CN109785272A - A kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction - Google Patents
A kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction Download PDFInfo
- Publication number
- CN109785272A CN109785272A CN201811551730.5A CN201811551730A CN109785272A CN 109785272 A CN109785272 A CN 109785272A CN 201811551730 A CN201811551730 A CN 201811551730A CN 109785272 A CN109785272 A CN 109785272A
- Authority
- CN
- China
- Prior art keywords
- wavelet
- mineral
- decomposition
- mineral spectra
- optimal
- 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
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention belongs to spectral signal processing and high spectrum mineral map plotting technical fields, and in particular to a kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction；The purpose of the present invention is to provide the methods that Optimal wavelet bases in a kind of mineral spectra feature extraction are chosen, comprising the following steps: step 1 measures Mineral spectra；Step 2 carries out denoising to mineral spectra by averaging method；Step 3 determines some column wavelet basis included in Wavelet Cluster type to be selected and every kind of Wavelet Cluster；Step 4 carries out wavelet decomposition to target mineral spectral signal using the wavelet basis series in different Wavelet Clusters；Step 5 calculates the energy and aromatic entropy of decomposition scales at different levels according to coefficient of wavelet decomposition；Step 6 calculates the ratio of energy and aromatic entropy；Step 7 extracts mineral spectra feature according to selected Optimal wavelet bases decomposition coefficient.
Description
Technical field
The invention belongs to spectral signal processing and high spectrum mineral map plotting technical fields, and in particular to a kind of mineral spectra is special
The method that Optimal wavelet bases are chosen in sign extraction.
Background technique
The reflection of rock forming mineral spectrum, Absorption Characteristics cover entire visible light to infrared range of spectrum, examine according to mineral
Disconnected property spectral signature can be with Direct Recognition mineral type and mineral constituent, quantitative inversion atural object constituent content.Therefore, with suitable
Method accurately extract and assaying spectral signature it is most important.
Spectra feature extraction refers to from the mineral original spectral data obtained, removes extra interference information and retains
Effectively it is different from the diagnostic characteristic information of other atural objects.Spectroscopy differential method is a kind of traditional Spectra feature extraction method, main
It is used to extract the multiple spectrums parameter such as band wavelength position, depth.Spectroscopy differential method and principal component analysis, discriminant analysis method
Deng the Spectra feature extraction method based on statistics, all it is a kind of Time Domain Analysis for being based purely on waveform, does not account for spectrum
The frequency domain character of signal.
Wavelet transformation has simultaneously in the ability of time domain and frequency domain characterization signal local feature, can be from signal in different rulers
Characteristics of signals is analyzed in wavelet transform result under degree, is widely used in field of signal processing.There was only one kind with Fourier transformation
Basic function is different, has unlimited a variety of wavelet basis in wavelet transformation theory, can satisfy the needs of various problems.Different wavelet basis
With different time-frequency characteristics, shape differences are larger, and bearing length and systematicness are also different.Therefore, for the same mine
Object light spectrum signal is chosen different wavelet basis and is handled, and often difference is very big for obtained result, how to choose Optimal wavelet bases use
It is key problems-solving of the present invention in mineral spectra feature extraction.
Summary of the invention
The purpose of the present invention is to provide the methods that Optimal wavelet bases in a kind of mineral spectra feature extraction are chosen, using not
Wavelet decomposition is carried out to Target scalar spectral signal with wavelet basis a series of in Wavelet Cluster, decomposition at different levels are calculated based on decomposition coefficient
The energy of scale and aromatic entropy are used for using the ratio size of energy and aromatic entropy as the criterion for choosing Optimal wavelet bases
Mineral spectra feature extraction.
The technical scheme is that
A kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction, comprising the following steps:
Step 1 measures Mineral spectra；
Step 2 carries out mineral spectra by averaging method to remove dryness processing；
Step 3 determines some column wavelet basis included in Wavelet Cluster type to be selected and every kind of Wavelet Cluster；
Step 4 carries out wavelet decomposition to target mineral spectral signal using the wavelet basis series in different Wavelet Clusters；
Step 5 calculates the energy and aromatic entropy of decomposition scales at different levels according to coefficient of wavelet decomposition；
Step 6 calculates the ratio of energy and aromatic entropy；
Step 7 extracts mineral spectra feature according to selected Optimal wavelet bases decomposition coefficient.
In the step 1, certain mineral spectra curve that measurement obtains is denoted as Xi(λ), wherein λ is spectral band sequence
Number, i is the number of repeatedly measurement mineral spectra.
In the step 2, denoising is carried out to mineral spectra by averaging method, certain mineral light obtained to measurement
Compose Xi(λ) is averaged after summing according to corresponding wave band, it is assumed that has carried out n times spectral measurement to mineral altogether, averaging method is to mineral
The formula of spectrum denoising are as follows:
In the step 3, a series of wavelet basis for including in the type and every kind of Wavelet Cluster of Wavelet Cluster to be selected are determined,
Wavelet Cluster type and the wavelet basis for being included series are chosen according to Wavelet Cluster property.
In the step 4, small wavelength-division is carried out to target mineral spectral signal using the wavelet basis series in different Wavelet Clusters
Solution, set wavelet basis function asDecomposition order j, applicationWavelet basis carries out j layers to the mineral spectra signal Z (λ) after denoising
Discrete wavelet transformation, obtaining decomposition coefficient sequence is Cj(k) (k=1,2 ... 2j), represent k-th of decomposition coefficient of jth layer.
In the step 5, the energy and aromatic entropy of decomposition scales at different levels are calculated according to coefficient of wavelet decomposition, wherein each
Decomposition layer energy can be calculated by the quadratic sum of this layer of decomposition coefficient, i.e.,
Probability distribution P (k) calculation formula of wavelet decomposition jth k-th of node energy of layer is
Pj(k)=| C (k) |2/E(j) (3)
Aromatic entropy F (j) is used to describe the Energy distribution of coefficient of wavelet decomposition, its calculation formula is
In the step 6, calculate energy and aromatic entropy ratio R (j), using the corresponding wavelet basis of maximum R (j) value as
Optimal wavelet bases；Its calculation formula is
R (j)=E (j)/F (j) (5).
The beneficial effects of the present invention are:
A kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction is proposed, this method is simple and efficient, realizes
The fast selecting of Optimal wavelet bases improves the efficiency of mineral spectra feature extraction effect and signal processing, reduces artificial warp
The interference tested, it is applied widely.
Detailed description of the invention
Fig. 1 is the method flow diagram that Optimal wavelet bases are chosen in a kind of mineral spectra feature extraction.
Specific embodiment
The present invention is further introduced below with reference to embodiment:
A kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction, comprising the following steps:
Step 1 measures Mineral spectra；
Step 2 carries out denoising to mineral spectra by averaging method；
Step 3 determines some column wavelet basis included in Wavelet Cluster type to be selected and every kind of Wavelet Cluster；
Step 4 carries out wavelet decomposition to target mineral spectral signal using the wavelet basis series in different Wavelet Clusters；
Step 5 calculates the energy and aromatic entropy of decomposition scales at different levels according to coefficient of wavelet decomposition；
Step 6 calculates the ratio of energy and aromatic entropy；
Step 7 extracts mineral spectra feature according to selected Optimal wavelet bases decomposition coefficient.
In the step 1, the kind mineral spectra curve that measurement obtains is denoted as Xi(λ), wherein λ is spectral band serial number,
I is the number of repeatedly measurement mineral spectra.
In the step 2, denoising is carried out to mineral spectra by averaging method, certain mineral light obtained to measurement
Compose Xi(λ) is averaged after summing according to corresponding wave band, it is assumed that has carried out n times spectral measurement to mineral altogether, averaging method is to mineral
The formula of spectrum denoising are as follows:
In the step 3, a series of wavelet basis for including in the type and every kind of Wavelet Cluster of Wavelet Cluster to be selected are determined,
Wavelet Cluster type and the wavelet basis for being included series are chosen according to Wavelet Cluster property.
In the step 4, small wavelength-division is carried out to target mineral spectral signal using the wavelet basis series in different Wavelet Clusters
Solution, set wavelet basis function asDecomposition order j, applicationWavelet basis carries out j layers to the mineral spectra signal Z (λ) after denoising
Discrete wavelet transformation, obtaining decomposition coefficient sequence is Cj(k) (k=1,2 ... 2j), represent k-th of decomposition coefficient of jth layer.
In the step 5, the energy and aromatic entropy of decomposition scales at different levels are calculated according to coefficient of wavelet decomposition, wherein each
Decomposition layer energy can be calculated by the quadratic sum of this layer of decomposition coefficient, i.e.,
Probability distribution P (k) calculation formula of wavelet decomposition jth k-th of node energy of layer is
Pj(k)=| C (k) |2/E(j) (3)
Aromatic entropy F (j) is used to describe the Energy distribution of coefficient of wavelet decomposition, its calculation formula is
In the step 6, calculate energy and aromatic entropy ratio R (j), using the corresponding wavelet basis of maximum R (j) value as
Optimal wavelet bases；Its calculation formula is
R (j)=E (j)/F (j) (5).
Embodiment
1 the present invention will be described in detail with reference to the accompanying drawing.
The present invention is a kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction, the specific steps are as follows:
Step 1, measurement Mineral spectra, by ASD spectrometer measurement mineral spectra, certain mineral light for measuring
Spectral curve is denoted as Xi(λ), wherein λ is spectral band serial number, and λ sum is 2151 in this example, and i is to measure same mineral spectra
Number, measurement sum i is 30 in this example.
Step 2 carries out denoising to mineral spectra by averaging method, certain mineral spectra X obtained to measurementi(λ)
It is averaged after summing according to corresponding wave band, it is assumed that n times spectral measurement has been carried out to mineral, averaging method denoises mineral spectra
Formula are as follows:
A series of wavelet basis for including in step 3, the type for determining Wavelet Cluster to be selected and every kind of Wavelet Cluster, in invention
Any restriction do not done to Wavelet Cluster type, but as an example, the Wavelet Cluster in this example selects db, symlet and coiflet, it is small
The wavelet basis quantity that wave cluster includes can be adjusted according to demand, in this example db small echo series include db1, db2, db3,
Db4, db5, db6, db7, db8, db9 and db10；Symlet small echo series include sym1, sym2, sym3, sym4, sym5,
Sym6, sym7, sym8, sym9 and sym10；Coiflet small echo series include coif1, coif2, coif3, coif4, coif5 and
coif6。
Step 4, the wavelet basis series determined using step 3 carry out wavelet decomposition to target mineral spectral signal respectively, will
Wavelet basis function is denoted asDecomposition order j is set as 5 layers, applicationWavelet basis carries out j to the mineral spectra signal Z (λ) after denoising
Layer scattering wavelet decomposition, obtaining decomposition coefficient sequence is Cj(k) (k=1,2 ... 2j), represent k-th of decomposition coefficient of jth layer.
Step 5, the energy and aromatic entropy that decomposition scales at different levels are calculated according to coefficient of wavelet decomposition, wherein each decomposition layer energy
Amount can be calculated by the quadratic sum of this layer of decomposition coefficient, i.e.,
Probability distribution P (k) calculation formula of wavelet decomposition jth k-th of node energy of layer is
Pj(k)=| C (k) |2/E(j) (3)
Aromatic entropy F (j) is used to describe the Energy distribution of coefficient of wavelet decomposition, its calculation formula is
Step 6, the ratio R (j) for calculating energy and aromatic entropy, using the corresponding wavelet basis of maximum R (j) value as optimal wavelet
Base.Its calculation formula is
R (j)=E (j)/F (j) (5)
Step 7 extracts mineral spectra feature according to selected Optimal wavelet bases decomposition coefficient.Using Optimal wavelet bases to mineral
Spectral signal carries out 5 layers of wavelet decomposition, obtains detail wavelet coefficients (D1,D2,D3,D4,D5) and approximation wavelet coefficients A1As light
The result that spectrum signature is extracted.
Claims (7)
1. a kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction, it is characterised in that: the following steps are included:
Step 1 measures Mineral spectra；
Step 2 carries out mineral spectra by averaging method to remove dryness processing；
Step 3 determines some column wavelet basis included in Wavelet Cluster type to be selected and every kind of Wavelet Cluster；
Step 4 carries out wavelet decomposition to target mineral spectral signal using the wavelet basis series in different Wavelet Clusters；
Step 5 calculates the energy and aromatic entropy of decomposition scales at different levels according to coefficient of wavelet decomposition；
Step 6 calculates the ratio of energy and aromatic entropy；
Step 7 extracts mineral spectra feature according to selected Optimal wavelet bases decomposition coefficient.
2. the method that Optimal wavelet bases are chosen in a kind of mineral spectra feature extraction according to claim 1, feature exist
In: wherein in the step 1, the mineral spectra curve that measurement obtains is denoted as Xi(λ), wherein λ is spectral band serial number, and i is
The repeatedly number of measurement mineral spectra.
3. the method that Optimal wavelet bases are chosen in a kind of mineral spectra feature extraction according to claim 1, feature exist
In: in the step 2, denoising is carried out to mineral spectra by averaging method, the mineral spectra X obtained to measurementi(λ) is pressed
Photograph is averaged after answering wave band to sum, it is assumed that has carried out n times spectral measurement to mineral altogether, averaging method denoises mineral spectra
Formula are as follows:
4. the method that Optimal wavelet bases are chosen in a kind of mineral spectra feature extraction according to claim 1, feature exist
In: in the step 3, a series of wavelet basis for including in the type and every kind of Wavelet Cluster of Wavelet Cluster to be selected are determined, according to small
Wave cluster property chooses Wavelet Cluster type and the wavelet basis for being included series.
5. the method that Optimal wavelet bases are chosen in a kind of mineral spectra feature extraction according to claim 1, feature exist
In: in the step 4, wavelet decomposition is carried out to target mineral spectral signal using the wavelet basis series in different Wavelet Clusters, if
Determining wavelet basis function isDecomposition order j, applicationWavelet basis carries out j layer scattering to the mineral spectra signal Z (λ) after denoising
Wavelet decomposition, obtaining decomposition coefficient sequence is Cj(k) (k=1,2 ... 2j), represent k-th of decomposition coefficient of jth layer.
6. the method that Optimal wavelet bases are chosen in a kind of mineral spectra feature extraction according to claim 1, feature exist
In: in the step 5, the energy and aromatic entropy of decomposition scales at different levels are calculated according to coefficient of wavelet decomposition, wherein each decomposition layer
Energy can be calculated by the quadratic sum of this layer of decomposition coefficient, i.e.,
Probability distribution P (k) calculation formula of wavelet decomposition jth k-th of node energy of layer is
Pj(k)=| C (k) |2/E(j) (3)
Aromatic entropy F (j) is used to describe the Energy distribution of coefficient of wavelet decomposition, its calculation formula is
7. the method that Optimal wavelet bases are chosen in a kind of mineral spectra feature extraction according to claim 1, feature exist
In: in the step 6, the ratio R (j) of energy and aromatic entropy is calculated, using the corresponding wavelet basis of maximum R (j) value as optimal small
Wave base；Its calculation formula is
R (j)=E (j)/F (j) (5).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811551730.5A CN109785272A (en) | 2018-12-19 | 2018-12-19 | A kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811551730.5A CN109785272A (en) | 2018-12-19 | 2018-12-19 | A kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109785272A true CN109785272A (en) | 2019-05-21 |
Family
ID=66497267
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811551730.5A Pending CN109785272A (en) | 2018-12-19 | 2018-12-19 | A kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109785272A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110765881A (en) * | 2019-09-25 | 2020-02-07 | 哈尔滨工程大学 | Wavelet basis selection method based on principal component analysis |
| CN111027488A (en) * | 2019-12-11 | 2020-04-17 | 深圳先进技术研究院 | Signal classification method and device |
| CN112711986A (en) * | 2020-12-09 | 2021-04-27 | 核工业北京地质研究院 | Hyperspectral remote sensing altered mineral extraction method |
| CN112711980A (en) * | 2020-11-27 | 2021-04-27 | 核工业北京地质研究院 | Method for selecting wavelet base in mineral spectral feature extraction |
| CN113034645A (en) * | 2021-03-23 | 2021-06-25 | 中国地质科学院地质力学研究所 | Lithologic map filling method, device and storage medium |
| CN113324918A (en) * | 2021-01-26 | 2021-08-31 | 核工业北京地质研究院 | Rock spectrum denoising method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101101338A (en) * | 2007-07-10 | 2008-01-09 | 廊坊开发区中油油田科技工贸有限责任公司 | Exploration method and system for oil and gas |
| CN106126938A (en) * | 2016-06-27 | 2016-11-16 | 中国科学院电子学研究所 | A kind of Optimal wavelet bases choosing method and wavelet threshold denoising method |
| CN104076003B (en) * | 2014-07-04 | 2017-01-11 | 核工业北京地质研究院 | Extraction method of mineral spectrum absorption characteristic parameters |
-
2018
- 2018-12-19 CN CN201811551730.5A patent/CN109785272A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101101338A (en) * | 2007-07-10 | 2008-01-09 | 廊坊开发区中油油田科技工贸有限责任公司 | Exploration method and system for oil and gas |
| CN104076003B (en) * | 2014-07-04 | 2017-01-11 | 核工业北京地质研究院 | Extraction method of mineral spectrum absorption characteristic parameters |
| CN106126938A (en) * | 2016-06-27 | 2016-11-16 | 中国科学院电子学研究所 | A kind of Optimal wavelet bases choosing method and wavelet threshold denoising method |
Non-Patent Citations (1)
| Title |
|---|
| MEHDIABDOLMALEKI 等: ""Selecting optimum base wavelet for extracting spectral alteration features associated with porphyry copper mineralization using hyperspectral images"", 《INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATIONS AND GEOINFORMATION》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110765881A (en) * | 2019-09-25 | 2020-02-07 | 哈尔滨工程大学 | Wavelet basis selection method based on principal component analysis |
| CN110765881B (en) * | 2019-09-25 | 2022-08-02 | 哈尔滨工程大学 | Wavelet basis selection method based on principal component analysis |
| CN111027488A (en) * | 2019-12-11 | 2020-04-17 | 深圳先进技术研究院 | Signal classification method and device |
| WO2021115059A1 (en) * | 2019-12-11 | 2021-06-17 | 深圳先进技术研究院 | Signal classification method and device |
| CN112711980A (en) * | 2020-11-27 | 2021-04-27 | 核工业北京地质研究院 | Method for selecting wavelet base in mineral spectral feature extraction |
| CN112711986A (en) * | 2020-12-09 | 2021-04-27 | 核工业北京地质研究院 | Hyperspectral remote sensing altered mineral extraction method |
| CN113324918A (en) * | 2021-01-26 | 2021-08-31 | 核工业北京地质研究院 | Rock spectrum denoising method |
| CN113324918B (en) * | 2021-01-26 | 2023-01-24 | 核工业北京地质研究院 | Rock spectrum denoising method |
| CN113034645A (en) * | 2021-03-23 | 2021-06-25 | 中国地质科学院地质力学研究所 | Lithologic map filling method, device and storage medium |
| CN113034645B (en) * | 2021-03-23 | 2021-09-03 | 中国地质科学院地质力学研究所 | Lithologic map filling method, device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109785272A (en) | A kind of method that Optimal wavelet bases are chosen in mineral spectra feature extraction | |
| Wang | Multichannel matching pursuit for seismic trace decomposition | |
| Klinkenberg | Fractals and morphometric measures: is there a relationship? | |
| Pannekoucke et al. | Background‐error correlation length‐scale estimates and their sampling statistics | |
| CN110110738A (en) | A kind of Recognition Method of Radar Emitters based on multi-feature fusion | |
| CN101986139A (en) | Method for inverting vegetation parameters by remote sensing based on reflection spectrum wavelet transform | |
| CN105137373B (en) | A kind of denoising method of exponential signal | |
| Zhang et al. | Separation of magnetotelluric signals based on refined composite multiscale dispersion entropy and orthogonal matching pursuit | |
| Zhang et al. | Grey wolf optimization-based variational mode decomposition for magnetotelluric data combined with detrended fluctuation analysis | |
| Lyubushin | Global seismic noise entropy | |
| Safarkhani et al. | Improving C 1 and C 3 empirical Green’s functions from ambient seismic noise in NW Iran using RMS ratio stacking method | |
| CN107356963B (en) | A kind of adaptive seismic signal coherence body property analysis method of data-driven | |
| Tian et al. | Application of fractional differential calculation in pretreatment of saline soil hyperspectral reflectance data | |
| CN109932682B (en) | Two-dimensional multi-snapshot non-grid compressed beam forming sound source identification method | |
| Kononov et al. | Model-associated forest parameter retrieval using VHF SAR data at the individual tree level | |
| He et al. | Local wavelet packet decomposition of soil hyperspectral for SOM estimation | |
| Li et al. | A multispectral remote sensing data spectral unmixing algorithm based on variational Bayesian ICA | |
| CN113720952B (en) | Method, device, equipment and medium for generating image plate for reservoir interpretation evaluation | |
| CN106054247B (en) | Method for calculating high-precision reflection coefficient based on converted wave seismic data | |
| Zhang et al. | Characterization of fresh milk products based on multidimensional Raman spectroscopy | |
| Garrison | Modeling and simulation of bin-bin correlations in GNSS-R waveforms | |
| CN111398735A (en) | Transformer substation grounding grid fault detection method based on information entropy | |
| Huang et al. | Improved 3-D GPR detection by NUFFT combined with MPD method | |
| CN109782346A (en) | A kind of acquisition footprint drawing method based on anatomic element analysis | |
| Li et al. | Simulation Study on Estimation Method of Spectrally Pure Material Proportion of a Space Target Based on Cost Function Optimization |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190521 |