CN111047528A - High-spectrum image unmixing method based on goblet sea squirt group - Google Patents
High-spectrum image unmixing method based on goblet sea squirt group Download PDFInfo
- Publication number
- CN111047528A CN111047528A CN201911181901.4A CN201911181901A CN111047528A CN 111047528 A CN111047528 A CN 111047528A CN 201911181901 A CN201911181901 A CN 201911181901A CN 111047528 A CN111047528 A CN 111047528A
- Authority
- CN
- China
- Prior art keywords
- goblet
- sea squirt
- individual
- goblet sea
- squirt group
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 241000251555 Tunicata Species 0.000 title claims abstract description 39
- 238000000034 method Methods 0.000 title claims abstract description 26
- 238000001228 spectrum Methods 0.000 title description 4
- 241000251557 Ascidiacea Species 0.000 claims description 8
- 239000000126 substance Substances 0.000 claims description 6
- 230000003044 adaptive effect Effects 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 4
- 230000003595 spectral effect Effects 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 230000008901 benefit Effects 0.000 abstract description 3
- 238000005457 optimization Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
Images
Classifications
-
- G06T5/94—
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N2021/1793—Remote sensing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
- Y02A40/10—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in agriculture
Abstract
The invention discloses a hyperspectral image unmixing method based on a goblet sea squirt group. Meanwhile, the goblet sea squirt group algorithm is adopted to solve the constructed objective function, and the goblet sea squirt group algorithm is utilized to solve the advantage of high accuracy, so that higher unmixing accuracy is obtained.
Description
Technical Field
The invention relates to the field of image processing, in particular to a hyperspectral image unmixing method based on a goblet sea squirt group.
Background
The remote sensing image can provide natural resource information and environment information, and the hyperspectral image unmixing technology is an important technology for analyzing the ground feature information in the remote sensing image. Due to the fact that ground surface objects in the natural environment are complex, the hyperspectral image acquired by the spectrometer is easily influenced by the atmosphere, and the hyperspectral image analysis process becomes very complex. In order to identify the ground surface and ground objects more accurately, the hyperspectral image unmixing technology becomes more important.
The goblet sea squirt group algorithm is a novel intelligent optimization algorithm, is simple and understandable in mechanism and easy to realize, and can be used for solving the optimization problem. A Shadow-based multiple linear Mixing Model (SMLM) is a general Model for the problem of hyperspectral unmixing, which not only retains the general attributes of the mixed spectrum, but also considers the Shadow effect in a real scene.
In order to obtain higher unmixing precision and accurately estimate ground feature proportion information, researchers at home and abroad propose a plurality of unmixing methods. The unmixing method based on the nonlinear model is mainly divided into a Bayes method and a gradient method, but the Bayes method is complex in calculation, low in unmixing efficiency, prone to falling into local extreme values in the gradient method and low in unmixing precision. Therefore, the hyperspectral image unmixing effect under the nonlinear model needs to be improved.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provide a hyperspectral image unmixing method based on a goblet sea squirt group. Meanwhile, the goblet sea squirt group algorithm is adopted to solve the constructed objective function, and the goblet sea squirt group algorithm is utilized to solve the advantage of high accuracy, so that higher unmixing accuracy is obtained.
The purpose of the invention is realized by the following technical scheme:
the hyperspectral image unmixing method based on the goblet sea squirt group comprises the following steps:
firstly, extracting image end members by adopting a geometric-based end member extraction algorithm VCA to obtain an end member spectral curve and an end member number R;
determining a search space dimension D and a position code by utilizing a hyperspectral image SMLM mixed model according to the number R of end members;
step three, setting the population scale, the iteration times and the upper and lower boundaries of the search space of the goblet sea squirt group;
step four, initializing the position of the individual goblet sea squirt group in the search space range;
step five, constructing an objective function, and calculating the current position X of each individual goblet sea squirt groupiAn adapted value for the objective function;
sixthly, the leader individual in the goblet sea squirt group moves according to the position updating formula of the leader, and the follower individual in the goblet sea squirt group moves according to the position updating formula of the follower;
step seven, calculating the individual adaptive value of each goblet ascidian after updating the position according to the objective function, if the individual adaptive value is less than the current position XiThe current position X is replaced by the updated positioniOtherwise, the current position X is still keptiThe change is not changed;
step nine, judging whether the current iteration times reach the preset iteration times or not, if so, ending the iteration, and outputting the position of the optimal individual in the current goblet sea squirt groupThereby obtaining abundance vectorNon-linear parameterAnd shadow weightsOtherwise, returning to the step six;
step ten, judging whether to perform unmixing on all pixel points of the hyperspectral image, and if so, finishing the calculation; otherwise, returning to the step four, and calculating the next pixel.
Further, the second step specifically comprises:
the expression of the hyperspectral image SMLM mixed model is as follows:
r is the number of end members, miSpectral curve representing the ith end-member, aiRepresenting the ith end-memberAbundance value of (a), abundance vector ai=[a1,…,aR]TSatisfying the constraint of nonnegativity sum, and the nonlinear parameter P is belonged to [0,1 ∈]The shadow weight Q ∈ [0,1 ]]。
Further, the fifth step specifically comprises: and constructing an objective function by using the reconstruction error, wherein the expression is as follows:
wherein, | | · | | represents a two-norm, abundanceThe non-negative sum must be satisfied as a constraint,representing data reconstructed using the SMLM model, y representing observed data, non-linear parametersShadow weights
Further, the sixth step specifically comprises:
the location update formula for the leader is as follows:
wherein the content of the first and second substances,representing the first individual of goblet sea squirt (leader), at the position of dimension j, FjRepresenting the position of the food source in the j dimension, will search the upper boundary ub of the spacejIs 1, lower boundary lbjIs 0, coefficient c1The definition is as follows:
where L represents the current iteration number, L represents the total iteration number, and parameter c2And c3Take [0,1]The random number of (a) is set,
the follower's location update formula is as follows:
wherein the content of the first and second substances,the location of the ith goblet ascidian (follower) in dimension j is shown.
Compared with the prior art, the technical scheme of the invention has the following beneficial effects:
the invention provides a goblet sea squirt group-based hyperspectral image unmixing method, which introduces an optimization thought into hyperspectral image unmixing on the basis of a shadow model, reconstructs a target function, converts the unmixing problem into an optimization problem, and avoids complex steps of controlling constraint conditions in the traditional unmixing method. By utilizing the advantage of high convergence precision of the goblet sea squirt group algorithm and solving the objective function by adopting the goblet sea squirt group algorithm, the defect that the traditional gradient optimization algorithm is easy to fall into a local extreme value is overcome, so that higher unmixing precision is achieved. Meanwhile, the search space of the improved goblet sea squirt group algorithm is controlled to meet two constraint conditions of abundance, and complicated steps in the unmixing method are simplified.
Drawings
FIG. 1 is a flow chart of the hyperspectral image unmixing method based on the goblet sea squirt group of the present invention.
FIG. 2 is a pseudo-color image of a real hyperspectral image Japser Ridge.
FIG. 3 is an end-member spectral plot extracted from Japser Ridge data using the VCA end-member extraction algorithm.
FIG. 4 is an abundance plot of Japser Ridge data obtained using the method of the invention.
Detailed Description
The invention is described in further detail below with reference to the figures and specific examples. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
With reference to fig. 1 and 2, the following embodiments are described:
step 1: inputting a real hyperspectral image Japser Ridge, and performing end member extraction on Japser Ridge data by adopting VCA (virtual container architecture), so as to obtain an end member spectral curve (shown in FIG. 3) and an end member number R which is 4.
Step 2: determining a search space dimension D of 6 and a position code by using a hyperspectral image SMLM mixed model according to the end member number R of 4; the expression of the hyperspectral image SMLM mixed model is as follows:
r is the number of end members, miSpectral curve representing the ith end-member, aiRepresenting the abundance value, abundance vector a, of the ith end-memberi=[a1,…,aR]TSatisfying the constraint of nonnegativity sum, and the nonlinear parameter P is belonged to [0,1 ∈]The shadow weight Q ∈ [0,1 ]]。
And step 3: setting the population size N of the goblet sea squirt group to be 50, the iteration number I to be 500, the upper boundary of a search space to be 1 and the lower boundary of the search space to be 0;
and 4, step 4: initializing the position of individual goblet sea squirt group in the search space range;
and 5: constructing an objective function, and calculating the current position X of each individual of the goblet sea squirt groupiAn adapted value for the objective function;
in step 5, a target function is constructed by using the reconstruction error, and the expression is as follows:
wherein, | | · | | represents a two-norm, abundanceThe non-negative sum must be satisfied as a constraint,representing reconstructed data, y representing observed data, non-linearity parametersShadow weights
Step 6: the leader individual in the goblet sea squirt group moves according to the position updating formula of the leader, and the follower individual in the goblet sea squirt group moves according to the position updating formula of the follower;
in step 6, the position updating formula of the leader is as follows:
wherein the content of the first and second substances,representing the first individual of goblet sea squirt (leader), at the position of dimension j, FjRepresenting the position of the food source in the j dimension, will search the upper boundary ub of the spacejIs 1, lower boundary lbjIs 0, coefficient c1The definition is as follows:
where L denotes the current iteration number, and L-500 denotes the total iteration number. Parameter c2And c3Take [0,1]The random number of (2).
The follower's location update formula is as follows:
wherein the content of the first and second substances,the location of the ith goblet ascidian (follower) in dimension j is shown.
And 7: calculating the individual adaptive value of each goblet ascidian after updating the position according to the objective function, if less than the current position XiThe current position X is replaced by the updated positioniOtherwise, the current position X is still keptiThe change is not changed;
and step 9: judging whether the current iteration times reach the preset iteration times, if so, ending the iteration, and outputting the position of the optimal individual in the current goblet sea squirt groupThereby obtaining abundance vectorNon-linear parameterAnd shadow weightsOtherwise, returning to execute the step 6;
step 10: judging whether to mix all pixel points of the hyperspectral image, and if so, finishing the calculation; otherwise, returning to execute the step 4 and calculating the next pixel.
The resulting abundance plot from experiments using the unmixing method of the present invention on Japser Ridge data is shown in FIG. 4. The method is compared with a Bayes method and an FCLS method, and a reconstruction error RE and a spectrum angle distribution SAM are used as performance indexes of experimental results. The results of the experiment are shown in table 1.
TABLE 1 comparison of unmixing results
The present invention is not limited to the above-described embodiments. The foregoing description of the specific embodiments is intended to describe and illustrate the technical solutions of the present invention, and the above specific embodiments are merely illustrative and not restrictive. Those skilled in the art can make many changes and modifications to the invention without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (4)
1. A hyperspectral image unmixing method based on a goblet sea squirt group is characterized by comprising the following steps:
firstly, extracting image end members by adopting a geometric-based end member extraction algorithm VCA to obtain an end member spectral curve and an end member number R;
determining a search space dimension D and a position code by utilizing a hyperspectral image SMLM mixed model according to the number R of end members;
step three, setting the population scale, the iteration times and the upper and lower boundaries of the search space of the goblet sea squirt group;
step four, initializing the position of the individual goblet sea squirt group in the search space range;
step five, constructing an objective function, and calculating the current position X of each individual goblet sea squirt groupiAn adapted value for the objective function;
sixthly, the leader individual in the goblet sea squirt group moves according to the position updating formula of the leader, and the follower individual in the goblet sea squirt group moves according to the position updating formula of the follower;
step seven, calculating the individual adaptive value of each goblet ascidian after updating the position according to the objective function, if the individual adaptive value is less than the current position XiThe current position X is replaced by the updated positioniOtherwise, the current position X is still keptiThe change is not changed;
step nine, judging whether the current iteration times reach the preset iteration times or not, if so, ending the iteration, and outputting the position of the optimal individual in the current goblet sea squirt groupThereby obtaining abundance vectorNon-linear parameterAnd shadow weightsOtherwise, returning to the step six;
step ten, judging whether to perform unmixing on all pixel points of the hyperspectral image, and if so, finishing the calculation; otherwise, returning to the step four, and calculating the next pixel.
2. The method for unmixing hyperspectral images based on a goblet sea squirt group as claimed in claim 1, wherein the second step specifically comprises:
the expression of the hyperspectral image SMLM mixed model is as follows:
r is the number of end members, miSpectral curve representing the ith end-member, aiRepresenting the abundance value, abundance vector a, of the ith end-memberi=[a1,…,aR]TSatisfying the constraint of nonnegativity sum, and the nonlinear parameter P is belonged to [0,1 ∈]The shadow weight Q ∈ [0,1 ]]。
3. The hyperspectral image unmixing method based on the ascidian goblet group according to claim 1, wherein step five comprises: and constructing an objective function by using the reconstruction error, wherein the expression is as follows:
4. The hyperspectral image unmixing method based on the ascidian goblet group according to claim 1, wherein the sixth step comprises:
the location update formula for the leader is as follows:
wherein the content of the first and second substances,representing the first individual of goblet sea squirt (leader), at the position of dimension j, FjRepresenting the position of the food source in the j dimension, will search the upper boundary ub of the spacejIs 1, lower boundary lbjIs 0, coefficient c1The definition is as follows:
where L represents the current iteration number, L represents the total iteration number, and parameter c2And c3Take [0,1]The random number of (a) is set,
the follower's location update formula is as follows:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911181901.4A CN111047528B (en) | 2019-11-27 | 2019-11-27 | Hyperspectral image unmixing method based on goblet sea squirt group |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911181901.4A CN111047528B (en) | 2019-11-27 | 2019-11-27 | Hyperspectral image unmixing method based on goblet sea squirt group |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111047528A true CN111047528A (en) | 2020-04-21 |
CN111047528B CN111047528B (en) | 2023-07-07 |
Family
ID=70233795
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911181901.4A Active CN111047528B (en) | 2019-11-27 | 2019-11-27 | Hyperspectral image unmixing method based on goblet sea squirt group |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111047528B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112085705A (en) * | 2020-08-11 | 2020-12-15 | 温州大学 | Image segmentation method and device based on improved goblet sea squirt group algorithm |
CN113255138A (en) * | 2021-05-31 | 2021-08-13 | 河北工业大学 | Load distribution optimization method for power system |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6665438B1 (en) * | 1999-05-05 | 2003-12-16 | American Gnc Corporation | Method for hyperspectral imagery exploitation and pixel spectral unmixing |
CN105809185A (en) * | 2015-12-31 | 2016-07-27 | 天津大学 | High-spectrum image nonlinear demixing method based on neural network and differential search |
CN105975912A (en) * | 2016-04-27 | 2016-09-28 | 天津大学 | Hyperspectral image nonlinearity solution blending method based on neural network |
CN106056524A (en) * | 2016-05-25 | 2016-10-26 | 天津商业大学 | Hyper-spectral image nonlinear de-mixing method based on differential search |
CN109284860A (en) * | 2018-08-28 | 2019-01-29 | 温州大学 | A kind of prediction technique based on orthogonal reversed cup ascidian optimization algorithm |
-
2019
- 2019-11-27 CN CN201911181901.4A patent/CN111047528B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6665438B1 (en) * | 1999-05-05 | 2003-12-16 | American Gnc Corporation | Method for hyperspectral imagery exploitation and pixel spectral unmixing |
CN105809185A (en) * | 2015-12-31 | 2016-07-27 | 天津大学 | High-spectrum image nonlinear demixing method based on neural network and differential search |
CN105975912A (en) * | 2016-04-27 | 2016-09-28 | 天津大学 | Hyperspectral image nonlinearity solution blending method based on neural network |
CN106056524A (en) * | 2016-05-25 | 2016-10-26 | 天津商业大学 | Hyper-spectral image nonlinear de-mixing method based on differential search |
CN109284860A (en) * | 2018-08-28 | 2019-01-29 | 温州大学 | A kind of prediction technique based on orthogonal reversed cup ascidian optimization algorithm |
Non-Patent Citations (2)
Title |
---|
陈涛;王梦馨;黄湘松;: "基于樽海鞘群算法的无源时差定位" * |
陈雷;郭艳菊;葛宝臻;: "基于微分搜索的高光谱图像非线性解混算法" * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112085705A (en) * | 2020-08-11 | 2020-12-15 | 温州大学 | Image segmentation method and device based on improved goblet sea squirt group algorithm |
CN112085705B (en) * | 2020-08-11 | 2024-03-15 | 温州大学 | Image segmentation method and device based on improved goblet sea squirt swarm algorithm |
CN113255138A (en) * | 2021-05-31 | 2021-08-13 | 河北工业大学 | Load distribution optimization method for power system |
CN113255138B (en) * | 2021-05-31 | 2023-05-23 | 河北工业大学 | Load distribution optimization method for power system |
Also Published As
Publication number | Publication date |
---|---|
CN111047528B (en) | 2023-07-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108416353B (en) | Method for quickly segmenting rice ears in field based on deep full convolution neural network | |
CN105550649B (en) | Extremely low resolution ratio face identification method and system based on unity couping local constraint representation | |
CN109376772B (en) | Power load combination prediction method based on neural network model | |
CN111126662B (en) | Irrigation decision making method, device, server and medium based on big data | |
CN111047528A (en) | High-spectrum image unmixing method based on goblet sea squirt group | |
Ji et al. | Mapping land use/cover dynamics of the Yellow River Basin from 1986 to 2018 supported by Google Earth Engine | |
CN108470358B (en) | Point cloud registration method based on second-order oscillation artificial bee colony algorithm | |
CN104715024A (en) | Multimedia hotspot analysis method | |
CN107784361B (en) | Image recognition method for neural network optimization | |
CN112766155A (en) | Deep learning-based mariculture area extraction method | |
CN111161199A (en) | Spatial-spectral fusion hyperspectral image mixed pixel low-rank sparse decomposition method | |
CN111126549A (en) | Double-star spectrum fitting method based on strategy improved goblet and sea squirt intelligent algorithm | |
CN115272881B (en) | Long-tail remote sensing image target identification method based on dynamic relation distillation | |
Yakimov et al. | Multifractal analysis of neutral community spatial structure | |
CN115566689A (en) | Method for optimizing load peak-valley time division and peak-valley electricity price by improving skyhawk optimization algorithm | |
CN110008996B (en) | Heterogeneous domain adaptation method based on divergence calculation | |
CN108509840B (en) | Hyperspectral remote sensing image waveband selection method based on quantum memory optimization mechanism | |
CN107644230B (en) | Spatial relationship modeling method for remote sensing image object | |
CN109409407A (en) | A kind of industry monitoring data clustering method based on LE algorithm | |
CN110987751B (en) | Quantitative grading evaluation method for pore throat of compact reservoir in three-dimensional space | |
CN105334730B (en) | The IGA optimization T S of heating furnace oxygen content obscure ARX modeling methods | |
CN116415177A (en) | Classifier parameter identification method based on extreme learning machine | |
CN115641583A (en) | Point cloud detection method, system and medium based on self-supervision and active learning | |
CN110175639B (en) | Short-term wind power prediction method based on feature selection | |
CN113889233A (en) | Cell positioning and counting method based on manifold regression network and application |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |