CN103913765B - A kind of nucleic power spectrum Peak Search Method - Google Patents
A kind of nucleic power spectrum Peak Search Method Download PDFInfo
- Publication number
- CN103913765B CN103913765B CN201410109038.2A CN201410109038A CN103913765B CN 103913765 B CN103913765 B CN 103913765B CN 201410109038 A CN201410109038 A CN 201410109038A CN 103913765 B CN103913765 B CN 103913765B
- Authority
- CN
- China
- Prior art keywords
- peak
- power spectrum
- search method
- length
- follows
- 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.)
- Expired - Fee Related
Links
- 238000001228 spectrum Methods 0.000 title claims abstract description 53
- 238000000034 method Methods 0.000 title claims abstract description 31
- 230000003595 spectral effect Effects 0.000 claims abstract description 14
- 238000009499 grossing Methods 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 6
- 230000000877 morphologic effect Effects 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 238000005260 corrosion Methods 0.000 claims description 2
- 230000007797 corrosion Effects 0.000 claims description 2
- 230000010339 dilation Effects 0.000 claims description 2
- 238000001514 detection method Methods 0.000 abstract description 2
- 230000005855 radiation Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 8
- 230000001629 suppression Effects 0.000 description 2
Abstract
The present invention relates to nuclear radiation detection technical field, it is provided that a kind of nucleic power spectrum Peak Search Method, the method, according to the global feature of energy spectral peak, uses Mathematical Morphology Method to extract energy spectral peak, is not easily susceptible to local noise and the impact of distortion in power spectrum.The inventive method can overcome local noise and the impact of distortion in power spectrum, improves the stability of Peak Search Method.
Description
Technical field
The present invention relates to nuclear radiation detection technical field, specifically a kind of nucleic power spectrum Peak Search Method.
Background technology
During Radionuclide analysis and identification, sixty-four dollar question is the peak position positioning each nucleic peak accurately from power spectrum.Only search out each energy spectral peak tram, it could be further analyzed and identify.Constantly increasing and the impact of statistic fluctuation of power spectrum complexity, proposes bigger challenge to power spectrum peak-seeking.
The most frequently used power spectrum Peak Search Method is derivative method peak-seeking, and gamma-spectrometric data is regarded as a continuous print curve by it, positions power spectrum peak-to-peak position and border by calculating power spectrum single order, second order, the extreme value of three order derivatives, zero crossing.Wherein summit is positioned at first derivative by just to negative zero crossing, peak position be positioned at the negative local minimum of second dervative and three order derivatives by bearing positive zero crossing.Owing to differentiating, the saltus step to local data is the most sensitive, if therefore the noise in power spectrum is relatively big or occurs in that distortion locally, then one can may be detected multiple first derivative zero crossing in the range of spectral peak, derivative method peak-seeking was lost efficacy, thus causes leakage peak or a peak to be detected as multiple ghost peak.
Summary of the invention
The purpose of the present invention is contemplated to overcome above-mentioned technical problem, and provides a kind of power spectrum Peak Search Method, and the method, according to the global feature of energy spectral peak, uses Mathematical Morphology Method to extract energy spectral peak, is not easily susceptible to local noise and the impact of distortion in power spectrum.
In order to achieve the above object, the technical solution of the present invention is as follows.
A kind of nucleic power spectrum Peak Search Method, the method comprises the following steps:
Step 1, smooths original power spectrum, removes noise；
Step 2, carries out the conversion of morphology white cap to the power spectrum after smooth, power spectrum is divided into several null values and nonzero value region；
Step 3, adds up the length in each nonzero value region, filters the length ghost peak less than the minimum peak width threshold value set；
Step 4, in each nonzero value region, finds power spectrum peak-to-peak position and right boundary；
Step 5, according to calculated peak position, peak height and right boundary to can spectral peak judge.
What the Peak Search Method in the present invention was investigated is entirety morphology feature rather than the extreme value of local of energy spectral peak, therefore, it is possible to overcome local noise and the impact of distortion in power spectrum, improves the stability of Peak Search Method.
Accompanying drawing explanation
Fig. 1 is original gamma-spectrometric data pending in the present invention.
Fig. 2 is the gamma-spectrometric data schematic diagram in the present invention after energy spectrum smoothing.
Fig. 3 is the gamma-spectrometric data schematic diagram in the present invention after the white cap of morphology converts.
Fig. 4 is the gamma-spectrometric data schematic diagram after filtering ghost peak in the present invention.
Fig. 5 is the power spectrum peak-seeking result schematic diagram according to the embodiment of the present invention.
Detailed description of the invention
For making the object, technical solutions and advantages of the present invention clearer, below in conjunction with specific embodiment, and referring to the drawings, the present invention is described in more detail.
A kind of nucleic power spectrum Peak Search Method that the present invention proposes specifically includes following steps:
Step 1, carries out Gaussian smoothing, suppression noise to original power spectrum:
Power spectrum completes smoothly through gaussian filtering, and filtering can be expressed as a convolution process, and expression formula is:
WhereinfFor original power spectrum,yFor filtered power spectrum,mFor the length of gamma-spectrometric data,g i For Gaussian convolution core, expression formula is:
WhereinσFor Gaussian function standard deviation, the length of Gaussian convolution core is equal to 2K+1。σThe biggest, Gaussian convolution core is the longest, then the dynamics that smooths is the strongest, and the suppression to noise is the strongest, but can filter tiny energy spectral peak and cause leakage peak.Therefore, in actual power spectrum smoothing process, the noise characteristic according to original power spectrum is needed to select length and the Gaussian function variance of Gaussian convolution coreσValue.General convolution kernel length takes between 3 ~ 11,σValue takes between 1-3.Fig. 1 is pending original gamma-spectrometric data.Fig. 2 is the gamma-spectrometric data schematic diagram after energy spectrum smoothing, whereinK=3,σ=1。
Step 2, carries out the white cap of morphology according to the suitable morphological structuring elements of feature selection of energy spectral peak to be detected to the power spectrum after smoothing and converts:
Use morphological structuring elementsB, to power spectrumyCarry out morphology white cap conversion expression formula as follows:
WhereinFor morphology opening operation, expression formula is as follows:
WhereinWithBeing respectively morphological dilations and corrosion change, expression formula is as follows:
Structural elementBUse the straightway parallel with abscissa, the length of structural elementLFor odd number,LBe set as peak to be detected minimum peak width two extraordinarily 1.Fig. 3 show the gamma-spectrometric data schematic diagram after the white cap of morphology converts, whereinL=31.After converting through the white cap of morphology, power spectrum is divided into several null values and nonzero value region.
Step 3, in the data after converting through white cap, adds up the length in each nonzero value region, if this length is less than the minimum peak width threshold value set, then all of data in this region is set to null value, to filter ghost peak.Minimum peak width threshold value refers to the minimum widith at peak detected by this Peak Search Method, will be considered as ghost peak less than the peak of this width, and this minimum peak width threshold value typically sets manually.Such as, power spectrum peak-to-peak to be detected in a power spectrum is wide by both greater than 10, then this also threshold value be just set as 10.Fig. 4 is the gamma-spectrometric data schematic diagram after filtering ghost peak, and wherein minimum peak width threshold value is set as 20.
Step 4, in each nonzero value region, finds the maximum in this regiony p , the position at this value place is peak positionx p .With peak positionx p Peak left margin, i.e. x=is found to the left for starting pointx p , x p -1,
x p -2 ... 0, until any one in following boundary search condition is unsatisfactory for.
NowxValue is peak left marginx L 。
With peak positionx p Peak right margin, i.e. x=is found to the right for starting pointx p , x p +1,
x p +2,…N, until any one in following condition is unsatisfactory for.
NowxValue is peak left marginx R 。
WhereinCFor confidence level coefficient constant, typically take between 1-3.
Step 5, according to calculated peak position, peak height and right boundary to can spectral peak judge, it determines formula is as follows:
Wherein,H p For (x L , y L ) and (x R ,y R ) in peak positionx p The first-order linear interpolation at place.TFor peak-seeking threshold value set in advance, typically take between 1-3.If meeting discriminant, then it is assumed that this peak is real peak.It it is the power spectrum peak-seeking result schematic diagram of embodiment shown in Fig. 5.
Claims (5)
1. a nucleic power spectrum Peak Search Method, it is characterised in that the method comprises the following steps:
Step 1, smooths original power spectrum, removes noise；
Step 2, carries out the conversion of morphology white cap to the power spectrum after smooth, power spectrum is divided into several null values and nonzero value region；The suitable morphological structuring elements of feature selection according to energy spectral peak to be detected carries out the white cap of morphology to the power spectrum after smoothing and converts:
Use morphological structuring elementsB, to power spectrumyCarry out morphology white cap conversion expression formula as follows:
WhereinFor morphology opening operation, expression formula is as follows:
WhereinWithBeing respectively morphological dilations and corrosion change, expression formula is as follows:
Structural elementBUse the straightway parallel with abscissa, the length of structural elementLFor odd number,LBe set as peak to be detected minimum peak width two extraordinarily 1, through the white cap of morphology convert after, power spectrum is divided into several null values and nonzero value region；
Step 3, adds up the length in each nonzero value region, filters the length ghost peak less than the minimum peak width threshold value set；
Step 4, in each nonzero value region, finds power spectrum peak-to-peak position and right boundary；
Step 5, according to calculated peak position, peak height and right boundary to can spectral peak judge.
Nucleic power spectrum Peak Search Method the most according to claim 1, it is characterised in that described step 1 particularly as follows:
Power spectrum completes smoothly through gaussian filtering, and filtering is expressed as a convolution process, and expression formula is:
WhereinfFor original power spectrum,yFor filtered power spectrum,mFor the length of gamma-spectrometric data,g i For Gaussian convolution core, expression formula is:
WhereinσFor Gaussian function standard deviation, the length of Gaussian convolution core is equal to 2K+ 1, in actual power spectrum smoothing process, convolution kernel length takes between 3 ~ 11, and σ value takes between 1 ~ 3.
Nucleic power spectrum Peak Search Method the most according to claim 1, it is characterized in that, described step 3 particularly as follows: through white cap convert after data in, add up the length in each nonzero value region, if this length is less than the minimum peak width threshold value set, then all of data in this region are set to null value, to filter ghost peak；Described minimum peak width threshold value refers to the minimum widith at peak detected by this Peak Search Method, will be considered as ghost peak less than the peak of this width.
Nucleic power spectrum Peak Search Method the most according to claim 1, it is characterised in that described step 4 particularly as follows:
In each nonzero value region, find the maximum in this regiony p , the position at this value place is peak positionx p , with peak positionx p Peak left margin, i.e. x=is found to the left for starting pointx p , x p -1,
x p -2 ... 0, until any one in following boundary search condition is unsatisfactory for,
NowxValue is peak left marginx L ；
With peak positionx p Peak right margin, i.e. x=is found to the right for starting pointx p , x p +1,
x p +2,…N, until any one in following condition is unsatisfactory for,
NowxValue is peak left marginx R ,
WhereinCFor confidence level coefficient constant, take between 1-3.
Nucleic power spectrum Peak Search Method the most according to claim 1, it is characterised in that described step 5 particularly as follows:
According to calculated peak position, peak height and right boundary to can spectral peak judge, it determines formula is as follows:
Wherein,H p For (x L ,
y L ) and (x R ,y R ) in peak positionx p The first-order linear interpolation at place,TFor peak-seeking threshold value set in advance, take between 1-3, if meeting discriminant, then it is assumed that this peak is real peak.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410109038.2A CN103913765B (en) | 2014-03-24 | 2014-03-24 | A kind of nucleic power spectrum Peak Search Method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410109038.2A CN103913765B (en) | 2014-03-24 | 2014-03-24 | A kind of nucleic power spectrum Peak Search Method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103913765A CN103913765A (en) | 2014-07-09 |
| CN103913765B true CN103913765B (en) | 2016-08-31 |
Family
ID=51039590
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410109038.2A Expired - Fee Related CN103913765B (en) | 2014-03-24 | 2014-03-24 | A kind of nucleic power spectrum Peak Search Method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103913765B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107179310B (en) * | 2017-06-01 | 2018-02-27 | 温州大学 | Raman spectrum characteristic peak recognition methods based on robust noise variance evaluation |
| CN108333617B (en) * | 2018-01-12 | 2019-06-25 | 山东省科学院海洋仪器仪表研究所 | The quick Peak Search Method of radioactive material quality detection in a kind of seawater |
| CN108375783B (en) * | 2018-02-07 | 2019-06-25 | 山东省科学院海洋仪器仪表研究所 | A kind of automatic peak-seeking method of seawater radiological measuring |
| CN109696702B (en) * | 2019-01-22 | 2022-08-26 | 山东省科学院海洋仪器仪表研究所 | Overlapping peak judgment method for detection of seawater radionuclide K40 |
| CN111861944B (en) * | 2020-08-07 | 2023-06-30 | 温州理工学院 | Nuclide energy spectrum peak searching method based on morphology of multi-structure elements |
| CN113607867A (en) * | 2021-07-23 | 2021-11-05 | 清华大学合肥公共安全研究院 | Dual-fold-spectrum peak analysis method based on peak body mapping |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101692149A (en) * | 2009-09-30 | 2010-04-07 | 武汉理工大学 | Wave crest seeking method and wave crest seeking system |
| CN102098101A (en) * | 2011-01-04 | 2011-06-15 | 武汉理工大学 | Crest searching method and system |
| JP2011257206A (en) * | 2010-06-08 | 2011-12-22 | Shimadzu Corp | Peak detection method and device |
| CN102435214A (en) * | 2011-09-20 | 2012-05-02 | 华北电力大学（保定） | High-precision and fast optical fiber grating spectrum peak-searching method |
| CN102706372A (en) * | 2012-03-28 | 2012-10-03 | 中国航空工业集团公司北京长城计量测试技术研究所 | Optical fiber wavelength demodulating light spectrum peak value positioning method |
| WO2013082010A1 (en) * | 2011-11-30 | 2013-06-06 | Molecular Devices, Llc | System and method for determining a baseline measurement for a biological response curve |
| CN103487074A (en) * | 2013-10-12 | 2014-01-01 | 重庆邮电大学 | Method for processing FBG (fiber bragg grating) sensing signal by utilizing three-point peek-seeking algorithm |
| CN103604449A (en) * | 2013-11-15 | 2014-02-26 | 武汉理工光科股份有限公司 | FPGA-based quick peak searching method and system |
-
2014
- 2014-03-24 CN CN201410109038.2A patent/CN103913765B/en not_active Expired - Fee Related
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101692149A (en) * | 2009-09-30 | 2010-04-07 | 武汉理工大学 | Wave crest seeking method and wave crest seeking system |
| JP2011257206A (en) * | 2010-06-08 | 2011-12-22 | Shimadzu Corp | Peak detection method and device |
| CN102098101A (en) * | 2011-01-04 | 2011-06-15 | 武汉理工大学 | Crest searching method and system |
| CN102435214A (en) * | 2011-09-20 | 2012-05-02 | 华北电力大学（保定） | High-precision and fast optical fiber grating spectrum peak-searching method |
| WO2013082010A1 (en) * | 2011-11-30 | 2013-06-06 | Molecular Devices, Llc | System and method for determining a baseline measurement for a biological response curve |
| CN102706372A (en) * | 2012-03-28 | 2012-10-03 | 中国航空工业集团公司北京长城计量测试技术研究所 | Optical fiber wavelength demodulating light spectrum peak value positioning method |
| CN103487074A (en) * | 2013-10-12 | 2014-01-01 | 重庆邮电大学 | Method for processing FBG (fiber bragg grating) sensing signal by utilizing three-point peek-seeking algorithm |
| CN103604449A (en) * | 2013-11-15 | 2014-02-26 | 武汉理工光科股份有限公司 | FPGA-based quick peak searching method and system |
Non-Patent Citations (3)
| Title |
|---|
| Beam search for peak alignment of NMR signals;Geun-Cheol Lee 等;《Analytica Chimica Acta》;20041231;第513卷;第413-416页 * |
| 基于可调谐F-P滤波解调系统的直接寻峰算法与功率加权寻峰算法;孙金卫 等;《四川兵工学报》;20110630;第32卷(第6期);第83-85页 * |
| 复杂背景中小目标的探测与识别技术研究;刘希佳;《中国优秀硕士学位论文数据库 信息科技辑》;20130215(第02期);I138-I844 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103913765A (en) | 2014-07-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103913765B (en) | A kind of nucleic power spectrum Peak Search Method | |
| CN109035163B (en) | Self-adaptive image denoising method based on deep learning | |
| Kanagasundaram et al. | Improving short utterance i-vector speaker verification using utterance variance modelling and compensation techniques | |
| EP2945096A1 (en) | Character recognition method | |
| CN107220988B (en) | Part image edge extraction method based on improved canny operator | |
| CN103942797A (en) | Scene image text detection method and system based on histogram and super-pixels | |
| CN106685478B (en) | Frequency hopping signal parameter estimation method based on signal time-frequency image information extraction | |
| JP2012016453A5 (en) | ||
| CN105793892A (en) | Image processing method and apparatus and photographing device | |
| CN104112261A (en) | Norm ratio regularization based rapid image blind deblurring method | |
| KR100887183B1 (en) | Preprocessing apparatus and method for illumination-invariant face recognition | |
| DARUS | Coefficient estimates of classes of q-starlike and q-convex functions | |
| US20130094768A1 (en) | Face-Tracking Method with High Accuracy | |
| Dong et al. | Image retrieval based on improved Canny edge detection algorithm | |
| CN104036517B (en) | Image matting method based on gradient sampling | |
| CN103997592A (en) | Method and system for video noise reduction | |
| CN106770192B (en) | Laser induced breakdown spectroscopy continuous background bearing calibration based on interpolation method | |
| CN103136530A (en) | Method for automatically recognizing target images in video images under complex industrial environment | |
| CN102073850A (en) | Image processing based quasi-cylinder counting statistic method | |
| CN105488523A (en) | Data clustering analysis method based on Grassmann manifold | |
| CN102279925B (en) | Chain processing face recognition method and system | |
| CN112560629B (en) | Smart phone key detection method based on 1D-CNN | |
| CN102346830B (en) | Gradient histogram-based virus detection method | |
| CN111274858A (en) | Business license identification method in network transaction supervision | |
| CN111414821B (en) | Target detection method and related device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160831 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |