CN112971784A - Wearable bone conduction fatigue driving detection method and device - Google Patents
Wearable bone conduction fatigue driving detection method and device Download PDFInfo
- Publication number
- CN112971784A CN112971784A CN202110549907.3A CN202110549907A CN112971784A CN 112971784 A CN112971784 A CN 112971784A CN 202110549907 A CN202110549907 A CN 202110549907A CN 112971784 A CN112971784 A CN 112971784A
- Authority
- CN
- China
- Prior art keywords
- signals
- fatigue driving
- physiological sensing
- gain
- signal
- 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.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 77
- 210000000988 bone and bone Anatomy 0.000 title claims abstract description 39
- 210000004556 brain Anatomy 0.000 claims abstract description 32
- 230000001815 facial effect Effects 0.000 claims abstract description 27
- 230000010287 polarization Effects 0.000 claims abstract description 26
- 230000000694 effects Effects 0.000 claims abstract description 24
- 238000005516 engineering process Methods 0.000 claims abstract description 16
- 210000005252 bulbus oculi Anatomy 0.000 claims abstract description 14
- 230000008030 elimination Effects 0.000 claims abstract description 11
- 238000003379 elimination reaction Methods 0.000 claims abstract description 11
- 238000007781 pre-processing Methods 0.000 claims abstract description 9
- 230000008859 change Effects 0.000 claims abstract description 8
- 238000006243 chemical reaction Methods 0.000 claims abstract description 8
- 238000012935 Averaging Methods 0.000 claims abstract description 6
- 230000003321 amplification Effects 0.000 claims description 53
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 53
- 238000012706 support-vector machine Methods 0.000 claims description 22
- 230000004424 eye movement Effects 0.000 claims description 20
- 230000003595 spectral effect Effects 0.000 claims description 12
- 230000003071 parasitic effect Effects 0.000 claims description 10
- 125000004432 carbon atom Chemical group C* 0.000 claims description 9
- 230000007613 environmental effect Effects 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 8
- 210000001508 eye Anatomy 0.000 claims description 8
- 238000007637 random forest analysis Methods 0.000 claims description 6
- 238000001228 spectrum Methods 0.000 claims description 6
- 230000001133 acceleration Effects 0.000 claims description 5
- 210000001097 facial muscle Anatomy 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 5
- 238000004422 calculation algorithm Methods 0.000 claims description 3
- 238000007477 logistic regression Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 239000000284 extract Substances 0.000 claims description 2
- 230000000193 eyeblink Effects 0.000 claims 1
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 238000000926 separation method Methods 0.000 abstract description 2
- 238000000034 method Methods 0.000 description 10
- 206010039203 Road traffic accident Diseases 0.000 description 5
- 230000000007 visual effect Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 206010041349 Somnolence Diseases 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 238000013145 classification model Methods 0.000 description 2
- 238000003066 decision tree Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 208000010340 Sleep Deprivation Diseases 0.000 description 1
- 208000032140 Sleepiness Diseases 0.000 description 1
- 208000003443 Unconsciousness Diseases 0.000 description 1
- 230000000739 chaotic effect Effects 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000002401 inhibitory effect Effects 0.000 description 1
- 230000003211 malignant effect Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000013186 photoplethysmography Methods 0.000 description 1
- 230000006461 physiological response Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 229920006395 saturated elastomer Polymers 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000037321 sleepiness Effects 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/16—Devices for psychotechnics; Testing reaction times ; Devices for evaluating the psychological state
- A61B5/18—Devices for psychotechnics; Testing reaction times ; Devices for evaluating the psychological state for vehicle drivers or machine operators
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B3/00—Apparatus for testing the eyes; Instruments for examining the eyes
- A61B3/10—Objective types, i.e. instruments for examining the eyes independent of the patients' perceptions or reactions
- A61B3/113—Objective types, i.e. instruments for examining the eyes independent of the patients' perceptions or reactions for determining or recording eye movement
-
- 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
- A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
-
- 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
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Surgery (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Animal Behavior & Ethology (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Pathology (AREA)
- Signal Processing (AREA)
- Psychiatry (AREA)
- Physiology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Educational Technology (AREA)
- Child & Adolescent Psychology (AREA)
- Developmental Disabilities (AREA)
- Power Engineering (AREA)
- Hospice & Palliative Care (AREA)
- Psychology (AREA)
- Social Psychology (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Human Computer Interaction (AREA)
- Ophthalmology & Optometry (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention relates to a wearable bone conduction fatigue driving detection method, which comprises the following steps: dynamically adjusting analog gain according to the change amplitude of the preprocessed physiological sensing signals of different types, performing local oversampling and averaging on the acquired physiological sensing signals, and performing downsampling on the averaged physiological sensing signals; preprocessing a physiological sensing signal; eliminating signal noise; performing analog-to-electric conversion on the physiological sensing signal; extracting fatigue driving characteristics; extracted fatigue driving detection features. The invention has the beneficial effects that: according to the wearable bone conduction fatigue driving detection device, physiological characteristic analog signals of a person can be efficiently and accurately acquired, brain wave signals, eyeball movement polarization signals, facial electromyographic signals and skin-electricity activity signals which are closely related to fatigue driving are obtained by adopting an implemented noise elimination technology and a signal separation technology, and accurate fatigue driving detection is carried out on the basis of the four types of signals and an individualized layered model.
Description
Technical Field
The invention belongs to the field of driving safety protection, and particularly relates to a wearable bone conduction fatigue driving detection method and device.
Background
With the increase of vehicle reserves worldwide, the rapid increase of the number of traffic accidents has become a serious social problem. According to relevant data statistics, about one third of casualties in traffic accidents every year are caused by fatigue driving. The driving fatigue refers to drowsiness and sleepiness caused by long-time driving or insufficient sleep of a driver, and data of many countries indicate that the driving fatigue is one of important reasons for causing malignant traffic accidents; the driver fatigue is the same as that of drunk driving and becomes a main hidden danger of traffic accidents, but the driver fatigue is easy to detect after drunk driving and has certain concealment. Driving fatigue refers to a brief and involuntary loss of attention, often characterized by a loss of consciousness and frequent and involuntary closure of both eyes.
With the high development of intelligent sensors, intelligent pattern recognition, automotive electronics and vehicle dynamics technologies, the fatigue driving detection device and technology become a research hotspot in the field of traffic safety at home and abroad in recent years. Through the detection of the fatigue state of the driver, traffic accidents caused by fatigue driving can be greatly reduced. Particularly, for drivers who are engaged in business operations such as long-distance passenger transport, freight transport and the like, due to professional requirements, the drivers often drive continuously for a long time, and the drivers are difficult to keep a high-alert state during driving; it is therefore more important to detect the fatigue state in real time. The existing fatigue driving detection device and method can be divided into three categories:
(1) the fatigue driving detection device and method based on the visual characteristics have the technical means that the visual characteristic information of a driver is extracted by using a computer visual technology to carry out fatigue judgment, the visual characteristics are greatly influenced by illumination, and turning and line changing can greatly influence the detection based on the vehicle behaviors, so that the accuracy, reliability and practicability are not high.
(2) A fatigue driving detecting device and method based on the automobile driving state, the technical means is to use the vehicle-mounted sensor to detect the speed, the lateral acceleration, the transverse displacement, the lane departure, the change of the vehicle driving track and other characteristics of the vehicle to estimate the fatigue state of the driver. However, the accuracy of such a fatigue driving detection method is yet to be further improved due to limitations imposed by the specific model of the vehicle, the specific conditions of the road, and the individual driving habits, driving experiences, and driving conditions of the driver.
(3) The device and the method for detecting the fatigue driving based on the physiological characteristics of the driver have the technical means that whether the driver enters a fatigue state or not is judged by utilizing some physiological index sensors. Because related researches show that the physiological response of a human body becomes dull in a fatigue state, namely, the response of the physiological signal of the human body is delayed, and the index also deviates from a normal state. The method is a relatively reliable fatigue driving detection method at present, but the technology is not widely used and popularized due to the following three reasons: 1) the physiological index signal of a human body is weak, and the fatigue driving detection precision is not high due to the interference of movement and environmental noise on signal acquisition in the driving process of a driver; 2) most of the existing fatigue driving detection devices based on physiological characteristics adopt intelligent bracelets, utilize various sensors (such as skin temperature sensors, heart rate sensors, photoplethysmography sensors and the like) and various biological signals to detect the fatigue driving state, but still can not monitor the brain and eye movement functions, and the sensitivity and specificity are still not ideal enough; 3) most of the existing fatigue driving detection methods based on physiological characteristics adopt a preset comparison threshold value mode for judgment (if the heart rate is smaller than a certain set threshold value, fatigue driving is judged), and the difference of the individual body conditions of drivers causes that the threshold value comparison method has no universality.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provides a wearable bone conduction fatigue driving detection method and device.
The wearable bone conduction fatigue driving detection method comprises the following steps:
s1, physiological sensing signal acquisition: integrating a high-sensitivity sensing circuit in a bone conduction device, and acquiring physiological sensing signals of a driver through the bone conduction device integrated with the high-sensitivity sensing circuit;
s2, preprocessing a physiological sensing signal: dynamically adjusting analog gain according to the change amplitude of the preprocessed physiological sensing signals of different types, performing local oversampling and averaging on the acquired physiological sensing signals, and performing downsampling on the averaged physiological sensing signals;
s3, signal noise elimination: eliminating the environmental noise contained in the mixed physiological sensing signal collected in the step S1 by adjusting the triple cascade amplifying circuit corresponding to the unit gain amplification, the triple cascade amplifying circuit corresponding to the feedforward differential pre-amplification and the triple cascade amplifying circuit corresponding to the gain dynamics respectively;
s4, performing physiological sensing signal mode-to-electricity conversion: separating the collected mixed physiological sensing signals into brain wave signals, eyeball movement polarization signals, facial electromyographic signals and electrodermal activity signals by using a band-pass filter; adopting an analog-digital converter A to convert brain wave signals, eyeball movement polarization signals and facial muscle electrical signals into corresponding digital signals, and adopting an analog-digital converter B to convert electrodermal activity signals into corresponding digital signals;
s5, fatigue driving feature extraction: dividing time sequence data of brain wave signals, eye movement polarization signals, facial electromyographic signals and skin electrical activity signals into segments with fixed sizes, and extracting time domain features, frequency spectrum features and nonlinear features from each segment;
s6, based on the fatigue driving detection characteristics extracted in the step S5, adopting a hierarchical fatigue driving detection model consisting of three basic classifiers; a random forest classifier is adopted in the first layer of the hierarchical fatigue driving detection model, an Adaboost classifier is adopted in the second layer, and a support vector machine is adopted in the last layer; judging whether a prediction sample with the probability larger than a set value in a first layer and a second layer of a hierarchical fatigue driving detection model is fatigue driving, transferring the residual sample to a third layer, judging the fatigue driving of the residual sample by using a support vector machine, if the output of the support vector machine is 1, judging the residual sample to be the fatigue driving, and if the output of the support vector machine is 0, judging the residual sample to be the non-fatigue driving; and if the driver is judged to be in the fatigue driving state, early warning information is sent to the driver through the bone conduction device.
Preferably, the manner of dynamically adjusting the analog gain in step S2 is specifically:
1) under the condition that the acquired physiological sensing signals do not comprise facial electromyographic signals, the gains of the brain wave signals and the eye movement polarization signals are kept at the maximum value;
2) when the amplitude of the acquired physiological sensing signal is suddenly increased, rapidly detecting whether a facial electromyographic signal is found by adopting peak envelope;
3) when the acquired physiological sensing signals comprise facial electromyographic signals, slowly adjusting the reduction speed of the amplitude of the acquired physiological sensing signals, and meanwhile repairing missing physiological sensing signal acquisition samples by using light-weight linear interpolation.
Preferably, step S3 specifically includes the following steps:
s3.1, amplifying unit gain; from kirchhoff's law:
in the above formula, the first and second carbon atoms are,
is a signal source;
is the input voltage;
and
is the skin electrode contact impedance;
is the input impedance of each operational amplifier;
is a complex number of symbols, and is,
is-1;
is the angular frequency of the wave to be transmitted,
is the angular velocity of the beam of light,
is the frequency of an alternating current signal
Is/are as follows
Doubling;
is the parasitic capacitance of each operational amplifier;
is the output voltage of each operational amplifier;
an alternating current impedance representing a parasitic capacitance;
removing input voltage from equation (1)
To obtain the actual gain of the triple cascade amplification circuit
Skin electrode contact impedance
And inherent capacitance of signal line𝐶The relationship between w:
in the above formula, the first and second carbon atoms are,
is the actual gain of the triple cascaded amplification circuit,𝐺=
/
wherein
Is the output voltage of each operational amplifier and,
is a signal source;
and
is the skin electrode contact impedance;
is the input resistance of each operational amplifier;
is a complex number of symbols, and is,
is-1;
is the angular frequency;𝐶w is the inherent capacitance of the signal line;𝐴is the ideal voltage gain for each operational amplifier;
is the input capacitance of each operational amplifier; using an operational amplifier buffer for each electrode, converting the high impedance line to approximately zero, rewriting equation (2) as the form of the operational amplifier buffer in the unity gain amplification stage:
in the above formula, the first and second carbon atoms are,
the resistance of the operational amplifier during the gain amplification stage,
is the contact impedance of the skin electrode,
is a complex number of symbols, and is,
is-1;
is the angular frequency;
is the inherent capacitance of the signal line during the gain amplification stage;
is the input capacitance of the operational amplifier in the gain amplification stage;
is the ideal voltage gain of the operational amplifier during the gain amplification stage;
is the parasitic capacitance of the operational amplifier used in the gain amplification stage;𝐴is each oneThe ideal voltage gain of the operational amplifier;
fixing the connection between each electrode and its buffer in a stable configuration according to equation (3); shielding by using a micro coaxial shielding cable, and then driving the cable shielding by using the same voltage as the output signal of the operational amplifier;
s3.2, feedforward differential pre-amplification: pre-amplifying the post-aural signals by adopting a feedforward differential amplification technology before driving the cable to a triple cascade amplification circuit; the cross-connection technology is adopted, and the gain of the two feedforward instrument amplifiers is set by using a gain resistor to generate a fully differential and pre-amplified signal;
s3.3, gain dynamic adjustment: and dynamically adjusting the signal gain in real time based on the difference amplitude between different types of physiological sensing signals to obtain high-resolution brain wave signals, eye movement polarization signals, facial electromyographic signals and electrodermal activity signals.
Preferably, step S5 specifically includes the following steps:
s5.1, extracting time domain features;
s5.1.1, extracting typical characteristics from each physiological sensing signal time sequence segment;
s5.1.2, decomposing the polarized signal of eyeball movement with wavelet, extracting the eye jump and eyeball link characteristics,
s5.1.3, connecting the fatigue driving time domain features extracted in the step S5.1.1 and the step S5.1.2 to finally form a time domain detection feature vector;
s5.2, extracting spectral features: extracting spectral features to analyze the characteristics of the electroencephalogram signals, wherein the spectral features comprise power ratios and absolute powers;
s5.3, extracting nonlinear features; extracting four nonlinear characteristics of a relevant dimension, a lya probov index, an entropy and a fractal dimension based on brain wave signals;
s5.4, selecting fatigue driving detection characteristics; selecting the most relevant feature set from a time domain feature set, a spectral feature set and a nonlinear feature set, including recursive feature elimination, L1-based feature selection and tree-based feature selection;
s5.4.1, adopting a greedy optimization algorithm to eliminate recursive features and removing the features with the minimum influence on training errors;
s5.4.2, selecting a detection model for linear fatigue driving based on the characteristics of L1, wherein the detection model for linear fatigue driving comprises Logistic regression and SVM, and the detection model for linear fatigue driving uses the characteristics that the L1 norm removal coefficient is zero;
s5.4.3, feature importance ranking generated using a tree-based model to cull irrelevant descriptors.
Preferably, the analog-to-digital converter a in step S4 is 24 bits; the analog-to-digital converter B is 16 bits.
Preferably, the set value of the probabilities in the first and second layers of the hierarchical fatigue driving detection model in step S6 is 0.7.
Preferably, the typical features extracted in step s5.1.1 include: mean, variance, minimum, maximum, Hoss parameter, skewness, and kurtosis.
Preferably, the eye jump and eye link characteristics in step S5.1.2 include average speed, maximum speed, average acceleration, maximum acceleration, blink amplitude, average amplitude, peak closing speed, peak opening speed, average closing speed, and closing time.
Preferably, the bone conduction device in step S1 and step S6 are both bone conduction headphones.
Wearable bone conduction fatigue driving detection device includes: the system comprises a physiological sensing signal acquisition module, a physiological sensing signal preprocessing module, a signal noise elimination module, a physiological sensing signal analog-to-digital conversion module, a fatigue driving characteristic extraction module and a fatigue driving judgment module;
the physiological sensing signal acquisition module integrates a high-sensitivity sensing circuit in the bone conduction device, and acquires a physiological sensing signal of a driver through the bone conduction device integrated with the high-sensitivity sensing circuit;
the physiological sensing signal preprocessing module is used for dynamically adjusting analog gain according to the change amplitude of the preprocessed different types of physiological sensing signals, carrying out local oversampling and averaging on the acquired physiological sensing signals and carrying out downsampling on the averaged physiological sensing signals;
the signal noise elimination module is used for eliminating environmental noise contained in the acquired mixed physiological sensing signal by adjusting a triple cascade amplification circuit corresponding to unit gain amplification, a triple cascade amplification circuit corresponding to feedforward differential pre-amplification and a triple cascade amplification circuit corresponding to gain dynamics respectively;
the physiological sensing signal module-to-electricity conversion module separates the collected mixed physiological sensing signals into brain wave signals, eyeball movement polarization signals, facial electromyographic signals and electrodermal activity signals by using a band-pass filter; adopting a 24-bit analog-digital converter to convert brain wave signals, eye movement polarization signals and facial muscle electrical signals into corresponding digital signals, and adopting a 16-bit analog-digital converter to convert electrodermal activity signals into corresponding digital signals;
the fatigue driving feature extraction module divides time sequence data of brain wave signals, eye movement polarization signals, facial electromyographic signals and skin electrical activity signals into segments with fixed sizes, and extracts time domain features, frequency spectrum features and nonlinear features from each segment;
the fatigue driving judging module adopts a hierarchical fatigue driving detection model consisting of three basic classifiers based on the fatigue driving detection features extracted by the fatigue driving feature extracting module; a random forest classifier is adopted in the first layer of the hierarchical fatigue driving detection model, an Adaboost classifier is adopted in the second layer, and a support vector machine is adopted in the last layer; judging whether a prediction sample with the probability larger than a set value in a first layer and a second layer of a hierarchical fatigue driving detection model is fatigue driving, transferring the residual sample to a third layer, judging the fatigue driving of the residual sample by using a support vector machine, if the output of the support vector machine is 1, judging the residual sample to be the fatigue driving, and if the output of the support vector machine is 0, judging the residual sample to be the non-fatigue driving; and if the driver is judged to be in the fatigue driving state, early warning information is sent to the driver through the bone conduction device.
The invention has the beneficial effects that: according to the wearable bone conduction fatigue driving detection device, physiological characteristic analog signals of a person can be efficiently and accurately acquired, brain wave signals, eyeball movement polarization signals, facial electromyographic signals and skin-electricity activity signals which are closely related to fatigue driving are obtained by adopting an implemented noise elimination technology and a signal separation technology, and accurate fatigue driving detection is carried out on the basis of the four types of signals and an individualized layered model.
Drawings
FIG. 1 is a flow chart of a method of detecting fatigue driving in an embodiment of the present invention;
FIG. 2 is a flow chart of physiological sensor signal preprocessing according to an embodiment of the present invention;
FIG. 3 is a diagram of a triple cascade amplifier for signal noise cancellation according to an embodiment of the present invention;
FIG. 4 is a flow chart of physiological sensing signal noise cancellation according to an embodiment of the present invention;
fig. 5 is a flowchart of fatigue driving detection feature extraction in the embodiment of the present invention.
Detailed Description
The present invention will be further described with reference to the following examples. The following examples are set forth merely to aid in the understanding of the invention. It should be noted that, for a person skilled in the art, several modifications can be made to the invention without departing from the principle of the invention, and these modifications and modifications also fall within the protection scope of the claims of the present invention.
Example 1:
as shown in fig. 1, a wearable bone conduction fatigue driving detection method includes the following steps:
s1, physiological sensing signal acquisition: integrating a high-sensitivity sensing circuit in a bone conduction earphone, and acquiring physiological sensing signals of a driver through the bone conduction earphone integrated with the high-sensitivity sensing circuit;
s2, as shown in fig. 2, the physiological sensing signal preprocessing is performed: dynamically adjusting analog gain according to the change amplitude of the preprocessed physiological sensing signals of different types, performing local oversampling and averaging on the acquired physiological sensing signals, and performing downsampling on the averaged physiological sensing signals;
the mode of dynamically adjusting the analog gain is specifically as follows:
1) under the condition that the acquired physiological sensing signals do not comprise facial electromyographic signals, the gains of the brain wave signals and the eye movement polarization signals are kept at the maximum value;
2) when the amplitude of the acquired physiological sensing signal is suddenly increased, rapidly detecting whether a facial electromyographic signal is found by adopting peak envelope;
3) when the acquired physiological sensing signals comprise facial electromyographic signals, slowly adjusting the reduction speed of the amplitude of the acquired physiological sensing signals to avoid gain oscillation, and meanwhile, repairing missing physiological sensing signal acquisition samples by using light-weight linear interpolation.
S3, as shown in fig. 4, performs signal noise removal: eliminating the environmental noise contained in the mixed physiological sensing signal collected in step S1 by adjusting the triple cascade amplifier circuit corresponding to the unit gain amplification, the triple cascade amplifier circuit corresponding to the feedforward differential pre-amplification, and the triple cascade amplifier circuit corresponding to the gain dynamics, respectively, as shown in fig. 3;
s4, performing physiological sensing signal mode-to-electricity conversion: separating the collected mixed physiological sensing signals into four physiological sensing signals for fatigue driving detection, such as brain wave signals, eye movement polarization signals, facial electromyographic signals, skin electrical activity signals and the like by using different band-pass filters; adopting a 24-bit analog-digital converter to convert brain wave signals, eye movement polarization signals and facial muscle electrical signals into corresponding digital signals, and adopting a 16-bit analog-digital converter to convert electrodermal activity signals into corresponding digital signals;
s5, as shown in fig. 5, fatigue driving feature extraction is performed: dividing time sequence data of brain wave signals, eye movement polarization signals, facial electromyographic signals and skin electrical activity signals into segments with fixed sizes, extracting time domain features, frequency spectrum features and nonlinear features from each segment, and performing feature selection to obtain fatigue driving detection;
s5.1, extracting time domain features;
s5.1.1, extracting 7 typical characteristics from each physiological sensing signal time sequence segment;
s5.1.2, carrying out wavelet decomposition on the eyeball motion polarization signal, extracting 10 eye jumps and eyeball link characteristics,
s5.1.3, connecting the fatigue driving time domain features extracted in the step S5.1.1 and the step S5.1.2 to finally form a 38-dimensional time domain detection feature vector;
s5.2, extracting spectral features: because brain wave signals generally exist in discrete frequency ranges at different stages, spectral features are extracted to analyze the features of the brain wave signals, and the spectral features comprise two spectral features of power ratio and absolute power;
s5.3, extracting nonlinear features; the bioelectricity signals show various complex nonlinear behaviors, particularly electroencephalogram chaotic parameters can be used for human fatigue detection, and four nonlinear characteristics of relevant dimensions, Riya Ponugh indexes, entropies and fractal dimensions are extracted based on brain wave signals;
s5.4, selecting fatigue driving detection characteristics; when all features are used together for fatigue driving detection, irrelevant relevant features or feature redundancy may degrade performance. Therefore, three feature selection methods are adopted to improve the detection accuracy, including recursive feature elimination, feature selection based on L1 and feature selection based on trees, and the most relevant feature set is selected from a time domain feature set, a frequency spectrum feature set and a nonlinear feature set;
s5.4.1, adopting a greedy optimization algorithm to eliminate recursive features and removing the features with the minimum influence on training errors;
s5.4.2, selecting a detection model for linear fatigue driving based on the characteristics of L1, wherein the detection model for linear fatigue driving comprises Logistic regression and SVM, and the detection model for linear fatigue driving uses the characteristics that the L1 norm removal coefficient is zero;
s5.4.3, feature importance ranking generated using a tree-based model to cull irrelevant descriptors.
S6, based on the fatigue driving detection characteristics extracted in the step S5, adopting a hierarchical fatigue driving detection model consisting of three basic classifiers; a random forest classifier is adopted in the first layer of the hierarchical fatigue driving detection model, an Adaboost classifier is adopted in the second layer, and a support vector machine is adopted in the last layer; the first layer is used for randomly selecting a feature subset containing more than 2 attributes from a candidate attribute set of each node of each decision tree by adopting a random forest classifier comprising a plurality of decision trees, and then selecting the optimal attribute from the subset to calculate the probability that a sample to be detected is fatigue driving, which is recorded as P1.
And the second layer adopts an Adaboost classifier and adopts a plurality of decision table weak classifiers to jointly train as a strong classifier, and weights are given to the classification error rate obtained based on each weak classifier and the sample weights are updated. Training the next weak classifier based on the weight matrix of all samples, circulating in sequence until the error rate is 0 or convergence, weighting and summing the results of all the weak classifiers to form a strong classifier, and calculating the probability that the sample to be detected is fatigue driving based on the strong classifier, wherein the probability is recorded as P2.
Judging that a prediction sample with the probability greater than a set value (P1 + P2> 0.7) in a first layer and a second layer of a hierarchical fatigue driving detection model is fatigue driving, training a two-classification support vector machine classification model by adopting 70% of sample number, judging the fatigue driving state of a detection sample with the probability less than 0.7 of the fatigue driving state estimated by the first two layers by adopting the support vector machine classification model, wherein each sample to be detected comprises 23-dimensional features after the features are selected, and using the vector after the features are selected as the input of a support vector machine classifier; if the output of the support vector machine is 1, determining the driver to be in fatigue driving, and if the output of the support vector machine is 0, determining the driver to be in non-fatigue driving; and if the driver is in the fatigue driving state, early warning information is sent to the driver through the bone conduction earphone.
Example 2:
based on embodiment 1, the step S3 of performing signal noise cancellation specifically includes:
s3.1, amplifying unit gain; because the physiological sensing signal is extremely weak, enhancement processing is required, which is obtained by kirchhoff's law:
in the above formula, the first and second carbon atoms are,
is a signal source;
is the input voltage;
and
is the skin electrode contact impedance;
is the input impedance of each operational amplifier;
is a complex number of symbols, and is,
is-1;
is the angular frequency (angular velocity),
is the angular velocity of the beam of light,
is the frequency of an alternating current signal
Is/are as follows
Doubling;
is the parasitic capacitance of each operational amplifier;
is the output voltage of each operational amplifier;
an alternating current impedance representing a parasitic capacitance;
removing input voltage from equation (1)
To obtain the actual gain of the triple cascade amplification circuit
Skin electrode contact impedance
And inherent capacitance of signal line𝐶The relationship between w:
in the above formula, the first and second carbon atoms are,
is the actual gain of the triple cascaded amplification circuit,𝐺=
/
wherein
Is the output voltage of each operational amplifier and,
is a signal source;
and
is a skin electrode jointA contact resistance;
is the input resistance of each operational amplifier;
is a complex number of symbols, and is,
is-1;
is the angular frequency (angular velocity);𝐶w is the inherent capacitance of the signal line;𝐴is the ideal voltage gain for each operational amplifier;
is the input capacitance of each operational amplifier,𝑍𝑜is the output impedance of each operational amplifier;
when motion occurs, the cable swings and the electrode moves respectively on the inherent capacitance of the signal wire𝐶w, skin electrode contact impedance
And
which results in a practical gain of the triple cascade amplification circuit
Fluctuating. In order to make the inherent capacitance of the signal line𝐶The ripple effect of w, which results from the change in the triboelectric process and the parasitic capacitance in the measurement network, is minimized; using an operational amplifier buffer for each electrode, converting the high impedance line to approximately zero, rewriting equation (2) as the form of the operational amplifier buffer in the unity gain amplification stage (first stage):
in the above formula, the first and second carbon atoms are,
the resistance of the operational amplifier during the gain amplification stage,
is the contact impedance of the skin electrode,
is a complex number of symbols, and is,
is-1;
is the angular frequency (angular velocity);
is the inherent capacitance of the signal line during the gain amplification stage;
is the input capacitance of the operational amplifier in the gain amplification stage;
is the ideal voltage gain of the operational amplifier during the gain amplification stage;
is the parasitic capacitance of the operational amplifier used in the gain amplification stage;
fixing the connection between each electrode and its buffer in a stable configuration to avoid triboelectric noise according to equation (3); and using micro-coaxial shielding cable for shielding, and then using the same voltage as the output signal of operational amplifier to drive the cable for shielding, thereby effectively reducing the inherent capacitance of the signal line in the gain amplification stage
Contact impedance to skin electrode
The unit gain amplification of the physiological sensing signal is realized;
s3.2, feedforward differential pre-amplification: in order to ensure the robustness to environmental interference, weak and overlapped behind-the-ear signals are pre-amplified in advance by adopting a feed-forward differential amplification technology before a cable is driven to a triple cascade amplification circuit (a physiological characteristic sensing circuit); inspired by the robustness of a balanced audio system to noise, a feedforward differential amplification technology is adopted, the common-mode rejection ratio is obviously improved, namely, the capacity of inhibiting noise from the environment is improved, a cross connection technology is adopted, the gains of two feedforward instrument amplifiers are set by only one gain resistor, and completely differential and pre-amplified signals are generated, so that the post-aural signals can resist environmental interference while driving a cable to a sensing circuit;
s3.3, gain dynamic adjustment: because the amplitude range difference between brain wave signals, eye movement polarization signals, facial electromyographic signals and electrodermal activity signals is large, the difference can cause the signals at the analog-to-digital converter on the sensing circuit to be saturated; and dynamically adjusting the signal gain in real time based on the difference amplitude between different types of physiological sensing signals to obtain high-resolution brain wave signals, eye movement polarization signals, facial electromyographic signals and electrodermal activity signals.
Example 3:
based on the embodiments 1 and 2, further, in step S4, the collected physiological characteristic simulation signals have frequencies in the ranges of 4 to 35 Hz, 0.1 to 10 Hz, and 10 to 100Hz, and the mixed physiological characteristic simulation signals are separated into brain wave signals, eye movement polarization signals, facial electromyogram signals, and electrodermal activity signals by applying different band-pass filters. Specifically, the method comprises the following steps: 1) respectively using 4-8 Hz, 8-12 Hz and 12-35 Hz band-pass filters to extract brain wave frequency band signals related to sleep; 2) extracting horizontal eye movement signals and vertical eye movement signals by using a 0.3-10 Hz filter; 3) then, performing 10-100Hz band-pass filtering and median filtering on the mixed signal, extracting the frequency band of the facial electromyographic signal, and removing peaks and other redundant components; 4) the electrodermal activity signal is a superposition of two different components, and a non-negative deconvolution technology is adopted to extract skin conductance response and skin conductance levels within frequency ranges of 0.05-1.5 Hz and 0-0.05 Hz.
Claims (10)
1. A wearable bone conduction fatigue driving detection method is characterized by comprising the following steps:
s1, physiological sensing signal acquisition: integrating a high-sensitivity sensing circuit in a bone conduction device, and acquiring physiological sensing signals of a driver through the bone conduction device integrated with the high-sensitivity sensing circuit;
s2, preprocessing a physiological sensing signal: dynamically adjusting analog gain according to the change amplitude of the preprocessed physiological sensing signals of different types, performing local oversampling and averaging on the acquired physiological sensing signals, and performing downsampling on the averaged physiological sensing signals;
s3, signal noise elimination: eliminating the environmental noise contained in the mixed physiological sensing signal collected in the step S1 by adjusting the triple cascade amplifying circuit corresponding to the unit gain amplification, the triple cascade amplifying circuit corresponding to the feedforward differential pre-amplification and the triple cascade amplifying circuit corresponding to the gain dynamics respectively;
s4, performing physiological sensing signal mode-to-electricity conversion: separating the collected mixed physiological sensing signals into brain wave signals, eyeball movement polarization signals, facial electromyographic signals and electrodermal activity signals by using a band-pass filter; adopting an analog-digital converter A to convert brain wave signals, eyeball movement polarization signals and facial muscle electrical signals into corresponding digital signals, and adopting an analog-digital converter B to convert electrodermal activity signals into corresponding digital signals;
s5, fatigue driving feature extraction: dividing time sequence data of brain wave signals, eye movement polarization signals, facial electromyographic signals and skin electrical activity signals into segments with fixed sizes, and extracting time domain features, frequency spectrum features and nonlinear features from each segment;
s6, based on the fatigue driving detection characteristics extracted in the step S5, adopting a hierarchical fatigue driving detection model consisting of three basic classifiers; a random forest classifier is adopted in the first layer of the hierarchical fatigue driving detection model, an Adaboost classifier is adopted in the second layer, and a support vector machine is adopted in the last layer; judging whether a prediction sample with the probability larger than a set value in a first layer and a second layer of a hierarchical fatigue driving detection model is fatigue driving, transferring the residual sample to a third layer, judging the fatigue driving of the residual sample by using a support vector machine, if the output of the support vector machine is 1, judging the residual sample to be the fatigue driving, and if the output of the support vector machine is 0, judging the residual sample to be the non-fatigue driving; and if the driver is judged to be in the fatigue driving state, early warning information is sent to the driver through the bone conduction device.
2. The wearable bone conduction fatigue driving detection method according to claim 1, wherein the manner of dynamically adjusting the analog gain in step S2 is specifically:
1) under the condition that the acquired physiological sensing signals do not comprise facial electromyographic signals, the gains of the brain wave signals and the eye movement polarization signals are kept at the maximum value;
2) when the amplitude of the acquired physiological sensing signal is suddenly increased, rapidly detecting whether a facial electromyographic signal is found by adopting peak envelope;
3) when the acquired physiological sensing signals comprise facial electromyographic signals, slowly adjusting the reduction speed of the amplitude of the acquired physiological sensing signals, and meanwhile repairing missing physiological sensing signal acquisition samples by using light-weight linear interpolation.
3. The wearable bone conduction fatigue driving detection method according to claim 1, characterized in that: step S3 specifically includes the following steps:
s3.1, amplifying unit gain; from kirchhoff's law:
in the above formula, the first and second carbon atoms are,
is a signal source;
is the input voltage;
and
is the skin electrode contact impedance;
is the input impedance of each operational amplifier;
is a complex number of symbols, and is,
is-1;
is the angular frequency of the wave to be transmitted,
is the angular velocity of the beam of light,
is the frequency of an alternating current signal
Is/are as follows
Doubling;
is the parasitic capacitance of each operational amplifier;
is the output voltage of each operational amplifier;
an alternating current impedance representing a parasitic capacitance;
removing input voltage from equation (1)
To obtain the actual gain of the triple cascade amplification circuit
Skin electrode contact impedance
And inherent capacitance of signal line𝐶The relationship between w:
in the above formula, the first and second carbon atoms are,
is the actual gain of the triple cascaded amplification circuit,𝐺=
/
wherein
Is each oneThe output voltage of the operational amplifier is set,
is a signal source;
and
is the skin electrode contact impedance;
is the input resistance of each operational amplifier;
is a complex number of symbols, and is,
is-1;
is the angular frequency;𝐶w is the inherent capacitance of the signal line;𝐴is the ideal voltage gain for each operational amplifier;
is the input capacitance of each operational amplifier; using an operational amplifier buffer for each electrode, converting the high impedance line to approximately zero, rewriting equation (2) as the form of the operational amplifier buffer in the unity gain amplification stage:
in the above formula, the first and second carbon atoms are,
for operation amplificationThe resistance of the amplifier during the gain amplification stage,
is the contact impedance of the skin electrode,
is a complex number of symbols, and is,
is-1;
is the angular frequency;
is the inherent capacitance of the signal line during the gain amplification stage;
is the input capacitance of the operational amplifier in the gain amplification stage;
is the ideal voltage gain of the operational amplifier during the gain amplification stage;
is the parasitic capacitance of the operational amplifier used in the gain amplification stage;𝐴is the ideal voltage gain for each operational amplifier;
fixing the connection between each electrode and its buffer in a stable configuration according to equation (3); shielding by using a micro coaxial shielding cable, and then driving the cable shielding by using the same voltage as the output signal of the operational amplifier;
s3.2, feedforward differential pre-amplification: pre-amplifying the post-aural signals by adopting a feedforward differential amplification technology before driving the cable to a triple cascade amplification circuit; the cross-connection technology is adopted, and the gain of the two feedforward instrument amplifiers is set by using a gain resistor to generate a fully differential and pre-amplified signal;
s3.3, gain dynamic adjustment: and dynamically adjusting the signal gain in real time based on the difference amplitude between different types of physiological sensing signals to obtain high-resolution brain wave signals, eye movement polarization signals, facial electromyographic signals and electrodermal activity signals.
4. The wearable bone conduction fatigue driving detection method according to claim 1, characterized in that: step S5 specifically includes the following steps:
s5.1, extracting time domain features;
s5.1.1, extracting typical characteristics from each physiological sensing signal time sequence segment;
s5.1.2, decomposing the polarized signal of eyeball movement with wavelet, extracting the eye jump and eyeball link characteristics,
s5.1.3, connecting the fatigue driving time domain features extracted in the step S5.1.1 and the step S5.1.2 to finally form a time domain detection feature vector;
s5.2, extracting spectral features: extracting spectral features to analyze the characteristics of the electroencephalogram signals, wherein the spectral features comprise power ratios and absolute powers;
s5.3, extracting nonlinear features; extracting four nonlinear characteristics of a relevant dimension, a lya probov index, an entropy and a fractal dimension based on brain wave signals;
s5.4, selecting fatigue driving detection characteristics; selecting the most relevant feature set from a time domain feature set, a spectral feature set and a nonlinear feature set, including recursive feature elimination, L1-based feature selection and tree-based feature selection;
s5.4.1, adopting a greedy optimization algorithm to eliminate recursive features and removing the features with the minimum influence on training errors;
s5.4.2, selecting a detection model for linear fatigue driving based on the characteristics of L1, wherein the detection model for linear fatigue driving comprises Logistic regression and SVM, and the detection model for linear fatigue driving uses the characteristics that the L1 norm removal coefficient is zero;
s5.4.3, feature importance ranking generated using a tree-based model to cull irrelevant descriptors.
5. The wearable bone conduction fatigue driving detection method according to claim 1, characterized in that: in step S4, the analog-to-digital converter a is 24 bits; the analog-to-digital converter B is 16 bits.
6. The wearable bone conduction fatigue driving detection method according to claim 1, characterized in that: in step S6, the probability set values in the first layer and the second layer of the hierarchical fatigue driving detection model are 0.7.
7. The wearable bone conduction fatigue driving detection method according to claim 4, wherein the typical features extracted in step S5.1.1 include: mean, variance, minimum, maximum, Hoss parameter, skewness, and kurtosis.
8. The wearable bone conduction fatigue driving detection method of claim 4, wherein the eye jump and eye link characteristics in step S5.1.2 comprise average velocity, maximum velocity, average acceleration, maximum acceleration, eye blink amplitude, average amplitude, peak close velocity, peak open velocity, average close velocity, and close time.
9. The wearable bone conduction fatigue driving detection method according to claim 1, characterized in that: the bone conduction device in step S1 and step S6 is a bone conduction headset.
10. A wearable bone conduction fatigue driving detection device, comprising: the system comprises a physiological sensing signal acquisition module, a physiological sensing signal preprocessing module, a signal noise elimination module, a physiological sensing signal analog-to-digital conversion module, a fatigue driving characteristic extraction module and a fatigue driving judgment module;
the physiological sensing signal acquisition module integrates a high-sensitivity sensing circuit in the bone conduction device, and acquires a physiological sensing signal of a driver through the bone conduction device integrated with the high-sensitivity sensing circuit;
the physiological sensing signal preprocessing module is used for dynamically adjusting analog gain according to the change amplitude of the preprocessed different types of physiological sensing signals, carrying out local oversampling and averaging on the acquired physiological sensing signals and carrying out downsampling on the averaged physiological sensing signals;
the signal noise elimination module is used for eliminating environmental noise contained in the acquired mixed physiological sensing signal by adjusting a triple cascade amplification circuit corresponding to unit gain amplification, a triple cascade amplification circuit corresponding to feedforward differential pre-amplification and a triple cascade amplification circuit corresponding to gain dynamics respectively;
the physiological sensing signal module-to-electricity conversion module separates the collected mixed physiological sensing signals into brain wave signals, eyeball movement polarization signals, facial electromyographic signals and electrodermal activity signals by using a band-pass filter; adopting a 24-bit analog-digital converter to convert brain wave signals, eye movement polarization signals and facial muscle electrical signals into corresponding digital signals, and adopting a 16-bit analog-digital converter to convert electrodermal activity signals into corresponding digital signals;
the fatigue driving feature extraction module divides time sequence data of brain wave signals, eye movement polarization signals, facial electromyographic signals and skin electrical activity signals into segments with fixed sizes, and extracts time domain features, frequency spectrum features and nonlinear features from each segment;
the fatigue driving judging module adopts a hierarchical fatigue driving detection model consisting of three basic classifiers based on the fatigue driving detection features extracted by the fatigue driving feature extracting module; a random forest classifier is adopted in the first layer of the hierarchical fatigue driving detection model, an Adaboost classifier is adopted in the second layer, and a support vector machine is adopted in the last layer; judging whether a prediction sample with the probability larger than a set value in a first layer and a second layer of a hierarchical fatigue driving detection model is fatigue driving, transferring the residual sample to a third layer, judging the fatigue driving of the residual sample by using a support vector machine, if the output of the support vector machine is 1, judging the residual sample to be the fatigue driving, and if the output of the support vector machine is 0, judging the residual sample to be the non-fatigue driving; and if the driver is judged to be in the fatigue driving state, early warning information is sent to the driver through the bone conduction device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110549907.3A CN112971784B (en) | 2021-05-20 | 2021-05-20 | Wearable bone conduction fatigue driving detection method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110549907.3A CN112971784B (en) | 2021-05-20 | 2021-05-20 | Wearable bone conduction fatigue driving detection method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112971784A true CN112971784A (en) | 2021-06-18 |
| CN112971784B CN112971784B (en) | 2021-08-03 |
Family
ID=76337054
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110549907.3A Active CN112971784B (en) | 2021-05-20 | 2021-05-20 | Wearable bone conduction fatigue driving detection method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112971784B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114036808A (en) * | 2022-01-10 | 2022-02-11 | 中铁大桥科学研究院有限公司 | Intelligent identification and monitoring method for fatigue parameters of ballastless track of high-speed rail large-span cable-stayed bridge |
Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003290180A (en) * | 2002-04-04 | 2003-10-14 | Rakken:Kk | Apparatus and method for measuring degree of fatigue |
| CN101127510A (en) * | 2006-08-17 | 2008-02-20 | 松下电器产业株式会社 | Differential input class D amplifier |
| CN101281507A (en) * | 2008-05-07 | 2008-10-08 | 李明荣 | USB interface type DSP real-time simulation development system |
| CN102647164A (en) * | 2012-04-26 | 2012-08-22 | 中国科学院微电子研究所 | Automatic gain control loop of ultralow-power-consumption medical equipment |
| CN202843628U (en) * | 2012-08-31 | 2013-04-03 | 漳州师范学院 | Audio-transported fatigue driving physiological signal collecting glasses |
| CN105266799A (en) * | 2015-09-16 | 2016-01-27 | 广东工业大学 | Electrocardiogram amplifier automatic gain control method based on blind separation technology |
| CN206007247U (en) * | 2016-06-24 | 2017-03-15 | 深圳市宇通联发科技有限公司 | A kind of bone conduction feedback device based on the anti-fatigue monitoring of brain wave |
| US20170336641A1 (en) * | 2017-08-07 | 2017-11-23 | Maximilian Ralph Peter von und zu Liechtenstein | Apparatus und Method for Rendering a Virtual Monitor on Smart Ophthalmic Devices in Augmented Reality Environments |
| CN108092636A (en) * | 2017-12-15 | 2018-05-29 | 佛山市晴天视觉数字科技有限公司 | Faint brain electro-detection numerical model analysis control chip system in high precision |
| CN108937923A (en) * | 2018-08-02 | 2018-12-07 | 扬州市紫麓信息技术有限公司 | A kind of real-time driving fatigue monitoring system based on EEG signals, electro-ocular signal and electromyography signal |
| CN208371800U (en) * | 2017-10-29 | 2019-01-15 | 吉林大学 | A kind of spectacle driving fatigue monitoring warning device based on multi-feature fusion |
| CN209220256U (en) * | 2018-01-13 | 2019-08-09 | 翟玉莲 | Fatigue driving based on pulse blood oxygen real-time monitoring monitors alarm set |
| CN110151203A (en) * | 2019-06-06 | 2019-08-23 | 常熟理工学院 | Fatigue driving recognition methods based on multistage avalanche type convolution Recursive Networks EEG analysis |
| CN111329497A (en) * | 2020-02-21 | 2020-06-26 | 华南理工大学 | Wearable fatigue driving monitoring system and method based on forehead electroencephalogram signals |
| CN111768620A (en) * | 2020-06-17 | 2020-10-13 | 浙大城市学院 | Road anomaly detection method based on window division and deformation clustering |
| US20200330017A1 (en) * | 2019-04-17 | 2020-10-22 | Bose Corporation | Fatigue and drowsiness detection |
| CN111860267A (en) * | 2020-07-13 | 2020-10-30 | 浙大城市学院 | Multichannel body-building movement identification method based on human body bone joint point positions |
-
2021
- 2021-05-20 CN CN202110549907.3A patent/CN112971784B/en active Active
Patent Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003290180A (en) * | 2002-04-04 | 2003-10-14 | Rakken:Kk | Apparatus and method for measuring degree of fatigue |
| CN101127510A (en) * | 2006-08-17 | 2008-02-20 | 松下电器产业株式会社 | Differential input class D amplifier |
| CN101281507A (en) * | 2008-05-07 | 2008-10-08 | 李明荣 | USB interface type DSP real-time simulation development system |
| CN102647164A (en) * | 2012-04-26 | 2012-08-22 | 中国科学院微电子研究所 | Automatic gain control loop of ultralow-power-consumption medical equipment |
| CN202843628U (en) * | 2012-08-31 | 2013-04-03 | 漳州师范学院 | Audio-transported fatigue driving physiological signal collecting glasses |
| CN105266799A (en) * | 2015-09-16 | 2016-01-27 | 广东工业大学 | Electrocardiogram amplifier automatic gain control method based on blind separation technology |
| CN206007247U (en) * | 2016-06-24 | 2017-03-15 | 深圳市宇通联发科技有限公司 | A kind of bone conduction feedback device based on the anti-fatigue monitoring of brain wave |
| US20170336641A1 (en) * | 2017-08-07 | 2017-11-23 | Maximilian Ralph Peter von und zu Liechtenstein | Apparatus und Method for Rendering a Virtual Monitor on Smart Ophthalmic Devices in Augmented Reality Environments |
| CN208371800U (en) * | 2017-10-29 | 2019-01-15 | 吉林大学 | A kind of spectacle driving fatigue monitoring warning device based on multi-feature fusion |
| CN108092636A (en) * | 2017-12-15 | 2018-05-29 | 佛山市晴天视觉数字科技有限公司 | Faint brain electro-detection numerical model analysis control chip system in high precision |
| CN209220256U (en) * | 2018-01-13 | 2019-08-09 | 翟玉莲 | Fatigue driving based on pulse blood oxygen real-time monitoring monitors alarm set |
| CN108937923A (en) * | 2018-08-02 | 2018-12-07 | 扬州市紫麓信息技术有限公司 | A kind of real-time driving fatigue monitoring system based on EEG signals, electro-ocular signal and electromyography signal |
| US20200330017A1 (en) * | 2019-04-17 | 2020-10-22 | Bose Corporation | Fatigue and drowsiness detection |
| CN110151203A (en) * | 2019-06-06 | 2019-08-23 | 常熟理工学院 | Fatigue driving recognition methods based on multistage avalanche type convolution Recursive Networks EEG analysis |
| CN111329497A (en) * | 2020-02-21 | 2020-06-26 | 华南理工大学 | Wearable fatigue driving monitoring system and method based on forehead electroencephalogram signals |
| CN111768620A (en) * | 2020-06-17 | 2020-10-13 | 浙大城市学院 | Road anomaly detection method based on window division and deformation clustering |
| CN111860267A (en) * | 2020-07-13 | 2020-10-30 | 浙大城市学院 | Multichannel body-building movement identification method based on human body bone joint point positions |
Non-Patent Citations (3)
| Title |
|---|
| HAHM, SANGWOO: "Drowsiness Driving Prevention System using Bone Conduction Device", 《KSII TRANSACTIONS ON INTERNET AND INFORMATION SYSTEMS》 * |
| 张有涛: "单通道8bit 1.4 GS/s折叠内插ADC", 《固体电子学研究与进展》 * |
| 徐久钦: "运动健将的得力助手 韶音Xtrainerz骨传导耳机体验评测", 《消费电子》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114036808A (en) * | 2022-01-10 | 2022-02-11 | 中铁大桥科学研究院有限公司 | Intelligent identification and monitoring method for fatigue parameters of ballastless track of high-speed rail large-span cable-stayed bridge |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112971784B (en) | 2021-08-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ma et al. | Driving fatigue detection from EEG using a modified PCANet method | |
| CN108693973B (en) | Emergency condition detection system fusing electroencephalogram signals and environmental information | |
| CN109498041B (en) | Driver road rage state identification method based on electroencephalogram and pulse information | |
| CN102119857B (en) | Electroencephalogram detecting system and method for fatigue driving on basis of matching pursuit algorithm | |
| CN112754431A (en) | Respiration and heartbeat monitoring system based on millimeter wave radar and lightweight neural network | |
| AlSharabi et al. | A DWT-entropy-ANN based architecture for epilepsy diagnosis using EEG signals | |
| CN108836324B (en) | Fatigue driving early warning method and system based on electroencephalogram signal monitoring | |
| CN113100756A (en) | Stacking-based Parkinson tremor detection method | |
| CN112971784B (en) | Wearable bone conduction fatigue driving detection method and device | |
| Zhou et al. | Vigilance detection method for high‐speed rail using wireless wearable EEG collection technology based on low‐rank matrix decomposition | |
| CN109481164B (en) | Electric wheelchair control system based on electroencephalogram signals | |
| CN108013884A (en) | A kind of fatigue of automobile driver drives active safety monitoring system and its monitoring method | |
| CN110584597A (en) | Multi-channel electroencephalogram signal monitoring method based on time-space convolutional neural network and application | |
| Suman et al. | EOG based vigilance monitoring system | |
| Jiang et al. | Denoising and chaotic feature extraction of electrocardial signals for driver fatigue detection by Kolmogorov entropy | |
| Hu et al. | Application of energy entropy in motor imagery EEG classification | |
| Lei et al. | Study on driving fatigue evaluation system based on short time period ECG signal | |
| Pramanick et al. | Deep learning based urban sound classification and ambulance siren detector using spectrogram | |
| Zhu et al. | Heavy truck driver's drowsiness detection method using wearable eeg based on convolution neural network | |
| Zhu et al. | Eeg-based system using deep learning and attention mechanism for driver drowsiness detection | |
| Dey et al. | Automatic detection of drowsiness in EEG records based on time analysis | |
| Chai et al. | Classification of driver fatigue in an electroencephalography-based countermeasure system with source separation module | |
| Qin et al. | Classification of driving fatigue based on EEG signals | |
| Patrick et al. | An algorithm for automatic detection of drowsiness for use in wearable EEG systems | |
| Postalcioglu | Wavelet transform based feature extraction for EEG signal classification |
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 | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20231201 Address after: Room 521, Unit 2, Building 4, Dexin Yunchuan Business Center, Xihu District, Hangzhou City, Zhejiang Province, China (self declared) Patentee after: Hangzhou Zhenwei Food Collection Technology Co.,Ltd. Address before: 310015 No. 51 Huzhou street, Hangzhou, Zhejiang, Gongshu District Patentee before: HANGZHOU City University |
|
| TR01 | Transfer of patent right |