CN110188735A - A kind of instruction plant recognition methods based on EO-1 hyperion - Google Patents
A kind of instruction plant recognition methods based on EO-1 hyperion Download PDFInfo
- Publication number
- CN110188735A CN110188735A CN201910496996.2A CN201910496996A CN110188735A CN 110188735 A CN110188735 A CN 110188735A CN 201910496996 A CN201910496996 A CN 201910496996A CN 110188735 A CN110188735 A CN 110188735A
- Authority
- CN
- China
- Prior art keywords
- instruction plant
- instruction
- collection
- recognition methods
- environment
- 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 73
- 230000009467 reduction Effects 0.000 claims abstract description 25
- 238000001228 spectrum Methods 0.000 claims abstract description 15
- 230000005856 abnormality Effects 0.000 claims abstract description 9
- 230000008569 process Effects 0.000 claims abstract description 9
- 230000003595 spectral effect Effects 0.000 claims description 12
- 238000004458 analytical method Methods 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 9
- 238000007619 statistical method Methods 0.000 claims description 8
- 230000009466 transformation Effects 0.000 claims description 7
- 238000009499 grossing Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- 238000000513 principal component analysis Methods 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000011160 research Methods 0.000 claims description 4
- 238000013480 data collection Methods 0.000 claims description 3
- 241000196324 Embryophyta Species 0.000 description 76
- 238000001514 detection method Methods 0.000 description 9
- 238000000605 extraction Methods 0.000 description 8
- 239000011159 matrix material Substances 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 230000009545 invasion Effects 0.000 description 3
- 238000004611 spectroscopical analysis Methods 0.000 description 3
- 238000012935 Averaging Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000004069 differentiation Effects 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 239000004615 ingredient Substances 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000002203 pretreatment Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- 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
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/194—Terrestrial scenes using hyperspectral data, i.e. more or other wavelengths than RGB
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Multimedia (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- General Health & Medical Sciences (AREA)
- Biochemistry (AREA)
- Health & Medical Sciences (AREA)
- Pathology (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Astronomy & Astrophysics (AREA)
- Remote Sensing (AREA)
- Immunology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The instruction plant recognition methods based on EO-1 hyperion that the invention discloses a kind of, the recognition methods include the following steps: to acquire the high spectrum image of the instruction plant under laboratory environment and under field environment respectively, are denoted as source sample;Rejecting abnormalities sample process is carried out to the instruction plant sample obtained under two kinds of environment, is denoted as and does not make pretreatment sample;The instruction plant data of two kinds of data sources are pre-processed respectively, corresponding instruction plant feature set is obtained, is denoted as primary features collection;Dimensionality reduction is carried out to the primary features collection of the instruction plant under laboratory environment, obtained new feature collection is denoted as standard feature Ji Ku;Dimension-reduction treatment is carried out with all band and local wave band to the primary features collection of instruction plant under field environment;Vector machine modeling is supported to all band of instruction plant mid-level features collection in the field environment and local wave band respectively, optimal preprocess method is obtained by comparison;Primary features collection is optimized.Recognition methods in the present invention is high-efficient.
Description
Technical field
The present invention relates to the Eigenvalue Extraction of a kind of pair of Alien Invasive Plants, specially a kind of entering based on EO-1 hyperion
Plants identification method is invaded, the characteristics extraction applied technical field of Alien Invasive Plants is belonged to.
Background technique
Instruction plant is low with visual recognition degree under attached main plant visible light, and irregular distribution, field environment is complicated and changeable, mesh
Preceding artificial on-site survey is one of the method the most universal to instruction plant detection, and artificial on-site survey is the subjective judgement by people,
Instruction plant can not be detected and make accurate judgement, and artificial on-site survey mode low efficiency, time-consuming, accuracy rate is low etc.;But
The present invention is high with spectral resolution using high spectrum image monitoring technology, wave band is more, nicety of grading is high, accurate positioning, information
The features such as amount is abundant, has great advantage to the detection of instruction plant, small, cost which overcome the coverage areas manually trampleed on
The disadvantages of time is more, to improve to instruction plant detection efficiency.
The Image Acquisition of EO-1 hyperion is mainly satellite remote sensing and unmanned plane, for satellite remote sensing technology, is suitable for advising greatly
The high spectrum image of mould acquires, but for the often ignored of middle and small scale.The present invention using unmanned rack carry EO-1 hyperion at
As instrument obtains instruction plant high spectrum image, compared with satellite remote sensing, the features such as flexibility is high, high resolution, for invasion
The detection of plant is advantageously.
Past detects instruction plant using satellite remote sensing, but due to the small-scale instruction plant of satellite remote sensing centering
Ignore, miss the early prevention stage of instruction plant, existing instruction plant is largely invaded, and the production for arriving crops is seriously endangered
Amount.Therefore it is very important for the detection of instruction plant.
It to sum up, is at present by artificial and carry out using satellite remote sensing technology, but manually to the detection of instruction plant
All there is relatively large detection error with satellite remote sensing technology detection;And the present invention is planted using instruction plant as object in conjunction with invasion
The characteristics of object, and unmanned air vehicle technique is used, thus the method for identifying instruction plant feature;The present invention is by a variety of common method knots
It closes, forms the method more effectively new to instruction plant detection.Therefore, a kind of entering based on EO-1 hyperion is proposed regarding to the issue above
Invade plants identification method.
Summary of the invention
The object of the invention is that providing a kind of instruction plant identification based on EO-1 hyperion to solve the above-mentioned problems
Method.
The present invention is achieved through the following technical solutions above-mentioned purpose, a kind of instruction plant identification side based on EO-1 hyperion
Method, the recognition methods include the following steps:
Step 1. acquires the high spectrum image of the instruction plant under laboratory environment and under field environment respectively, is denoted as source sample
This;
Step 2. carries out rejecting abnormalities sample process to the instruction plant sample obtained under two kinds of environment, is denoted as and does not do pre- place
Manage sample;
Step 3. respectively pre-processes the instruction plant data of two kinds of data sources, obtains corresponding instruction plant feature
Collection, is denoted as primary features collection;Wherein the instruction plant preprocess method under laboratory environment includes putting down smoothly and after first differential
Sliding, the instruction plant volume preprocess method under field environment includes not pre-processing, and 9 gliding smoothings, S-G is smooth, and single order is micro-
Divide, second-order differential, polynary scatter correction, standard normal variable transformation, after first differential smooth (first differential+S-G is smooth), first marks
Quasi- normal variate transformation goes trend to handle nine kinds of preprocess methods again;
Step 4. carries out dimensionality reduction to the primary features collection of the instruction plant under laboratory environment, and obtained new feature collection is denoted as
Standard feature Ji Ku;
Step 5. carries out dimension-reduction treatment to the primary features collection of instruction plant under field environment with all band and local wave band,
Wherein dimensionality reduction is carried out using primary features collection of the principal component analysis to all band instruction plant;And to the instruction plant of local wave band
Primary features collection dimensionality reduction be pass through instruction plant bloom Spectral characteristics analysis and combine statistical analysis technique realize;It is obtained after dimensionality reduction
Data be denoted as stage data collection;
Step 6. respectively props up all band of instruction plant mid-level features collection in the field environment and local wave band
Vector machine modeling is held, optimal preprocess method is obtained by comparison;
Step 7. optimizes primary features collection.
Preferably, the step 2 is to the instruction plant sample mahalanobis distance rejecting abnormalities sample obtained under two kinds of environment
Processing, to improve the differentiation quality of established model.
Preferably, the processing of primary features collection is carried out based on pretreated data in the step 4, and its dimensionality reduction
Processing is obtained by bloom Spectral characteristics analysis and statistical analysis technique.
Preferably, the standard feature collection that dimensionality reduction obtains under laboratory environment in the step 4 is in order to research field ring
Feature set is extracted under border and does basis, and the data under laboratory environment are only involved in the analysis of local wave band under field environment, number
According to as standard set.
Preferably, during carrying out dimensionality reduction for the mid-level features collection of local wave band instruction plant in the step 5, ginseng
According to the standard feature Ji Ku in step 4, corresponding mid-level features collection is obtained.
Preferably, the step 5,6 are to be based respectively on the primary features collection that nine kinds of preprocess methods obtain to be handled.
Preferably, instruction plant feature set is optimized in the step 7, i.e., the feature in instruction plant feature set carries out single
Feature ordering optimization.
Preferably, it is excellent to be that the primary features collection converted for the quasi- normal variate of optimal pretreatment carries out for the step 7
Change.
The beneficial effects of the present invention are: instruction plant data of the present invention pass through laboratory non-imaged bloom spectra system and field
High spectrum image system acquisition, the mainly identification to field instruction plant.Instruction plant identification in field is based on full wave
Instruction plant feature extraction and instruction plant feature extraction two schemes based on local wave band.The method of the present invention is planted to invasion
Object feature set is simplified comprising many algorithms, that is, preprocess method, Method of Data with Adding Windows, data classification and optimization method lead to
The method of the present invention is crossed, identification instruction plant can be quickly obtained, and discrimination is high.
Detailed description of the invention
Fig. 1 is that the present invention is based on the flow chart of data processing figure of local wave band;
Fig. 2 be the present invention is based under full wave flow chart of data processing to the collecting flowchart of instruction plant high-spectral data
Figure.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Embodiment:
A kind of instruction plant recognition methods based on EO-1 hyperion, the recognition methods include the following steps:
Step 1. acquires the high spectrum image of the instruction plant under laboratory environment and under field environment respectively, is denoted as source sample
This;
Step 2. carries out rejecting abnormalities sample process to the instruction plant sample obtained under two kinds of environment, is denoted as and does not do pre- place
Manage sample;
Step 3. respectively pre-processes the instruction plant data of two kinds of data sources, obtains corresponding instruction plant feature
Collection, is denoted as primary features collection;Wherein the instruction plant preprocess method under laboratory environment includes putting down smoothly and after first differential
Sliding, the instruction plant volume preprocess method under field environment includes not pre-processing, and 9 gliding smoothings, S-G is smooth, and single order is micro-
Divide, second-order differential, polynary scatter correction, standard normal variable transformation, after first differential smooth (first differential+S-G is smooth), first marks
Quasi- normal variate transformation goes trend to handle nine kinds of preprocess methods again;
Step 4. carries out dimensionality reduction to the primary features collection of the instruction plant under laboratory environment, and obtained new feature collection is denoted as
Standard feature Ji Ku;
Step 5. carries out dimension-reduction treatment to the primary features collection of instruction plant under field environment with all band and local wave band,
Wherein dimensionality reduction is carried out using primary features collection of the principal component analysis to all band instruction plant;And to the instruction plant of local wave band
Primary features collection dimensionality reduction be pass through instruction plant bloom Spectral characteristics analysis and combine statistical analysis technique realize;It is obtained after dimensionality reduction
Data be denoted as stage data collection;
Step 6. respectively props up all band of instruction plant mid-level features collection in the field environment and local wave band
Vector machine modeling is held, optimal preprocess method is obtained by comparison;
Step 7. optimizes primary features collection.
Further, the step 2 is to the instruction plant sample mahalanobis distance rejecting abnormalities sample obtained under two kinds of environment
Present treatment, to improve the differentiation quality of established model.
Further, the processing of primary features collection is to be carried out based on pretreated data, and it is dropped in the step 4
Dimension processing is obtained by bloom Spectral characteristics analysis and statistical analysis technique.
Further, the standard feature collection that dimensionality reduction obtains under laboratory environment in the step 4 is in order to research field
Feature set is extracted under environment and does basis, and the data under laboratory environment are only involved in the analysis of local wave band under field environment,
Data are as standard set.
Further, during carrying out dimensionality reduction for the mid-level features collection of local wave band instruction plant in the step 5,
Referring to the standard feature Ji Ku in step 4, corresponding mid-level features collection is obtained.
Further, the step 5,6 are to be based respectively on the primary features collection that nine kinds of preprocess methods obtain to be handled
's.
Further, instruction plant feature set is optimized in the step 7, i.e., the feature in instruction plant feature set carries out
Single feature ordering optimization.
Further, the step 7 is that the primary features collection converted for the optimal quasi- normal variate of pretreatment carries out
Optimization.
Acquisition by two approach (under laboratory environment and under field environment) to instruction plant high-spectral data, and it is right
Data source is used based on the feature extraction of all band instruction plant and based on the instruction plant feature two of local wave band under field environment
The characteristics extraction of kind mode, describes in terms of two individually below, such as Figure of description 1-2:
1. acquiring the instruction plant high-spectral data sample of two kinds of data sources.
2. pair two kinds of data source samples carry out rejecting abnormalities sample (be denoted as and do not make pretreatment sample)
Achieve the purpose that rejecting abnormalities sample using mahalanobis distance (MD), the specific method is as follows:
Wherein, k=1 ... ..., n are sample number, XkFor the spectral value of each sample, μ is that the spectrum of sample set is average
Value, ∑-1For covariance matrix.Calculation formula are as follows:
Wherein, k=1 ... ..., n are sample number, are every XkThe spectral value of a sample, μ are that the spectrum of sample set is average
Value, ∑-1For covariance matrix.
Threshold value is arranged to mahalanobis distance, is exceptional sample beyond threshold value, and rejected.
3. two kinds of data source samples are pre-processed
Data sample under 3.1 laboratory environments is pre-processed
Using smooth two kinds and processing method after first differential processing and first differential.
High spectrum image sample preprocessing under 3.2 field environments
3.2.1 being based on the feature extraction of all band instruction plant, carry out as follows:
3.2.1.1 smooth (two kinds of preprocess methods)
Achieved the purpose that by two kinds of algorithms it is smooth, it is specific as follows:
Rolling average exponential smoothing: averaging to single sample data gliding smoothing window, to carry out noise reduction to data.
Savitzky-Golay convolution exponential smoothing (S-G is smooth) is that one kind is based on local multinomial least square in time domain
The filtering method of method fitting.Its operational formula are as follows:
Wherein, yiFor the value filtered out in point i;yi+jFor the value not filtered out in point i;J specifies smoothed data on the left of the point
Points;M is to specify data points smooth on the right side of the point;cjFor deconvolution parameter.
3.2.1.2 differential (two kinds of pretreatments)
The differential method mainly has first differential (FDR) and second-order differential (SDR), and calculation formula respectively indicates are as follows:
3.2.1.3 multiplicative scatter correction (MSC)
The summation of all spectroscopic datas is averaged again first, using counted average spectral data as " ideal spectrum ", is calculated
Formula are as follows:
Then the intercept constant and regression coefficient of every spectroscopic data and average spectral data are calculated:
Finally calculate the relative tilt of the amendment curve of spectrum:
3.2.1.4 standard normal variable converts
By original spectral data standard normal calculation formula such as formula:
3.2.1.5 after first differential smooth (first differential+S-G is smooth)
3.2.1.6 first standard normal variable transformation goes trend to handle again
4. dimensionality reduction
4.1 dimensionality reduction based on all band instruction plant feature --- principal component analysis
By original higher-dimension variable data (x1, x2..., xp) it is converted into new generalized variable data (y according to the following formula1,
y2..., ym) i.e. principal component, it is as follows to solve formula:
Wherein, it is necessary to meet following condition for solution formula solution procedure:
A) principal component (y1, y2..., ym) between it is irrelevant each other, with achieve the effect that reduce data redundancy.
B) according to the descending sequence of the variance yields of principal component, i.e. Var (y1) > Var (y2) > ... > Var (ym), it will
After main feature information aggregation to fraction principal component analysis in original high dimensional data in data.
C) coefficient in formula need to meet condition bm1+bm2+...+bmp=1.
Based on there are m n-dimensional vector x of certain correlationiExtract the PCA algorithm concrete implementation of its main feature information
Process is as follows:
Step1: sample data centralization, i.e., to by m n-dimensional vector xiThe matrix by rows of composition carries out zero averaging,
Step2: it calculatesCovariance matrix,
Step3: diagonal matrix Λ and eigenmatrix are arrived to covariance C feature decomposition, wherein Λ=Diag [λ1,
λ2..., λm], λiIt is characterized value, U=[U1, U2..., Um], UiFor feature vector;
K rank principal component before Step4 takes,Wherein Y is exactly the main feature information for needing to extract.
Comparison obtains 10 dimension instruction plant feature sets.
The 4.2 instruction plant Feature Dimension Reductions based on local wave band
Pass through instruction plant bloom Spectral characteristics analysis and statistical analysis technique is combined to realize, first to pretreated experiment
Spectroscopic data under room environmental is analyzed, so that selected characteristic parameter, is denoted as standard database.Afterwards in standard database
On the basis of to the high-spectral data selected characteristic parameter under field environment.
5. single eigenvalue support vector machines
Vector machine classification is supported to feature set first, is realized by following calculating formula:
Classification problem is nonlinear relevant issues in this research, needs to classify using Kernel Function Transformation.
Formula is as follows:
To basic function (RBF):
5.1 are based on full wave instruction plant feature
Feature set based on 1,2,3,4 is classified using support vector machines, thus the result obtained to principal component analysis
It is verified.
The 5.2 instruction plant features based on local wave band
Feature set based on 1,2,3,4 is classified using statistical analysis, to verify to the result that dimensionality reduction obtains.
6. optimization
Former 138 dimensional feature collection of data is optimized, using single feature set sort method according to master i.e. inside feature set
The descending sequence of the variance yields of ingredient, therefore above method is verified.
7. obtaining result
This method is feasible to the extraction of instruction plant characteristic value, and accuracy rate is high.
It is obvious to a person skilled in the art that invention is not limited to the details of the above exemplary embodiments, Er Qie
In the case where without departing substantially from spirit or essential attributes of the invention, the present invention can be realized in other specific forms.Therefore, no matter
From the point of view of which point, the present embodiments are to be considered as illustrative and not restrictive, and the scope of the present invention is by appended power
Benefit requires rather than above description limits, it is intended that all by what is fallen within the meaning and scope of the equivalent elements of the claims
Variation is included within the present invention.Any reference signs in the claims should not be construed as limiting the involved claims.
In addition, it should be understood that although this specification is described in terms of embodiments, but not each embodiment is only wrapped
Containing an independent technical solution, this description of the specification is merely for the sake of clarity, and those skilled in the art should
It considers the specification as a whole, the technical solutions in the various embodiments may also be suitably combined, forms those skilled in the art
The other embodiments being understood that.
Claims (8)
1. a kind of instruction plant recognition methods based on EO-1 hyperion, it is characterised in that: the recognition methods includes the following steps:
Step 1. acquires the high spectrum image of the instruction plant under laboratory environment and under field environment respectively, is denoted as source sample;
Step 2. carries out rejecting abnormalities sample process to the instruction plant sample obtained under two kinds of environment, is denoted as and does not make pretreatment sample
This;
Step 3. respectively pre-processes the instruction plant data of two kinds of data sources, obtains corresponding instruction plant feature set,
It is denoted as primary features collection;Wherein the instruction plant preprocess method under laboratory environment include after smooth and first differential smoothly,
Instruction plant volume preprocess method under field environment includes not pre-processing, and 9 gliding smoothings, S-G is smooth, first differential,
Second-order differential, polynary scatter correction, standard normal variable convert, smooth (first differential+S-G is smooth), first standard after first differential
Normal variate transformation goes trend to handle nine kinds of preprocess methods again;
Step 4. carries out dimensionality reduction to the primary features collection of the instruction plant under laboratory environment, and obtained new feature collection is denoted as standard
Feature set library;
Step 5. carries out dimension-reduction treatment to the primary features collection of instruction plant under field environment with all band and local wave band, wherein
Dimensionality reduction is carried out using primary features collection of the principal component analysis to all band instruction plant;And to the first of the instruction plant of local wave band
Grade feature set dimensionality reduction is to pass through instruction plant bloom Spectral characteristics analysis and statistical analysis technique is combined to realize;The number obtained after dimensionality reduction
According to being denoted as stage data collection;
Step 6. respectively to all band of instruction plant mid-level features collection in the field environment and local wave band be supported to
The modeling of amount machine obtains optimal preprocess method by comparison;
Step 7. optimizes primary features collection.
2. a kind of instruction plant recognition methods based on EO-1 hyperion according to claim 1, it is characterised in that: the step
The instruction plant sample mahalanobis distance rejecting abnormalities sample process obtained under 2 pairs of two kinds of environment, to improve established model
Differentiate quality.
3. a kind of instruction plant recognition methods based on EO-1 hyperion according to claim 1, it is characterised in that: the step
The processing of primary features collection is to be carried out based on pretreated data, and its dimension-reduction treatment is by bloom spectrum signature point in 4
What analysis and statistical analysis technique obtained.
4. a kind of instruction plant recognition methods based on EO-1 hyperion according to claim 1, it is characterised in that: the step
The standard feature collection that dimensionality reduction obtains under laboratory environment in 4 be in order to research field environment under extract feature set do basis, and
Data under laboratory environment are only involved in the analysis of local wave band under field environment, and data are as standard set.
5. a kind of instruction plant recognition methods based on EO-1 hyperion according to claim 1, it is characterised in that: the step
During carrying out dimensionality reduction for the mid-level features collection of local wave band instruction plant in 5, referring to the standard feature collection in step 4
Library obtains corresponding mid-level features collection.
6. a kind of instruction plant recognition methods based on EO-1 hyperion according to claim 1, it is characterised in that: the step
5,6 be to be based respectively on the primary features collection that nine kinds of preprocess methods obtain to be handled.
7. a kind of instruction plant recognition methods based on EO-1 hyperion according to claim 1, it is characterised in that: the step
Instruction plant feature set is optimized in 7, i.e., the feature in instruction plant feature set carries out single feature ordering optimization.
8. a kind of instruction plant recognition methods based on EO-1 hyperion according to claim 1, it is characterised in that: the step
7 be that the primary features collection converted for the optimal quasi- normal variate of pretreatment optimizes.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910496996.2A CN110188735A (en) | 2019-06-10 | 2019-06-10 | A kind of instruction plant recognition methods based on EO-1 hyperion |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910496996.2A CN110188735A (en) | 2019-06-10 | 2019-06-10 | A kind of instruction plant recognition methods based on EO-1 hyperion |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110188735A true CN110188735A (en) | 2019-08-30 |
Family
ID=67721039
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910496996.2A Pending CN110188735A (en) | 2019-06-10 | 2019-06-10 | A kind of instruction plant recognition methods based on EO-1 hyperion |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110188735A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111257242A (en) * | 2020-02-27 | 2020-06-09 | 西安交通大学 | High-spectrum identification method for pollutant components of insulator |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070044445A1 (en) * | 2005-08-01 | 2007-03-01 | Pioneer Hi-Bred International, Inc. | Sensor system, method, and computer program product for plant phenotype measurement in agricultural environments |
CN108764199A (en) * | 2018-06-06 | 2018-11-06 | 中国农业科学院深圳农业基因组研究所 | The automatic identifying method and system of instruction plant Mikania micrantha |
CN108875620A (en) * | 2018-06-06 | 2018-11-23 | 中国农业科学院深圳农业基因组研究所 | The monitoring method and system of instruction plant |
-
2019
- 2019-06-10 CN CN201910496996.2A patent/CN110188735A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070044445A1 (en) * | 2005-08-01 | 2007-03-01 | Pioneer Hi-Bred International, Inc. | Sensor system, method, and computer program product for plant phenotype measurement in agricultural environments |
CN108764199A (en) * | 2018-06-06 | 2018-11-06 | 中国农业科学院深圳农业基因组研究所 | The automatic identifying method and system of instruction plant Mikania micrantha |
CN108875620A (en) * | 2018-06-06 | 2018-11-23 | 中国农业科学院深圳农业基因组研究所 | The monitoring method and system of instruction plant |
Non-Patent Citations (2)
Title |
---|
胡佳等: "基于WorldView-2的薇甘菊信息精细提取", 《中南林业科技大学学报》 * |
艾金泉等: "入侵种互花米草的光谱分层分析方法", 《测绘科学》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111257242A (en) * | 2020-02-27 | 2020-06-09 | 西安交通大学 | High-spectrum identification method for pollutant components of insulator |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2022160771A1 (en) | Method for classifying hyperspectral images on basis of adaptive multi-scale feature extraction model | |
Li et al. | Identifying blueberry fruit of different growth stages using natural outdoor color images | |
US7899625B2 (en) | Method and system for robust classification strategy for cancer detection from mass spectrometry data | |
Safren et al. | Detection of green apples in hyperspectral images of apple-tree foliage using machine vision | |
CN104990892B (en) | The spectrum picture Undamaged determination method for establishing model and seeds idenmtification method of seed | |
CN107527326A (en) | A kind of wheat scab diagnostic method based on high light spectrum image-forming | |
CN108537751B (en) | Thyroid ultrasound image automatic segmentation method based on radial basis function neural network | |
CN104680185B (en) | Hyperspectral image classification method based on boundary point reclassification | |
CN109190698B (en) | Classification and identification system and method for network digital virtual assets | |
CN104713835A (en) | Online numerical recognition method for colors of tobacco leaves | |
Zhou et al. | Identification of the variety of maize seeds based on hyperspectral images coupled with convolutional neural networks and subregional voting | |
US20230243744A1 (en) | Method and system for automatically detecting and reconstructing spectrum peaks in near infrared spectrum analysis of tea | |
CN112200042B (en) | Method for analyzing ecological change trend by using space-time ecological environment remote sensing fractal dimension | |
CN102072767A (en) | Wavelength similarity consensus regression-based infrared spectrum quantitative analysis method and device | |
Huang et al. | Hyperspectral imaging for identification of an invasive plant Mikania micrantha Kunth | |
CN115292334A (en) | Intelligent planting method and system based on vision, electronic equipment and storage medium | |
Supekar et al. | Multi-parameter based mango grading using image processing and machine learning techniques | |
CN104990891B (en) | A kind of seed near infrared spectrum and spectrum picture qualitative analysis model method for building up | |
CN110188735A (en) | A kind of instruction plant recognition methods based on EO-1 hyperion | |
LU500715B1 (en) | Hyperspectral Image Classification Method Based on Discriminant Gabor Network | |
CN115393631A (en) | Hyperspectral image classification method based on Bayesian layer graph convolution neural network | |
CN114863286A (en) | Mixed waste plastic classification method based on multi-algorithm collaborative optimization | |
CN114971259A (en) | Method for analyzing quality consistency of formula product by using near infrared spectrum | |
CN109726641B (en) | Remote sensing image cyclic classification method based on automatic optimization of training samples | |
CN111595802A (en) | Construction method and application of Clinacanthus nutans seed source place classification model based on NIR (near infrared spectroscopy) |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190830 |