CN111709927A - Automatic identification method and system for type abnormal pattern spots - Google Patents

Automatic identification method and system for type abnormal pattern spots Download PDF

Info

Publication number
CN111709927A
CN111709927A CN202010534638.9A CN202010534638A CN111709927A CN 111709927 A CN111709927 A CN 111709927A CN 202010534638 A CN202010534638 A CN 202010534638A CN 111709927 A CN111709927 A CN 111709927A
Authority
CN
China
Prior art keywords
image
spots
index
spot
category
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202010534638.9A
Other languages
Chinese (zh)
Other versions
CN111709927B (en
Inventor
李冲
李昊霖
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.)
Sichuan Surveying And Mapping Product Quality Supervision And Inspection Station Ministry Of Natural Resources (sichuan Surveying And Mapping Product Quality Supervision And Inspection Station)
Original Assignee
Sichuan Surveying And Mapping Product Quality Supervision And Inspection Station Ministry Of Natural Resources (sichuan Surveying And Mapping Product Quality Supervision And Inspection Station)
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 Sichuan Surveying And Mapping Product Quality Supervision And Inspection Station Ministry Of Natural Resources (sichuan Surveying And Mapping Product Quality Supervision And Inspection Station) filed Critical Sichuan Surveying And Mapping Product Quality Supervision And Inspection Station Ministry Of Natural Resources (sichuan Surveying And Mapping Product Quality Supervision And Inspection Station)
Priority to CN202010534638.9A priority Critical patent/CN111709927B/en
Priority to PCT/CN2020/100451 priority patent/WO2021248599A1/en
Publication of CN111709927A publication Critical patent/CN111709927A/en
Priority to AU2021102441A priority patent/AU2021102441A4/en
Application granted granted Critical
Publication of CN111709927B publication Critical patent/CN111709927B/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
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • 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
    • G06T2207/10036—Multispectral image; Hyperspectral image

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Quality & Reliability (AREA)
  • Image Processing (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a method and a system for automatically identifying type abnormal pattern spots. The method comprises the following steps: acquiring a multispectral remote sensing image to be identified; cutting the multispectral remote sensing image to obtain an image of the image spot with the category easy to make mistakes or confuse; calculating the comprehensive spectral index of each image spot image; according to the size of the comprehensive spectral index of each image spot image, sorting each image spot image according to element classes; determining whether the set number of image spot images before and after the image spot image sequence have the category abnormal image spots; if the image spots exist, a set number of image spots are added before and/or after the image spot image sequence, the added image spots are subjected to category abnormity identification, and if the added image spots exist in the image spots with category abnormity, the step of identifying the image spots with the set number before and/or after the image spot image sequence is skipped to until no category abnormity image spots exist in the added image spots. The invention can realize automatic identification of the category abnormal pattern spots.

Description

Automatic identification method and system for type abnormal pattern spots
Technical Field
The invention relates to the technical field of image processing, in particular to a method and a system for automatically identifying type abnormal pattern spots.
Background
With the development of society, in the field of geographic national condition monitoring, higher requirements are put forward on the accuracy and fineness of earth surface coverage data. However, due to the factors of complex earth surface conditions, inconsistent remote sensing image quality, large subjective influence of interpretation by personnel and the like in China, the problem of map spot classification errors often occurs in earth surface coverage data, the quality of the geographical national condition monitoring result is seriously influenced, and each production unit and quality inspection unit need to spend a large amount of manpower, material resources and financial resources to check the large map spots and the same spectrum foreign matter errors in the earth surface coverage data every year.
Disclosure of Invention
The invention aims to provide a method and a system for automatically identifying type abnormal pattern spots.
In order to achieve the purpose, the invention provides the following scheme:
a method for automatically identifying type abnormal image spots comprises the following steps:
acquiring a multispectral remote sensing image to be identified;
based on the geometric range of the image spots, cutting the multispectral remote sensing image to obtain the image of the image spots with the categories easy to make mistakes or confuse;
calculating a comprehensive spectral index of each image spot image, wherein the comprehensive spectral index is formed by weighted combination of a vegetation index, a water body index and a difference building index and is used for distinguishing the image spot element types;
sorting the image spots by element classes according to the size of the comprehensive spectral index of the image spots;
determining whether the set number of image spot images before and after the image spot image sequence have the category abnormal image spots;
if the image spots with abnormal category exist, increasing the image spots with set number before and/or after the image spot image sequence, and carrying out the identification of the abnormal category on the increased image spots, if the image spots with abnormal category exist in the increased image spots, skipping to the step of identifying the image spots with the set number before and/or after the image spot image sequence until the abnormal category image spots do not exist in the increased image spots,
and outputting the abnormal image spot image.
Optionally, before the multispectral remote sensing image is cropped based on the geometric range of the pattern spots, the method further includes:
and determining the weight coefficient of the vegetation index, the water body index and the differential building index in the comprehensive spectrum index for distinguishing the error-prone or confusable element classes according to the historical data.
Optionally, the sorting the image spots by element classification according to the size of the integrated spectral index of the image spots specifically includes:
determining the median or average value of the comprehensive spectral index of each image spot image;
and sorting the image spots according to the element classes according to the magnitude of the median or the average value of the comprehensive spectral indexes of the image spots.
Optionally, the method for detecting the category abnormal pattern spots includes:
judging whether the difference value of the comprehensive spectral indexes of the adjacent image spot images is larger than a set threshold value or not;
if so, indicating that the category abnormal image spots exist, wherein the category abnormal image spots are all the image spots between the adjacent image spot image and the first end of the image spot sequence, and the first end is the end of the image spot sequence which is closer to the adjacent image spot image.
Optionally, before the calculating the comprehensive spectral index of each image spot image, the method further includes:
carrying out normalization processing on the vegetation index and the water body index;
and stretching the differential building index, the normalized vegetation index and the normalized water body index.
Optionally, the method further includes:
and outputting the class abnormal image spot identified by the computer to a human-computer interaction end for manual secondary identification.
Optionally, the cutting process of the multispectral remote sensing image is performed in a computer memory.
The invention also provides a system for automatically identifying the type of the abnormal pattern spots, which comprises the following components:
the multispectral remote sensing image acquisition module to be identified is used for acquiring the multispectral remote sensing image to be identified;
the image cutting module of the image spot is used for cutting the multispectral remote sensing image based on the geometric range of the image spot so as to obtain the image of the image spot with the category which is easy to make mistakes or confuse;
the comprehensive spectral index calculation module is used for calculating a comprehensive spectral index of each image spot image, and the comprehensive spectral index is formed by weighted combination of a vegetation index, a water body index and a difference building index and is used for distinguishing the image spot element types;
the sorting module is used for sorting the image spots by element classes according to the size of the comprehensive spectral index of the image spots;
and the category anomaly determination module is used for determining whether category anomaly image spots exist in the image spots of which the number is set before and after the image spot image sequence and determining whether category anomaly image spots exist in the added image spot images.
Optionally, the system further includes:
and the initial parameter determining module is used for determining the error-prone or confusable element types and weight coefficients of the vegetation index, the water body index and the difference building index in the comprehensive spectrum index for distinguishing the error-prone or confusable element types according to the historical data.
Optionally, the sorting module specifically includes:
the numerical value determining unit is used for determining the median or average value of the comprehensive spectral index of each image spot image;
and the sorting unit is used for sorting the image spot images according to the element classes according to the magnitude of the median or the average value of the comprehensive spectral indexes of the image spot images.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects: the method and the system for automatically identifying the type abnormal pattern spots extract the pattern spots which are easy to make mistakes or are easy to confuse in the multispectral remote sensing image to be identified, calculate the spectral indexes of the pattern spots which can distinguish the pattern spots which are easy to make mistakes or are easy to confuse, sort the pattern spots according to the size of the spectral indexes, and finally determine the type abnormal pattern spots by carrying out type abnormal detection on the pattern spots with the set number at the two ends of the sequence. The invention realizes the automatic identification of the pattern spot type abnormity and improves the detection efficiency.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a schematic flow chart of an automatic identification method for pattern spots with type anomalies according to embodiment 1 of the present invention;
fig. 2 is a schematic structural diagram of an automatic type anomaly pattern recognition system provided in embodiment 2 of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Example 1
Fig. 1 is a schematic flow chart of an automatic identification method for a type-anomaly image spot according to embodiment 1 of the present invention, and referring to fig. 1, the automatic identification method for a type-anomaly image spot according to this embodiment includes the following steps:
step 101: acquiring a multispectral remote sensing image to be identified;
step 102: based on the geometric range of the image spots, cutting the multispectral remote sensing image to obtain the image of the image spots with the categories easy to make mistakes or confuse;
step 103: calculating a comprehensive spectral index of each image spot image, wherein the comprehensive spectral index is formed by weighted combination of a vegetation index, a water body index and a difference building index and is used for distinguishing the image spot element types;
step 104: sorting the image spots by element classes according to the size of the comprehensive spectral index of the image spots;
step 105: determining whether the set number of image spot images before and after the image spot image sequence have the category abnormal image spots;
step 106: if the image spots with abnormal category exist, increasing a set number of image spots before and/or after the image spot image sequence, and performing category abnormal identification on the increased image spots, if the increased image spots have abnormal category image spots, jumping to the step of increasing the set number of image spots before and/or after the image spot image sequence for identification until no category abnormal image spots exist in the increased image spots;
step 107: and outputting the image of the category abnormal image spots.
In this embodiment, the geometric range of each pattern spot is sequentially obtained from the earth surface coverage vector data, an image cropping function is defined, the cropping range is the geometric range of the pattern spot, the cropped image is the multispectral remote sensing image to be identified, the number of the bands of the multispectral remote sensing image to be identified is not less than 4, the bands at least include red, green, blue, near infrared and the like, and the image cropping function defines the processing mode of the image data in terms of the pattern spot spectral index, wherein preferably, the processing process is performed in a computer memory, and the cropping result is not output to a computer hard disk. And calculating a comprehensive spectrum index of each image spot image, wherein the comprehensive spectrum index is formed by weighting and combining a vegetation index, a water body index and a difference building index and is used for distinguishing the image spot element types. Of course, in other embodiments, other spectral feature data that can distinguish the types of spot elements may be used. After the comprehensive spectrum index of each image spot is obtained through calculation, the middle value of the comprehensive spectrum index of the image spot can be selected to represent that the image spots are sorted according to the size. Of course, in other embodiments, the average value of the integrated spectral index of the spot image may be selected to represent the spot for spot sorting. In the sorting process, the element classes are sorted, that is, the patches of all the element classes are not mixed together for sorting, but the patches belonging to the same element class are sorted separately, for example, the patches of all the element classes with the attribute of paddy field are sorted into one group, and the patches of all the element classes with the attribute of lake are sorted into one group. After obtaining the image spot sequence, firstly, judging whether the image spot images of a set number (for example, 2) before and after the image spot sequence have the category abnormal image spots, if not, determining that the image spots in the whole image spot sequence are normal and the category abnormal image spots do not exist, if the image spot images of the set number before and after the image spot sequence have the category abnormal image spots, for example, the image spot images of the set number at the front end (rear end) of the image spot sequence have the category abnormal image spots, increasing the set number (for example, 2-3) of the image spots at the front end (rear end) of the image spot sequence to perform category abnormal detection until the increased set number of the image spots are normal image spots, at the moment, determining that the last increased image spots are the category normal image spots, and the image spots identified before are the category abnormal image spots.
The determination method of the intermediate value of the comprehensive spectral index of the image spot image is as follows: when the number of the pixel points in the image spot is an odd number, the most middle numerical value in the pixel comprehensive spectrum index sequence is the intermediate value of the image spot comprehensive spectrum index of the image spot, when the number of the pixel points in the image spot is an even number, the most middle two numerical values in the pixel comprehensive spectrum index sequence are taken out, and the average value of the two numerical values is calculated, namely the intermediate value of the image spot comprehensive spectrum index of the image spot.
In this embodiment, after the abnormal pattern spots are determined, the category abnormal pattern spots may be directly output, or the category abnormal pattern spots may be output to a human-computer interaction terminal, and secondary recognition may be performed manually.
In this embodiment, before step 102, the method may further include: and determining the weight coefficient of the vegetation index, the water body index and the differential building index in the comprehensive spectrum index for distinguishing the error-prone or confusable element classes according to the historical data. For example, the quality problems of large pattern spots and homospectral foreign matters found in national basic geographic national condition monitoring data over the years are counted, spectral information of non-visible light bands and visible light bands of images of the pattern spots is researched and analyzed aiming at easily-error and easily-confused element types, such as paddy fields, dry lands, orchards, forest lands and the like, and a weight coefficient of a vegetation index, a water body index and a difference building index in a pattern spot comprehensive spectral index capable of distinguishing the easily-error and easily-confused element types is obtained.
As an embodiment, the method for detecting a category-specific abnormal pattern spot may specifically be:
judging whether the difference value of the comprehensive spectral indexes of the adjacent image spot images is larger than a set threshold value or not;
if so, indicating that the category abnormal image spots exist, wherein the category abnormal image spots are all the image spots from the adjacent image spot image to the first end of the image spot sequence, and the first end is the end, close to the adjacent image spot image, in the image spot sequence.
For example, whether a category abnormal image spot exists in the first 5 image spots of the image spot sequence is judged, and the specific operation method is as follows: judging whether the difference value of the comprehensive spectral indexes (such as the middle value of the comprehensive spectral indexes) of two adjacent spots in the 5 spots is larger than a set threshold value, wherein the set threshold value can be generally set to be 0.04, and if the difference value of the comprehensive spectral indexes (such as the middle value of the comprehensive spectral indexes) of the 4 th spot and the 5 th spot is larger than the set threshold value, considering that the first 4 spots are all classified abnormal spots.
In this embodiment, when calculating the integrated spectral index of each pattern spot, each index needs to be preprocessed, and the preprocessing process may include normalization and stretching, and the specific process is as follows:
firstly, the normalized vegetation index NDVI of the image spot is calculated by using the following formula, so that the vegetation information in the image is highlighted.
Figure BDA0002536580630000061
Wherein, BnirNear infrared band, B, of remote-sensing imagesredIs a red band of the remote sensing image.
Then, the normalized water body index NDWI of the image spot is calculated by using the following formula, and the water body information in the image is highlighted.
Figure BDA0002536580630000062
Wherein, BnirNear infrared band, B, of remote-sensing imagesgreenIs the green wave band of the remote sensing image.
Then, the difference building index DSBI of the pattern spot is calculated using the following formula.
DSBI＝k*(Bblue-Bred)+(1-k)*(Bblue-Bgreen)
Where k is a calculation coefficient, which can be set to 0.5, BblueIn the blue band, B, of the remotely sensed imageredIn the red band, B, of the remote-sensing imagegreenIs the green wave band of the remote sensing image.
And finally, stretching the normalized vegetation index, the normalized water body index and the difference building index to a value range of 0,1, and stretching the normalized vegetation index and the normalized water body index according to the following formulas.
Figure BDA0002536580630000071
Wherein, PixelvPixel's on the original normalized water or vegetation index'vIs the pixel value after the stretching process.
The differential building index is subjected to stretching treatment according to the following formula.
Figure BDA0002536580630000072
Wherein, PixelvIs originalPixel values on differential building index, Pixel'vFor Pixel values after stretchingv_minPixel value minimum for the original difference building index of the entire scenev_maxThe maximum pixel value of the whole scene original difference building index.
And comprehensively weighting the stretched vegetation index NDVI, the water body index NDWI and the building index NSBI according to the following formula, and calculating the comprehensive spectral index NCI of the pattern spots.
Figure BDA0002536580630000073
Wherein k is1Is the coefficient of NDVI, k2Coefficient of NDWI, k3Is the coefficient of NSBI. k is a radical of1、k2、k3Can be determined according to the statistical analysis result of the spectrum information of each class of elements, when the NDVI value is easier to distinguish a certain element class from other element classes, then the NCI of the element class is calculated, k1Set to a value of 1, k2And k3Is set to 0; when the value of NDWI more easily distinguishes one element class from other element classes, then the NCI of that element class is calculated, k2Set to a value of 1, k1And k3Is set to 0; when the value of NSBI more easily distinguishes one element class from other element classes, then when calculating the NCI of that element class, k3Set to a value of 1, k1And k2The value of (d) is set to 0.
Example 2
Fig. 2 is a schematic structural diagram of an automatic identification system for type-specific anomaly patches according to embodiment 2 of the present invention, and referring to fig. 2, the automatic identification system for type-specific anomaly patches according to this embodiment includes:
the multispectral remote sensing image to be identified acquiring module 201 is used for acquiring the multispectral remote sensing image to be identified;
the image cutting module 202 is configured to cut the multispectral remote sensing image based on a geometric range of the image spots to obtain an image of the image spots with categories easy to make mistakes or confuse;
the comprehensive spectral index calculation module 203 is used for calculating a comprehensive spectral index of each image spot image, wherein the comprehensive spectral index is formed by weighted combination of a vegetation index, a water body index and a difference building index and is used for distinguishing the image spot element types;
the sorting module 204 is used for sorting the image spot images according to the comprehensive spectral indexes of the image spot images in element classes;
the category anomaly determination module 205 is configured to determine whether category anomaly patches exist in a set number of patch images before and after the patch image sequence, and determine whether category anomaly patches exist in an added patch image.
As an embodiment, the system further comprises:
and the initial parameter determining module is used for determining the error-prone or confusable element types and weight coefficients of the vegetation index, the water body index and the difference building index in the comprehensive spectrum index for distinguishing the error-prone or confusable element types according to the historical data.
As an embodiment, the sorting module 204 specifically includes:
the numerical value determining unit is used for determining the median or average value of the comprehensive spectral index of each image spot image;
and the sorting unit is used for sorting the image spot images according to the element classes according to the magnitude of the median or the average value of the comprehensive spectral indexes of the image spot images.
As an embodiment, the category anomaly determination module 205 specifically includes:
the judging unit is used for judging whether the difference value of the comprehensive spectral indexes of the adjacent pattern spot images is larger than a set threshold value or not;
and the category abnormal pattern spot determining unit is used for determining the category abnormal pattern spots of all pattern spots between the adjacent pattern spot images and the first end of the pattern spot sequence when the difference value of the comprehensive spectral indexes of the adjacent pattern spot images is larger than a set threshold value, wherein the first end is the end, which is closer to the adjacent pattern spot images, in the pattern spot sequence.
As an embodiment, the system further comprises:
the normalization processing module is used for performing normalization processing on the vegetation index and the water body index;
and the stretching processing module is used for stretching the differential building index, the normalized vegetation index and the normalized water body index.
The automatic identification method and the system for the type abnormal pattern spots realize the automatic extraction of large pattern spots and same-spectrum foreign matter errors, improve the efficiency and the reliability of quality detection of surface coverage data, and effectively improve the quality of geographical national condition achievements.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.

Claims (10)

1. A method for automatically identifying type of abnormal pattern spots is characterized by comprising the following steps:
acquiring a multispectral remote sensing image to be identified;
based on the geometric range of the image spots, cutting the multispectral remote sensing image to obtain the image of the image spots with the categories easy to make mistakes or confuse;
calculating a comprehensive spectral index of each image spot image, wherein the comprehensive spectral index is formed by weighted combination of a vegetation index, a water body index and a difference building index and is used for distinguishing the image spot element types;
sorting the image spots by element classes according to the size of the comprehensive spectral index of the image spots;
determining whether the set number of image spot images before and after the image spot image sequence have the category abnormal image spots;
if the image spots with abnormal category exist, increasing the image spots with set number before and/or after the image spot image sequence, and carrying out the identification of the abnormal category on the increased image spots, if the image spots with abnormal category exist in the increased image spots, skipping to the step of identifying the image spots with the set number before and/or after the image spot image sequence until the abnormal category image spots do not exist in the increased image spots,
and outputting the image of the category abnormal image spots.
2. The method according to claim 1, wherein before the step of cropping the multispectral remote-sensing image based on the geometric range of the pattern spots, the method further comprises:
and determining the weight coefficient of the vegetation index, the water body index and the differential building index in the comprehensive spectrum index for distinguishing the error-prone or confusable element classes according to the historical data.
3. The method according to claim 1, wherein the step of sorting the spot images according to their comprehensive spectral indexes by element classes comprises:
determining the median or average value of the comprehensive spectral index of each image spot image;
and sorting the image spots according to the element classes according to the magnitude of the median or the average value of the comprehensive spectral indexes of the image spots.
4. The method according to claim 1, wherein the method for detecting the category abnormal pattern spot comprises:
judging whether the difference value of the comprehensive spectral indexes of the adjacent image spot images is larger than a set threshold value or not;
if so, indicating that the category abnormal image spots exist, wherein the category abnormal image spots are all the image spots between the adjacent image spot image and the first end of the image spot sequence, and the first end is the end of the image spot sequence which is closer to the adjacent image spot image.
5. The method for automatically identifying type anomaly spots according to claim 1, wherein before the calculating the comprehensive spectral index of each spot image, the method further comprises:
carrying out normalization processing on the vegetation index and the water body index;
and stretching the differential building index, the normalized vegetation index and the normalized water body index.
6. The method for automatically identifying type anomaly patches according to claim 1, characterized in that the method further comprises:
and outputting the class abnormal image spot identified by the computer to a human-computer interaction end for manual secondary identification.
7. The method for automatically identifying the type abnormal image spots according to claim 1, wherein the cutting process of the multispectral remote sensing image is performed in a computer memory.
8. An automatic identification system for type anomaly patches, comprising:
the multispectral remote sensing image acquisition module to be identified is used for acquiring the multispectral remote sensing image to be identified;
the image cutting module of the image spot is used for cutting the multispectral remote sensing image based on the geometric range of the image spot so as to obtain the image of the image spot with the category which is easy to make mistakes or confuse;
the comprehensive spectral index calculation module is used for calculating a comprehensive spectral index of each image spot image, and the comprehensive spectral index is formed by weighted combination of a vegetation index, a water body index and a difference building index and is used for distinguishing the image spot element types;
the sorting module is used for sorting the image spots by element classes according to the size of the comprehensive spectral index of the image spots;
and the category anomaly determination module is used for determining whether category anomaly image spots exist in the image spots of which the number is set before and after the image spot image sequence and determining whether category anomaly image spots exist in the added image spot images.
9. The system according to claim 8, wherein the system further comprises:
and the initial parameter determining module is used for determining the error-prone or confusable element types and weight coefficients of the vegetation index, the water body index and the difference building index in the comprehensive spectrum index for distinguishing the error-prone or confusable element types according to the historical data.
10. The system according to claim 8, wherein the sorting module specifically includes:
the numerical value determining unit is used for determining the median or average value of the comprehensive spectral index of each image spot image;
and the sorting unit is used for sorting the image spot images according to the element classes according to the magnitude of the median or the average value of the comprehensive spectral indexes of the image spot images.
CN202010534638.9A 2020-06-12 2020-06-12 Automatic identification method and system for type abnormal pattern spots Active CN111709927B (en)

Priority Applications (3)

Application Number Priority Date Filing Date Title
CN202010534638.9A CN111709927B (en) 2020-06-12 2020-06-12 Automatic identification method and system for type abnormal pattern spots
PCT/CN2020/100451 WO2021248599A1 (en) 2020-06-12 2020-07-06 Method and system for automatically identifying spot of which category is abnormal
AU2021102441A AU2021102441A4 (en) 2020-06-12 2021-05-10 Method for automatically identifying abnormal category pattern spots and system thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010534638.9A CN111709927B (en) 2020-06-12 2020-06-12 Automatic identification method and system for type abnormal pattern spots

Publications (2)

Publication Number Publication Date
CN111709927A true CN111709927A (en) 2020-09-25
CN111709927B CN111709927B (en) 2022-05-27

Family

ID=72540346

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010534638.9A Active CN111709927B (en) 2020-06-12 2020-06-12 Automatic identification method and system for type abnormal pattern spots

Country Status (3)

Country Link
CN (1) CN111709927B (en)
AU (1) AU2021102441A4 (en)
WO (1) WO2021248599A1 (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113484245B (en) * 2021-07-05 2022-11-22 重庆市规划和自然资源调查监测院 Remote sensing rapid monitoring method and system for paddy field planting pattern in hilly and mountainous areas and computer readable storage medium
CN114491187B (en) * 2022-01-20 2022-11-18 重庆市规划和自然资源调查监测院 Intelligent natural resource monitoring integrated system
CN118155005A (en) * 2024-05-13 2024-06-07 成都坤舆空间科技有限公司 Ecological restoration map spot matching classification method based on RAFT-Stereo algorithm

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103593852A (en) * 2013-11-29 2014-02-19 中国科学院光电研究院 Hyperspectral image abnormality detection method based on homogeneous patches
CN103593853A (en) * 2013-11-29 2014-02-19 武汉大学 Remote-sensing image multi-scale object-oriented classification method based on joint sparsity representation
CN104408463A (en) * 2014-10-15 2015-03-11 中国土地勘测规划院 High-resolution construction land graph spot identification method based on PanTex and linear characteristic
US20180322656A1 (en) * 2015-11-30 2018-11-08 Delphi Technologies, Llc Method for identification of candidate points as possible characteristic points of a calibration pattern within an image of the calibration pattern
CN110363798A (en) * 2019-07-24 2019-10-22 宁波市测绘设计研究院 A kind of generation method of remote sensing image interpretation sample set
CN111144249A (en) * 2019-12-16 2020-05-12 广州地理研究所 Ground surface coverage type determination method based on automatic optimization MESMA

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120271609A1 (en) * 2011-04-20 2012-10-25 Westerngeco L.L.C. Methods and computing systems for hydrocarbon exploration
KR101756380B1 (en) * 2017-03-06 2017-07-10 한국지질자원연구원 Detection methods of porphyry copper deposits using malachite hyperspectral imagery
CN109726679B (en) * 2018-12-28 2020-11-27 常州市星图测绘科技有限公司 Remote sensing classification error spatial distribution mapping method
CN110135310A (en) * 2019-04-30 2019-08-16 云南财经大学 A kind of crops remote sensing recognition method based on single argument feature selection approach
CN111178169B (en) * 2019-12-12 2023-03-10 广州地理研究所 Urban surface covering fine classification method and device based on remote sensing image

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103593852A (en) * 2013-11-29 2014-02-19 中国科学院光电研究院 Hyperspectral image abnormality detection method based on homogeneous patches
CN103593853A (en) * 2013-11-29 2014-02-19 武汉大学 Remote-sensing image multi-scale object-oriented classification method based on joint sparsity representation
CN104408463A (en) * 2014-10-15 2015-03-11 中国土地勘测规划院 High-resolution construction land graph spot identification method based on PanTex and linear characteristic
US20180322656A1 (en) * 2015-11-30 2018-11-08 Delphi Technologies, Llc Method for identification of candidate points as possible characteristic points of a calibration pattern within an image of the calibration pattern
CN110363798A (en) * 2019-07-24 2019-10-22 宁波市测绘设计研究院 A kind of generation method of remote sensing image interpretation sample set
CN111144249A (en) * 2019-12-16 2020-05-12 广州地理研究所 Ground surface coverage type determination method based on automatic optimization MESMA

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
刘炜: "土地利用/覆被变化信息遥感图像自动分类识别与提取方法研究", 《中国博士学位论文全文数据库 (信息科技辑)》 *
吴林梅: "主题模型建模框架下的高分辨率遥感影像尺度自适应分类", 《中国优秀硕士学位论文全文数据库 (基础科学辑)》 *
李亮 等: "基于像斑直方图G统计量的遥感影像分类", 《国土资源遥感》 *

Also Published As

Publication number Publication date
AU2021102441A4 (en) 2021-06-24
WO2021248599A1 (en) 2021-12-16
CN111709927B (en) 2022-05-27

Similar Documents

Publication Publication Date Title
CN111709927B (en) Automatic identification method and system for type abnormal pattern spots
US7184890B2 (en) Cloud shadow detection: VNIR-SWIR
Leckie et al. Detection and assessment of trees with Phellinus weirii (laminated root rot) using high resolution multi-spectral imagery
CN112101159B (en) Multi-temporal forest remote sensing image change monitoring method
CN112766075B (en) Hyperspectral remote sensing black and odorous water body grading method based on semi-supervised learning strategy
CN104990892B (en) The spectrum picture Undamaged determination method for establishing model and seeds idenmtification method of seed
CN110376202B (en) Tea tree anthracnose lesion identification method based on imaging hyperspectral technology
CN111007013B (en) Crop rotation fallow remote sensing monitoring method and device for northeast cold region
CN108446584A (en) A kind of unmanned plane scouting video image target automatic testing method
CN112699756B (en) Hyperspectral image-based tea origin identification method and system
US20050111692A1 (en) Cloud cover assessment: VNIR-SWIR
KR102177985B1 (en) Apparatus and Method for Automatically Detecting Landslide Area
CN113128388B (en) Optical remote sensing image change detection method based on space-time spectrum characteristics
Font et al. An image processing method for in-line nectarine variety verification based on the comparison of skin feature histogram vectors
CN107330882A (en) Foreign matter online test method after a kind of cut based on machine vision
CN111060455B (en) Northeast cold-cool area oriented remote sensing image crop marking method and device
CN116559111A (en) Sorghum variety identification method based on hyperspectral imaging technology
CN116563704A (en) Multispectral vision epidemic wood identification method, multispectral vision epidemic wood identification system, multispectral vision epidemic wood identification equipment and multispectral vision epidemic wood storage medium
CN116124710A (en) Early pine tree identification method for pine wood nematode disease
CN114398595A (en) Fire point remote sensing identification method based on BP neural network
US20050114026A1 (en) Sub-visible cloud cover assessment: VNIR-SWIR
CN116310826B (en) High-resolution remote sensing image forest land secondary classification method based on graphic neural network
Costa et al. Classification of macaw palm fruits from colorimetric properties for determining the harvest moment
CN112183489A (en) Pine color-changing standing tree identification and positioning method, device, equipment and storage medium
Busuioceanu et al. Evaluation of the CASSI-DD hyperspectral compressive sensing imaging system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant