CN109272030A - Apple surface earlier damage quick nondestructive recognition methods based on fiber spectrum technology - Google Patents
Apple surface earlier damage quick nondestructive recognition methods based on fiber spectrum technology Download PDFInfo
- Publication number
- CN109272030A CN109272030A CN201811021625.0A CN201811021625A CN109272030A CN 109272030 A CN109272030 A CN 109272030A CN 201811021625 A CN201811021625 A CN 201811021625A CN 109272030 A CN109272030 A CN 109272030A
- Authority
- CN
- China
- Prior art keywords
- apple
- sample
- test samples
- surface earlier
- damage
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
-
- 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
- G06V10/58—Extraction of image or video features relating to hyperspectral data
Abstract
The apple surface earlier damage quick nondestructive recognition methods based on fiber spectrum technology that the invention discloses a kind of has slight damage apple sample to be randomly assigned, establishes calibration samples collection and test samples collection method includes the following steps: collecting intact and surface；The spectral reflectivity that apple sample is concentrated using spectra collection system acquisition calibration samples collection and test samples, obtains correction and test samples collection original spectral data；Original spectral data is pre-processed using standard normal transformation (SNV), and dimensionality reduction is carried out to pretreated spectroscopic data using principal component analytical method, to extract the characteristic spectrum that can reflect apple surface earlier damage, correction and test samples collection property data base are established；Finally, establishing the identification model of apple surface earlier damage using simplified K arest neighbors (SKNN) mode identification method.The present invention is based on fiber spectrum technology combination Chemical Measurement, can quickly, non-damage drive go out the apple that there is slight damage on surface.
Description
Technical field
The present invention relates to fruit surface earlier damage technical field of nondestructive testing, more particularly to one kind to be based on fiber spectrum skill
The apple surface earlier damage quick nondestructive recognition methods of art.
Background technique
Apple is in world's fruit in the market in occupation of very important status.Because it is rich in vitamin abundant, it is
The big apples in the world four, grape, citrus and banana hat.However, apple is in picking or transportational process, because colliding and squeezing
Caused by surface earlier damage be difficult visually to be identified, this will result directly in makes during preservation and freshness because of surface earlier damage
At large area fester or infect and bring economic loss.It can be seen that being detected to the surface earlier damage of fresh apple aobvious
It obtains particularly important.Traditional detection method is manual operation mostly, is taken time and effort, and low efficiency, is unable to satisfy extensive life
The demand of production.Therefore, quick, lossless, the efficient apple surface earlier damage detection method of one kind is developed in fruit grading
In field with good application prospect.
Fiber spectrum technology has quick, lossless, easy, green ring as a kind of novel qualitative and quantitative analysis technology
The advantages that guarantor, has been widely used in the industries such as agricultural, food, medicine, petroleum.Fiber spectrum technology is by optical fiber technology and spectrum
Technology is dexterously combined together, and is used optical fibers as sample to be tested detection probe, is carried out non-destructive testing to sample to be tested.According to
The apple spectral reflectivity difference that the intact and surface measured is had damage, establishes characteristic spectrum database, establishes apple table
The identification model of face earlier damage, to realize that the quick nondestructive of the apple surface earlier damage based on fiber spectrum technology is known
Not.
Summary of the invention
The purpose of the present invention is to provide a kind of, and the apple surface earlier damage quick nondestructive based on fiber spectrum technology is known
Other method, it is intended to realize quick, lossless identification.
A kind of technical solution of the present invention: apple surface earlier damage quick nondestructive identification side based on fiber spectrum technology
Method, comprising the following steps:
1) foundation of sample spectrum:
Collecting intact and surface has slight damage apple sample to be randomly assigned, and establishes calibration samples collection and test samples collection；
The spectral reflectivity that sample is concentrated using the correction of spectra collection system acquisition and test samples, obtains correction and test samples collection is former
Beginning spectroscopic data；
2) sample original spectral data pre-processes:
Correction and test samples collection original spectrum are pre-processed using standard normal transformation (SNV), to eliminate surface scattering
And influence of the change in optical path length to spectrum；
3) foundation of sample characteristics spectra database:
Dimensionality reduction is carried out to pretreated spectroscopic data using principal component analytical method, can reflect that apple surface early stage damages to extract
The characteristic spectrum of wound establishes correction and test samples collection sample characteristics spectra database；
4) identification model is established:
The identification of apple surface earlier damage is established using simplified K arest neighbors (SKNN) mode identification method combination Chemical Measurement
Model, then to model correction and forecast assessment；
5) verifying of identification model:
Testing model is distinguished to the correct recognition rata of apple surface earlier damage using calibration set sample and inspection set sample.
Spectra collection system in the step 1) include: high-performance optical spectrometer QEPro, optical fiber R600-7-VIS-125F,
Halogen tungsten lamp light source HL-2000, standard reflection blank WS-1, reflection probe bracket RPH-ADP and RPH-1 and computer.
The time of integration of spectra collection system in the step 1) is 110ms, Multiple-Scan average out to 8, sliding average
Width is 1, and spectra collection range is 200 ~ 1000nm, and spectral band is 1024.
The step 2) and step 3) is using MATLAB R2016b software realization to the pre- place of sample original spectral data
Reason and sample characteristics spectra database.
It is 0.99 preceding 7 principal components as sample that the step 3), which uses Principal Component Analysis to select contribution rate of accumulative total,
Collect characteristic spectrum data.
Principal component analytical method provided by the invention effectively eliminates correlation present in original spectral data and letter
Redundancy is ceased, realizes the dimensionality reduction of spectroscopic data well, is conducive to carry out depth excavation to original spectral data.
The present invention establishes the normal fruit of characterization apple and surface has damage the feature of fruit by the spectral reflectivity of acquisition apple
Database establishes apple surface early stage damage using simplified K arest neighbors (SKNN) mode identification method combination Chemical Measurement
The identification model of wound provides a kind of quick, lossless, accurate method for the identification of apple surface earlier damage.
Detailed description of the invention
Fig. 1 is the apple surface earlier damage non-damage drive method provided in an embodiment of the present invention based on fiber spectrum technology
Flow chart；
Fig. 2 is the original spectrum curve graph of apple sample provided in an embodiment of the present invention；
Fig. 3 is provided in an embodiment of the present invention to convert (SNV) treated relative spectral reflectivity curve graph by standard normal.
Specific embodiment
The invention will be further described with reference to the accompanying drawings and examples, and what is be exemplified below is only specific reality of the invention
Example is applied, but protection scope of the present invention is not limited to that.
Referring to Fig. 1, a kind of apple surface earlier damage quick nondestructive recognition methods based on fiber spectrum technology, including with
Lower step:
1. the foundation of sample spectrum:
Collecting intact and surface has slight damage apple sample to be randomly assigned, and establishes calibration samples collection and test samples collection；
The spectral reflectivity of sample is concentrated using the correction of spectra collection system acquisition and test samples, wherein the time of integration is 110ms, more
Secondary scanning average out to 8, sliding average width are 1, acquire 200 ~ 1000nm band spectrum reflectivity, are corrected and test samples
Collect original spectral data；
2. sample original spectral data pre-processes:
Correction and test samples collection original spectral data are pre-processed using standard normal transformation (SNV), to eliminate surface
The influence of scattering and change in optical path length to spectrum；
3. the foundation of sample characteristics spectra database:
Dimensionality reduction is carried out to pretreated spectroscopic data using principal component analytical method, can reflect that apple surface early stage damages to extract
The characteristic spectrum of wound establishes correction and test samples collection sample characteristics spectra database；
4. establishing identification model:
The identification of apple surface earlier damage is established using simplified K arest neighbors (SKNN) mode identification method combination Chemical Measurement
Model, then to model correction and forecast assessment；
5. the verifying of identification model:
Testing model is distinguished to the correct recognition rata of apple surface earlier damage using calibration set sample and inspection set sample.
Embodiment:
1. the foundation of sample spectrum:
Choosing a batch intact and surface uniform in size in local supermarket of Wal-Mart in the present embodiment has slight damage apple sample
This (totally 160) are randomly assigned, and establish calibration samples collection (120) and test samples collection (40)；Utilize spectra collection system acquisition
Correction and test samples concentrate the spectral reflectivity of sample, and wherein the time of integration is 110ms, and Multiple-Scan average out to 8 slides flat
Equal width is 1.
Spectral reflectivity collection process is by OceanView(Ocean Optics, copyright 2013) software control, light
Spectrum acquisition range is 200 ~ 1000nm, within this range totally 1024 wave bands, obtains correction and test samples collection original spectrum number
According to；The original spectrum curve graph of the collected apple sample of the present embodiment is as shown in Figure 2；
2. sample original spectral data pre-processes:
Correction and test samples collection original spectral data are pre-processed using standard normal transformation (SNV), to eliminate surface
The influence of scattering and change in optical path length to spectrum；By standard normal transformation (SNV) treated relative spectral reflectivity curve
Figure is as shown in Figure 3；
3. the foundation of sample characteristics spectra database:
Dimensionality reduction is carried out to pretreated spectroscopic data using principal component analytical method, can reflect that apple surface early stage damages to extract
The characteristic wavelength of wound, by dimensionality reduction technology multiple variable compressions at a few main variables, these principal components can not only
Enough reflect most information of initial data, and it is irrelevant between each principal component, so as to reduce well noise,
Reduce the interference of redundancy.The method has chosen preceding 7 principal components as characteristic variable, accumulates contribution rate to 0.99；
4 establish identification model:
The identification of apple surface earlier damage is established using simplified K arest neighbors (SKNN) mode identification method combination Chemical Measurement
Model, then to model correction and forecast assessment；
The verifying of 5 identification models:
Correct recognition rata using calibration set sample and inspection set sample difference testing model to apple surface earlier damage, school
The correct recognition rata just collected is 100%, the correct recognition rata 95% of inspection set, and whole correct recognition rata is 98.8%.
As can be seen from the above embodiments, the present invention not only can using fiber spectrum technology identification apple surface earlier damage
Realize Fast nondestructive evaluation, and recognition effect is fine.
Finally, the embodiment above of the invention can only all be considered the description of the invention and cannot limit the present invention.
Claims indicate protection scope of the present invention, therefore, with the comparable meaning and scope of claims of the present invention
Interior any change, is all considered as being included within the scope of the claims.
Claims (5)
1. a kind of apple surface earlier damage quick nondestructive recognition methods based on fiber spectrum technology, it is characterised in that: including
Following steps:
1) foundation of sample spectrum:
Collecting intact and surface has slight damage apple sample to be randomly assigned, and establishes calibration samples collection and test samples collection；
The spectral reflectivity that sample is concentrated using the correction of spectra collection system acquisition and test samples, obtains correction and test samples collection is former
Beginning spectroscopic data；
2) sample original spectral data pre-processes:
Correction and test samples collection original spectrum are pre-processed using standard normal transformation (SNV), to eliminate surface scattering
And influence of the change in optical path length to spectrum；
3) foundation of sample characteristics spectra database:
Dimensionality reduction is carried out to pretreated spectroscopic data using principal component analytical method, can reflect that apple surface early stage damages to extract
The characteristic spectrum of wound establishes correction and test samples collection sample characteristics spectra database；
4) identification model is established:
The identification of apple surface earlier damage is established using simplified K arest neighbors (SKNN) mode identification method combination Chemical Measurement
Model, then to model correction and forecast assessment；
5) verifying of identification model:
Testing model is distinguished to the correct recognition rata of apple surface earlier damage using calibration set sample and inspection set sample.
2. the apple surface earlier damage quick nondestructive identification side according to claim 1 based on fiber spectrum technology
Method, it is characterised in that: the spectra collection system in the step 1) includes: high-performance optical spectrometer QEPro, optical fiber R600-7-
VIS-125F, halogen tungsten lamp light source HL-2000, standard reflection blank WS-1, reflection probe bracket RPH-ADP and RPH-1 and meter
Calculation machine.
3. the apple surface earlier damage quick nondestructive identification side according to claim 1 based on fiber spectrum technology
Method, it is characterised in that: the time of integration of the spectra collection system in the step 1) is 110ms, and Multiple-Scan average out to 8 is sliding
Dynamic mean breadth is 1, and spectra collection range is 200 ~ 1000nm, and spectral band is 1024.
4. the apple surface earlier damage quick nondestructive identification side according to claim 1 based on fiber spectrum technology
Method, it is characterised in that: the step 2) and step 3) is using MATLAB R2016b software realization to sample original spectral data
Pretreatment and sample characteristics spectra database foundation.
5. the apple surface earlier damage quick nondestructive identification side according to claim 1 based on fiber spectrum technology
Method, it is characterised in that: the step 3) uses Principal Component Analysis that contribution rate of accumulative total is selected to make for 0.99 preceding 7 principal components
For sample set characteristic spectrum data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811021625.0A CN109272030A (en) | 2018-09-03 | 2018-09-03 | Apple surface earlier damage quick nondestructive recognition methods based on fiber spectrum technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811021625.0A CN109272030A (en) | 2018-09-03 | 2018-09-03 | Apple surface earlier damage quick nondestructive recognition methods based on fiber spectrum technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109272030A true CN109272030A (en) | 2019-01-25 |
Family
ID=65187406
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811021625.0A Pending CN109272030A (en) | 2018-09-03 | 2018-09-03 | Apple surface earlier damage quick nondestructive recognition methods based on fiber spectrum technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109272030A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109946243A (en) * | 2019-03-26 | 2019-06-28 | 贵阳学院 | The method of Fast nondestructive evaluation plum soluble solid content |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102967578A (en) * | 2012-11-08 | 2013-03-13 | 中国农业科学院农业质量标准与检测技术研究所 | Method for obtaining near-infrared spectrum of beef sample online and application thereof in evaluating beef quality |
| CN104181170A (en) * | 2014-09-05 | 2014-12-03 | 熊菊莲 | Fruit appearance detection method based on spectrum image analysis |
| CN104359838A (en) * | 2014-11-11 | 2015-02-18 | 上海理工大学 | Non-destructive detection method for hardness of pear |
| US20170160200A1 (en) * | 2014-05-26 | 2017-06-08 | Hitachi, Ltd. | Optical analysis device |
| CN107064047A (en) * | 2017-03-02 | 2017-08-18 | 兰州大学 | A kind of Fuji apple quality damage-free detection method based near infrared spectrum |
-
2018
- 2018-09-03 CN CN201811021625.0A patent/CN109272030A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102967578A (en) * | 2012-11-08 | 2013-03-13 | 中国农业科学院农业质量标准与检测技术研究所 | Method for obtaining near-infrared spectrum of beef sample online and application thereof in evaluating beef quality |
| US20170160200A1 (en) * | 2014-05-26 | 2017-06-08 | Hitachi, Ltd. | Optical analysis device |
| CN104181170A (en) * | 2014-09-05 | 2014-12-03 | 熊菊莲 | Fruit appearance detection method based on spectrum image analysis |
| CN104359838A (en) * | 2014-11-11 | 2015-02-18 | 上海理工大学 | Non-destructive detection method for hardness of pear |
| CN107064047A (en) * | 2017-03-02 | 2017-08-18 | 兰州大学 | A kind of Fuji apple quality damage-free detection method based near infrared spectrum |
Non-Patent Citations (3)
| Title |
|---|
| 樊书祥: "基于可见/近红外光谱及成像技术的苹果可溶性固形物检测研究", 《中国博士学位论文全文数据库 农业科技辑》 * |
| 谈文艺: "基于高光谱成像技术的苹果外部损伤精确识别与分级方法研究", 《中国优秀硕士学位论文全文数据库 农业科技辑》 * |
| 黄星奕，丁然，史嘉辰: "霉变出芽花生的近红外光谱无损检测研究", 《中国农业科技导报》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109946243A (en) * | 2019-03-26 | 2019-06-28 | 贵阳学院 | The method of Fast nondestructive evaluation plum soluble solid content |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5441703B2 (en) | Visible / Near-Infrared Spectroscopy and Grape Brewing Method | |
| CN103837492B (en) | A kind of Kiwi berry based on near-infrared spectrum technique expand fruit lossless detection method | |
| CN101308086B (en) | Fruit internal quality on-line checking apparatus based on near infrared spectra technology | |
| CN103674854A (en) | Apple grading method based on synergic interval partial least square discriminant analysis | |
| CN104965973B (en) | A kind of Apple Mould Core multiple-factor Non-Destructive Testing discrimination model and method for building up thereof | |
| CN104020129A (en) | Method for discriminating fermentation quality of congou black tea based on near-infrared-spectroscopy-combined amino acid analysis technology | |
| CN108872129A (en) | A kind of insulating paper near infrared spectrum analytic method based on Partial Least Squares | |
| CN104280349A (en) | Method for identifying hollowness of white radishes based on hyperspectral image | |
| CN107219184A (en) | A kind of meat discrimination method and device traced to the source applied to the place of production | |
| CN105548070A (en) | Apple soluble solid near-infrared detection part compensation method and system | |
| CN102128805A (en) | Method and device for near infrared spectrum wavelength selection and quick quantitative analysis of fruit | |
| CN105738311A (en) | Apple sweetness non-damage quick detection method based on near-infrared spectrum technology | |
| CN109272030A (en) | Apple surface earlier damage quick nondestructive recognition methods based on fiber spectrum technology | |
| CN110376154A (en) | Fruit online test method and system based on spectrum correction | |
| CN204630927U (en) | The navel orange place of production, south, a kind of Jiangxi device for quickly identifying | |
| CN105911000A (en) | Characteristic wave band based blood spot egg on-line detection method | |
| CN109946246A (en) | A kind of detection method and device of apple soluble solid | |
| CN113030011A (en) | Rapid nondestructive testing method and system for sugar content of fruits | |
| PUANGSOMBUT et al. | Evaluation of internal quality of fresh‐cut pomelo using VIS/NIR transmittance | |
| CN106198423A (en) | A kind of method differentiating ham sausage grade based on visible and near infrared spectrum analytical technology | |
| CN114324195A (en) | Multichannel agricultural product nondestructive testing method | |
| CN106872396B (en) | A kind of method of difference nir instrument measurement grape pol model conversion | |
| CN104181125A (en) | Method for rapidly determining Kol-bach value of beer malt | |
| CN106442400B (en) | A kind of method that near infrared spectrum quickly determines soil type fresh tea leaves | |
| CN106323888B (en) | Ultra-low emission flue gas measuring method |
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: 20190125 |