CN110473205A - Remote sensing image information extracting method and system based on arrow bar phantom - Google Patents
Remote sensing image information extracting method and system based on arrow bar phantom Download PDFInfo
- Publication number
- CN110473205A CN110473205A CN201910617812.3A CN201910617812A CN110473205A CN 110473205 A CN110473205 A CN 110473205A CN 201910617812 A CN201910617812 A CN 201910617812A CN 110473205 A CN110473205 A CN 110473205A
- Authority
- CN
- China
- Prior art keywords
- data
- vector
- module
- image
- remote sensing
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 26
- 230000008859 change Effects 0.000 claims abstract description 31
- 238000004364 calculation method Methods 0.000 claims abstract description 24
- 230000002452 interceptive effect Effects 0.000 claims abstract description 17
- 238000012549 training Methods 0.000 claims abstract description 10
- 238000000605 extraction Methods 0.000 claims description 25
- 230000000903 blocking effect Effects 0.000 claims description 17
- 230000003595 spectral effect Effects 0.000 claims description 15
- 230000011218 segmentation Effects 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 8
- 238000000638 solvent extraction Methods 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 6
- 230000002194 synthesizing effect Effects 0.000 claims description 6
- 230000015572 biosynthetic process Effects 0.000 claims description 3
- 238000003786 synthesis reaction Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 239000011159 matrix material Substances 0.000 description 8
- 230000008569 process Effects 0.000 description 8
- 230000000694 effects Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000012790 confirmation Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000012805 post-processing Methods 0.000 description 2
- 238000012163 sequencing technique Methods 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000009966 trimming Methods 0.000 description 1
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
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- 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
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Geometry (AREA)
- Image Processing (AREA)
Abstract
The invention discloses remote sensing image information extracting methods and system based on arrow bar phantom, comprising the following steps: building arrow bar phantom；Object each in the arrow bar phantom is subjected to feature calculation；Training sample is acquired, the characteristic dimension after calculating is exercised supervision classification；The difference that two phase images of front and back divide figure spot in characteristic dimension is calculated, the big figure spot of difference is determined as doubtful region of variation, variation discovery information is extracted by arrow grid interactive editor.The beneficial effects of the present invention are: displaying live view checks arrow bar phantom information；Real-time update neighborhood topology index file realizes the automatic building of topological relation；It applies it to high spatial resolution remote sense image satellite images interpretation, in remote sensing image extracting change information, meets the application demand that different high score remotely-sensed data source information are extracted.
Description
Technical Field
The invention relates to the technical field of remote sensing, in particular to a method and a system for extracting remote sensing image information based on a vector grid model.
Background
With the continuous emission of remote sensing satellites, remote sensing images and historical land coverage data are increasing day by day, interactive editing cannot be avoided in the post-processing process of remote sensing image information, a traditional GIS platform cannot meet the strong interactive requirement of information extraction post-processing, and a huge vector grid data body is difficult to support.
Compared with the traditional medium-low resolution remote sensing image, the high-spatial resolution remote sensing image has fewer spectral features and richer spatial texture and geometric information, and the traditional method takes pixels as an analysis unit to extract high-resolution remote sensing information, cannot effectively utilize the spatial texture and the geometric features of the high-resolution image, takes a homogeneous object as the analysis unit, and can completely combine the spectral features, the texture features and the geometric features of the image to extract high-resolution data information.
An effective solution to the problems in the related art has not been proposed yet.
Disclosure of Invention
Aiming at the technical problems in the related art, the invention provides a remote sensing image information extraction method and system based on a vector grid model, which can avoid the problems of weak grid interactive editing capacity, topological problems easily occurring in the process of editing vector data by using a GIS means and conversion problems caused by different types of vector and grid data.
In order to achieve the technical purpose, the technical scheme of the invention is realized as follows:
a remote sensing image information extraction method based on a vector grid model comprises the following steps:
constructing a vector grid model;
performing feature calculation on each object in the vector grid model;
collecting training samples, and carrying out supervision and classification on the calculated feature dimensions;
calculating the difference value of the segmentation image spots of the images in the front period and the back period on the characteristic dimension, judging the image spots with large difference values as suspected change areas, and extracting change discovery information through vector grid interactive editing.
Further, the constructing the vector grid model includes:
acquiring data, wherein the data is raster data or input raster data and vector data;
setting parameters, wherein the parameters comprise a region merging parameter and a waveband weight parameter;
partitioning the image;
merging the block data according to the similar areas;
and synthesizing the block region combination result, and generating the vector grid model by using the block region combination result.
Further, the blocking the image comprises:
partitioning the image;
traversing block data of each image;
performing image first region merging on each block of data;
and generating a region merging result of each block of data.
Further, the information in the vector grid model includes grid data, vector data, region merging parameter setting information, a region merging result, and a topology index file.
Further, the feature dimensions include vector features, geometric features, spectral features, texture features, and custom features.
In another aspect of the present invention, a remote sensing image information extraction system based on multi-level and multi-scale segmentation is provided, which includes:
the construction module is used for constructing a vector grid model;
the characteristic calculation module is used for carrying out characteristic calculation on each object in the vector grid model;
the supervision and classification module is used for collecting training samples and carrying out supervision and classification on the calculated characteristic dimensions;
and the difference value calculation module is used for calculating the difference value of the segmented image spots of the front and rear images on the characteristic dimension, judging the image spot with the large difference value as a suspected change area, and extracting change discovery information through vector grid interactive editing.
Further, the building module comprises:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring data, and the data is raster data or input raster data and vector data;
the device comprises a parameter setting module, a parameter setting module and a parameter setting module, wherein the parameter setting module is used for setting parameters, and the parameters comprise region merging parameters and waveband weight parameters;
the image blocking module is used for blocking the image;
the merging module is used for merging the block data according to the similar area;
and the synthesis module is used for synthesizing the block region combination result and generating the vector grid model.
Further, the image blocking module includes:
the blocking module is used for blocking the image;
the traversal module is used for traversing the block data of each image;
the first region merging module is used for performing first region merging on images of each block of data;
and the generating module is used for generating the area merging result of each piece of data.
Further, the information in the vector grid model includes grid data, vector data, region merging parameter setting information, a region merging result, and a topology index file.
Further, the feature dimensions include vector features, geometric features, spectral features, texture features, and custom features.
The invention has the beneficial effects that:
the vector grid model information is browsed and checked in real time, meanwhile, the vector grid model is used as basic data, the whole process of information extraction application to interactive editing, roaming and fast browsing is supported, and time consumption of data format conversion caused by different data types is saved; the vector grid model records the topological relation between the graphs in real time, and the map spot objects can update the neighborhood topology index file in real time through a series of editing of cutting, digging, connecting in series, combining and assigning values, so that the automatic construction of the topological relation is realized;
the vector grid model is used as basic data, functions of feature calculation, supervision and classification, change discovery, vector grid interactive editing and the like of the vector grid model are achieved, the vector grid model is applied to high-spatial-resolution remote sensing image remote sensing information interpretation and remote sensing image change information extraction, and application requirements of different high-resolution remote sensing data source information extraction are met.
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 diagram of a process for constructing a vector grid model according to an embodiment of the present invention;
FIG. 2 is a graph showing the effect of the vector grid model according to the embodiment of the present invention;
FIG. 3 is a schematic flowchart of applying the vector grid model to information interpretation of high-resolution remote sensing images according to an embodiment of the present invention;
FIG. 4(a) is a GF-1 remote sensing image + historical road network data effect plot in accordance with an embodiment of the present invention;
FIG. 4(b) is a diagram illustrating the effect of SVM supervised classification results according to an embodiment of the present invention;
fig. 5 is a schematic flowchart of a method for extracting information from a remote sensing image based on a vector grid model according to an embodiment of the present invention;
fig. 6(a) is a diagram illustrating the GF1 remote sensing image + change extraction result in 2014 according to the embodiment of the present invention;
fig. 6(b) is a diagram illustrating the GF1 remote sensing image + change extraction result in 2017 according to the embodiment of the present invention;
fig. 7 is a schematic structural diagram of a remote sensing image information extraction system based on a vector grid model according to an embodiment 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 that can be derived by one of ordinary skill in the art from the embodiments given herein are intended to be within the scope of the present invention.
As shown in fig. 3 and 5, the method for extracting information of remote sensing images based on a vector grid model according to the embodiment of the present invention includes the following steps:
constructing a vector grid model;
specifically, there are two situations for creating the vector grid model: 1) merging adjacent similar areas into a spot object by taking the raster data as input, recording the index relation of merging adjacent areas each time, and generating a data set of a homogeneous spot object; 2) and combining adjacent similar areas into a spot object by taking the raster data and the vector data as input, inheriting the shape and the attribute characteristics of the vector data by the spot object in the combining process, and simultaneously recording the combining index relation of each adjacent area to generate a spot object data set covering raster and vector information.
Performing feature calculation on each object in the vector grid model;
specifically, with the vector grid model as basic data, feature calculation is performed on each object in the vector grid model, and the feature calculation includes: the system comprises vector features, geometric features, spectral features, textural features and custom features, wherein the vector features are attribute features of corresponding vector data when input data are grid and vector data; the number of the geometric features is 5, which is mainly described from the shape of the image spot object, and the area, the circumference, the compactness, the length-width and the length-width of the image spot are the area/the outer area; the number of the spectral features is 5, the spectral features are mainly described from the spectral features of the divided image spots, and the spectral features comprise spectral values, standard deviations, maximum values, minimum values and brightness values of the image spots; the texture features are mainly gray level co-occurrence matrix features, and the gray level co-occurrence matrix is a common method for describing textures by researching the spatial correlation characteristics of gray levels; the user-defined characteristic mainly provides a characteristic calculator for user-defining new characteristics, and can perform operation processing such as addition, subtraction, multiplication, division, power operation and the like on the existing characteristics to obtain new user-defined characteristics.
Collecting training samples, and carrying out supervision and classification on the calculated feature dimensions;
specifically, the method provides sample management and supervision classification capability, wherein the sample management has the functions of sample collection, sample deletion and sample export; the supervised classification method comprises Bayes, neural network, support vector machine, decision tree, random forest and the like, and according to the ground surface coverage category, the ground feature category samples are collected, the characteristics of different dimensions are input, and the remote sensing image automatic interpretation is realized by using a supervised classification classifier.
Calculating the difference value of the segmentation image spots of the images in the front period and the back period on the characteristic dimension, judging the image spots with large difference values as suspected change areas, and extracting change discovery information through vector grid interactive editing.
Specifically, a remote sensing image change detection method is provided as a difference method, common methods include Euclidean distance, Manhattan distance, Chebyshev distance and the like, the difference value of the image spots segmented on the characteristic dimension of the images in the two stages before and after is calculated, the image spots with large difference values are judged as suspected change areas, and change discovery information is obtained; providing vector grid interactive editing confirmation, and obtaining a final change discovery result, wherein the final change discovery result comprises the capabilities of pattern spot cutting, hole digging, serial connection, combination, assignment and the like, and the pattern spot cutting is a process of cutting one pattern spot into a plurality of pattern spots; the pattern spot holing is to carry out holing treatment on the existing pattern spots; the image spot concatenation is to carry out trimming processing on the boundary of the existing image spots; the pattern spot combination is to combine a plurality of adjacent pattern spots; the image spot assignment is a process of reassigning the attribute of the image spot.
In a specific embodiment of the present invention, the constructing the vector grid model includes:
acquiring data, wherein the data is raster data or input raster data and vector data;
setting parameters, wherein the parameters comprise a region merging parameter and a waveband weight parameter;
partitioning the image;
merging the block data according to the similar areas;
and synthesizing the block region combination result, and generating the vector grid model by using the block region combination result.
Specifically, the S1 inputting data, wherein the inputting data includes inputting raster data, or inputting raster data and vector data; s2 sets parameters, wherein the parameters include region merging parameters and band weight parameters, and the specific parameters are shown in table 1.
Table 1 set parameter description
And S3 homogeneous region merging, namely, blocking the input images, traversing block data of each image, performing first region merging on each block data of the images, stopping merging and controlling by the scale parameter, and generating a region merging result of each block data.
S4 merging the result of the region merging of the blocks, and mosaic-synthesizing the result of the region merging between different blocks.
S5, outputting a vector grid model, where the vector grid model shows an effect as shown in fig. 2, and the output vector grid model file includes: 1) the method comprises the steps that data input information comprises raster data and vector data information, wherein the raster data information comprises basic information such as geographic coordinates, wave band numbers, bit depths and data types, and the vector data information comprises basic information such as geographic coordinates and vector attribute characteristics; 2) region merging parameter setting information; 3) the region merging result is a pattern spot object result in a grid format; 4) and the topology index file records the neighborhood relation of the adjacent pattern spot objects.
In an embodiment of the present invention, the partitioning the image comprises:
partitioning the image;
traversing block data of each image;
performing image first region merging on each block of data;
and generating a region merging result of each block of data.
In a specific embodiment of the present invention, the information in the vector grid model includes grid data, vector data, region merging parameter setting information, a region merging result, and a topology index file.
In a specific embodiment of the present invention, the feature dimensions include vector features, geometric features, spectral features, texture features, and custom features.
In another aspect of the present invention, a remote sensing image information extraction system based on multi-level and multi-scale segmentation is provided, which includes:
the construction module is used for constructing a vector grid model;
the characteristic calculation module is used for carrying out characteristic calculation on each object in the vector grid model;
the supervision and classification module is used for collecting training samples and carrying out supervision and classification on the calculated characteristic dimensions;
and the difference value calculation module is used for calculating the difference value of the segmented image spots of the front and rear images on the characteristic dimension, judging the image spot with the large difference value as a suspected change area, and extracting change discovery information through vector grid interactive editing.
In a specific embodiment of the present invention, the building block includes:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring data, and the data is raster data or input raster data and vector data;
the device comprises a parameter setting module, a parameter setting module and a parameter setting module, wherein the parameter setting module is used for setting parameters, and the parameters comprise region merging parameters and waveband weight parameters;
the image blocking module is used for blocking the image;
the merging module is used for merging the block data according to the similar area;
and the synthesis module is used for synthesizing the block region combination result and generating the vector grid model.
In an embodiment of the present invention, the image blocking module includes:
the blocking module is used for blocking the image;
the traversal module is used for traversing the block data of each image;
the first region merging module is used for performing first region merging on images of each block of data;
and the generating module is used for generating the area merging result of each piece of data.
In a specific embodiment of the present invention, the information in the vector grid model includes grid data, vector data, region merging parameter setting information, a region merging result, and a topology index file.
In a specific embodiment of the present invention, the feature dimensions include vector features, geometric features, spectral features, texture features, and custom features.
And performing vector grid model and high-resolution remote sensing image information extraction description by using two embodiments of supervision, classification, remote sensing interpretation and change detection.
In order to facilitate understanding of the above-described technical aspects of the present invention, the above-described technical aspects of the present invention will be described in detail below in terms of specific usage.
Vector grid model applied to high-resolution remote sensing image information interpretation
Applying the vector grid model to high-resolution remote sensing image information interpretation, specifically referring to a flow chart in fig. 3, firstly, selecting a Er Hai region GF1 multispectral remote sensing image and historical road network data as input data, referring to a flow chart in fig. 4(a), secondly, creating the vector grid model, then, performing characteristic calculation by taking the vector grid model as basic data, and calculating the characteristics of each pattern spot object, wherein the specific calculation characteristics comprise:
(1) vector characteristics: category characteristics of historical road network data;
(2) mean value: calculating layer average values of all n pixels forming an image object;
(3) standard deviation: calculating the layer values of all n pixels forming an image object to obtain a standard deviation;
(4) brightness: an average of the spectral averages of an image object;
(5) maximum value: maximum value obtained by sequencing layer values of all n pixels forming an image object;
(6) minimum value: the minimum value obtained by sequencing the layer values of all n pixels forming an image object;
(7) NDWI: NDWI (Normalized Difference Water Index), formula: NDWI ═ (p (green) -p (nir))/(p (green) + p (nir));
(8) NDVI: NDVI (Normalized Difference Vegetation Index) is given by the formula: NDVI ═ p (nir) -p (r))/(p (nir)) + p (r));
inheriting the existing road type on the basis of the characteristic calculation; then, four kinds of training samples of forest land, cultivated land, water body and building are collected, multi-dimensional features are input, and supervised classification by a Support vector machine (SVM for short) is carried out.
The adopted SVM supervised classification is a binary classification model, a basic model of the SVM supervised classification is defined as a linear classifier with the maximum interval on a feature space, and a learning strategy of the SVM supervised classification is interval maximization and can be finally converted into the solution of a convex quadratic programming problem. The principle is as follows:
assume input training samples { (x)1，y1)，...，(xl，yl)}(xi∈Rn，yiE { -1, 1}, i { -1, 2.·, l) may be hyperplane<w，x>The + b-0 (b belongs to R) is linearly divided into two types, and the hyperplane must satisfy yi(<w，x>+b)≥1(i＝1，2，...，l),<*，*>Representing the vector inner product, and solving a quadratic programming problem equivalently when the classification interval of the optimal hyperplane solved by the SVM is maximum:
s.t.yi(<w，xi>+b)≥1 i＝1，2，....，l
the decision function based on the optimal classification hyperplane is:
wherein,is the only solution to the quadratic programming problem described above.
