CN114756823B - Method for improving prediction capability of pepper spectrum model - Google Patents
Method for improving prediction capability of pepper spectrum model Download PDFInfo
- Publication number
- CN114756823B CN114756823B CN202210403774.3A CN202210403774A CN114756823B CN 114756823 B CN114756823 B CN 114756823B CN 202210403774 A CN202210403774 A CN 202210403774A CN 114756823 B CN114756823 B CN 114756823B
- Authority
- CN
- China
- Prior art keywords
- prediction
- pepper
- samples
- value
- spectrum model
- 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.)
- Active
Links
- 238000001228 spectrum Methods 0.000 title claims abstract description 69
- 235000002566 Capsicum Nutrition 0.000 title claims abstract description 63
- 239000006002 Pepper Substances 0.000 title claims abstract description 60
- 235000016761 Piper aduncum Nutrition 0.000 title claims abstract description 60
- 235000017804 Piper guineense Nutrition 0.000 title claims abstract description 60
- 235000008184 Piper nigrum Nutrition 0.000 title claims abstract description 60
- 238000000034 method Methods 0.000 title claims abstract description 34
- 244000203593 Piper nigrum Species 0.000 title 1
- 241000722363 Piper Species 0.000 claims abstract description 59
- 238000012549 training Methods 0.000 claims abstract description 47
- 238000012937 correction Methods 0.000 claims abstract description 20
- 238000012360 testing method Methods 0.000 claims abstract description 9
- 230000003595 spectral effect Effects 0.000 claims abstract description 8
- 238000005259 measurement Methods 0.000 claims abstract description 4
- 235000007650 Aralia spinosa Nutrition 0.000 claims description 10
- 241000949456 Zanthoxylum Species 0.000 claims description 10
- 239000000341 volatile oil Substances 0.000 claims description 9
- 244000089698 Zanthoxylum simulans Species 0.000 claims description 8
- 238000004458 analytical method Methods 0.000 claims description 8
- 238000002329 infrared spectrum Methods 0.000 claims description 7
- 238000013178 mathematical model Methods 0.000 claims description 6
- 238000009614 chemical analysis method Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 4
- 241000758706 Piperaceae Species 0.000 description 3
- 238000010924 continuous production Methods 0.000 description 2
- 238000004128 high performance liquid chromatography Methods 0.000 description 2
- 238000004949 mass spectrometry Methods 0.000 description 2
- 238000010238 partial least squares regression Methods 0.000 description 2
- 229930003827 cannabinoid Natural products 0.000 description 1
- 239000003557 cannabinoid Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004476 mid-IR spectroscopy Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- 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
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Pure & Applied Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Algebra (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Operations Research (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Probability & Statistics with Applications (AREA)
- Health & Medical Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
- Other Investigation Or Analysis Of Materials By Electrical Means (AREA)
Abstract
The invention relates to the field of pepper spectrum model prediction, in particular to a method for improving the prediction capability of a pepper spectrum model. The technical proposal comprises: establishing an original pepper spectrum model; collecting spectral data of a plurality of pepper samples with known component calibration values, and setting the spectral data as a prediction set; performing blind measurement on the prediction set by adopting an original pepper spectrum model to obtain a corresponding prediction value; calculating a predicted deviation value by combining the predicted value of the predicted set with the corresponding calibration value; and defined as an accurate sample and an erroneous sample; obtaining a change training set and a change prediction set according to the error sample set; calculating the prediction accuracy T 1 of the change prediction set; establishing a change spectrum model by adopting a change training set, and carrying out blind test on a change prediction set to obtain a corresponding prediction value; calculating a prediction accuracy T 2 by combining the predicted value of the change predicted set and the corresponding calibration value thereof; and comparing the prediction accuracy T 1 with T 2, and performing the iterative correction according to the comparison result. The method is suitable for improving the prediction capability of the pepper spectrum model.
Description
Technical Field
The invention relates to the field of pepper spectrum model prediction, in particular to a method for improving the prediction capability of a pepper spectrum model.
Background
For quality detection and classification of Chinese prickly ash, the current main detection technologies include a gas-mass spectrometry method, a high performance liquid chromatography method, a mid-infrared spectrometry method and the like, but the methods are mainly applied in laboratories, the detection cost of the gas-mass spectrometry method and the detection cost of the high performance liquid chromatography method are relatively expensive, the sample treatment is complicated, the requirement on experimental operation is very high, the quick measurement cannot be carried out, and great difficulty is brought to Chinese prickly ash detection and classification. Therefore, the simple, quick and lossless method for distinguishing the quality of the peppers is realized, and the method has important practical significance.
Compared with other chemical analysis technologies, the portable near infrared spectrum technology has the characteristics of rapidness, accuracy, no need of sample pretreatment, no damage to samples, no pollution and the like, is a very suitable quality detection technology for the peppers, and meanwhile, the portable near infrared spectrometer is low in cost, simple to operate and convenient to carry, and can be purchased in a large amount to meet the detection requirements of various peppers. However, in the practical application of the portable near infrared spectrum analysis of the pepper sample, the data acquisition and analysis of the pepper sample are always a continuous process, and the original spectrum model obtained in the early modeling analysis process is often not suitable for the subsequent spectrum analysis due to the self-volatility of the pepper sample, so that the problems of poor prediction effect and low prediction accuracy of the original pepper spectrum model are caused, and therefore, if the prediction capability of the original spectrum model is required to be improved, the original spectrum model needs to be corrected by adopting a continuous process, and the prediction capability of the pepper spectrum model is further effectively improved.
Disclosure of Invention
The invention aims to overcome the defects in the prior art, and provides a method for improving the prediction capability of a pepper spectrum model, which can effectively improve the prediction capability of a training set sample after the alternation to establish the spectrum model.
The invention adopts the following technical scheme to realize the aim and improves the prediction capability of the pepper spectrum model, and the method comprises the following steps:
Step 1, collecting spectrum data of Zanthoxylum bungeanum samples with X known component calibration values to form an original training set X 1, and establishing an original Zanthoxylum bungeanum spectrum model M 1, wherein X is an integer larger than 0;
Step 2, collecting spectral data of Y pepper samples with known component calibration values, and setting the Y pepper samples as a prediction set Y 1, wherein Y is an integer larger than 0;
Step 3, blind testing is carried out on the prediction set Y 1 by adopting an original pepper spectrum model M 1 to obtain a corresponding prediction value;
Step 4, calculating a predicted deviation value by combining the predicted value of the pepper sample predicted set Y 1 with the corresponding calibration value;
step 5, defining samples with prediction deviation values smaller than the fault tolerance threshold as accurate samples, and defining samples with prediction deviation values larger than or equal to the fault tolerance threshold as error samples;
Step 6, screening all error samples to form an error sample set Y 2;
Step 7, randomly selecting half of samples in the error sample set Y 2, adding the samples into the original training set X 1 to form a change training set X 2, removing the randomly selected half of samples from the prediction set Y 1, and forming a change prediction set Y 3 by the residual samples in the prediction set Y 1;
Step 8, calculating a first prediction accuracy T 1 of the change prediction set Y 3;
Step 9, a modified spectrum model M 2 is established by adopting a modified training set X 2, and blind measurement is carried out on a modified prediction set Y 3 to obtain a corresponding prediction value;
Step 10, calculating a second prediction accuracy T 2 by combining the predicted value of the change prediction set Y 3 and the corresponding calibration value;
Step 11, comparing and judging the first prediction accuracy T 1 and the second prediction accuracy T 2, and if T 1≥T2, not correcting the original spectrum model; if T 1<T2, the change training set X 2 is used as the original training set for the next spectral model correction.
Further, the known component calibration value is a pepper sample volatile oil content value.
The specific method for establishing the original pepper spectrum model M 1 comprises the following steps:
And (3) selecting the volatile oil content value of the pricklyash peel sample obtained by a chemical analysis method as a known calibration value, and establishing a mathematical model relation between near infrared spectrum data and the known calibration value by adopting a partial least square regression method, wherein the mathematical model is the original pricklyash peel spectrum model M 1.
Further, the method further comprises: and step 12, after the original training set is subjected to repeated correction, if the condition of T 1≥T2 appears in the correction of three continuous times, the repeated correction is completed, the training set after the repeated correction is completed is used as a final training set, and the pricklyash peel spectrum model after the repeated correction is completed is the final pricklyash peel spectrum model. The final pepper spectrum model is used for blind test analysis of pepper samples with unknown calibration values.
Further, in step 4, the specific method for calculating the predicted deviation value includes:
the prediction set Y 1 contains Y pepper samples, and the corresponding predicted values are also Y, and the predicted value b= (B 1,B2,…,BY) of the pepper sample is set, and the calibration value is w= (W 1,W2,…,WY), and the predicted deviation value g= |b-w|= (|b 1-W1|,|B2-W2|,...,|BY-WY |).
According to the invention, the prediction accuracy of the original pepper spectrum model and the corrected pepper spectrum model is taken as a reference to judge the prediction capacity of the spectrum model, the original training set is further overlapped by combining with the mispredicted sample, and the prediction capacity of the spectrum model established by the overlapped training set sample is effectively improved through continuous overlapped correction.
Drawings
Fig. 1 is a flowchart of a method for improving the prediction capability of a pepper spectrum model according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. It will be apparent that the described embodiments are some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention is described in further detail below with reference to the drawings and the detailed description.
The invention relates to a method for improving the prediction capability of a pepper spectrum model, which is shown in figure 1 and comprises the following steps:
Step 1, collecting spectrum data of Zanthoxylum bungeanum samples with X known component calibration values to form an original training set X 1, and establishing an original Zanthoxylum bungeanum spectrum model M 1, wherein X is an integer larger than 0;
In the embodiment of the invention, the daily quick detection of the pepper mainly detects the content of volatile oil and cannabinoid, wherein the content of the volatile oil is higher in correlation with the quality of the pepper, so the component content of the invention is selected as the volatile oil content, and the specific modeling method is as follows: and (3) selecting a pepper sample volatile oil content value obtained by a chemical analysis method as a known calibration value, and establishing a mathematical model relation between near infrared spectrum data and the known calibration value by adopting a partial least squares regression (PLS), wherein the mathematical model is a near infrared spectrum model M 1 of the pepper volatile oil.
Step 2, collecting spectral data of Y pepper samples with known component calibration values, and setting the Y pepper samples as a prediction set Y 1, wherein Y is an integer larger than 0;
Step 3, blind testing is carried out on the prediction set Y 1 by adopting an original pepper spectrum model M 1 to obtain a corresponding prediction value;
Step 4, calculating a predicted deviation value by combining the predicted value of the pepper sample predicted set Y 1 with the corresponding calibration value;
in the embodiment of the present invention, the prediction set Y 1 includes Y pepper samples, and the corresponding predicted values are also Y, and the predicted value b= (B 1,B2,…,BY) of the pepper samples is set, and the calibration value is w= (W 1,W2,…,WY), and the predicted deviation value g= |b-w|= (|b 1-W1|,|B2-W2|,...,|BY-WY |).
Step 5, defining samples with prediction deviation values smaller than the fault tolerance threshold as accurate samples, and defining samples with prediction deviation values larger than or equal to the fault tolerance threshold as error samples;
In the embodiment of the invention, due to the convenience of the portable near infrared spectrum equipment, the spectrum performance is greatly influenced, and when the prediction analysis is carried out on the same sample, a small amount of deviation between the predicted value and the calibration value is also considered as accurate prediction. According to the characteristics, the predicted deviation value is considered as a reasonable deviation value within a certain threshold range, and the threshold range is the fault tolerance threshold. Setting the fault tolerance threshold value as H, comparing the magnitude of the Y predicted deviation values with the magnitude of the fault tolerance threshold value one by one, if |b n-Wn | < H, n=1, 2, …, Y, then the sample is an accurate sample, otherwise, the sample is an error sample, comparing the predicted deviation values one by one, and counting the number of the accurate samples as Z and the number of the error samples as C.
Step 6, screening all error samples to form an error sample set Y 2;
Step 7, randomly selecting half of samples in the error sample set Y 2, adding the samples into the original training set X 1 to form a change training set X 2, removing the randomly selected half of samples from the prediction set Y 1, and forming a change prediction set Y 3 by the residual samples in the prediction set Y 1;
In the embodiment of the invention, Y 2 samples are set in the error sample set Y 2, when half of the samples selected randomly are added into the original training set X 1 to form the change training set X 2, the number of the samples contained in the change training set is known to be (x+y 2/2), half of the samples selected in the embodiment are removed from the prediction set Y 1, and the remaining samples form the change prediction set Y 3, the number of the samples contained in the change prediction set is known to be (Y-Y 2/2).
Step 8, calculating a first prediction accuracy T 1 of the change prediction set Y 3;
in this embodiment, if the number of accurate samples is Z, the prediction accuracy is: t 1=100%*Z/(Y-y2/2).
Step 9, a change training set X 2 is adopted to establish a change spectrum model M 2, and a change prediction set Y 3 is subjected to blind test to obtain a corresponding prediction value;
In the embodiment of the invention, the same spectrum data processing mode as that of the embodiment of the step 1 is adopted to carry out spectrum modeling on the change training set X 2, and the spectrum model M 2 is used for blind test of the change prediction set Y 3 to obtain a corresponding prediction value.
Step 10, calculating a second prediction accuracy T 2 by combining the predicted value of the change prediction set Y 3 and the corresponding calibration value;
In this embodiment, for the change prediction set Y 3, which includes (Y-Y 2/2) samples, the corresponding prediction value number is (Y-Y 2/2), and the pepper sample prediction value is set The calibration value isThe predicted bias value/>, can be calculatedFurther comparing (Y-Y 2/2) predicted deviation values to a fault tolerance threshold H: r i-Fi|<H,i=(1,2,…,Y-y2/2);
If the number of samples meeting the above condition is V, the number of accurate samples in the change prediction set Y 3 is V, and further calculating the prediction accuracy: t 2=100%*V/(Y-y2/2).
Step 11, comparing and judging the prediction accuracy T 1 with T 2, and if T 1≥T2, not correcting the original spectrum model; if T 1<T2, the change training set X 2 is used as the original training set for the next spectral model correction.
In this embodiment, if T 1≥T2 indicates that adding the error sample to the original training set does not improve the prediction accuracy of the spectrum model, the original training set does not need to be changed, and the original training set is reserved as the secondary spectrum model to correct the original training set; if T 1<T2, the step of adding the error sample to the original training set and improving the prediction precision of the spectrum model is described, and the error sample is further put into a change training set X 2 formed by the original training set to be used as the original training set for correcting the secondary spectrum model.
In one development, after the original training set is subjected to repeated correction, if the condition of T 1≥T2 appears in the correction of three continuous times, the correction of the current time is completed, the training set after the correction of the current time is used as a final training set, and the pricklyash peel spectrum model after the correction of the current time is completed is the final pricklyash peel spectrum model. The final pepper spectrum model is used for blind test analysis of pepper samples with unknown calibration values.
In summary, according to the method and the device for predicting the spectrum model, the original training set is subjected to data alternation through the prediction error samples, so that the original training set contains more samples with larger prediction deviation, the fault tolerance of the original training set is improved, the fault tolerance of the spectrum model after alternation is further improved, and the prediction capability of the spectrum model is effectively improved.
Claims (6)
1. The method for improving the prediction capability of the pepper spectrum model is characterized by comprising the following steps:
Step 1, collecting spectrum data of Zanthoxylum bungeanum samples with X known component calibration values to form an original training set X 1, and establishing an original Zanthoxylum bungeanum spectrum model M 1, wherein X is an integer larger than 0;
Step 2, collecting spectral data of Y pepper samples with known component calibration values, and setting the Y pepper samples as a prediction set Y 1, wherein Y is an integer larger than 0;
Step 3, blind testing is carried out on the prediction set Y 1 by adopting an original pepper spectrum model M 1 to obtain a corresponding prediction value;
Step 4, calculating a predicted deviation value by combining the predicted value of the pepper sample predicted set Y 1 with the corresponding calibration value;
step 5, defining samples with prediction deviation values smaller than the fault tolerance threshold as accurate samples, and defining samples with prediction deviation values larger than or equal to the fault tolerance threshold as error samples;
Step 6, screening all error samples to form an error sample set Y 2;
Step 7, randomly selecting half of samples in the error sample set Y 2, adding the samples into the original training set X 1 to form a change training set X 2, removing the randomly selected half of samples from the prediction set Y 1, and forming a change prediction set Y 3 by the residual samples in the prediction set Y 1;
Step 8, calculating a first prediction accuracy T 1 of the change prediction set Y 3;
Step 9, a modified spectrum model M 2 is established by adopting a modified training set X 2, and blind measurement is carried out on a modified prediction set Y 3 to obtain a corresponding prediction value;
Step 10, calculating a second prediction accuracy T 2 by combining the predicted value of the change prediction set Y 3 and the corresponding calibration value;
Step 11, comparing and judging the first prediction accuracy T 1 and the second prediction accuracy T 2, and if T 1≥T2, not correcting the original spectrum model; if T 1<T2, the change training set X 2 is used as the original training set for the next spectral model correction.
2. The method for improving the prediction capacity of the pepper spectrum model as claimed in claim 1, wherein the known component calibration value is the volatile oil content value of the pepper sample.
3. The method for improving the prediction capability of the pepper spectrum model as claimed in claim 2, wherein the specific method for establishing the original pepper spectrum model M 1 comprises the following steps:
And (3) selecting the volatile oil content value of the pricklyash peel sample obtained by a chemical analysis method as a known calibration value, and establishing a mathematical model relation between near infrared spectrum data and the known calibration value by adopting a partial least square regression method, wherein the mathematical model is the original pricklyash peel spectrum model M 1.
4. The method for improving the predictive power of a pepper spectrum model as claimed in claim 1, characterized in that said method further comprises: and step 12, after the original training set is subjected to repeated correction, if the condition of T 1≥T2 appears in the correction of three continuous times, the repeated correction is completed, the training set after the repeated correction is completed is used as a final training set, and the pricklyash peel spectrum model after the repeated correction is completed is the final pricklyash peel spectrum model.
5. The method for improving the predictive power of a pepper spectrum model as claimed in claim 4, characterized in that the final pepper spectrum model is used for blind test analysis of pepper samples of unknown calibration value.
6. The method for improving the prediction capacity of the pepper spectrum model as claimed in claim 1, wherein in step 4, the specific method for calculating the prediction bias value comprises:
the prediction set Y 1 contains Y pepper samples, and the corresponding predicted values are also Y, and the predicted value b= (B 1,B2,…,BY) of the pepper sample is set, and the calibration value is w= (W 1,W2,…,WY), and the predicted deviation value g= |b-w|= (|b 1-W1|,|B2-W2|,...,|BY-WY |).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210403774.3A CN114756823B (en) | 2022-04-18 | 2022-04-18 | Method for improving prediction capability of pepper spectrum model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210403774.3A CN114756823B (en) | 2022-04-18 | 2022-04-18 | Method for improving prediction capability of pepper spectrum model |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114756823A CN114756823A (en) | 2022-07-15 |
CN114756823B true CN114756823B (en) | 2024-06-11 |
Family
ID=82330590
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210403774.3A Active CN114756823B (en) | 2022-04-18 | 2022-04-18 | Method for improving prediction capability of pepper spectrum model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114756823B (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103942562A (en) * | 2014-03-27 | 2014-07-23 | 河海大学 | Hyperspectral image classifying method based on multi-classifier combining |
CN104732246A (en) * | 2015-03-05 | 2015-06-24 | 重庆邮电大学 | Semi-supervised cooperative training hyperspectral image classification method |
CN107024439A (en) * | 2017-03-23 | 2017-08-08 | 西北农林科技大学 | A kind of paddy rice different growing chlorophyll content EO-1 hyperion estimating and measuring method |
CN109711470A (en) * | 2018-12-27 | 2019-05-03 | 陕西师范大学 | Merge spectrum-spatial information (si) DBM hyperspectral image classification method |
CN113092407A (en) * | 2021-04-20 | 2021-07-09 | 深圳市国科光谱技术有限公司 | Method for identifying original place varieties of Wuchang rice fragrance rice based on near infrared spectrum |
CN113657276A (en) * | 2021-08-18 | 2021-11-16 | 江苏天汇空间信息研究院有限公司 | Model migration training method for semantic segmentation of remote sensing image |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2018018038A1 (en) * | 2016-07-22 | 2018-01-25 | The Regents Of The University Of California | System and method for small molecule accurate recognition technology ("smart") |
-
2022
- 2022-04-18 CN CN202210403774.3A patent/CN114756823B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103942562A (en) * | 2014-03-27 | 2014-07-23 | 河海大学 | Hyperspectral image classifying method based on multi-classifier combining |
CN104732246A (en) * | 2015-03-05 | 2015-06-24 | 重庆邮电大学 | Semi-supervised cooperative training hyperspectral image classification method |
CN107024439A (en) * | 2017-03-23 | 2017-08-08 | 西北农林科技大学 | A kind of paddy rice different growing chlorophyll content EO-1 hyperion estimating and measuring method |
CN109711470A (en) * | 2018-12-27 | 2019-05-03 | 陕西师范大学 | Merge spectrum-spatial information (si) DBM hyperspectral image classification method |
CN113092407A (en) * | 2021-04-20 | 2021-07-09 | 深圳市国科光谱技术有限公司 | Method for identifying original place varieties of Wuchang rice fragrance rice based on near infrared spectrum |
CN113657276A (en) * | 2021-08-18 | 2021-11-16 | 江苏天汇空间信息研究院有限公司 | Model migration training method for semantic segmentation of remote sensing image |
Non-Patent Citations (2)
Title |
---|
DOSC在花椒挥发油含量近红外光谱分析中的应用;祝诗平;王刚;尹雄;兰雪冬;任德齐;;农业机械学报;20080425(04);全文 * |
基于近红外光谱的花椒麻味物质快速检测方法;祝诗平;王刚;杨飞;阚建全;郭静;;红外与毫米波学报;20080415(02);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN114756823A (en) | 2022-07-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111863147B (en) | SF6And concentration detection method and device for decomposed component gas | |
CN114372063A (en) | Fault detection method based on chromatograph and electronic equipment | |
CN107966499B (en) | Method for predicting crude oil carbon number distribution by near infrared spectrum | |
CN114756823B (en) | Method for improving prediction capability of pepper spectrum model | |
CN102542284B (en) | Method for identifying spectrum | |
CN113567605A (en) | Method and device for constructing automatic interpretation model of mass chromatogram and electronic equipment | |
KR20090045941A (en) | Apparatus for determining information in order to temporally align two information signals | |
CN108267422B (en) | Abnormal sample removing method based on near infrared spectrum analysis | |
CN113484400B (en) | Mass spectrogram molecular formula calculation method based on machine learning | |
CN102057261B (en) | Method and apparatus for automatic calibration of spectrometers in chemometry by means of a bayes iterative estimation method | |
CN112949169B (en) | Coal sample test value prediction method based on spectral analysis | |
CN115436342A (en) | Method and device for reducing LIBS detection uncertainty among multiple batches of samples | |
CN111474124B (en) | Spectral wavelength selection method based on compensation | |
CN114067169A (en) | Raman spectrum analysis method based on convolutional neural network | |
CN114324712A (en) | Transformer oil chromatographic fault judgment method and device, terminal and storage medium | |
CN114694771A (en) | Sample classification method, training method of classifier, device and medium | |
CN117633445A (en) | Adulteration identification method, system and storage medium of natural rose essential oil | |
CN116106464B (en) | Control system, evaluation system and method for mass spectrum data quality degree or probability | |
CN116453437B (en) | Display screen module testing method, device, equipment and storage medium | |
CN111126496B (en) | Method for determining type of solid insulating material of transformer | |
CN111832145B (en) | Fault diagnosis method and system for oil-immersed power transformer | |
CN108388965B (en) | Method for detecting adulteration content of grease, terminal equipment and computer readable storage medium | |
CN113762759A (en) | Multi-index system evaluation method suitable for food detection | |
CN115270611A (en) | Model building method and device for nonlinear index of sample in petrochemical industry | |
CN113744814A (en) | Mass spectrum data library searching method and system based on Bayesian posterior probability model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant |