CN114445717A - Remote sensing image processing method and system for land resource management - Google Patents
Remote sensing image processing method and system for land resource management Download PDFInfo
- Publication number
- CN114445717A CN114445717A CN202210370667.5A CN202210370667A CN114445717A CN 114445717 A CN114445717 A CN 114445717A CN 202210370667 A CN202210370667 A CN 202210370667A CN 114445717 A CN114445717 A CN 114445717A
- Authority
- CN
- China
- Prior art keywords
- sample points
- point
- sample
- false color
- calculating
- 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
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Spectrometry And Color Measurement (AREA)
Abstract
The invention relates to the technical field of image processing, in particular to a remote sensing image processing method and a system for land resource management, wherein the method comprises the following steps: marking the ground object on the remote sensing image, and determining a sample point of the initial sample classification of the ground object; acquiring a ground object spectrum curve corresponding to a ground object, and performing false color synthesis on wave bands corresponding to a peak point, a valley point and a maximum curvature point in the ground object spectrum curve to obtain a false color image; obtaining a false color band combination of a false color image, wherein each band of the false color image corresponds to a RGB channel, and reclassifying the sample points according to the variation consistency of the same type of sample points in different false color band combinations to obtain intermediate sample points; determining a representative test vector according to the intermediate sample point; and determining a final sample point by using the representative test vector as a parameter of the PPI algorithm, and performing final sample classification on the remote sensing image according to the final sample point to obtain a classification result.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a remote sensing image processing method and system for land resource management.
Background
The land resources are generally managed by acquiring remote sensing images through satellites, and reflecting the land coverage utilization, environment monitoring, agricultural estimation and forest fire according to the remote sensing images, so that the identification of land types in the remote sensing images is an indispensable part, and the corresponding land resources can be accurately managed from the remote sensing images only by accurately classifying the land types.
With the development of computer technology, methods for identifying surface features through manual visual observation are gradually eliminated, and the existing methods for identifying surface features usually adopt a supervision classification method, namely, sample points of different types of surface features on an image are selected firstly, and then a computer classifies pixel points on the image according to the similarity between other pixel points and the sample points.
However, the selection of the classification samples is based on the RGB image, so the accuracy of the supervision samples and the purity of the pixels have a great influence on the classification results, and meanwhile, when the reflectivities of three RGB bands of different land types are similar, the situations of wrong classification and missed classification of the land types can occur.
Therefore, a method and a system for processing remote sensing images for land resource management are needed.
Disclosure of Invention
The invention provides a remote sensing image processing method and system for land resource management, which aim to solve the existing problems.
The invention relates to a remote sensing image processing method and a system for land resource management, which adopt the following technical scheme: the method comprises the following steps:
marking each ground object on the remote sensing image, and selecting a corresponding number of pixels as sample points of the initial sample classification for each ground object;
acquiring a ground object spectrum curve corresponding to each ground object, determining a peak point and a valley point in the ground object spectrum curve, calculating a maximum curvature point of the ground object spectrum curve, and performing false color synthesis on a wave band corresponding to the peak point, a wave band corresponding to the valley point and a wave band corresponding to the maximum curvature point in the ground object spectrum curve to obtain a false color image;
obtaining a false color band combination corresponding to each band in a false color image to an RGB channel, and reclassifying the sample points according to the variation consistency of the same type of sample points in different false color band combinations to obtain intermediate sample points;
calculating the similar distance between every two middle sample points in each similar middle sample point, and determining a representative test vector according to the similar middle sample point with the largest similar distance;
and determining a final sample point by using the representative test vector as a parameter of the PPI algorithm, and performing final sample classification on the remote sensing image according to the final sample point to obtain a classification result.
Further, the step of calculating the maximum curvature point of the surface feature spectrum curve comprises:
acquiring each point on a surface feature spectrum curve;
acquiring principal component directions of the points, wherein each principal component direction corresponds to a characteristic value;
acquiring a maximum principal component direction corresponding to the maximum characteristic value, and recording as a connected domain principal direction;
and calculating the rear principal component direction after each point on the ground object spectrum curve is removed, acquiring the difference value between the rear principal component direction and the maximum principal component direction, and taking the point corresponding to the maximum difference value as the maximum curvature point.
Further, the step of obtaining the false color band combination corresponding to each band in the false color image to the RGB channel includes:
recording wave bands corresponding to a peak point, wave bands corresponding to a valley point and wave bands corresponding to a maximum curvature point in the surface feature spectrum curve as a wave band, b wave band and c wave band in sequence;
RGB is used as a fixed channel, and the a wave band, the b wave band and the c wave band are respectively corresponding to the RGB channel to obtain 6 false color wave band combinations.
Further, the step of reclassifying the sample points according to the variation consistency of the same type of sample points in different false color band combinations to obtain intermediate sample points comprises:
calculating the tone value and the gray value of the same type of sample points in each false color waveband combination;
calculating a first direction chain code corresponding to the tone value of each sample point in the same type of sample points and a second direction chain code corresponding to the gray value;
calculating a first similarity of first direction chain codes of different sample points in the same type of initial samples and a second similarity of second direction chain codes;
and reclassifying the sample points according to the first similarity and the second similarity to obtain intermediate sample points.
Further, the step of calculating a first direction chain code corresponding to the hue value and a second direction chain code corresponding to the gray value of each sample point in the same type of sample points includes:
transferring the image of the first false color band combination to an HIS space, extracting H, namely a hue channel, and acquiring a first hue value corresponding to the image of the first false color band combination according to the hue channel;
calculating a second hue value corresponding to the image of the first false color band combination;
calculating a direction value of a connection line of the second hue value and the first hue value;
calculating a chain code containing 5 direction values and 6 hue values by combining all false color wave bands, wherein the chain code is called a first direction chain code;
and calculating a second direction chain code according to the method for calculating the first direction chain code.
Further, the step of reclassifying the sample points according to the first similarity and the second similarity to obtain intermediate sample points includes:
obtaining similar sample points with the first similarity of every two first direction chain codes being greater than 90%;
calculating the average first direction chain code of all the similar sample points with the first similarity larger than 90%;
calculating average direction chain codes of other types of sample points according to a method for calculating average first direction chain codes of similar sample points;
calculating the first final similarity of the first direction chain codes corresponding to the sample points with the similarity smaller than 90% in the same type of sample points and the average direction chain codes of the sample points in other types;
acquiring a second final similarity corresponding to the second similarity according to the mode of acquiring the first final similarity according to the first similarity;
and obtaining sample points corresponding to the first direction chain codes with the first final similarity larger than 90% and sample points corresponding to the second direction chain codes with the second final similarity larger than 90%, adding the sample points corresponding to the first direction chain codes with the first final similarity larger than 90% and the sample points corresponding to the second direction chain codes with the second final similarity larger than 90% into the corresponding sample points of other classes, completing the re-classification of the sample points, obtaining the classified sample points and recording the classified sample points as intermediate sample points.
Further, the step of calculating the similarity distance between every two middle sample points in each similar middle sample point and determining the representative test vector according to the similar middle sample point with the maximum similarity distance includes:
obtaining the similarity between every two middle sample points in the same kind of middle sample points, wherein the similarity is the similar distance;
adding the similar distances between each intermediate sample point and all other intermediate sample points in the same type of intermediate sample points to obtain an intermediate sample point corresponding to the maximum distance sum value;
and taking the intermediate sample points corresponding to the maximum distance sum value as pixels with highest purity in each type of intermediate sample points, and taking vector values of the pixels as representative test vectors in an n-dimensional sample space, wherein n is the number of wave bands corresponding to the satellite for acquiring the remote sensing image.
Further, the step of determining the final sample point using the representative test vector as a parameter of the PPI algorithm includes:
taking the representative test vector as a parameter of a PPI algorithm, and processing the remote sensing image by utilizing the PPI algorithm of the representative test vector to obtain a PPI image;
and selecting the pixel point with the maximum pixel gray value in each ground feature in the PPI image as a final sample point.
The invention also discloses a processing system of the remote sensing image processing method for land resource management, which comprises the following steps:
the information extraction unit is used for marking each ground object on the remote sensing image, and each ground object selects a corresponding number of pixels as sample points of the initial sample classification;
the image synthesis unit is used for acquiring a ground object spectrum curve corresponding to each ground object in the spectrum library, determining a peak point and a valley point in the ground object spectrum curve, calculating a maximum curvature point of the ground object spectrum curve, and performing false color synthesis on a wave band corresponding to the peak point, a wave band corresponding to the valley point and a wave band corresponding to the maximum curvature point in the ground object spectrum curve to obtain a false color image;
the first information processing unit is used for acquiring a false color waveband combination in a false color image, wherein each waveband corresponds to an RGB channel, and calculating the hue value and the gray value of the same type of sample point in each false color waveband combination;
the second information processing unit is used for reclassifying the sample points according to the consistency of the hue value changes and the consistency of the gray value changes of the same type of sample points in different false color band combinations;
the third information processing unit is used for calculating the maximum homogeneous sample point with the maximum distance from each sample point to the homogeneous sample points after reclassification and determining a representative test vector according to the maximum homogeneous sample point;
and the classification unit is used for determining a final sample point according to the representative test vector as a parameter of the PPI algorithm, and performing final sample classification on the remote sensing image according to the final sample point to obtain a classification result.
The invention has the beneficial effects that: the invention relates to a remote sensing image processing method and a system for land resource management, which classify samples of remote sensing images by sample points for initially classifying samples of each terrain, reclassify the sample points by utilizing the change consistency of similar sample points in different false color wave band combinations, determine a representative test vector by an intermediate sample point, determine a final sample point by taking the representative test vector as a parameter of a PPI algorithm to classify the samples of the remote sensing images, reduce the calculated amount of the PPI algorithm, separate the terrain of other classes mixed in the sample points, obtain the final sample points of different terrain, and realize the classification of the samples, thereby improving the classification precision and reducing the occurrence of wrong-classification and missing-classification conditions.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art 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 for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a flowchart illustrating the general steps of an embodiment of a method and system for remote sensing image processing for land resource management according to the present invention;
fig. 2 is a flowchart of the step S2 of obtaining the maximum curvature point of the feature spectrum curve;
FIG. 3 is a flow chart of the intermediate sample acquisition of FIG. 1;
FIG. 4 is a flow chart of the reclassification in FIG. 2;
FIG. 5 is a flow chart of the method of FIG. 1 for obtaining representative test vectors.
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.
An embodiment of the remote sensing image processing method for land resource management of the invention is shown in fig. 1, and the method comprises the following steps:
s1, before classifying the surface features, researchers determine the surface features to be classified, namely the surface features on the remote sensing image need to be classified into which categories, then mark the corresponding surface feature categories on the image, and each surface feature selects a corresponding number of pixels as sample points.
S2, selecting the samples on RGB channels corresponding to the same wave band, if the classified ground objects have ground objects with similar reflectivity on the corresponding wave band, because the ground objects with similar reflectivity are represented as similar color and gray value on the image, the situation of wrong division is easy to generate, in order to avoid directly influencing the subsequent classification result and ensure the purity of the manually selected sample point, firstly obtaining the ground object spectrum curve corresponding to each ground object in the spectrum library, determining the peak point and the valley point in the ground object spectrum curve, calculating the maximum curvature point of the ground object spectrum curve, and performing false color synthesis on the wave band corresponding to the peak point, the wave band corresponding to the valley point and the wave band corresponding to the maximum curvature point in the ground object spectrum curve to obtain a false color image.
Specifically, as shown in fig. 2, the step of calculating the maximum curvature point of the surface feature spectrum curve includes: s21, firstly, acquiring each point on the ground object wave spectrum curve; s22, acquiring principal component directions of the points by using a PCA algorithm to obtain a plurality of principal component directions, wherein each principal component direction is a 2-dimensional unit vector, and each principal component direction corresponds to a characteristic value; s23, acquiring the maximum principal component direction corresponding to the maximum characteristic value, and recording as the principal direction of the connected domain; and S24, calculating the rear principal component direction after each point on the ground object wave spectrum curve is removed, acquiring the difference value between the rear principal component direction and the maximum principal component direction, and taking the point corresponding to the maximum difference value as the maximum curvature point.
S3, obtaining false color band combination corresponding to each band in the false color image in the RGB channel, and reclassifying the sample points to obtain intermediate sample points according to tone value variation consistency of the same type of sample points in different false color band combinations.
Specifically, the step of S31, acquiring a false color image in which each band corresponds to a false color band combination in the RGB channel, includes: s311, recording wave bands corresponding to peak points, wave bands corresponding to valley points and wave bands corresponding to maximum curvature points in the surface feature spectrum curve as a wave band, b wave band and c wave band in sequence; s312, using RGB as a fixed channel, and respectively corresponding the a wave band, the b wave band and the c wave band to the RGB channel to obtain 6 false color wave band combinations.
S32, because the similar reflectivity of different ground objects in the RGB corresponding wave bands can cause that other ground objects are mixed in the same ground object sample point when selecting the classification sample, thereby affecting the accuracy of the classification result, in order to ensure the accuracy of the classification result, because the change conditions of the tone value and the gray value of the same ground object in the false color wave band combinations should be consistent, the step of reclassifying the sample point according to the tone value change consistency of the same sample point in different false color wave band combinations to obtain the intermediate sample point comprises: as shown in fig. 3, in particular, S321, calculating hue values and gray values of the same-type sample points in each pseudo-color band combination; s322, calculating a first direction chain code corresponding to the tone value of each sample point in the same type of sample points and a second direction chain code corresponding to the gray value; specifically, the image of the first false color band combination is transferred to the HIS space, H, namely a hue channel, is extracted, and a first hue value corresponding to the image of the first false color band combination is obtained according to the hue channel; calculating a second hue value corresponding to the image of the first false color band combination; calculating a direction value of a connection line of the second hue value and the first hue value; calculating all false color waveband combinations to obtain a chain code containing 5 direction values and 6 hue values, wherein the chain code is called a first direction chain code; s323, calculating a second direction chain code according to the method for calculating the first direction chain code; s324, calculating a first similarity of the first direction chain codes and a second similarity of the second direction chain codes of different sample points in the same type of sample points; s325, reclassifying the sample points according to the first similarity and the second similarity to obtain intermediate sample points, specifically, as shown in fig. 4, S3251, obtaining similar sample points with the first similarity of every two first direction chain codes being greater than 90%; s3252, calculating average first direction chain codes of all similar sample points with the first similarity larger than 90%; s3253, calculating average direction chain codes of other types of sample points according to a method for calculating average first direction chain codes of similar sample points; s3254, calculating the first final similarity of the first direction chain codes corresponding to the sample points with the similarity smaller than 90% in the same type of sample points and the average direction chain codes of the other type of sample points; s3255, acquiring a second final similarity corresponding to the second similarity according to the mode of acquiring the first final similarity according to the first similarity; s3256, obtaining a sample point corresponding to a first direction chain code with a first final similarity larger than 90% and a sample point corresponding to a second direction chain code with a second final similarity larger than 90%, adding the sample point corresponding to the first direction chain code with the first final similarity larger than 90% and the sample point corresponding to the second direction chain code with the second final similarity larger than 90% into the corresponding sample points of other classes, completing the reclassification of the sample points, obtaining the classified sample points and recording the classified sample points as intermediate sample points.
S4, calculating the similar distance between every two middle sample points in each middle sample point of the same type, and determining a representative test vector according to the middle sample point of the same type with the largest similar distance; as shown in fig. 5, specifically, S41, obtaining a similarity between every two middle sample points in the middle sample points of the same type, where the similarity is a similar distance; s42, summing the similar distances between each middle sample point and all other middle sample points in the same type of middle sample points to obtain a middle sample point corresponding to the maximum distance sum value; s43, taking the intermediate sample point corresponding to the maximum distance sum value as the highest-purity pixel in each type of intermediate sample point, and taking the vector value of these pixels as a representative test vector in an n-dimensional sample space, where n is the number of bands corresponding to the satellite that acquires the remote sensing image, for example, if an OLI sensor on a Landsat-8 satellite has 9 bands, n = 9.
S5, determining a final sample point by taking the representative test vector as a parameter of the PPI algorithm, carrying out final sample classification on the remote sensing image according to the final sample point to obtain a classification result, specifically, taking the representative test vector as a parameter of the PPI algorithm, and processing the remote sensing image by utilizing the PPI algorithm of the representative test vector to obtain a PPI image; and selecting the pixel point with the maximum pixel grey value in each ground feature in the PPI image as a final sample point, and classifying the final sample point by using a classifier to obtain a classification result.
The invention also discloses a processing system of the remote sensing image processing method for land resource management, which comprises the following steps: the system comprises an information extraction unit, an image synthesis unit, a first information processing unit, a second information processing unit, a third information processing unit and a classification unit, wherein the information extraction unit is used for marking each ground object on a remote sensing image, and each ground object selects a corresponding number of pixels as sample points of an initial sample classification; the image synthesis unit is used for acquiring a ground object spectrum curve corresponding to each ground object in the spectrum library, determining a peak point and a valley point in the ground object spectrum curve, calculating a maximum curvature point of the ground object spectrum curve, and performing false color synthesis on a wave band corresponding to the peak point, a wave band corresponding to the valley point and a wave band corresponding to the maximum curvature point in the ground object spectrum curve to obtain a false color image; the first information processing unit is used for acquiring a false color waveband combination in a false color image, wherein each waveband corresponds to an RGB channel, and calculating the hue value and the gray value of the same type of sample point in each false color waveband combination; the second information processing unit is used for reclassifying the sample points according to the consistency of the hue value changes and the consistency of the gray value changes of the same type of sample points in different false color band combinations; the third information processing unit is used for calculating the maximum homogeneous sample point with the maximum distance from each sample point to the homogeneous sample points after reclassification and determining a representative test vector according to the maximum homogeneous sample point; and the classification unit is used for determining a final sample point according to the representative test vector as a parameter of the PPI algorithm, and performing final sample classification on the remote sensing image according to the final sample point to obtain a classification result.
In summary, the invention provides a method and a system for processing remote sensing images for land resource management, which perform sample classification on each ground object by using sample points for initial sample classification, reclassify the sample points by using the variation consistency of the same type of sample points in different false color band combinations, determine a representative test vector by using intermediate sample points, determine a final sample point by using the representative test vector as a parameter of a PPI algorithm to perform sample classification on the remote sensing images, reduce the calculated amount of the PPI algorithm, separate the ground objects mixed with other types in the sample points, obtain the final sample points of different ground objects, and realize sample classification, thereby improving the classification precision and reducing the occurrence of wrong-classification and missing-classification conditions.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (9)
1. A remote sensing image processing method for land resource management is characterized by comprising the following steps:
marking each ground object on the remote sensing image, and selecting a corresponding number of pixels as sample points of the initial sample classification for each ground object;
acquiring a ground object spectrum curve corresponding to each ground object, determining a peak point and a valley point in the ground object spectrum curve, calculating a maximum curvature point of the ground object spectrum curve, and performing false color synthesis on a wave band corresponding to the peak point, a wave band corresponding to the valley point and a wave band corresponding to the maximum curvature point in the ground object spectrum curve to obtain a false color image;
obtaining a false color band combination corresponding to each band in a false color image to an RGB channel, and reclassifying the sample points according to the variation consistency of the same type of sample points in different false color band combinations to obtain intermediate sample points;
calculating the similar distance between every two middle sample points in each similar middle sample point, and determining a representative test vector according to the similar middle sample point with the largest similar distance;
and determining a final sample point by using the representative test vector as a parameter of the PPI algorithm, and performing final sample classification on the remote sensing image according to the final sample point to obtain a classification result.
2. A method of remote sensing image processing for land resource management as recited in claim 1 in which the step of calculating the point of maximum curvature of the feature spectral curve comprises:
acquiring each point on a surface feature spectrum curve;
acquiring principal component directions of the points, wherein each principal component direction corresponds to a characteristic value;
acquiring a maximum principal component direction corresponding to the maximum characteristic value, and recording as a connected domain principal direction;
and calculating the rear principal component direction after each point on the earth feature spectrum curve is removed, acquiring the difference value between the rear principal component direction and the maximum principal component direction, and taking the point corresponding to the maximum difference value as the maximum curvature point.
3. A method as claimed in claim 1, wherein the step of obtaining a false color band combination corresponding to each band in the false color image in the RGB channel comprises:
recording wave bands corresponding to a peak point, wave bands corresponding to a valley point and wave bands corresponding to a maximum curvature point in the surface feature spectrum curve as a wave band, b wave band and c wave band in sequence;
RGB is used as a fixed channel, and the a wave band, the b wave band and the c wave band are respectively corresponding to the RGB channel to obtain 6 false color wave band combinations.
4. The method of claim 1, wherein the step of reclassifying the sample points according to the consistency of the variation of the same type of sample points in different pseudo-color band combinations to obtain intermediate sample points comprises:
calculating the tone value and the gray value of the same type of sample points in each false color waveband combination;
calculating a first direction chain code corresponding to the tone value of each sample point in the same type of sample points and a second direction chain code corresponding to the gray value;
calculating a first similarity of first direction chain codes of different sample points in the same type of initial samples and a second similarity of second direction chain codes;
and reclassifying the sample points according to the first similarity and the second similarity to obtain intermediate sample points.
5. The remote sensing image processing method for land resource management according to claim 4, wherein the step of calculating a first direction chain code corresponding to the tone value and a second direction chain code corresponding to the gray value of each sample point in the same type of sample points comprises:
transferring the image of the first false color band combination to an HIS space, extracting H, namely a hue channel, and acquiring a first hue value corresponding to the image of the first false color band combination according to the hue channel;
calculating a second hue value corresponding to the image of the first false color band combination;
calculating a direction value of a connection line of the second hue value and the first hue value;
calculating a chain code containing 5 direction values and 6 hue values by combining all false color wave bands, wherein the chain code is called a first direction chain code;
and calculating a second direction chain code according to the method for calculating the first direction chain code.
6. The method as claimed in claim 4, wherein the step of reclassifying the sample points according to the first similarity and the second similarity to obtain intermediate sample points comprises:
obtaining similar sample points with the first similarity of every two first direction chain codes being greater than 90%;
calculating the average first direction chain code of all the similar sample points with the first similarity larger than 90%;
calculating average direction chain codes of other types of sample points according to a method for calculating average first direction chain codes of the same type of sample points;
calculating the first final similarity of the first direction chain codes corresponding to the sample points with the similarity smaller than 90% in the same type of sample points and the average direction chain codes of the sample points in other types;
acquiring a second final similarity corresponding to the second similarity according to the mode of acquiring the first final similarity according to the first similarity;
and obtaining sample points corresponding to the first direction chain codes with the first final similarity larger than 90% and sample points corresponding to the second direction chain codes with the second final similarity larger than 90%, adding the sample points corresponding to the first direction chain codes with the first final similarity larger than 90% and the sample points corresponding to the second direction chain codes with the second final similarity larger than 90% into the corresponding sample points of other classes, completing the re-classification of the sample points, obtaining the classified sample points and recording the classified sample points as intermediate sample points.
7. The remote sensing image processing method for land resource management according to claim 1, wherein the step of calculating the similar distance between every two middle sample points in each similar middle sample point, and determining the representative test vector according to the similar middle sample point with the largest similar distance comprises:
obtaining the similarity between every two middle sample points in the same kind of middle sample points, wherein the similarity is the similar distance;
adding the similar distances between each intermediate sample point and all other intermediate sample points in the same type of intermediate sample points to obtain an intermediate sample point corresponding to the maximum distance sum value;
and taking the intermediate sample points corresponding to the maximum distance sum value as pixels with highest purity in each type of intermediate sample points, and taking vector values of the pixels as representative test vectors in an n-dimensional sample space, wherein n is the number of wave bands corresponding to the satellite for acquiring the remote sensing image.
8. A method of remote sensing image processing for land resource management as defined in claim 1 wherein the step of determining a final sample point using a representative test vector as a parameter of the PPI algorithm comprises:
taking the representative test vector as a parameter of a PPI algorithm, and processing the remote sensing image by using the PPI algorithm of the representative test vector to obtain a PPI image;
and selecting the pixel point with the maximum pixel gray value in each ground feature in the PPI image as a final sample point.
9. A processing system for a method of remote sensing image processing for land resource management according to any one of claims 1 to 8, comprising:
the information extraction unit is used for marking each ground object on the remote sensing image, and each ground object selects a corresponding number of pixels as sample points of the initial sample classification;
the image synthesis unit is used for acquiring a ground object spectrum curve corresponding to each ground object in the spectrum library, determining a peak point and a valley point in the ground object spectrum curve, calculating a maximum curvature point of the ground object spectrum curve, and performing false color synthesis on a wave band corresponding to the peak point, a wave band corresponding to the valley point and a wave band corresponding to the maximum curvature point in the ground object spectrum curve to obtain a false color image;
the first information processing unit is used for acquiring a false color waveband combination in a false color image, wherein each waveband corresponds to an RGB channel, and calculating the hue value and the gray value of the same type of sample point in each false color waveband combination;
the second information processing unit is used for reclassifying the sample points according to the consistency of the hue value changes and the consistency of the gray value changes of the same type of sample points in different false color band combinations;
the third information processing unit is used for calculating the maximum homogeneous sample point with the maximum distance from each sample point to the homogeneous sample points after reclassification and determining a representative test vector according to the maximum homogeneous sample point;
and the classification unit is used for determining a final sample point according to the representative test vector as a parameter of the PPI algorithm, and performing final sample classification on the remote sensing image according to the final sample point to obtain a classification result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210370667.5A CN114445717B (en) | 2022-04-11 | 2022-04-11 | Remote sensing image processing method and system for land resource management |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210370667.5A CN114445717B (en) | 2022-04-11 | 2022-04-11 | Remote sensing image processing method and system for land resource management |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114445717A true CN114445717A (en) | 2022-05-06 |
| CN114445717B CN114445717B (en) | 2022-06-28 |
Family
ID=81359883
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210370667.5A Active CN114445717B (en) | 2022-04-11 | 2022-04-11 | Remote sensing image processing method and system for land resource management |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114445717B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6700939B1 (en) * | 1997-12-12 | 2004-03-02 | Xtremespectrum, Inc. | Ultra wide bandwidth spread-spectrum communications system |
| CN102222238A (en) * | 2011-05-17 | 2011-10-19 | 中国科学院遥感应用研究所 | Automatic simulation method of natural-color products of high-space-resolution remote sensing images |
| US20150071528A1 (en) * | 2013-09-11 | 2015-03-12 | Digitalglobe, Inc. | Classification of land based on analysis of remotely-sensed earth images |
| CN104751166A (en) * | 2013-12-30 | 2015-07-01 | 中国科学院深圳先进技术研究院 | Spectral angle and Euclidean distance based remote-sensing image classification method |
| CN105608473A (en) * | 2015-12-31 | 2016-05-25 | 中国资源卫星应用中心 | High-precision land cover classification method based on high-resolution satellite image |
| WO2020015326A1 (en) * | 2018-07-19 | 2020-01-23 | 山东科技大学 | Remote sensing image cloud shadow detection method supported by earth surface type data |
| CN113222054A (en) * | 2021-05-28 | 2021-08-06 | 珠江水利委员会珠江水利科学研究院 | Remote sensing image fusion method, system, equipment and medium based on characteristic ratio index |
-
2022
- 2022-04-11 CN CN202210370667.5A patent/CN114445717B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6700939B1 (en) * | 1997-12-12 | 2004-03-02 | Xtremespectrum, Inc. | Ultra wide bandwidth spread-spectrum communications system |
| CN102222238A (en) * | 2011-05-17 | 2011-10-19 | 中国科学院遥感应用研究所 | Automatic simulation method of natural-color products of high-space-resolution remote sensing images |
| US20150071528A1 (en) * | 2013-09-11 | 2015-03-12 | Digitalglobe, Inc. | Classification of land based on analysis of remotely-sensed earth images |
| CN104751166A (en) * | 2013-12-30 | 2015-07-01 | 中国科学院深圳先进技术研究院 | Spectral angle and Euclidean distance based remote-sensing image classification method |
| CN105608473A (en) * | 2015-12-31 | 2016-05-25 | 中国资源卫星应用中心 | High-precision land cover classification method based on high-resolution satellite image |
| WO2020015326A1 (en) * | 2018-07-19 | 2020-01-23 | 山东科技大学 | Remote sensing image cloud shadow detection method supported by earth surface type data |
| CN113222054A (en) * | 2021-05-28 | 2021-08-06 | 珠江水利委员会珠江水利科学研究院 | Remote sensing image fusion method, system, equipment and medium based on characteristic ratio index |
Non-Patent Citations (3)
| Title |
|---|
| NA WANG 等: "On the land resource management and sustainable development", 《IEEE》 * |
| 朱长明等: "地物波谱数据辅助的SPOT影像模拟真彩色方法研究", 《测绘学报》 * |
| 秦方普等: "基于谱聚类与类间可分性因子的高光谱波段选择", 《光谱学与光谱分析》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114445717B (en) | 2022-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Renza et al. | A new approach to change detection in multispectral images by means of ERGAS index | |
| Ke et al. | Adaptive change detection with significance test | |
| CN109063754B (en) | Remote sensing image multi-feature joint classification method based on OpenStreetMap | |
| US4912770A (en) | Method of detecting change using image | |
| KR102170260B1 (en) | Apparatus and method for fusing synthetic aperture radar image and multispectral image, method for detecting change using it | |
| EP2157544A1 (en) | System for adaptive removal of speckle noise in digital images and generation of a colour composite product based on automated analysis of speckle characteristics | |
| CN111814771A (en) | Image processing method and device | |
| CN111638565A (en) | Method for monitoring rainstorm in mountainous area | |
| Touati et al. | A reliable mixed-norm-based multiresolution change detector in heterogeneous remote sensing images | |
| CN113239736B (en) | Land coverage classification annotation drawing acquisition method based on multi-source remote sensing data | |
| EP4296706A1 (en) | Method and system for lulc guided sar visualization | |
| Verma et al. | Development of LR-PCA based fusion approach to detect the changes in mango fruit crop by using landsat 8 OLI images | |
| Erener et al. | A methodology for land use change detection of high resolution pan images based on texture analysis | |
| Calota et al. | Estimating NDVI from SAR images using DNN | |
| CN113689412A (en) | Thyroid image processing method and device, electronic equipment and storage medium | |
| CN114445717B (en) | Remote sensing image processing method and system for land resource management | |
| JP4285640B2 (en) | Object identification method, apparatus and program | |
| Aytekin et al. | Automatic and unsupervised building extraction in complex urban environments from multi spectral satellite imagery | |
| Richards et al. | Image classification methodologies | |
| US11300681B2 (en) | Weather radar apparatus and severe rain prediction method | |
| Luo et al. | Unsupervised classification of hyperspectral images by using linear unmixing algorithm | |
| Clark et al. | Finding a good segmentation strategy for tree crown transparency estimation | |
| Desclée et al. | Object-based method for automatic forest change detection | |
| CN113434621B (en) | ArcGIS-based water and soil resource thematic map generation method, device, equipment and medium | |
| Muren et al. | Automatic deforestation detectors based on frequentist statistics and their extensions for other spatial objects |
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 |