CN108836325A - A method of organoleptic substances classification is carried out based on smell brain wave and random forest - Google Patents
A method of organoleptic substances classification is carried out based on smell brain wave and random forest Download PDFInfo
- Publication number
- CN108836325A CN108836325A CN201810320413.6A CN201810320413A CN108836325A CN 108836325 A CN108836325 A CN 108836325A CN 201810320413 A CN201810320413 A CN 201810320413A CN 108836325 A CN108836325 A CN 108836325A
- Authority
- CN
- China
- Prior art keywords
- carried out
- random forest
- brain
- smell
- brain wave
- 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
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
- A61B5/377—Electroencephalography [EEG] using evoked responses
- A61B5/381—Olfactory or gustatory stimuli
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7246—Details of waveform analysis using correlation, e.g. template matching or determination of similarity
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Surgery (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Veterinary Medicine (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Psychiatry (AREA)
- Molecular Biology (AREA)
- Physics & Mathematics (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Psychology (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Signal Processing (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a kind of methods for carrying out organoleptic substances classification based on smell brain wave and random forest, include the following steps：S1, the acquisition of electroencephalogram spectrum information is carried out to candidate using brain-computer interface system, that is, electroencephalograph；S2, acquired electroencephalogram modal data is pre-processed；S3, feature extraction is carried out to pretreated profile information is completed based on linear characteristic analysis, including：Peak value, mean value, variance, centre frequency, maximum power and the power of α, β, θ frequency range and totally 72 dimension datas are as the brain electrical feature in EEG signals research；S4, pattern-recognition is carried out using RF random forest.The present invention can be pervasive in the sensory evaluation of substance, to the adjustment of new product development, basic research, ingredient and technique, reduces the appraisals such as cost, quality guarantee and product optimization and plays a significant role.
Description
Technical field
The present invention relates to organoleptic substances sorting technique fields, and in particular to one kind based on smell brain wave and random forest into
The method of row organoleptic substances classification.
Background technique
Sensory evaluation is the intersection for collecting the subjects such as modern physiology, psychology, statistics and gradually developing, growing up
Frontier branch of science, sense organ professional is in the tune to new product development, basic research, ingredient and technique in entire appraisement system
It is whole, reduce in the appraisals such as cost, quality guarantee and product optimization and play decisive role.With the hair of science and technology
Exhibition, more and more precision instruments for analyzing flavor substance come into being, can not be complete but rely solely on instrument analysis
The truly feels of full response human body, therefore human body evaluation still occupies an important position.During sensory evaluation, need
Personnel are participated in person in the sensory utilization such as vision, smell, sense of taste, are easy by daily life habit and diet feelings
The interference of condition, the sensory evaluation result provided pass through the thinking of brain, are doped with personal subjective factor, therefore have centainly
Subjectivity, poor repeatability.
EEG signals are a kind of basic bio signals of human body, are the effective means for recording brain activity, can be objective, true
Reflect human body physiological state on the spot, can be used for the auxiliary diagnosis of mental disease, including Parkinson's disease, Alzheimer disease, Weir
Inferior disease, epilepsy, brain tumor and schizophrenia etc..As the objective indicator of nervous function and physiological evaluation, by observing brain
Electric signal (EEG) can help us directly to understand electrophysiological change related with Xin Li, physiological status, by Novel presentation
Observation come determine may generation lesion.
Summary of the invention
Based on above-mentioned analysis, the present invention provides one kind carries out organoleptic substances classification based on smell brain wave and random forest
Method.
To achieve the above object, the technical scheme adopted by the invention is as follows：
A method of organoleptic substances classification is carried out based on smell brain wave and random forest, is included the following steps：
S1, the acquisition of electroencephalogram spectrum information is carried out to candidate using one machine interface system of brain, that is, electroencephalograph；
S2, acquired electroencephalogram modal data is pre-processed；
S3, feature extraction is carried out to pretreated profile information is completed based on linear characteristic analysis, including：α, β, θ frequency range
Peak value, mean value, variance, centre frequency, maximum power and power and totally 72 dimension datas as EEG signals research in brain
Electrical feature；
S4, pattern-recognition is carried out using RF random forest.
Wherein, the pretreatment of the eeg data, which includes at least, deletes bad block processing；Filter out 50Hz Hz noise；Stacked data
Add average treatment and wavelet transformation noise reduction process.
Wherein, the position of electrode is according to international standard lead 10-20 system rest, select Fp1, F3 relevant to smell,
Electroencephalogram modal data corresponding to F7, Fz electrode.
The present invention can be pervasive in the sensory evaluation of substance, sensory evaluation process is more succinct, more normative, preciseness and
Science to the adjustment of new product development, basic research, ingredient and technique, reduces cost, quality guarantee and product optimization etc. and comments
Valence work plays a significant role.
Detailed description of the invention
Fig. 1 is the flow chart of data processing figure in the embodiment of the present invention.
Fig. 2 is to filter out 50Hz Hz noise effect picture.
Fig. 3 is 7 groups of data investigation effect pictures in the embodiment of the present invention.
Fig. 4 is the result figure after wavelet transformation.
Fig. 5 is the influence that decision sets to classification performance in random forest.
Specific embodiment
In order to which objects and advantages of the present invention are more clearly understood, the present invention is carried out with reference to embodiments further
It is described in detail.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not used to limit this hair
It is bright.
Embodiment
Material and method：The beer in 3 kinds of different brands (Qingdao, blue zone, Harbin) is selected, alcohol concentration and raw material are such as
Shown in the following table 1：
The beer sample of 13 kinds of different brands of table
In terms of smell electroencephalogramsignal signal analyzing, used EEG signals are laboratories from data are adopted, and collect original brain
After electric signal, progress brain electricity pretreatment (including deletes bad block, filters out 50Hz Hz noise, data investigation is average, small echo first
Transformation and etc.)；In view of the electrode of acquisition includes therefore different regions has carried out electrode screening；Then it is carried out linearly
Analysis, by the peak value of α, β, θ frequency range, mean value, variance, centre frequency, maximum power and power and as characteristic value；Finally will
The characteristic value of selection is sent to corresponding classifier, obtains final classification accuracy.This patent flow chart of data processing figure such as Fig. 1
It is shown.
This experiment places brain using NCERP series electroencephalogram and evoked potentuial measuring system, according to the world 10-20 brain electric system
Electric frequency acquisition DC-50Hz, sample frequency 256Hz can at most obtain 36 channel datas (this experiment obtains 24 channel datas),
The bioelectrical signals of subject Different brain region are respectively represented.For the completion experiment that subject can be concentrated one's energy, acquisition brain electricity
Signal needs stringent experimental situation, such as quiet and to keep faint light.
1. signal acquisition experiment flow
This experiment in, select the age between 23-26 years old, dextro manuality, without any respiratory disorder, mental disease and
The subject of chronic disease.Allow subject in comfortable state, the EEG signals of observation subject, are in steady to its EEG signals in real time
When determining state, olfactory stimulation gesture is issued, experimenter randomly selects one from 3 kinds of samples and is gently placed on subject underthe nose
At 1-2cm and the stimulation of 2s is kept, makes subject that can sufficiently receive olfactory stimulation, EEG signals make corresponding response.Once
After experiment terminates, laboratory windowing ventilation is tested abundant rest 1-2min and repeats above-mentioned experimental implementation later.
It should be noted that following points for attention during experiment acquisition：
Before acquisition, subject keeps clearheaded, and cleans hair with neutrality shampoo before starting eeg signal acquisition,
In order to avoid making to make wave distortion due to scalp resistance is excessive because grease is excessively high.Meanwhile experiment content simply is introduced to subject, it gets across
This experiment is not damaged no pain, is eliminated because its state of mind influences the influence of brain wave acquisition result.
In acquisition, it is tested hyperphoria with fixed eyeballs cover, earplug, so that it is reduced the movements such as blink, eyeball movement, and it is comfortable to be tested holding
Posture, no any limb action occur, and prevent from being mixed with excessive myoelectricity in brain electricity.Meanwhile experimenter controls experiment and carries out
Time, prevent subject occur fatigue or boredom.During trial interval rest, actively linked up with subject, it is ensured that quilt
Examination keeps pleasant state.
2. the pretreatment of eeg data
Since collected EEG signals mix a variety of brains electricity (evoked brain potential, artefact, spontaneous brain electricity, noise etc.), so
Before the identification for carrying out smell brain electricity, needs to pre-process original EEG signals, thus obtain useful EEG signals.
The pretreatment of eeg data includes that (1) filters out 50Hz Hz noise, reduces it by pollution of the noise background to signal,
Improve signal-to-noise ratio, retains the authenticity of original signal, the result after filtering out 50Hz Hz noise is as shown in Figure 2；(2) data investigation
It is average, it, can will be random by superposed average since EEG signals are a kind of complex physiologic EEG signals of nonlinear and nonstationary
The spontaneous EEG signals of rule are offset, and regular evoked brain potential signal is enhanced, and realize that EEG signals time-domain information is visual
Change.In this experiment, the data that experiment is obtained carry out 7 groups of superposed averages, and data investigation effect picture is as shown in Figure 3；(3) small echo becomes
It changes method and carries out noise reduction, noise reduction effect is as shown in Figure 4.
3. the selection of electrode
The position of electrode is according to international standard lead 10-20 system rest, as shown in table 2.It, can in smell and sense of taste memory
To observe the interaction between prefrontal and marginal convolution.During carrying out long-term odor identification, socket of the eye frontal lobe region and bilateral
Cortex of temporal lobe is activated.In short term odor identification, right side temporal lobe is activated.In this experiment, it selects relevant to smell
Fp1, F3, F7, Fz electrode are analyzed.
2 ten-twenty electrode system electrode of table matches table name
4. the feature extraction of EEG signals
The data obtained after aforesaid operations are subjected to linear analysis, that is, extract its Linear Eigenvalue, including：α, β, θ frequency range
Peak value, mean value, variance, centre frequency, maximum power and power and totally 72 dimension datas as EEG signals research in brain
Electrical feature.
5. mode identification method
In this experiment, pattern-recognition is carried out using RF random forest, test set classification accuracy is up to 91.65%.Operation knot
Fruit is as shown in figure 5, since the quantity for the decision tree for including in random forest has a certain impact to its Generalization Capability.From figure
As can be seen that comprehensively considering the decision tree number for including in random forest for the EEG signals data that this experiment acquires
With the speed of modeling, more satisfactory when selecting in random forest comprising 25-100 decision tree, accuracy rate is up to 100%.
It can thus be seen that the method based on smell brain wave can classify the beer of different brands, therefore should
Method can be pervasive in the sensory evaluation of substance, to the adjustment of new product development, basic research, ingredient and technique, reduce cost,
The appraisals such as quality guarantee and product optimization play a significant role.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, without departing from the principle of the present invention, it can also make several improvements and retouch, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (3)
1. a kind of method for carrying out organoleptic substances classification based on smell brain wave and random forest, which is characterized in that including as follows
Step：
S1, the acquisition of electroencephalogram spectrum information is carried out to candidate using brain-computer interface system, that is, electroencephalograph；
S2, acquired electroencephalogram modal data is pre-processed；
S3, feature extraction is carried out to pretreated profile information is completed based on linear characteristic analysis, including：The peak of α, β, θ frequency range
Value, mean value, variance, centre frequency, maximum power and power and totally 72 dimension datas are special as the brain electricity in EEG signals research
Sign；
S4, pattern-recognition is carried out using RF random forest.
2. a kind of method that organoleptic substances classification is carried out based on smell brain wave and random forest as described in claim 1,
It is characterized in that, the pretreatment of the eeg data, which includes at least, deletes bad block processing；Filter out 50Hz Hz noise；Data investigation is flat
Processing and wavelet transformation noise reduction process.
3. a kind of method that organoleptic substances classification is carried out based on smell brain wave and random forest as described in claim 1,
It is characterized in that, the position of electrode selects Fp1, F3, F7, Fz relevant to smell according to international standard lead 10-20 system rest
Electroencephalogram modal data corresponding to electrode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810320413.6A CN108836325A (en) | 2018-04-02 | 2018-04-02 | A method of organoleptic substances classification is carried out based on smell brain wave and random forest |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810320413.6A CN108836325A (en) | 2018-04-02 | 2018-04-02 | A method of organoleptic substances classification is carried out based on smell brain wave and random forest |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108836325A true CN108836325A (en) | 2018-11-20 |
Family
ID=64212003
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810320413.6A Pending CN108836325A (en) | 2018-04-02 | 2018-04-02 | A method of organoleptic substances classification is carried out based on smell brain wave and random forest |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108836325A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112493998A (en) * | 2020-12-09 | 2021-03-16 | 北京意图科技有限公司 | Olfactory sensory evaluation method and system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1632816A (en) * | 2003-12-23 | 2005-06-29 | 广州可夫医疗科技有限公司 | Method for analyzing EEG fluctuation signal and equipment thereof |
| CN102106730A (en) * | 2011-03-16 | 2011-06-29 | 上海交通大学 | Method for processing electroencephalogram signal and detecting alertness based on fractal characteristics |
| CN102274022A (en) * | 2011-05-10 | 2011-12-14 | 浙江大学 | Sleep state monitoring method based on electroencephalogram signals |
| CN105852885A (en) * | 2016-03-23 | 2016-08-17 | 陕西师范大学 | N400 evoked potential lie detection method based on improved extreme learning machine |
| CN106919956A (en) * | 2017-03-09 | 2017-07-04 | 温州大学 | Brain wave age forecasting system based on random forest |
| CN107080545A (en) * | 2017-03-23 | 2017-08-22 | 兰州大学 | A kind of lie detection system based on brain electricity |
| CN107292296A (en) * | 2017-08-04 | 2017-10-24 | 西南大学 | A kind of human emotion wake-up degree classifying identification method of use EEG signals |
-
2018
- 2018-04-02 CN CN201810320413.6A patent/CN108836325A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1632816A (en) * | 2003-12-23 | 2005-06-29 | 广州可夫医疗科技有限公司 | Method for analyzing EEG fluctuation signal and equipment thereof |
| CN102106730A (en) * | 2011-03-16 | 2011-06-29 | 上海交通大学 | Method for processing electroencephalogram signal and detecting alertness based on fractal characteristics |
| CN102274022A (en) * | 2011-05-10 | 2011-12-14 | 浙江大学 | Sleep state monitoring method based on electroencephalogram signals |
| CN105852885A (en) * | 2016-03-23 | 2016-08-17 | 陕西师范大学 | N400 evoked potential lie detection method based on improved extreme learning machine |
| CN106919956A (en) * | 2017-03-09 | 2017-07-04 | 温州大学 | Brain wave age forecasting system based on random forest |
| CN107080545A (en) * | 2017-03-23 | 2017-08-22 | 兰州大学 | A kind of lie detection system based on brain electricity |
| CN107292296A (en) * | 2017-08-04 | 2017-10-24 | 西南大学 | A kind of human emotion wake-up degree classifying identification method of use EEG signals |
Non-Patent Citations (1)
| Title |
|---|
| ANURADHA SAHA ET AL.: "Olfaction Recognition by EEG Analysis Using Differential Evolution Induced Hopfield Neural Net", 《THE 2013 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN)》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112493998A (en) * | 2020-12-09 | 2021-03-16 | 北京意图科技有限公司 | Olfactory sensory evaluation method and system |
| CN112493998B (en) * | 2020-12-09 | 2021-12-21 | 北京意图科技有限公司 | Olfactory sensory evaluation method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Khosla et al. | A comparative analysis of signal processing and classification methods for different applications based on EEG signals | |
| Retter et al. | Uncovering the neural magnitude and spatio-temporal dynamics of natural image categorization in a fast visual stream | |
| US7570991B2 (en) | Method for real time attitude assessment | |
| Li et al. | Emotion classification based on gamma-band EEG | |
| Hajcak et al. | Attending to affect: appraisal strategies modulate the electrocortical response to arousing pictures. | |
| CN108403112A (en) | The method for carrying out organoleptic substances classification based on smell brain wave and GS-SVM | |
| CN108542385A (en) | A method of carrying out sense organ flavor substance classification using smell brain wave | |
| Akdeniz | Brain activity underlying face and face pareidolia processing: an ERP study | |
| CN111466931A (en) | Emotion recognition method based on EEG and food picture data set | |
| CN112426162A (en) | Fatigue detection method based on electroencephalogram signal rhythm entropy | |
| CN109567831A (en) | The Preference image identifying system and method for product under a kind of visual cognition neuromechanism | |
| Daneshi Kohan et al. | EEG/PPG effective connectivity fusion for analyzing deception in interview | |
| CN113576498B (en) | Visual and auditory aesthetic evaluation method and system based on electroencephalogram signals | |
| Eulitz et al. | Magnetic brain activity evoked and induced by visually presented words and nonverbal stimuli | |
| CN108836325A (en) | A method of organoleptic substances classification is carried out based on smell brain wave and random forest | |
| CN108836320A (en) | A method of organoleptic substances classification is carried out based on smell brain wave | |
| CN111671421A (en) | Electroencephalogram-based children demand sensing method | |
| CN108420430A (en) | A kind of organoleptic substances sorting technique based on smell brain wave and PSO-SVM | |
| Lin et al. | Neurophysiological markers of identifying regret by 64 channels EEG signal | |
| Cho et al. | An investigation of the influences of noise on EEG power bands and visual cognitive responses for human-oriented product design | |
| CN108814596A (en) | A method of classified based on the organoleptic substances of smell brain wave and RF | |
| Bjørge et al. | Identification of EEG-based signature produced by visual exposure to the primary colors RGB | |
| CN108836326A (en) | A method of organoleptic substances classification is carried out based on smell brain wave and wavelet packet | |
| Ross et al. | Violinists Playing with and without Music Notation: Investigating Hemispheric Brainwave Activity | |
| Plechawska-Wójcik et al. | Analysis of applied reference leads influence on an EEG spectrum |
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: 20181120 |