Introducing a relaxation variable ζ when training data in the original feature space is linearly irreversibleiEqual to or greater than 0(i ═ 1, 2.. multidot.l), so that the presence of misclassified samples is allowed, with the corresponding optimization problems:
s.t.yi(<w，xi>+b)≥1-ζi ζi≥0；i＝1，2，....，l
wherein, C > 0 is a penalty factor used for controlling the penalty degree of the wrong sample.
On the basis of the supervised classification result, the vector grid interactive editing cutting, hole digging, concatenation, combination and assignment functions are utilized to finely modify the supervised classification result, and the final information extraction result is obtained, which is shown in fig. 4 (b).
Vector grid model applied to extraction of change information of remote sensing image
In this embodiment, the vector grid model is applied to the extraction of the change information of the remote sensing image, and the specific flow is shown in fig. 5. GF1 multispectral remote sensing images in 2014 and 2017 of Qilian mountain are selected, and the images are shown in figure 6. Firstly, performing wave band superposition on two-stage remote sensing image data as input data to create a vector grid model; secondly, carrying out feature calculation by taking the vector grid model as basic data, wherein the calculation features are as follows:
(1) homogeneity: the homogeneity of the image texture is reflected, and the local change of the image texture is measured. If the value is large, the image texture lacks variation among different regions, the local part is very uniform, and the formula is as follows:wherein i and j represent rows, respectivelyAnd columns, N representing the number of rows or columns, Pi，jNormalized values for pixel (i, j).
(2) Mean value: the mean reflects the degree of regularity of the texture. Texture is disordered, difficult to describe and small in value; the regularity is strong, and the value which is easy to describe is larger. The formula is as follows:wherein i and j represent rows and columns, respectively, N represents the number of rows or columns, and Pi，jNormalized values for pixel (i, j).
(3) Standard deviation: the variance and standard deviation reflect the measure of the pixel value and the mean deviation, and when the gray scale change of the image is large, the variance and standard deviation are large. The formula is as follows:wherein i and j represent rows and columns, respectively, N represents the number of rows or columns, and Pi，jNormalized value, u, for pixel (i, j)i，jIs the mean value of the gray level co-occurrence matrix.
(4) Angular second moment: the sum of squares of the gray level co-occurrence matrix elements is also called energy, and the uniformity degree and the texture thickness degree of the image gray level distribution are reflected. Coarse texture, large energy, fine texture and small energy. The formula is as follows:wherein i and j represent rows and columns, respectively, N represents the number of rows or columns, and Pi，jNormalized values for pixel (i, j).
(5) Contrast ratio: the values of the metric matrix are how distributed and how much of the local variation in the image reflects the sharpness of the image and the depth of the texture. The deeper the furrows of the texture, the greater the contrast, the clearer the effect; otherwise, if the contrast value is small, the grooves are shallow and the effect is blurred. The formula is as follows:wherein i and j represent rows and columns, respectively, N represents the number of rows or columns, and Pi，jMarking the pixel (i, j)Normalized value, ui，jIs the mean value of the gray level co-occurrence matrix, deltai，jIs the gray level co-occurrence matrix standard deviation.
(6) Correlation: the degree of similarity of the gray levels of the image in the row or column direction is measured, so that the magnitude of the value reflects the local gray level correlation, and the larger the value, the larger the correlation. The formula is as follows:wherein i and j represent rows and columns, respectively, N represents the number of rows or columns, and Pi，jNormalized values for pixel (i, j).
(7) Non-similarity: similar to contrast but linearly increasing. If the local contrast is higher, the dissimilarity is also higher. The formula is as follows:wherein i and j represent rows and columns, respectively, N represents the number of rows or columns, and Pi，jNormalized values for pixel (i, j).
(8) Entropy: the image comprises randomness measurement of information quantity, and when all values in the co-occurrence matrix are equal or a pixel value shows the maximum randomness, the entropy is maximum; therefore, the entropy value indicates the complexity of the image gray level distribution, and the larger the entropy value, the more complex the image. The formula is as follows:wherein i and j represent rows and columns, respectively, N represents the number of rows or columns, and Pi，jNormalized values for pixel (i, j).
On the basis of the above feature calculation, it is found that the adopted method is a direct difference method, in this embodiment, the euclidean distance is used to calculate the feature difference of the images in two phases, and the principle of the euclidean distance is as follows:
the euclidean distance represents the true distance between two points in an n-dimensional space, n being the formula for the space:
wherein n represents a characteristic dimension, X ═ X1，x2，，...，xnDenotes the feature of the first point, Y ═ Y1，y2Yn represents the characteristics of the second point.
Sorting the Euclidean distance calculation results, outputting the areas with large difference values as suspected change areas, obtaining the final change finding result through manual interactive editing and confirmation, and finding a new road in 2017 years in 2014-materials in Qilian mountains as shown in figures 6a and 6 b.
In conclusion, by means of the technical scheme, the vector grid model information is browsed and checked in real time, meanwhile, the vector grid model is used as basic data, the whole process of information extraction application to interactive editing, roaming and fast browsing is supported, and time consumption of data format conversion caused by different data types is saved; the vector grid model records the topological relation between the graphs in real time, and the map spot objects can update the neighborhood topology index file in real time through a series of editing of cutting, digging, connecting in series, combining and assigning values, so that the automatic construction of the topological relation is realized; the vector grid model is used as basic data, functions of feature calculation, supervision and classification, change discovery, vector grid interactive editing and the like of the vector grid model are achieved, the vector grid model is applied to high-spatial-resolution remote sensing image remote sensing information interpretation and remote sensing image change information extraction, and application requirements of different high-resolution remote sensing data source information extraction are met.
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 (10)
1. A remote sensing image information extraction method based on a vector grid model is characterized by comprising the following steps:
constructing a vector grid model;
performing feature calculation on each object in the vector grid model;
collecting training samples, and carrying out supervision and classification on the calculated feature dimensions;
calculating the difference value of the segmentation image spots of the images in the front period and the back period on the characteristic dimension, judging the image spots with large difference values as suspected change areas, and extracting change discovery information through vector grid interactive editing.
2. The method for extracting remote sensing image information based on the vector grid model according to claim 1, wherein the constructing the vector grid model comprises:
acquiring data, wherein the data is raster data or input raster data and vector data;
setting parameters, wherein the parameters comprise a region merging parameter and a waveband weight parameter;
partitioning the image;
merging the block data according to the similar areas;
and synthesizing the block region combination result, and generating the vector grid model by using the block region combination result.
3. The method for extracting information of remote sensing images based on the vector grid model according to claim 1, wherein the step of partitioning the images comprises:
partitioning the image;
traversing block data of each image;
performing image first region merging on each block of data;
and generating a region merging result of each block of data.
4. The method for extracting information of remote sensing images based on the vector grid model as claimed in claim 1, wherein the information in the vector grid model includes raster data, vector data, region merging parameter setting information, region merging results and topology index files.
5. The method for extracting information of remote sensing images based on the vector grid model according to any one of claims 1-4, wherein the feature dimensions include vector features, geometric features, spectral features, texture features and custom features.
6. A remote sensing image information extraction system based on multi-level and multi-scale segmentation is characterized by comprising:
the construction module is used for constructing a vector grid model;
the characteristic calculation module is used for carrying out characteristic calculation on each object in the vector grid model;
the supervision and classification module is used for collecting training samples and carrying out supervision and classification on the calculated characteristic dimensions;
and the difference value calculation module is used for calculating the difference value of the segmented image spots of the front and rear images on the characteristic dimension, judging the image spot with the large difference value as a suspected change area, and extracting change discovery information through vector grid interactive editing.
7. The remote sensing image information extraction system based on multi-level and multi-scale segmentation of claim 6, wherein the construction module comprises:
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring data, and the data is raster data or input raster data and vector data;
the device comprises a parameter setting module, a parameter setting module and a parameter setting module, wherein the parameter setting module is used for setting parameters, and the parameters comprise region merging parameters and waveband weight parameters;
the image blocking module is used for blocking the image;
the merging module is used for merging the block data according to the similar area;
and the synthesis module is used for synthesizing the block region combination result and generating the vector grid model.
8. The remote sensing image information extraction system based on multi-level and multi-scale segmentation of claim 6, wherein the image blocking module comprises:
the blocking module is used for blocking the image;
the traversal module is used for traversing the block data of each image;
the first region merging module is used for performing first region merging on images of each block of data;
and the generating module is used for generating the area merging result of each piece of data.
9. The remote sensing image information extraction system based on multi-level and multi-scale segmentation of claim 6, wherein the information in the vector grid model comprises grid data, vector data, region merging parameter setting information, region merging results and a topology index file.
10. The remote sensing image information extraction system based on multi-level and multi-scale segmentation according to any one of claims 6-9, wherein the feature dimensions include vector features, geometric features, spectral features, texture features and custom features.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910617812.3A CN110473205A (en) | 2019-07-10 | 2019-07-10 | Remote sensing image information extracting method and system based on arrow bar phantom |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910617812.3A CN110473205A (en) | 2019-07-10 | 2019-07-10 | Remote sensing image information extracting method and system based on arrow bar phantom |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110473205A true CN110473205A (en) | 2019-11-19 |
Family
ID=68507496
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910617812.3A Pending CN110473205A (en) | 2019-07-10 | 2019-07-10 | Remote sensing image information extracting method and system based on arrow bar phantom |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110473205A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112800992A (en) * | 2021-02-03 | 2021-05-14 | 成都鼎安华智慧物联网股份有限公司 | Extraction method based on remote sensing data high-temperature abnormal information |
| CN112945352A (en) * | 2021-02-03 | 2021-06-11 | 成都鼎安华智慧物联网股份有限公司 | Extraction method based on remote sensing data water level abnormal information |
| CN113240026A (en) * | 2021-05-24 | 2021-08-10 | 中国科学院重庆绿色智能技术研究院 | Vector-grid combined inland water surface floater batch identification and extraction method |
| CN113822900A (en) * | 2021-07-09 | 2021-12-21 | 武汉大学 | Vector constraint object-oriented new image sample automatic selection method and system |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080009966A1 (en) * | 2006-07-05 | 2008-01-10 | Battelle Energy Alliance, Llc | Occupancy Change Detection System and Method |
| CN101510374A (en) * | 2009-03-12 | 2009-08-19 | 中国科学院遥感应用研究所 | Method for automatically updating land-use map based on historical data and remote sensing data |
| CN101644572A (en) * | 2009-06-24 | 2010-02-10 | 中国科学院地理科学与资源研究所 | Detection method of ocean eddy variation based on historical similarity cases |
| CN103500344A (en) * | 2013-09-02 | 2014-01-08 | 中国测绘科学研究院 | Method and module for extracting and interpreting information of remote-sensing image |
| CN104751478A (en) * | 2015-04-20 | 2015-07-01 | 武汉大学 | Object-oriented building change detection method based on multi-feature fusion |
| CN104834006A (en) * | 2015-04-28 | 2015-08-12 | 中国地质大学（北京） | Processing method and apparatus of structural cause deposition parameter diagram |
| CN105225227A (en) * | 2015-09-07 | 2016-01-06 | 中国测绘科学研究院 | The method and system that remote sensing image change detects |
| CN108846832A (en) * | 2018-05-30 | 2018-11-20 | 理大产学研基地(深圳)有限公司 | A kind of change detecting method and system based on multi-temporal remote sensing image and GIS data |
| CN108986113A (en) * | 2018-07-06 | 2018-12-11 | 航天星图科技（北京）有限公司 | A kind of block parallel multi-scale division algorithm based on LLTS frame |
-
2019
- 2019-07-10 CN CN201910617812.3A patent/CN110473205A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080009966A1 (en) * | 2006-07-05 | 2008-01-10 | Battelle Energy Alliance, Llc | Occupancy Change Detection System and Method |
| CN101510374A (en) * | 2009-03-12 | 2009-08-19 | 中国科学院遥感应用研究所 | Method for automatically updating land-use map based on historical data and remote sensing data |
| CN101644572A (en) * | 2009-06-24 | 2010-02-10 | 中国科学院地理科学与资源研究所 | Detection method of ocean eddy variation based on historical similarity cases |
| CN103500344A (en) * | 2013-09-02 | 2014-01-08 | 中国测绘科学研究院 | Method and module for extracting and interpreting information of remote-sensing image |
| CN104751478A (en) * | 2015-04-20 | 2015-07-01 | 武汉大学 | Object-oriented building change detection method based on multi-feature fusion |
