CN112801978A - Multispectral remote sensing image change detection method and device and storage medium - Google Patents
Multispectral remote sensing image change detection method and device and storage medium Download PDFInfo
- Publication number
- CN112801978A CN112801978A CN202110119376.4A CN202110119376A CN112801978A CN 112801978 A CN112801978 A CN 112801978A CN 202110119376 A CN202110119376 A CN 202110119376A CN 112801978 A CN112801978 A CN 112801978A
- Authority
- CN
- China
- Prior art keywords
- image
- band
- matrix
- multispectral
- waveband
- 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
- 230000008859 change Effects 0.000 title claims abstract description 105
- 238000001514 detection method Methods 0.000 title claims abstract description 63
- 238000012545 processing Methods 0.000 claims abstract description 35
- 238000004458 analytical method Methods 0.000 claims abstract description 32
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 19
- 238000000605 extraction Methods 0.000 claims abstract description 14
- 238000007499 fusion processing Methods 0.000 claims abstract description 12
- 238000007621 cluster analysis Methods 0.000 claims abstract description 7
- 230000004927 fusion Effects 0.000 claims abstract description 6
- 239000011159 matrix material Substances 0.000 claims description 103
- 238000000034 method Methods 0.000 claims description 30
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 5
- 238000003384 imaging method Methods 0.000 description 8
- 230000005855 radiation Effects 0.000 description 8
- 230000006870 function Effects 0.000 description 6
- 238000000926 separation method Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000012937 correction Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 239000000126 substance Substances 0.000 description 3
- 238000001914 filtration Methods 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000003064 k means clustering 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
- 239000004065 semiconductor Substances 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 230000017105 transposition Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
- G06T2207/10036—Multispectral image; Hyperspectral image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Abstract
The invention discloses a multispectral remote sensing image change detection method, a multispectral remote sensing image change detection device and a storage medium, wherein the multispectral remote sensing image change detection method comprises the following steps: performing single-band image extraction processing on the first multispectral image and the second multispectral image to obtain a plurality of first single-band images corresponding to the first multispectral image and a plurality of second single-band images corresponding to the second multispectral image; performing single-band slow feature analysis processing on a first single-band image and a second single-band image corresponding to each band to sufficiently suppress the difference of unchanged pixels in each band image so as to obtain an optimal feature difference map of each band containing more real change information; performing fusion processing on the optimal feature difference images to obtain variation intensity images corresponding to the optimal feature difference images; and carrying out binary clustering analysis on the change intensity image according to a preset clustering algorithm to obtain a change detection result. The device comprises: the device comprises a single-waveband image extraction processing module, a slow feature analysis module, a fusion module and a binary cluster analysis module.
Description
Technical Field
The invention relates to the field of multispectral remote sensing images, in particular to a multispectral remote sensing image change detection method, a multispectral remote sensing image change detection device and a storage medium.
Background
In recent years, remote sensing image change detection technology has been applied to many fields, and by detecting changes of a plurality of remote sensing images corresponding to different time phases (usually, the same month in different years) in the same region, the change characteristics and the change process of all ground features in the region can be known. However, due to more or less deviation in the imaging system, "pseudo-variation" is easy to occur in the multispectral remote sensing variation detection, so that the detection result is often not satisfactory. The most important errors are caused by the radiation difference of the same ground object caused by different external imaging conditions, and include: atmospheric and radiation conditions, sun angle, sensor calibration, soil moisture, etc. This results in that the same feature, if unchanged, will reflect different spectral values in the remote sensing images of different phases.
For this problem, at present, when two multispectral images corresponding to different regions are subjected to change detection, a slow feature analysis algorithm is usually used for the change detection of the two multispectral images. However, when the slow feature analysis algorithm is used for performing change detection, it cannot be guaranteed that each wave band can obtain a feature difference map with the minimum radiation error, so that real changes and 'pseudo changes' cannot be effectively separated by using the algorithm, and the problems of plaque breakage, difficulty in identifying slight changes, high undetected false detection rate and the like still exist.
Therefore, how to effectively reduce the "pseudo-variation" caused by the imaging conditions and construct a high-quality difference map to improve the separation degree between the real variation and the "pseudo-variation" and further improve the variation detection accuracy is a problem to be solved urgently for the variation detection.
Disclosure of Invention
Aiming at the problems that the change detection of the multispectral remote sensing image is easy to have plaque breakage and the fine change is difficult to identify, so that the accuracy rate of the detection result is low, the invention provides a method, a device and a storage medium for detecting the change of the multispectral remote sensing image, the internal information of an unchanged area in two single-band images corresponding to each band is fully mined by the optimal characteristic difference image of each band obtained by the processing of the invention, the separation degree between real change and 'pseudo change' is effectively improved, and the change detection precision is greatly improved; the invention is little affected by radiation difference, time cost is low, the current situation that subtle changes are difficult to accurately identify is improved, and the detection result has higher precision, which is described in detail in the following:
in a first aspect, a method for detecting changes in a multispectral remote sensing image, the method comprising:
performing single-band image extraction processing on the first multispectral image and the second multispectral image to obtain a plurality of first single-band images corresponding to the first multispectral image and a plurality of second single-band images corresponding to the second multispectral image;
performing single-band slow feature analysis processing on a first single-band image and a second single-band image corresponding to each band to sufficiently suppress the difference of unchanged pixels in each band image so as to obtain an optimal feature difference map of each band containing more real change information;
performing fusion processing on the optimal feature difference images to obtain variation intensity images corresponding to the optimal feature difference images;
and carrying out binary clustering analysis on the change intensity image according to a preset clustering algorithm to obtain a change detection result.
In one implementation, the performing slow single-band feature analysis on the first single-band image and the second single-band image corresponding to each band specifically includes:
respectively generating a first image matrix and a second image matrix according to a first single-band image and a second single-band image corresponding to the ith band; respectively calculating variance matrixes corresponding to the first image matrix and the second image matrix and a variance matrix of the difference value of the first image matrix and the second image matrix; all three variance matrixes are substituted into a first preset formula, and the ith waveband projection matrix is calculated and obtained;
calculating to obtain an ith wave band optimal characteristic difference value matrix according to the first image matrix, the second image matrix and the ith wave band projection matrix and a second preset formula;
and obtaining an ith wave band optimal characteristic difference value map according to the ith wave band optimal characteristic difference value matrix.
In a second aspect, a multispectral remote sensing image change detection apparatus, the apparatus comprising:
the single-waveband image extraction processing module is used for carrying out single-waveband image extraction processing on the first multispectral image and the second multispectral image so as to obtain a plurality of first single-waveband images corresponding to the first multispectral image and a plurality of second single-waveband images corresponding to the second multispectral image;
the slow feature analysis module is used for carrying out single-band slow feature analysis processing on the first single-band image and the second single-band image corresponding to each band so as to fully inhibit the difference of unchanged pixels in each band image and obtain an optimal feature difference map of each band containing more real change information;
the fusion module is used for carrying out fusion processing on the optimal feature difference images to obtain variation intensity images corresponding to the optimal feature difference images;
and the binary cluster analysis module is used for carrying out binary cluster analysis on the change intensity image according to a preset cluster algorithm so as to obtain a change detection result.
In a third aspect, a multispectral remote sensing image change detection apparatus comprises: a processor and a memory, the memory having stored therein program instructions, the processor calling the program instructions stored in the memory to cause the apparatus to perform the method steps of the first aspect.
In a fourth aspect, a computer-readable storage medium, storing a computer program comprising program instructions which, when executed by a processor, cause the processor to carry out the method steps of any of the first aspects.
The technical scheme provided by the invention has the beneficial effects that:
1. compared with the prior art that a slow characteristic analysis algorithm is adopted to carry out change detection on two multispectral images which are different and corresponding to the same region, the method starts from the angle of minimizing the radiation difference of the unchanged region in the two single-band images corresponding to each band, can fully excavate the internal information of the unchanged region in the two single-band images corresponding to each band, and effectively inhibits the radiation difference caused by different external imaging conditions, so that the separability between a changed pixel and an unchanged pixel is improved, and the change detection precision is effectively improved;
2. because the ground features have different reflection spectra in different wave bands, and each wave band contains certain change information when the ground features change, the optimal characteristic difference graph and the like of each wave band obtained by processing are fused into a change intensity graph in an equal weight mode and are subjected to Gaussian filtering processing, so that the change information contained in each wave band is fully utilized, pseudo change caused by imaging conditions is further eliminated, the current situation that fine change is difficult to accurately identify is improved, and the method has great application potential.
Drawings
FIG. 1 is a flow chart of a method for detecting changes in a multi-spectral remote sensing image;
FIG. 2 is another flow chart of a method for detecting changes in a multi-spectral remote sensing image;
FIG. 3 is a first multi-spectral image to be detected;
FIG. 4 is a second multispectral image to be detected;
FIG. 5 is a modified reference diagram;
FIG. 6 is a graph showing the results of change detection;
FIG. 7 is a schematic structural diagram of a multispectral remote sensing image change detection device;
FIG. 8 is a schematic structural diagram of a slow feature analysis module;
fig. 9 is another structural diagram of the multispectral remote sensing image change detection device.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention are described in further detail below.
The embodiment of the invention provides a multispectral remote sensing image change detection method, as shown in figure 1, the method comprises the following steps:
101: acquiring a first multispectral image and a second multispectral image;
in an embodiment of the present invention, the first multispectral image and the second multispectral image are multispectral images that do not correspond to the same region at the same time.
102: performing single-band image extraction processing on the first multispectral image and the second multispectral image to obtain a plurality of first single-band images corresponding to the first multispectral image and a plurality of second single-band images corresponding to the second multispectral image;
in the embodiment of the present invention, since the multispectral image is an image synthesized from a plurality of single-band images of different bands, after the first multispectral image and the second multispectral image are obtained, the single-band image extraction processing may be performed on the first multispectral image and the second multispectral image, so as to obtain a plurality of first single-band images corresponding to the first multispectral image and a plurality of second single-band images corresponding to the second multispectral image, that is, a first single-band image corresponding to each band is extracted from the first multispectral image and a second single-band image corresponding to each band is extracted from the second multispectral image. And (3) performing centralization and standardization operation on the extracted single-band image, and processing to obtain single-band image data with an average value of 0 (namely centralization operation) and a standard deviation of 1 (namely standardization operation). The operation can simplify subsequent calculation and reduce radiation difference to a certain extent, and provides convenience for subsequent generation of a high-quality change difference map.
103: performing single-band slow feature analysis processing on a first single-band image and a second single-band image corresponding to each band to sufficiently suppress the difference of unchanged pixels in each band image so as to obtain an optimal feature difference map of each band containing more real change information;
104: performing fusion processing on the optimal feature difference images to obtain variation intensity images corresponding to the optimal feature difference images;
in the embodiment of the invention, in order to further eliminate the 'pseudo change' caused by the imaging condition and improve the change detection precision, a plurality of obtained optimal characteristic difference images can be subjected to equal weight fusion to generate a change intensity image, so that the aim of obtaining a high-quality change difference image by fully utilizing the change information contained in each wave band is achieved.
105: and carrying out binary clustering analysis on the change intensity image according to a preset clustering algorithm to obtain a change detection result.
The preset clustering algorithm may be specifically a K-means clustering algorithm.
In summary, the embodiment of the present invention effectively improves the separation degree between the real change and the "pseudo change" through the steps 101-105, so as to greatly improve the change detection precision.
In the following, with reference to fig. 2 and a specific calculation formula, a multispectral remote sensing image change detection method in the above embodiment is detailed and expanded, and the method includes the following steps:
201: acquiring a first multispectral image and a second multispectral image;
the multispectral image is an image synthesized by a plurality of single-waveband images of different wavebands.
202: performing geometric correction and registration processing on the first multispectral image and the second multispectral image;
in the embodiment of the present invention, in order to ensure that a plurality of pixel points included in the first multispectral image correspond to a plurality of pixel points included in the second multispectral image one to one, after the first multispectral image and the second multispectral image are obtained, geometric correction and registration processing need to be performed on the first multispectral image and the second multispectral image.
Specifically, in this step, the remote sensing image processing software ENVI may be used to perform geometric correction and registration processing on the first multispectral image and the second multispectral image, but is not limited thereto.
203: performing single-band image extraction processing on the first multispectral image and the second multispectral image to obtain a plurality of first single-band images corresponding to the first multispectral image and a plurality of second single-band images corresponding to the second multispectral image;
204: carrying out image centralization and standardization processing on the first single-waveband image and the second single-waveband image corresponding to each waveband;
in the embodiment of the invention, in order to further eliminate the 'pseudo change' caused by the imaging condition and improve the change detection precision, the first single-waveband image and the second single-waveband image corresponding to each waveband can be subjected to image centralization and standardization processing, so that convenience is brought to the subsequent generation of a high-quality change difference map.
205: performing single-band slow feature analysis processing on a first single-band image and a second single-band image corresponding to each band to sufficiently suppress the difference of unchanged pixels in each band image so as to obtain an optimal feature difference map of each band containing more real change information;
in the embodiment of the present invention, the key of the single-band slow feature analysis algorithm is to minimize the variance of the feature difference image of each band as an optimization target, so as to project the first single-band image and the second single-band image corresponding to each band to an optimal feature space, and define a target function as:
wherein P is the total pixel number of the single-band image, wiIn order to project a first single-waveband image and a second single-waveband image corresponding to the ith waveband to the optimal feature space optimal projection matrix:
wherein the content of the first and second substances,
extracting the kth pixel value in the ith single-band image (namely the kth pixel value of the first single-band image corresponding to the ith band) from the first multispectral image,
extracting the kth pixel value in the ith single-band image (namely the kth pixel value of the second single-band image corresponding to the ith band) from the second multispectral image, wherein T is a transposition,
is a variance matrix of a difference image obtained according to a first single-band image and a second single-band image corresponding to the ith band, AiThe variance matrix of the difference image is obtained according to the first single-band image and the second single-band image corresponding to the ith band.
The objective function may also include the following constraints: the first is zero mean value constraint, which is to simplify algorithm calculation and improve operation speed, and the second is unit variance constraint, which is mainly to avoid constant solution.
Wherein the content of the first and second substances,
wherein the content of the first and second substances,
a variance matrix of the first one-band image corresponding to the ith band,
and the variance matrix of the second single-waveband image corresponding to the ith waveband.
Specifically, in this step, the process of performing single-band slow feature analysis processing on the first single-band image and the second single-band image corresponding to each band is as follows:
2051: generating a first image matrix corresponding to the first single-waveband image according to the first single-waveband image and the second single-waveband image corresponding to the ith waveband
A second image matrix corresponding to the second single-band image
2052: according to a first image matrix xiAnd a second image matrix yiSeparately calculating a first image matrix xiCorresponding variance matrix
Second image matrix yiCorresponding variance matrix
First image matrix xiAnd a second image matrix yiVariance matrix of difference values
2053: the variance matrix
Variance matrix
Sum variance matrix
Substituting the first preset formula into the first preset formula to calculate and obtain the ith wave band projection matrix wiWherein, in the step (A),
the first preset formula is a generalized characteristic equation:
and Λ is a diagonal matrix of the generalized eigenvalue, and the value of the diagonal matrix represents the variance of the characteristic difference image corresponding to the ith waveband.
2054: according to a first image matrix xiAnd a second image matrix yiAnd the calculated projection matrix w of the ith wave bandiAccording to a second preset formula, calculating to obtain an i-th wave band optimal characteristic difference value matrix SFAiWherein, in the step (A),
the second predetermined formula is:
SFAi＝wixi-wiyi (7)
2055: according to the ith wave band optimal characteristic difference value matrix SFAiAnd obtaining the optimal characteristic difference map of the ith wave band.
206: performing fusion processing on the optimal feature difference images to obtain variation intensity images corresponding to the optimal feature difference images;
specifically, in this step 206, the process of performing fusion processing on the optimal feature difference images to obtain the variation intensity images corresponding to the optimal feature difference images is as follows:
2061: the obtained ith wave band optimal characteristic difference value matrix SFAiWriting the optimal characteristic difference matrix into a preset optimal characteristic difference matrix SFA, and obtaining a final optimal characteristic difference matrix SFA (SFA) after the writing of the optimal characteristic difference matrixes of all wave bands is finished1,SFA2，...SFAN]；
2062: performing equal-weight fusion processing on the optimal characteristic difference matrix SFA to obtain a plurality of change intensity matrixes corresponding to the optimal characteristic difference matrixes;
wherein, the equal-weight fusion processing may include: and performing equal weight fusion by adopting an Euclidean distance formula, namely calculating by using a third preset formula to obtain a change intensity matrix.
The third preset formula is:
2063: and obtaining a change intensity image according to the change intensity matrix, and performing Gaussian filtering on the change intensity image to further reduce pseudo change caused by an imaging system error so as to improve the change detection precision.
207: and carrying out binary clustering analysis on the change intensity image according to a preset clustering algorithm to obtain a change detection result.
In summary, the embodiment of the present invention effectively improves the separation degree between the real change and the "pseudo change" through the steps 201-207, so as to greatly improve the change detection precision.
The feasibility of the above method section was verified in combination with specific experimental data, as described in detail below:
because multispectral images acquired in different years in the same region are difficult to obtain accurate change reference images, the adoption of a simulated change detection data set to carry out subjective and objective index analysis and verification on a change detection method is more convenient. The first image used in the experiment was a multispectral image taken by landsat5-TM on day 9, 13 of 2007 as the first multispectral image to be detected, and fig. 3 shows the first multispectral image; the second image is a multispectral image (only having radiation difference with the first multispectral image to be detected and no change in terrain and features) shot by landsdat 5-TM at 29.9.2007, and a change region is manually added to the multispectral image, and the multispectral image after the change region is manually added is taken as the second multispectral image to be detected, the second multispectral image is given in fig. 4, fig. 5 is a change reference image, fig. 6 is a change detection result obtained by implementing the method of the present invention, and objective indexes specifically describing the detection result can be shown in the following table. The objective indexes comprise missed detection number, false detection number, total error number, accuracy rate, KAPPA value and time overhead. The KAPPA value is a parameter that can measure the classification accuracy more accurately, and is usually used to measure the similarity between the variation result graph and the reference image, and the ideal value is 1, which means that the detection result is completely consistent with the reference image.
TABLE 1
In order to further verify the universality of the method, 50 sets of landsat5-TM data sets are adopted to compare the method with various advanced algorithms, and the average values of the missed detection number, the false detection number, the total false detection number, the accuracy, the KAPPA value and the time overhead of the 50 sets of data sets are obtained as follows:
TABLE 2
The embodiment of the invention provides a change detection method of a multispectral remote sensing image, which has certain superiority in detection precision compared with the change detection of two multispectral images which are different and correspond to the same region by adopting various advanced algorithms in the prior art. The embodiment of the invention performs single-band slow feature analysis processing on the first single-band image and the second single-band image corresponding to each band from the viewpoint of minimizing the spectral change of the unchanged region of each band, so that the intrinsic information of the unchanged region in the two single-band images corresponding to each band can be more fully mined, the separation degree of real change and 'pseudo change' is improved, the optimal feature difference image corresponding to each band is obtained, a change intensity difference image with higher quality is further constructed, and a detection result with higher change detection precision is obtained.
Based on the same inventive concept, as an implementation of the above method, referring to fig. 7, an embodiment of the present invention further provides a multispectral remote sensing image change detection apparatus, including:
the single-waveband image extraction processing module 1 is used for performing single-waveband image extraction processing on the first multispectral image and the second multispectral image to obtain a plurality of first single-waveband images corresponding to the first multispectral image and a plurality of second single-waveband images corresponding to the second multispectral image;
the slow feature analysis module 2 is configured to perform single-band slow feature analysis processing on the first single-band image and the second single-band image corresponding to each band, so as to sufficiently suppress the difference of unchanged pixels in each band image, and obtain an optimal feature difference map of each band that contains more real change information;
the fusion module 3 is used for performing fusion processing on the optimal feature difference images to obtain variation intensity images corresponding to the optimal feature difference images;
and the binary cluster analysis module 4 is used for carrying out binary cluster analysis on the change intensity image according to a preset clustering algorithm so as to obtain a change detection result.
In one implementation, referring to fig. 8, the slow signature analysis module 2 includes:
the generating submodule 21 is configured to generate a first image matrix and a second image matrix according to a first single-band image and a second single-band image corresponding to an ith band;
the first calculating submodule 22 is configured to calculate a variance matrix corresponding to the first image matrix and a variance matrix corresponding to a difference between the first image matrix and the second image matrix; all three variance matrixes are substituted into a first preset formula, and the ith waveband projection matrix is calculated and obtained;
the second calculation submodule 23 is configured to calculate an ith wave band optimal characteristic difference matrix according to the first and second image matrices and the ith wave band projection matrix and according to a second preset formula;
and the obtaining submodule 24 is configured to obtain an i-th waveband optimal feature difference map according to the i-th waveband optimal feature difference matrix.
It should be noted that the device description in the above embodiments corresponds to the description of the method embodiments, and the embodiments of the present invention are not described herein again.
The execution main bodies of the modules and units can be devices with calculation functions, such as a computer, a single chip microcomputer and a microcontroller, and in the specific implementation, the execution main bodies are not limited in the embodiment of the invention and are selected according to the requirements in practical application.
Based on the same inventive concept, the embodiment of the present invention further provides a multispectral remote sensing image change detection device, referring to fig. 9, the device includes: a processor 5 and a memory 6, the memory 6 having stored therein program instructions, the processor 5 calling upon the program instructions stored in the memory 6 to cause the apparatus to perform the following method steps in an embodiment:
performing single-band image extraction processing on the first multispectral image and the second multispectral image to obtain a plurality of first single-band images corresponding to the first multispectral image and a plurality of second single-band images corresponding to the second multispectral image;
performing single-band slow feature analysis processing on a first single-band image and a second single-band image corresponding to each band to sufficiently suppress the difference of unchanged pixels in each band image so as to obtain an optimal feature difference map of each band containing more real change information;
performing fusion processing on the optimal feature difference images to obtain variation intensity images corresponding to the optimal feature difference images;
and carrying out binary clustering analysis on the change intensity image according to a preset clustering algorithm to obtain a change detection result.
In one implementation, the performing slow single-band feature analysis on the first single-band image and the second single-band image corresponding to each band specifically includes:
respectively generating a first image matrix and a second image matrix according to a first single-band image and a second single-band image corresponding to the ith band; respectively calculating variance matrixes corresponding to the first image matrix and the second image matrix and a variance matrix of the difference value of the first image matrix and the second image matrix; all three variance matrixes are substituted into a first preset formula, and the ith waveband projection matrix is calculated and obtained;
calculating to obtain an ith wave band optimal characteristic difference value matrix according to the first image matrix, the second image matrix and the ith wave band projection matrix and a second preset formula;
and obtaining an ith wave band optimal characteristic difference value map according to the ith wave band optimal characteristic difference value matrix.
It should be noted that the device description in the above embodiments corresponds to the method description in the embodiments, and the embodiments of the present invention are not described herein again.
The execution main bodies of the processor and the memory can be devices with calculation functions such as a computer, a single chip microcomputer and a microcontroller, and in the specific implementation, the execution main bodies are not limited in the embodiment of the invention and are selected according to the requirements in practical application.
The memory 6 and the processor 5 transmit data signals through the bus 7, which is not described in detail in the embodiment of the present invention.
Based on the same inventive concept, an embodiment of the present invention further provides a computer-readable storage medium, where the storage medium includes a stored program, and when the program runs, the apparatus on which the storage medium is located is controlled to execute the method steps in the foregoing embodiments.
The computer readable storage medium includes, but is not limited to, flash memory, hard disk, solid state disk, and the like.
It should be noted that the descriptions of the readable storage medium in the above embodiments correspond to the descriptions of the method in the embodiments, and the descriptions of the embodiments of the present invention are not repeated here.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. The procedures or functions according to the embodiments of the invention are brought about in whole or in part when the computer program instructions are loaded and executed on a computer.
The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored on or transmitted over a computer-readable storage medium. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium or a semiconductor medium, etc.
In the embodiment of the present invention, except for the specific description of the model of each device, the model of other devices is not limited, as long as the device can perform the above functions.
Those skilled in the art will appreciate that the drawings are only schematic illustrations of preferred embodiments, and the above-described embodiments of the present invention are merely provided for description and do not represent the merits of the embodiments.
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 (8)
1. A multispectral remote sensing image change detection method is characterized by comprising the following steps:
performing single-band image extraction processing on the first multispectral image and the second multispectral image to obtain a plurality of first single-band images corresponding to the first multispectral image and a plurality of second single-band images corresponding to the second multispectral image;
performing single-band slow feature analysis processing on a first single-band image and a second single-band image corresponding to each band to sufficiently suppress the difference of unchanged pixels in each band image so as to obtain an optimal feature difference map of each band containing more real change information;
performing fusion processing on the optimal feature difference images to obtain variation intensity images corresponding to the optimal feature difference images;
and carrying out binary clustering analysis on the change intensity image according to a preset clustering algorithm to obtain a change detection result.
2. The method for detecting the change of the multispectral remote sensing image according to claim 1, wherein the slow single-band feature analysis processing on the first single-band image and the second single-band image corresponding to each band specifically comprises the following steps:
respectively generating a first image matrix and a second image matrix according to a first single-band image and a second single-band image corresponding to the ith band; respectively calculating variance matrixes corresponding to the first image matrix and the second image matrix and a variance matrix of the difference value of the first image matrix and the second image matrix; all three variance matrixes are substituted into a first preset formula, and the ith waveband projection matrix is calculated and obtained;
calculating to obtain an ith wave band optimal characteristic difference value matrix according to the first image matrix, the second image matrix and the ith wave band projection matrix and a second preset formula;
and obtaining an ith wave band optimal characteristic difference value map according to the ith wave band optimal characteristic difference value matrix.
3. The method for detecting the change of the multispectral remote sensing image according to claim 2, wherein the first preset formula is a generalized characteristic equation:
wherein, Λ is a diagonal matrix of the generalized eigenvalue, the value of which characterizes the variance of the characteristic difference image corresponding to the ith waveband,
are all variance matrices; w is aiIs a projection matrix.
4. The method for detecting changes in the multispectral remote sensing image according to claim 2, wherein the second predetermined formula is:
SFAi＝wixi-wiyi
wherein, SFAiFor the optimal feature difference matrix, xiIs a first image matrix, yiIs a second image matrix.
5. A multi-spectral remote sensing image change detection apparatus, comprising:
the single-waveband image extraction processing module is used for carrying out single-waveband image extraction processing on the first multispectral image and the second multispectral image so as to obtain a plurality of first single-waveband images corresponding to the first multispectral image and a plurality of second single-waveband images corresponding to the second multispectral image;
the slow feature analysis module is used for carrying out single-band slow feature analysis processing on the first single-band image and the second single-band image corresponding to each band so as to fully inhibit the difference of unchanged pixels in each band image and obtain an optimal feature difference map of each band containing more real change information;
the fusion module is used for carrying out fusion processing on the optimal feature difference images to obtain variation intensity images corresponding to the optimal feature difference images;
and the binary cluster analysis module is used for carrying out binary cluster analysis on the change intensity image according to a preset cluster algorithm so as to obtain a change detection result.
6. The device according to claim 5, wherein the slow feature analysis module comprises:
the generation submodule is used for respectively generating a first image matrix and a second image matrix according to a first single-waveband image and a second single-waveband image corresponding to the ith waveband;
the first calculation submodule is used for calculating a variance matrix corresponding to the first image matrix and the second image matrix and a variance matrix of the difference value of the first image matrix and the second image matrix respectively; all three variance matrixes are substituted into a first preset formula, and the ith waveband projection matrix is calculated and obtained;
the second calculation submodule is used for calculating to obtain an ith wave band optimal characteristic difference value matrix according to the first image matrix, the second image matrix and the ith wave band projection matrix and a second preset formula;
and the obtaining submodule is used for obtaining an ith wave band optimal characteristic difference map according to the ith wave band optimal characteristic difference matrix.
7. A multi-spectral remote sensing image change detection apparatus, the apparatus comprising: a processor and a memory, the memory having stored therein program instructions, the processor calling upon the program instructions stored in the memory to cause the apparatus to perform the method steps of any of claims 1-4.
8. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program comprising program instructions which, when executed by a processor, cause the processor to carry out the method steps of any of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110119376.4A CN112801978A (en) | 2021-01-28 | 2021-01-28 | Multispectral remote sensing image change detection method and device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110119376.4A CN112801978A (en) | 2021-01-28 | 2021-01-28 | Multispectral remote sensing image change detection method and device and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112801978A true CN112801978A (en) | 2021-05-14 |
Family
ID=75812502
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110119376.4A Pending CN112801978A (en) | 2021-01-28 | 2021-01-28 | Multispectral remote sensing image change detection method and device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112801978A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130230250A1 (en) * | 2012-03-01 | 2013-09-05 | Harris Corporation | Systems and methods for efficient feature based image and video analysis |
| CN103632155A (en) * | 2013-12-16 | 2014-03-12 | 武汉大学 | Remote-sensing image variation detecting method based on slow characteristic analysis |
| CN108830828A (en) * | 2018-04-28 | 2018-11-16 | 新疆大学 | A kind of method for detecting change of remote sensing image and device |
| CN109903246A (en) * | 2019-02-22 | 2019-06-18 | 新疆大学 | The method and device of detection image variation |
| CN111208793A (en) * | 2020-02-20 | 2020-05-29 | 浙江大学 | State monitoring method of non-stationary industrial process based on slow characteristic analysis |
| CN112084837A (en) * | 2020-07-13 | 2020-12-15 | 江南大学 | Remote sensing image change detection method and system based on depth network |
-
2021
- 2021-01-28 CN CN202110119376.4A patent/CN112801978A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130230250A1 (en) * | 2012-03-01 | 2013-09-05 | Harris Corporation | Systems and methods for efficient feature based image and video analysis |
| CN103632155A (en) * | 2013-12-16 | 2014-03-12 | 武汉大学 | Remote-sensing image variation detecting method based on slow characteristic analysis |
| CN108830828A (en) * | 2018-04-28 | 2018-11-16 | 新疆大学 | A kind of method for detecting change of remote sensing image and device |
| CN109903246A (en) * | 2019-02-22 | 2019-06-18 | 新疆大学 | The method and device of detection image variation |
| CN111208793A (en) * | 2020-02-20 | 2020-05-29 | 浙江大学 | State monitoring method of non-stationary industrial process based on slow characteristic analysis |
| CN112084837A (en) * | 2020-07-13 | 2020-12-15 | 江南大学 | Remote sensing image change detection method and system based on depth network |
Non-Patent Citations (1)
| Title |
|---|
| 徐俊峰: "多特征融合的遥感影像变化检测技术研究", 《中国优秀硕士学位论文全文数据库基础科学辑》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108932496B (en) | Method and device for counting number of target objects in area | |
| Chang et al. | Remote sensing image registration based on modified SIFT and feature slope grouping | |
| WO2014151035A1 (en) | Computer-based method and system of dynamic category object recognition | |
| CN111369605B (en) | Infrared and visible light image registration method and system based on edge features | |
| WO2018192023A1 (en) | Method and device for hyperspectral remote sensing image classification | |
| US10007678B2 (en) | Image processing apparatus, image processing method, and recording medium | |
| CN107240130B (en) | Remote sensing image registration method, device and system | |
| CN110738222B (en) | Image matching method and device, computer equipment and storage medium | |
| CN113269201A (en) | Hyperspectral image band selection method and system based on potential feature fusion | |
| Jhan et al. | A generalized tool for accurate and efficient image registration of UAV multi-lens multispectral cameras by N-SURF matching | |
| CN114897676A (en) | Unmanned aerial vehicle remote sensing multispectral image splicing method, device and medium | |
| CN108921884A (en) | Based on the optics and SAR Image registration method, equipment and storage medium for improving SIFT | |
| CN112801978A (en) | Multispectral remote sensing image change detection method and device and storage medium | |
| CN115564808B (en) | Multi-resolution hyperspectral/SAR image registration method based on public space-spectrum subspace | |
| CN116109535A (en) | Image fusion method, device and computer readable storage medium | |
| CN115511928A (en) | Matching method of multispectral image | |
| CN111076817B (en) | Abnormal target detection method based on optical multi-dimensional information integrated perception system | |
| CN112329654B (en) | Hyperspectral image data classification method and system based on multi-manifold learning algorithm | |
| CN112184776A (en) | Target tracking method, device and storage medium | |
| CN114241022A (en) | Unmanned aerial vehicle image automatic registration method and system | |
| WO2018076137A1 (en) | Method and device for obtaining hyper-spectral image feature descriptor | |
| CN106682672A (en) | Method and device for acquiring feature descriptor of hyper-spectral image | |
| CN114463393A (en) | Image registration method, computer equipment and storage device | |
| CN113643369A (en) | Positioning information correction method for remote sensing image | |
| Li et al. | Reliable and fast mapping of keypoints on large-size remote sensing images by use of multiresolution and global information |
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: 20210514 |
|
| RJ01 | Rejection of invention patent application after publication |