal matching pursuit algorithm, or a nonlinear orthogonal matching pursuit algorithm.
Further, the collaborative representation in step S23 is specifically:
s2310, a typical landmark database is set as a dictionary D, the real-time image acquired in the step S2 is set as a test sample y, and the objective function of the collaborative representation is represented as
Figure GDA0002651982200000031
In the formula, λ2Is a regular parameter, alpha is a sparse coefficient to be estimated, the superscript CR represents a cooperative expression, alpha(CR)Co-representing coefficients for the to-be-estimated;
s2320, solving a sparse coefficient alpha;
s2330, calculating a product of each atom in the dictionary D and the sparse coefficient alpha to obtain an estimated value of the real-time image y;
s2340, calculating two norms of residual errors of the test sample y and the estimated value D & alpha of each atom, namely Res | | | y-D & alpha | |, selecting a threshold value, and if | | | | y-D & alpha | | <, successfully matching the real-time image with the landmark library.
Further, the sparse coefficient α, i.e., α, is solved by the least square method in step S232(CR)＝(DTD+λ2I)-1DTy in the formula, DTDenotes a transpose matrix of D, and I denotes an identity matrix.
After adopting the technical scheme, compared with the background technology, the invention has the following advantages:
the method is well adapted, and can obtain good detection effect on typical landmark detection under complex conditions such as complex background, noise interference, illumination change and the like;
the invention utilizes dictionary representation of typical object library, and the detection effect stability is good because the robustness of the typical object library is good and various complex conditions are included.
Drawings
FIG. 1 is a schematic overall flow diagram of the present invention;
FIG. 2 is a schematic flow chart of the present invention for generating a representative landmark database;
fig. 3 is a schematic flow chart of matching the obtained image with a typical landmark database by using a dictionary representation method according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Examples
Referring to fig. 1, the invention discloses a method for detecting a typical landmark of a satellite remote sensing image based on dictionary representation, which comprises the following steps:
s1, selecting a landmark area and generating a typical landmark database of the landmark area;
and S2, acquiring images of the landmark areas, and matching the acquired images with the typical landmark database acquired in S1 by using a dictionary representation method to acquire detection results.
Referring to fig. 2, step S1 (this step is an offline processing part) specifically includes:
s11, acquiring images of the selected landmark area by using a satellite or an unmanned aerial vehicle, and obtaining a prior image of the landmark area and geographic coordinate information of the landmark area;
s12, extracting the feature of the prior image of the landmark region;
and S13, generating a typical landmark database of the landmark area, wherein the typical landmark database comprises prior image information, characteristic information and geographical position information.
The typical landmark database consists of template image data and landmark attribute data of typical landmarks in the observation area of the sensor, and is an important component for satellite image landmark matching. The precision, size, type and other factors of the landmark template image directly relate to the precision and efficiency of landmark matching, and when the number of landmarks in the landmark template database is small, few landmarks are successfully matched with the satellite image to obtain the effective offset parameters of the image, and the precision of image geometric positioning, image navigation and the like is also influenced.
Therefore, the landmark region selected in step S11 should have abundant and obvious image features and be not easy to be confused, such as remote sensing ground control points with obvious features and being difficult to copy, such as skyscrapers, overpasses, rivers, reservoir dams, airports, and the like.
The landmark attribute data mainly includes 4 aspects: 1) describing geographical location information of the control point, i.e. the spatial coordinates of the control point; 2) some necessary auxiliary information describing the geographical coordinates, such as the adopted coordinate system, the projection mode, the ellipsoid parameters, etc.; 3) auxiliary information describing the control point image, such as the type of sensor, the wave band, the image height, the image resolution, etc.; 4) the control point selects characteristic description, such as intersection of road or bridge center point, and the information can be used as auxiliary conditions of inquiry.
Meanwhile, in order to reduce the influence of the complex background on the detection effect and improve the detection adaptability, the prior image in step S11 includes image information under different weather conditions, different time, different illumination or different backgrounds.
Referring to fig. 3, the step S2 specifically includes:
s21, acquiring a real-time image of the landmark region, and performing image preprocessing;
due to imaging errors of the satellite image, the ground feature contained in the image is deformed, and in addition, noise influence in the image, a point diffusion phenomenon of an optical system during imaging of the optical sensor and shielding of cloud-containing areas in the satellite image on the ground feature influence the accuracy of landmark matching in the satellite image. Therefore, data preprocessing, such as image denoising, image correction and registration, is required to be performed on the acquired remote sensing image.
S22, extracting the characteristics of the preprocessed image; the feature extraction is to perform corresponding transformation on the image, extract image information, and select an HOG feature, a Rough feature, a SIFT feature, and the like according to the actual situation of the detection object.
And S23, matching the obtained image with the typical landmark database obtained in S1 by utilizing dictionary representation (namely sparse representation or collaborative representation), and obtaining a detection result.
In step S21, the landmark region is image-captured by the satellite, and the typical landmark database obtained in step S1 is stored in the memory of the satellite. Due to the influence of the environment, sensor and platform interference, noise and various forms of interference exist in the image data obtained by the satellite, and in order to ensure the quality and performance of the subsequent operations, a preprocessing operation needs to be performed on the image in step S2.
In step S23, the dictionary-based representation detection criterion is: the test sample may be approximately linearly represented by the labeled sample, the weight factor being given by0Norm or l1Norm penalty constraint (called sparse representation) or l2Norm penalty constraints (called co-representation).
Specifically, whether the sparse representation model or the collaborative representation model is selected is determined according to factors such as dictionary characteristics and scale (the number of atoms contained in the dictionary). In the sparse representation model, the representation coefficients corresponding to only a few atoms in the dictionary are not zero (i.e. only a few atoms contribute to the representation), which is suitable for the case that the dictionary size is large (usually, an overcomplete dictionary) or the dictionary atoms have strong diversity. In the collaborative representation model, representation coefficients corresponding to most atoms in the dictionary are not zero (namely, most atoms contribute to representation), and the method is suitable for the condition that the scale of the dictionary is small or the atomic anisotropies in the dictionary are not obvious.
The Sparse Representation (SR) is specifically:
s231, a typical landmark database is set as a dictionary D, the real-time image acquired in the step S2 is set as a test sample y, and the sparsely represented objective function is represented as
Figure GDA0002651982200000061
In the formula, λ1As a regularization parameter, alpha is the sparse coefficient to be estimated,α is a column vector, SR stands for sparse representation, α(SR)For the sparse representation coefficient to be estimated, | | | | luminance2Representing a 2-norm, i.e. the norm or length of the vector, | | | | | | luminance1Represents the 1-norm, i.e., the sum of the absolute values of the atoms of the vector.
S232, solving a sparse coefficient alpha;
s233, calculating the product of each atom in the dictionary D and the sparse coefficient alpha to obtain an estimated value of the real-time image y;
and S234, calculating a two-norm of a residual error between the test sample y and the estimated value D & alpha of the dictionary, namely Res | | | y-D & alpha | |, selecting a proper threshold value, and if | | | y-D & alpha | <, successfully matching the real-time image with the landmark library.
The threshold selection method comprises the following steps: the proper threshold value is selected through experiments, namely, images with known information (including positive samples and negative samples, wherein the positive samples are the landmark images to be matched, and the negative samples are not the matched landmark images) are represented by a dictionary, and a threshold value capable of effectively separating the positive samples from the negative samples is selected through experiments.
In step S232, the sparse coefficient α is solved through an orthogonal matching pursuit algorithm (OMP), a synchronous orthogonal matching pursuit algorithm (SOMP), or a nonlinear orthogonal matching pursuit algorithm (Kernel-OMP, KOMP).
The Collaborative Representation (CR) described in step S23 specifically includes:
s2310, a typical landmark database is set as a dictionary D, the real-time image acquired in the step S2 is set as a test sample y, and the objective function of the collaborative representation is represented as
Figure GDA0002651982200000062
In the formula, λ2The method comprises the following steps that (1) a is a regular parameter, alpha is a sparse coefficient to be estimated, and alpha is a column vector; CR stands for co-expression, α(CR)For the co-expression coefficients to be estimated,
Figure GDA0002651982200000071
represents the square of the 2-norm of the vector, i.e., the square of the modulus of the vector;
s2320, step S232Solving the sparse coefficient alpha by the least square method, if alpha exists(CR)＝(DTD+λ2I)-1DTy; wherein CR represents a cooperative expression, DTA transposed matrix representing the matrix D, I representing an identity matrix;
s2330, calculating a product of each atom in the dictionary D and the sparse coefficient alpha to obtain an estimated value of the real-time image y;
s2340, calculating two norms of residual errors of the test sample y and the estimated value D & alpha of each atom, namely Res | | | y-D & alpha | |, selecting a proper threshold value, and if | | | | y-D & alpha | | <, successfully matching the real-time image with the landmark library.
The threshold selection method comprises the following steps: the proper threshold value is selected through experiments, namely, images with known information (including positive samples and negative samples, wherein the positive samples are the landmark images to be matched, and the negative samples are not the matched landmark images) are represented by a dictionary, and a threshold value capable of effectively separating the positive samples from the negative samples is selected through experiments.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (6)

1. The satellite remote sensing image typical landmark detection method based on dictionary representation is characterized by comprising the following steps:
s1, selecting a landmark area and generating a typical landmark database of the landmark area;
s2, collecting images of the landmark areas, and matching the obtained images with the typical landmark database obtained in the S1 by using a dictionary representation method to obtain detection results; the step S2 specifically includes:
s21, acquiring a real-time image of the landmark region, and performing image preprocessing;
s22, extracting the characteristics of the preprocessed image;
s23, matching the obtained image with the typical landmark database obtained in S1 by utilizing sparse representation or collaborative representation to obtain a detection result; the sparse representation in step S23 is specifically:
s231, a typical landmark database is set as a dictionary D, the real-time image acquired in the step S2 is set as a test sample y, and the sparsely represented objective function is represented as
Figure FDA0002651982190000011
In the formula, λ1Is a regular parameter, alpha is a sparse coefficient to be estimated, the superscript SR represents sparse representation, alpha(SR)Representing coefficients for the sparsity to be estimated;
s232, solving a sparse coefficient alpha;
s233, calculating the product of each atom in the dictionary D and the sparse coefficient alpha to obtain an estimated value of the real-time image y;
and S234, calculating a two-norm of a residual error between the test sample y and the estimated value D & alpha of the dictionary, namely Res | | | y-D & alpha | |, selecting a threshold value, and if | | | y-D & alpha | | <, successfully matching the real-time image with the landmark library.
2. The method for detecting the satellite remote sensing image typical landmark based on the dictionary representation as claimed in claim 1, wherein the step S1 specifically includes:
s11, carrying out image acquisition on the selected landmark region to obtain a prior image of the landmark region and geographic coordinate information of the landmark region;
s12, extracting the feature of the prior image of the landmark region;
and S13, generating a typical landmark database of the landmark area, wherein the typical landmark database comprises prior image information, characteristic information and geographical position information.
3. The method for detecting the satellite remote sensing image typical landmark based on the dictionary representation as claimed in claim 2, characterized in that: the prior images in step S11 include images under different weather conditions, different times, different lighting, or different backgrounds.
4. The method for detecting the satellite remote sensing image typical landmark based on the dictionary representation as claimed in claim 1, wherein: in step S232, the sparse coefficient α is solved by an orthogonal matching pursuit algorithm, a synchronous orthogonal matching pursuit algorithm, or a nonlinear orthogonal matching pursuit algorithm.
5. The method for detecting the satellite remote sensing image typical landmark based on the dictionary representation as claimed in claim 1, wherein the collaborative representation in the step S23 is specifically:
s2310, a typical landmark database is set as a dictionary D, the real-time image acquired in the step S2 is set as a test sample y, and the objective function of the collaborative representation is represented as
Figure FDA0002651982190000021
In the formula, λ2Is a regular parameter, alpha is a sparse coefficient to be estimated, the superscript CR represents a cooperative expression, alpha(CR)Co-representing coefficients for the to-be-estimated;
s2320, solving a sparse coefficient alpha;
s2330, calculating a product of each atom in the dictionary D and the sparse coefficient alpha to obtain an estimated value of the real-time image y;
s2340, calculating two norms of residual errors of the test sample y and the estimated value D & alpha of each atom, namely Res | | | y-D & alpha | |, selecting a threshold value, and if | | | | y-D & alpha | | <, successfully matching the real-time image with the landmark library.
6. The method for detecting the satellite remote sensing image typical landmark based on the dictionary representation as claimed in claim 5, wherein: in step S232, the sparse coefficient alpha, namely alpha, is solved by the least square method(CR)＝(DTD+λ2I)-1DTy in the formula, DTDenotes a transpose matrix of D, and I denotes an identity matrix.
CN201711380350.5A 2017-12-20 2017-12-20 Dictionary representation-based satellite remote sensing image typical landmark detection method Active CN108090898B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711380350.5A CN108090898B (en) 2017-12-20 2017-12-20 Dictionary representation-based satellite remote sensing image typical landmark detection method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711380350.5A CN108090898B (en) 2017-12-20 2017-12-20 Dictionary representation-based satellite remote sensing image typical landmark detection method

Publications (2)

Publication Number Publication Date
CN108090898A CN108090898A (en) 2018-05-29
CN108090898B true CN108090898B (en) 2020-11-10

Family

ID=62177404

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711380350.5A Active CN108090898B (en) 2017-12-20 2017-12-20 Dictionary representation-based satellite remote sensing image typical landmark detection method

Country Status (1)

Country Link
CN (1) CN108090898B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109375205B (en) * 2018-09-28 2020-11-03 清华大学 Method and device for recognizing scenes of multiple types of unmanned aerial vehicles based on dictionary learning
CN109614998A (en) * 2018-11-29 2019-04-12 北京航天自动控制研究所 Landmark database preparation method based on deep learning
CN114022784B (en) * 2021-11-09 2022-04-19 中国人民解放军61646部队 Method and device for screening landmark control points

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104200478B (en) * 2014-09-12 2017-03-22 广东财经大学 Low-resolution touch screen image defect detection method based on sparse representation
CN105631478A (en) * 2015-12-25 2016-06-01 天津科技大学 Plant classification method based on sparse expression dictionary learning
CN107133260A (en) * 2017-03-22 2017-09-05 新奥特（北京）视频技术有限公司 The matching and recognition method and device of a kind of landmark image

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于局部空间线性恢复模型的多光谱与全色图像融合算法;张易凡，何明一;《西北工业大学学报》;20080131(第1期);第57-63页 *

Also Published As

Publication number Publication date
CN108090898A (en) 2018-05-29

Similar Documents

Publication Publication Date Title
WO2019042232A1 (en) Fast and robust multimodal remote sensing image matching method and system
CN105761242B (en) Blind person walking positioning method based on computer binocular vision and inertial measurement
Zhang et al. Automated water classification in the Tibetan plateau using Chinese GF-1 WFV data
CN108090898B (en) Dictionary representation-based satellite remote sensing image typical landmark detection method
Huang et al. The QXS-SAROPT dataset for deep learning in SAR-optical data fusion
CN109597074B (en) SAR image geometric positioning parameter correction method and system
KR101439213B1 (en) Method for 3D Location determination in single image using Rational Polynomial Coefficients information of stereo satellite images
CN107679476A (en) A kind of Sea Ice Types Classification in Remote Sensing Image method
Kim et al. Investigating applicability of unmanned aerial vehicle to the tidal flat zone
Sevara Top secret topographies: recovering two and three-dimensional archaeological information from historic reconnaissance datasets using image-based modelling techniques
Wang et al. Large-scale orthorectification of GF-3 SAR images without ground control points for China’s land area
CN111144350B (en) Remote sensing image positioning accuracy evaluation method based on reference base map
Ye et al. Improved geometric modeling of 1960s KH-5 ARGON satellite images for regional Antarctica applications
Farella et al. The EuroSDR TIME benchmark for historical aerial images
Zhu et al. An efficient and effective approach for georeferencing AVHRR and GaoFen-1 imageries using inland water bodies
CN110927765A (en) Laser radar and satellite navigation fused target online positioning method
CN109886988A (en) A kind of measure, system, device and the medium of Microwave Imager position error
Chakravortty et al. Fusion of hyperspectral and multispectral image data for enhancement of spectral and spatial resolution
CN112200845A (en) Image registration method and device
KR101775124B1 (en) System and method for automatic satellite image processing for improvement of location accuracy
Geng et al. Generation of Large-Scale Orthophoto Mosaics Using MEX HRSC Images for the Candidate Landing Regions of China’s First Mars Mission
Coulter et al. Automated co-registration of multitemporal airborne frame images for near real-time change detection
Hou et al. A Landmark Matching Algorithm for the Geostationary Satellite Images Based on Multi-Level Grids
Son et al. Comparison of Geometric Correction Schemes for Geostationary Ocean Color Imager Slots Without Gcps
Wang et al. SAR image rectification based on Vector Map

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