| CN104834006A (en) * | 2015-04-28 | 2015-08-12 | 中国地质大学（北京） | Processing method and apparatus of structural cause deposition parameter diagram |
| CN105225227A (en) * | 2015-09-07 | 2016-01-06 | 中国测绘科学研究院 | The method and system that remote sensing image change detects |
| CN108846832A (en) * | 2018-05-30 | 2018-11-20 | 理大产学研基地(深圳)有限公司 | A kind of change detecting method and system based on multi-temporal remote sensing image and GIS data |
| CN108986113A (en) * | 2018-07-06 | 2018-12-11 | 航天星图科技（北京）有限公司 | A kind of block parallel multi-scale division algorithm based on LLTS frame |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112800992A (en) * | 2021-02-03 | 2021-05-14 | 成都鼎安华智慧物联网股份有限公司 | Extraction method based on remote sensing data high-temperature abnormal information |
| CN112945352A (en) * | 2021-02-03 | 2021-06-11 | 成都鼎安华智慧物联网股份有限公司 | Extraction method based on remote sensing data water level abnormal information |
| CN113240026A (en) * | 2021-05-24 | 2021-08-10 | 中国科学院重庆绿色智能技术研究院 | Vector-grid combined inland water surface floater batch identification and extraction method |
| CN113240026B (en) * | 2021-05-24 | 2022-03-25 | 中国科学院重庆绿色智能技术研究院 | Vector-grid combined inland water surface floater batch identification and extraction method |
| CN113822900A (en) * | 2021-07-09 | 2021-12-21 | 武汉大学 | Vector constraint object-oriented new image sample automatic selection method and system |
| CN113822900B (en) * | 2021-07-09 | 2023-09-22 | 武汉大学 | Method and system for automatically selecting new image sample based on vector constraint object-oriented |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110473205A (en) | Remote sensing image information extracting method and system based on arrow bar phantom | |
| CN108830870B (en) | Satellite image high-precision farmland boundary extraction method based on multi-scale structure learning | |
| CN106384081B (en) | Slope farmland extraction method and system based on high-resolution remote sensing image | |
| Jasiewicz et al. | Landscape similarity, retrieval, and machine mapping of physiographic units | |
| Blaschke et al. | Image segmentation methods for object-based analysis and classification | |
| CN107067405B (en) | Remote sensing image segmentation method based on scale optimization | |
| CN112381013B (en) | Urban vegetation inversion method and system based on high-resolution remote sensing image | |
| CN108428220B (en) | Automatic geometric correction method for ocean island reef area of remote sensing image of geostationary orbit satellite sequence | |
| CN105761238B (en) | A method of passing through gray-scale statistical data depth information extraction well-marked target | |
| CN109784192B (en) | Hyperspectral image classification method based on superpixel feature extraction neural network algorithm | |
| CN105608474A (en) | High-resolution-image-based regional adaptive cultivated land extraction method | |
| CN106446875A (en) | County remote-sensing scale-oriented crop planting area information extraction method and device | |
| CN104408733B (en) | Object random walk-based visual saliency detection method and system for remote sensing image | |
| CN102254326A (en) | Image segmentation method by using nucleus transmission | |
| CN102663757A (en) | Semi-automatic image cutting method based on nuclear transfer | |
| CN110796038A (en) | Hyperspectral remote sensing image classification method combined with rapid region growing superpixel segmentation | |
| CN112508936A (en) | Remote sensing image change detection method based on deep learning | |
| CN115272860B (en) | Determination method and system for rice planting area, electronic equipment and storage medium | |
| CN110598564A (en) | OpenStreetMap-based high-spatial-resolution remote sensing image transfer learning classification method | |
| CN111222539A (en) | Method for optimizing and expanding supervision classification samples based on multi-source multi-temporal remote sensing image | |
| He et al. | A Method of Identifying Thunderstorm Clouds in Satellite Cloud Image Based on Clustering. | |
| Wang et al. | Multispectral point cloud superpoint segmentation | |
| Engstrom et al. | Evaluating the Relationship between Contextual Features Derived from Very High Spatial Resolution Imagery and Urban Attributes: A Case Study in Sri Lanka | |
| CN117636160A (en) | Automatic high-resolution remote sensing cultivated land block updating method based on semi-supervised learning | |
| CN109299295B (en) | Blue printing layout database searching method |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191119 |
|
| RJ01 | Rejection of invention patent application after publication |