CN113370786A - Vehicle-mounted drunk driving comprehensive detection system for unit vehicle based on multi-source information fusion - Google Patents

Vehicle-mounted drunk driving comprehensive detection system for unit vehicle based on multi-source information fusion Download PDF

Info

Publication number
CN113370786A
CN113370786A CN202110649485.7A CN202110649485A CN113370786A CN 113370786 A CN113370786 A CN 113370786A CN 202110649485 A CN202110649485 A CN 202110649485A CN 113370786 A CN113370786 A CN 113370786A
Authority
CN
China
Prior art keywords
driver
alcohol
drunk driving
module
drunk
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
Application number
CN202110649485.7A
Other languages
Chinese (zh)
Other versions
CN113370786B (en
Inventor
赵红专
蒋燕
代静
吴浩
蔡介南
徐阳
黄华波
欧俊
莫家龙
邱琦骁
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guilin Liangzi Iot Technology Co ltd
Guilin University of Electronic Technology
Original Assignee
Guilin Liangzi Iot Technology Co ltd
Guilin University of Electronic Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Guilin Liangzi Iot Technology Co ltd, Guilin University of Electronic Technology filed Critical Guilin Liangzi Iot Technology Co ltd
Priority to CN202110649485.7A priority Critical patent/CN113370786B/en
Publication of CN113370786A publication Critical patent/CN113370786A/en
Application granted granted Critical
Publication of CN113370786B publication Critical patent/CN113370786B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60K—ARRANGEMENT OR MOUNTING OF PROPULSION UNITS OR OF TRANSMISSIONS IN VEHICLES; ARRANGEMENT OR MOUNTING OF PLURAL DIVERSE PRIME-MOVERS IN VEHICLES; AUXILIARY DRIVES FOR VEHICLES; INSTRUMENTATION OR DASHBOARDS FOR VEHICLES; ARRANGEMENTS IN CONNECTION WITH COOLING, AIR INTAKE, GAS EXHAUST OR FUEL SUPPLY OF PROPULSION UNITS IN VEHICLES
    • B60K28/00—Safety devices for propulsion-unit control, specially adapted for, or arranged in, vehicles, e.g. preventing fuel supply or ignition in the event of potentially dangerous conditions
    • B60K28/02—Safety devices for propulsion-unit control, specially adapted for, or arranged in, vehicles, e.g. preventing fuel supply or ignition in the event of potentially dangerous conditions responsive to conditions relating to the driver
    • B60K28/06—Safety devices for propulsion-unit control, specially adapted for, or arranged in, vehicles, e.g. preventing fuel supply or ignition in the event of potentially dangerous conditions responsive to conditions relating to the driver responsive to incapacity of driver
    • B60K28/063—Safety devices for propulsion-unit control, specially adapted for, or arranged in, vehicles, e.g. preventing fuel supply or ignition in the event of potentially dangerous conditions responsive to conditions relating to the driver responsive to incapacity of driver preventing starting of vehicles
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60K—ARRANGEMENT OR MOUNTING OF PROPULSION UNITS OR OF TRANSMISSIONS IN VEHICLES; ARRANGEMENT OR MOUNTING OF PLURAL DIVERSE PRIME-MOVERS IN VEHICLES; AUXILIARY DRIVES FOR VEHICLES; INSTRUMENTATION OR DASHBOARDS FOR VEHICLES; ARRANGEMENTS IN CONNECTION WITH COOLING, AIR INTAKE, GAS EXHAUST OR FUEL SUPPLY OF PROPULSION UNITS IN VEHICLES
    • B60K28/00—Safety devices for propulsion-unit control, specially adapted for, or arranged in, vehicles, e.g. preventing fuel supply or ignition in the event of potentially dangerous conditions
    • B60K28/02—Safety devices for propulsion-unit control, specially adapted for, or arranged in, vehicles, e.g. preventing fuel supply or ignition in the event of potentially dangerous conditions responsive to conditions relating to the driver
    • B60K28/06—Safety devices for propulsion-unit control, specially adapted for, or arranged in, vehicles, e.g. preventing fuel supply or ignition in the event of potentially dangerous conditions responsive to conditions relating to the driver responsive to incapacity of driver
    • B60K28/066—Safety devices for propulsion-unit control, specially adapted for, or arranged in, vehicles, e.g. preventing fuel supply or ignition in the event of potentially dangerous conditions responsive to conditions relating to the driver responsive to incapacity of driver actuating a signalling device

Abstract

The invention discloses a vehicle-mounted drunk driving comprehensive detection system for unit vehicles based on multi-source information fusion, which comprises identity verification and alcohol detection in the processes of starting and driving of the vehicles and is suitable for unit vehicles such as official vehicles, tourism charters, transport fleets and the like. The face recognition system comprises a main control module, a display module, a communication module, an audible and visual alarm module, a face recognition and alcohol detection module, an alcohol acquisition module and a positioning module, wherein the Facenet system is adopted to realize face recognition identity verification; recognizing drunk driving through a trained Nin network by adopting visual detection; adopting a plurality of alcohol detectors for joint detection, and comprehensively obtaining detection results by combining the values detected by the three sensors and the results of face image recognition through a deep learning training method; alcohol detection is carried out in the starting process of the vehicle, and the starting and stopping of the vehicle are controlled according to the detection result; and the driver identity and drunk driving detection are verified in real time in the driving process, the vehicle is controlled to give an alarm according to the result, and relevant information is uploaded to a superior management department and a traffic police head team command center.

Description

Vehicle-mounted drunk driving comprehensive detection system for unit vehicle based on multi-source information fusion
Technical Field
The invention belongs to the technical field of drunk driving detection, and particularly relates to a vehicle-mounted drunk driving comprehensive detection system for a unit vehicle based on multi-source information fusion.
Background
With the development of economy, the automobile holding capacity in the world is on an increasing trend. In China, the trend is more obvious, traffic accidents become problems all over the world, the traffic accidents caused by automobiles bring great personnel and property losses to the world, property losses and casualties caused by the traffic accidents become problems which cannot be ignored in the world at present, and the driving safety is more and more important. According to the investigation of the world health organization, drunk driving is an important cause of traffic accidents. Every year, the drunk driving causes nearly ten thousands of traffic accidents, and seriously threatens the life and property safety of people. The law in China clearly defines that: the alcohol content reaches 20mg/100ml but less than 80mg/100ml, and belongs to drinking driving. For drunk driving, the driving license of 6 months is temporarily deducted, and the penalty is more than 1000 Yuan and less than 2000 Yuan. Before that, once drunk driving is punished, drunk driving again is carried out, the drunk driving is detained in less than 10 days, and the drunk driving is carried out with a fine of more than 1000 Yuan and less than 2000 Yuan, so that the driving license is cancelled. After drinking, the commercial vehicle is held for 15 days and is subjected to 5000 yuan of fine, the driving license is suspended and sold, and the driving license cannot be obtained again within 5 years.
Although the punishment is increased all the time, the drunk driving problem is not completely eradicated. The reason is that the existing alcohol detection method is that after a traffic police stops a suspicious vehicle, whether a driver belongs to drunk driving or not is judged by detecting the content of alcohol contained in gas exhaled by the driver, the detection method is random passive, only a part of drunk driving drivers can be detected, most drunk driving drivers are not detected, the detection range is small, and in the detection process, even if drunk driving is detected, the drunk driving event also happens, the precaution can not be realized, and the traffic accident is easy to happen, so that detection means are urgently needed, the traffic accident is reduced, and the driving safety rate and the traffic efficiency are improved.
Disclosure of Invention
The invention provides a vehicle-mounted drunk driving comprehensive detection system for unit vehicles based on multi-source information fusion, which is suitable for unit vehicles such as official vehicles, touring vans, transport fleets and the like, can actively detect the alcohol content of a driver before the driver drives the vehicle and detect the alcohol content in real time during driving, and controls the vehicle according to a detection result to prevent the situations of drunk driving, midway drinking of the driver or driver replacement, so that drunk driving behaviors are restrained from the source, traffic accidents are reduced, and the driving safety rate and the traffic efficiency are improved.
In order to achieve the purpose, the technical scheme of the invention is as follows:
a vehicle-mounted drunk driving comprehensive detection system for unit vehicles based on multi-source information fusion mainly comprises a main control module, a display module, a communication module, a positioning module, an audible and visual alarm, a face recognition and alcohol detection module and an alcohol acquisition module, wherein the main control module, the display module, the communication module, the positioning module, the audible and visual alarm, the face recognition and alcohol detection module and the alcohol acquisition module are installed in an automobile; the main control module is provided with a first processor; the positioning module is provided with a GPS positioning system; the audible and visual alarm is provided with a buzzer and an alarm lamp; the display module is provided with a display screen; the face recognition and alcohol detection module comprises a camera, a storage unit and a second processor; the camera and the storage unit are respectively electrically connected with the second processor; the communication module comprises an information transmission unit, a voice prompt unit and a telephone prompt unit; the information transmission unit, the voice prompt unit, the telephone prompt unit, the second processor, the GPS, the buzzer, the alarm lamp and the display screen are respectively and electrically connected with the first processor; the first processor is connected with an ignition control module of the automobile through a relay; the first processor is also electrically connected with a lamp control module of the automobile; the alcohol collection module comprises a storage battery, a sweat alcohol sensor, a gas alcohol sensor and an infrared sensor; the storage battery, the sweat alcohol sensor, the gas alcohol sensor and the infrared sensor are respectively and electrically connected with the first processor.
The first processor and the ignition control part of the automobile form an AND gate through the relay, the engine is controlled to be started in a combined mode, the relay in the initial state is closed, and when the first processor receives the drunk driving signal, the relay is controlled to be disconnected, and the engine is closed.
As a further technical improvement, the sweat alcohol sensor is arranged on the left side of the steering wheel of the automobile; the infrared sensor is arranged on the right side of the automobile steering wheel; the gas alcohol sensor is arranged on the left side of the driver seat of the automobile.
As a further technical improvement, the camera is arranged at one side in front of the copilot of the automobile, faces to the driver seat of the automobile and monitors the head state of the driver.
As a further technical improvement, an electric heating piece and a chemical drug pilocarpine are further installed on the automobile steering wheel, and the electric heating piece is electrically connected with the first processor.
As a further technical improvement, the first processor and the second processor both adopt AT89C51 single-chip microcomputers.
As a further technical improvement, the positioning module is arranged at the position of an A column of the automobile.
The communication module comprises an information transmission unit, and the information transmission unit is used for uploading the vehicle position, the license plate, the alcohol degree and the identity information of the driver to a traffic police command center through a mobile network when the driver belongs to the drunk driving behavior.
The communication module further comprises a voice prompt unit, and the voice prompt unit is used for providing voice broadcast when the driver belongs to the drunk driving behavior.
The communication module further comprises a telephone prompting unit, and the telephone prompting unit is used for providing a telephone dialing service for the driver when the driver belongs to the drunk driving behavior.
The system adopts two methods of detecting the alcohol content in the driver by various sensors and identifying the facial features of the driver, such as congestion of eyes, flushing or tragic white face, stiff expression, tiredness, sleepiness and the like, by a camera to evaluate whether the driver drunk to drive, and comprehensively judges the alcohol content in the driver by a deep learning training method.
And (3) finishing the tasks of face detection and face alignment by adopting MTCNN, carrying out face recognition by adopting Facenet, and carrying out drunk driving judgment by adopting a Nin network recognition image.
The alcohol acquisition module is used for detecting the alcohol concentration of a driver and sending an alcohol concentration signal to the main control module.
The sweat alcohol sensor is a patch type alcohol sensor and is used for detecting whether sweat secreted on the hands of a driver contains alcohol components, and because the alcohol concentration in the sweat has a certain relation with the alcohol concentration in blood, the alcohol concentration in the blood of the driver can be known by monitoring the alcohol concentration in the sweat, so that the aim of monitoring whether the driver is drunk or not is fulfilled; the sweat alcohol sensor is arranged on the steering wheel, particularly on the left side of the steering wheel; detecting the alcohol concentration data in sweat on a driver holding the steering wheel, converting the alcohol concentration data into an electric signal, transmitting the electric signal to the main control module for processing, and comprehensively judging whether the driver drives drunk or not together with the other two sensors and the camera;
meanwhile, the measuring result is prevented from being influenced by insufficient sweat amount when the driver detects the sweat amount. An electric heating piece and a chemical drug Pilocarpine (Pilocarpine) are arranged on an automobile steering wheel, so that sweat can be promoted to be discharged, alcohol can be volatilized, and the detection precision is enhanced;
the gas alcohol sensor is a device for detecting alcohol gas around a driver seat based on the principle that the detected alcohol concentration in the gas is converted into a useful electric signal, and information related to the existence of the gas to be detected in the environment can be obtained according to the strength of the electric signals, when a driver sits on a seat and starts an ignition device, the gas alcohol sensor starts to work, and can be but not limited to an alcohol sensitive sensor, the alcohol sensor is used for detecting the alcohol concentration value P around the driver in a vehicle, after a certain threshold value is reached, the information is transmitted to the center of a main control module for judgment, the gas alcohol sensor is arranged on the left side of the driver seat, and the sensor, other two sensors and a camera are used for comprehensively judging whether the driver is drunk or not;
the infrared sensor is designed based on an infrared detection circuit system, is arranged on the right side of a steering wheel, takes the fingers of a driver as detection objects, when the driver holds the steering wheel, an alcohol detection module starts to circularly detect alcohol in the driver body within set time, an analog signal output by the infrared sensor is converted into a digital signal through an A/D (analog/digital) converter, and the digital signal is input into an AT89C51 singlechip of a main control module to carry out data analysis and processing, so that the alcohol content in the driver body is judged, and whether the driver is drunk or not is comprehensively judged by the infrared sensor, the other two sensors and a camera.
The face recognition and alcohol detection module comprises a camera, a storage unit and a processor; the camera is used for collecting head image information of a driver, the storage unit stores pre-stored registered face data information of the driver, the processor preprocesses the picture, and adopts a trained Facenet network to carry out face recognition and verify the identity of the driver; identifying facial features such as congestion of eyes, flushing or tragic white face, stiff expression, tiredness, drowsiness and the like by adopting a convolutional neural network (Nin network), evaluating whether a driver drives drunk or not, and transmitting a judgment result to a main control module; because the expressions such as distortion, confusion and the like can appear on the facial expression of the alcoholics, and the expressions such as lethargy and the like can appear at the same time, whether the driver drunk or not can be effectively evaluated through the facial expression.
The main control module adopts a circuit which takes a single chip microcomputer as main control, integrates the numerical values of concentration detection of the three alcohol sensors and the result of face image recognition through deep learning to obtain the result of alcohol detection and take corresponding measures, and is used for controlling and coordinating the work of other modules. The alcohol collection module can be controlled to detect the alcohol in the blood of the driver at regular time. When the automobile is started, after the identity authentication is successful, alcohol detection is carried out; if the detection result is drunk driving, the buzzer gives an alarm for prompting, and meanwhile, a voice prompt unit in the communication module is started to inquire whether a driver needs to make a call or search for drunk driving, so that convenience is provided for the drunk driver, wherein the call comprises a prestored parent call, a prestored drunk driving call and the like; in the driving process, the camera compares facial features of a driver in real time, simultaneously, the face recognition module, the alcohol detection module and the alcohol acquisition module perform alcohol detection, if the detection result is judged to be drunk driving, the positioning module and the communication module are started, license plates, position information, alcohol concentration and personal information of the driver are uploaded to a higher management department of a company and a traffic police head team command center in a network mode, meanwhile, the singlechip controls the acousto-optic alarm module to be started with the display module, the display screen can display warning information, the buzzer sends out harsh buzzing to prompt the driver to pay attention to own state, and the vehicle is stopped when being prompted by voice.
The positioning module uses a GPS positioning technology, and receives positioning signals through a GPS antenna so as to determine the position of the module, including longitude, latitude, time information and the like, and realize the positioning of the vehicle.
The communication module comprises a voice prompt unit, a telephone prompt unit and an information transmission unit; when a driver is found to drive drunk in the driving process of an automobile, the license plate, the positioning, the alcohol concentration and the personal information of the driver are uploaded to a traffic police headquarters command center in a mobile network mode, and meanwhile, a voice prompter broadcasts a result in voice when the driver is detected to drive drunk; in addition, the telephone prompting unit provides a telephone dialing service for the driver and dials a telephone prestored in advance.
When receiving the alarm signal of the main control module, the sound and light alarm module can control the alarm lamp and the buzzer to alarm so as to warn a driver.
The display module is used for displaying the judgment result of the main control module so that the driver can know the state of the driver.
The use method of the vehicle-mounted drunk driving comprehensive detection system based on multi-source information fusion comprises the alcohol detection in the automobile starting and driving processes, is suitable for unit vehicles such as official vehicles, tourism charters, transport fleets and the like, and comprises the following steps:
s1: the driver starts the automobile engine through a key, the camera is opened at the moment, the facial image of the driver is collected, the second processor adopts the Facenet network to verify the identity of the driver, and meanwhile, the obtained facial features of the driver are uploaded to a cloud end to be called by a manager;
s2: after the identity authentication is successful, the sweat alcohol sensor, the gas alcohol sensor and the infrared sensor are electrified to start alcohol detection of the driver, and signals are transmitted to the main control module for processing; meanwhile, the camera acquires a facial image of the driver and transmits the facial image to the second processor, the second processor adopts a Nin network to identify facial features such as congestion of eyes, flushing or tragic white face, stiff expression, fatigue, drowsiness and the like, whether the driver drives drunk or not is evaluated, and the result is transmitted to the main control module;
s3: the method comprises the steps that a main control module performs data fusion by adopting a prediction function obtained by deep learning training, a sweat alcohol sensor, a gas alcohol sensor, an infrared sensor and an image recognition result jointly judge whether a driver drives drunk or not, the data content and the image recognition result detected by the three sensors are input into the prediction function to obtain a prediction real value, and then the prediction real value is converted into 0 or 1 through a Sigmoid function, wherein 0 represents drunk driving behavior, and 1 represents drunk driving behavior, so that whether drunk driving behavior occurs or not is judged;
s4: in the starting process, if the automobile is detected to be qualified, the automobile normally runs, if drunk driving is detected, the engine is shut down, and meanwhile, a voice prompt unit in the communication module inquires whether a driver needs to make a call or search for drunk driving, wherein the call comprises a pre-stored parent call, a prestored drunk driving call and the like; in the driving process of the automobile, the camera compares facial features of the driver in real time to ensure that the information of the driver can be correctly matched, if the information of the driver cannot be successfully matched, the sound-light alarm module is started, the automobile starts alarming, double-flash is started to prompt parking, and meanwhile, the license plate, the positioning, the alcohol concentration and the personal information of the driver are uploaded to a company higher-level management department; in addition, drunk driving detection is carried out in real time in the driving process, if the detection result is drunk driving, the main control module receives the position information of the positioning module, the license plate, the positioning, the alcohol concentration and the personal information of a driver are uploaded to a higher management department of a company and a traffic police command center through the communication module, and meanwhile, the car gives an alarm and opens double-flash prompt to stop.
As a further technical improvement, the prediction function is obtained by a deep learning training method and comprises the following steps: arranging relevant drivers to perform experiments, measuring the in-vivo alcohol content of the drivers by using a blood alcohol content testing method, then arranging the drivers to get on the vehicles to measure three groups of sensor data and image recognition results, finding out the relationship among the three groups of sensor data, the image recognition results and the data obtained by using the blood alcohol content testing by using a deep learning training method to obtain a prediction function, wherein the measured values of the three sensors are respectively marked as x1,x2,x3And the image recognition result is marked as x4；
The specific algorithm is as follows: a set of data x obtained from the experiment1,x2,x3,x4Is marked as
Figure BDA0003111179450000051
The data measured by the blood alcohol content test method is recorded as yi(ii) a Multiple groups of data are obtained through multiple experiments and are recorded as a sample set
Figure BDA0003111179450000052
Figure BDA0003111179450000053
Obtaining a prediction function by linear regression
Figure BDA0003111179450000054
So that
Figure BDA0003111179450000055
Get the optimal solution
Figure BDA0003111179450000056
Predicting the value of linear regression
Figure BDA0003111179450000057
Is recorded as z, then
Figure BDA0003111179450000058
And z is the predicted real value, now to convert z to a y value close to 0 or 1, a Sigmoid function is introduced
Figure BDA0003111179450000059
The conversion of the measured values is completed, after which a conversion value of 0 or 1, i.e. y, is obtainediE (0,1), wherein 0 represents the drunk driving behavior, and 1 represents the drunk driving behavior, thereby judging whether the drunk driving behavior occurs.
As further technical improvement, the identity verification model and the image drunk driving recognition model are subjected to recognition training in a cloud, and the recognition training comprises Facenet, Nin and deep learning construction and training. The method comprises the following steps:
the invention adopts MTCNN to complete the tasks of face detection and face alignment, adopts Facenet to carry out face recognition, maps the face to the characteristic vector of Euclidean space through CNN (convolutional neural network), calculates the distance of the face characteristics of different pictures, and trains the network through the prior knowledge that the distance of the same individual face is always smaller than the distance of different individual faces.
According to the drunk driving identification method, drunk driving identification is carried out through a Nin network model, firstly, a large number of samples are collected by using a drunk driving detection system, and the drunk driving identification method is divided into drunk type and non-drunk type according to whether a driver is drunk, wherein most of the drunk driving identification method is used as a training sample, and the rest of the drunk driving identification method is used as a testing sample. And acquiring face region images through a camera to obtain training samples, and training the Nin network model by using the face region images. The purpose of image recognition is whether the driver drives with wine or not, and only 2 types of drunk driving and drunk-free driving are needed. The trained Nin network model can identify whether the driver is drunk driving in the image.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention is suitable for unit vehicles such as official vehicles, travelling vans, transport fleets and the like, comprises the alcohol detection during the starting process and the driving process of the automobile, and adopts a deep learning training method to integrate a plurality of sensors and cameras to identify the facial features such as hyperemia, flushing or tragic white, stiff expression, tiredness, drowsiness and the like of the eyes of a driver by an alcohol acquisition module and a face recognition and alcohol detection module so as to finish the alcohol detection during the starting process of the automobile and also improve the accuracy of the alcohol concentration judgment; meanwhile, real-time comparison and real-time alcohol detection of the facial features of the driver in the driving process are realized, and the detected abnormality is uploaded to a traffic administration gate for supervision. The invention can make different reactions according to the detection result in the starting process and the driving process, thereby realizing a double detection device before and during driving and greatly reducing the probability of traffic accidents.
2. The invention can acquire the facial features of the driver and compare and check the facial features with the archived facial features of the driver by the camera arranged in front of the driver seat, meanwhile, the acquired facial features of the driver are stored in the storage unit and uploaded to the cloud to be called by a manager, and the information of the driver is compared in real time in the driving process, so that the driver is prevented from being changed midway.
3. The invention adopts a plurality of sensors to detect the alcohol content in the driver and adopts the camera to identify the facial features such as hyperemia, flushing or tragic white, stiff expression, tiredness, dozing and the like of the eyes of the driver to comprehensively judge the alcohol detection, thereby improving the detection precision.
Drawings
Fig. 1 is a block diagram of a vehicle-mounted drunk driving comprehensive detection system based on multi-source information fusion according to an embodiment of the present invention.
FIG. 2 is a schematic view of the structure of the present invention.
Fig. 3 is a schematic diagram of the position structure of the invention installed in the automobile.
FIG. 4 is a flowchart illustrating an exemplary vehicle start-up process.
FIG. 5 is a flowchart illustrating a driving process of a vehicle according to an embodiment.
FIG. 6 is a schematic diagram of data flow according to an embodiment.
Fig. 7 is a block diagram of a communication module according to an embodiment of the present invention.
Wherein, each icon and the corresponding part name are as follows:
the system comprises a main control module, a 3-A/D (analog-to-digital) converter, a 4-gas alcohol sensor, a 5-infrared sensor, a 6-sweat alcohol sensor, a 7-face recognition and alcohol detection module, a 71-camera, an 8-communication module, a 9-positioning module, a 10-display screen, a 12-display module, a 14-alcohol collection module and a 15-audible and visual alarm.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Example (b):
as shown in fig. 1-5, the vehicle-mounted drunk driving comprehensive detection system for unit vehicles based on multi-source information fusion of the embodiment mainly comprises a main control module 1, a display module 12, a communication module 8, a positioning module 9, an audible and visual alarm 15, a face recognition and alcohol detection module 7 and an alcohol collection module 14, which are installed inside a vehicle; the main control module 1 is provided with a first processor; the positioning module 9 is provided with a GPS positioning system; the audible and visual alarm 15 is provided with a buzzer and an alarm lamp; the display module 12 is provided with a display screen 10; the face recognition and alcohol detection module 7 comprises a camera 71, a storage unit and a second processor; the camera 71 and the storage unit are respectively electrically connected with the second processor; the communication module 8 comprises an information transmission unit, a voice prompt unit and a telephone prompt unit; the information transmission unit, the voice prompt unit, the telephone prompt unit, the second processor, the GPS, the buzzer, the alarm lamp and the display screen 10 are respectively and electrically connected with the first processor; the first processor is connected with an ignition control module of the automobile through a relay; the first processor is also electrically connected with a lamp control module of the automobile; the alcohol collection module 14 comprises a storage battery, a sweat alcohol sensor 6, a gas alcohol sensor 4 and an infrared sensor 5; the storage battery, the sweat alcohol sensor 6, the gas alcohol sensor 4 and the infrared sensor 5 are respectively electrically connected with the first processor.
The sweat alcohol sensor 6 is arranged on the left side of the automobile steering wheel; the infrared sensor 5 is arranged on the right side of the automobile steering wheel; the gas alcohol sensor 4 is arranged on the left side of the driver seat of the automobile.
The camera 71 is arranged at one side in front of a copilot of the automobile, and the camera 71 faces to the driver seat of the automobile and monitors the head state of the driver.
The automobile steering wheel is further provided with an electric heating piece and a chemical drug pilocarpine, and the electric heating piece is electrically connected with the first processor.
The first processor and the second processor are both of AT89C51 single-chip microcomputer.
The positioning module 9 is arranged at the position of the A column of the automobile.
The vehicle-mounted drunk driving comprehensive detection system based on multi-source information fusion comprises an automobile starting and alcohol detection in the driving process, is suitable for unit vehicles such as a business car, a travel package car and a transport fleet, and comprises the following steps:
s1: the driver starts the automobile engine through a key, the camera 71 is opened at the moment, the facial image of the driver is collected, the second processor adopts the Facenet network to verify the identity of the driver, and meanwhile, the obtained facial features of the driver are uploaded to the cloud to be called by a manager;
s2: after the identity authentication is successful, the sweat alcohol sensor 6, the gas alcohol sensor 4 and the infrared sensor 5 are electrified to start alcohol detection of the driver, and signals are transmitted to the main control module 1 for processing; meanwhile, the camera 71 acquires a facial image of the driver and transmits the facial image to the second processor, the second processor adopts a Nin network to identify facial features such as congestion of eyes, flushing or tragic white face, stiff expression, fatigue, drowsiness and the like, whether the driver drives drunk or not is evaluated, and the result is transmitted to the main control module 1;
s3: the main control module 1 performs data fusion by adopting a prediction function obtained by deep learning training, the sweat alcohol sensor 6, the gas alcohol sensor 4, the infrared sensor 5 and an image recognition result jointly determine whether a driver drives drunk, the data content and the image recognition result detected by the three sensors are input into the prediction function to obtain a prediction real value, and then the prediction real value is converted into 0 or 1 through a Sigmoid function, wherein 0 represents drunk driving behavior, and 1 represents drunk driving behavior, so that whether drunk driving behavior occurs is determined;
s4: in the starting process, if the automobile is detected to be qualified, the automobile normally runs, if the automobile is detected to be drunk driven, the engine is shut down, and meanwhile, a voice prompt unit in the communication module 8 inquires whether a driver needs to make a call or search for a designated driver, wherein the call comprises a pre-stored parent call, a pre-stored designated driver call and the like; in the driving process of the automobile, the camera 71 compares facial features of the driver in real time to ensure that the information of the driver can be correctly matched, if the information of the driver cannot be successfully matched, the sound-light alarm module 15 is started, the automobile starts alarming, double-flash is started to prompt parking, and meanwhile, the license plate, the positioning, the alcohol concentration and the personal information of the driver are uploaded to a company superior management department; in addition, drunk driving detection is carried out in real time in the driving process, if the detection result is drunk driving, the main control module 1 receives the position information of the positioning module 9, license plates, positioning, alcohol concentration and driver personal information are uploaded to a company upper management department and a traffic police head team command center through the communication module 8, and meanwhile, the car gives an alarm and opens double-flash to prompt parking.
In this embodiment, an MTCNN + Facenet network is used to perform face recognition, most sample data are used to train a model, a small part of the sample data are used to verify the model, MTCNN is used to complete the tasks of face detection and face alignment, and Facenet is used to perform recognition.
The MTCNN is adopted to complete the tasks of face detection and face alignment, and the specific steps are as follows: firstly, zooming a picture of a driver by a zooming coefficient to generate an image pyramid, inputting P-Net, then generating a human face boundary box by a full convolution network, and screening by using non-maximum suppression (NMS), wherein the input image of the P-Net is 12x12x 3; scaling the image to 24x24x3, inputting R-Net, and performing secondary screening by using NMS to remove a large number of non-face frames; and (5) scaling the image output by the R-Net to 48 multiplied by 3, inputting the O-Net, and performing secondary screening by using NMS to obtain a face candidate frame.
Facenet is adopted for recognition, and is a classic face recognition. The embodiment specifically comprises the following steps: the embedding database is generated by first utilizing facenet to calculate embedding in the driver database and is stored in the storage unit. During recognition, inputting the face region image processed by MTCNN into a faceNet network, extracting features through a trained deep learning network, calculating the feature vector of Embedding after normalization and L2 norm, and mapping the image x to a d-dimensional Euclidean space through a function f; and comparing Euclidean distances among the characteristic vectors to judge whether the driver is effective or not.
According to the drunk driving identification method, drunk driving identification is carried out through a Nin network model, firstly, a large number of samples are collected by using a drunk driving detection system, and the drunk driving identification method is divided into drunk driving and non-drunk driving according to whether a driver is drunk or not, wherein most of the drunk driving identification method is used as a training sample, and the small part of the drunk driving identification method is used as a testing sample. And acquiring face region images through a camera to obtain training samples, and training the Nin network model by using the face region images. The purpose of image recognition is whether the driver drives with wine or not, and only 2 types of drunk driving and drunk-free driving are needed. The trained Nin network model can identify whether the driver is drunk driving in the image.
Specifically, a Nin network model architecture is constructed: the purpose of the Nin network in the invention is to identify whether the driver is drunk, and the images are classified into drunk images and non-drunk images according to the drunk images. The method is characterized by comprising 3 Mlpconv layers (Mlpconv1, Mlpconv2 and Mlpconv3), wherein each Mlpconv layer is followed by a pooling layer (Pool1, Pool2 and Pool3), except the last Mlpconv layer, each Mlpconv layer is followed by a dropout layer to prevent overfitting, and finally a global mean pooling layer (GAP) is followed to generate a feature map for two corresponding classes of drunk and drunk-free respectively, and then the average value of each feature map is calculated and directly input into a softmax layer of 2 labels for classification, wherein drunk and drunk-free are obtained.
The formula for the Mlpconv layer is as follows:
Figure BDA0003111179450000091
in the formula: (i, j) denotes a position index, x, of a pixel point of the imagei,jRepresenting an image block with position index (i, j) in the convolution window, k representing the index of the feature map channel, n representing the number of layers of the multi-layer perceptron, k representing the number of layers of the multi-layer perceptronbRepresenting the bias of the characteristic diagram, the activation function of the multilayer perceptron adopts a Relu function.
And (3) training a Nin network model: preprocessing the input face candidate image, inputting the face candidate image into the Nin network, and training by adopting a Caffe (conditional Architecture for Fast Feature embedding) deep learning framework. The number of training iterations is 10000, and the training learning rate is 0.01.
The prediction function is obtained by a deep learning training method and comprises the following steps: arranging relevant drivers to perform experiments, measuring the alcohol content in the body of the driver by using a blood alcohol content testing method, then arranging the drivers to get on the vehicle to measure three groups of sensor data, finding out the relation between the three groups of sensor data, picture recognition drunk driving results and data obtained by using the blood alcohol content testing by using a deep learning training method to obtain a prediction function, wherein the measured values of the three sensors are respectively marked as x1,x2,x3And the image recognition result is marked as x4；
The specific algorithm is as follows: a set of data x obtained from the experiment1,x2,x3,x4Is marked as
Figure BDA0003111179450000092
The data measured by the blood alcohol content test method is recorded as yi(ii) a Multiple groups of data are obtained through multiple experiments and are recorded as a sample set
Figure BDA0003111179450000093
Figure BDA0003111179450000094
Obtaining a prediction function by linear regression
Figure BDA0003111179450000095
So that
Figure BDA0003111179450000096
Get the optimal solution
Figure BDA0003111179450000097
Predicting the value of linear regression
Figure BDA0003111179450000098
Is recorded as z, then
Figure BDA0003111179450000099
And z is the predicted real value, now to convert z to a y value close to 0 or 1, a Sigmoid function is introduced
Figure BDA00031111794500000910
The conversion of the measured values is completed, after which a conversion value of 0 or 1, i.e. y, is obtainediE (0,1), wherein 0 represents the drunk driving behavior, and 1 represents the drunk driving behavior, thereby judging whether the drunk driving behavior occurs.
The identity verification model and the image drunk driving recognition model are subjected to recognition training in a cloud, and the recognition training comprises the construction and training of Facenet and Nin network and deep learning.
The alcohol detection module 14 is used for detecting the alcohol concentration of the driver and sending an alcohol concentration signal to the signal processing module 2. The sweat alcohol sensor 6 is used for detecting whether sweat secreted on the hands of a driver contains alcohol components, because the alcohol concentration in the sweat has a certain relation with the alcohol concentration in blood, the alcohol concentration in the blood of the driver can be known by monitoring the alcohol concentration in the sweat, and the purpose of monitoring whether the driver is drunk or not is achieved; the sweat alcohol sensor 6 is arranged on the steering wheel, particularly on the left side of the steering wheel; the data of the alcohol concentration in sweat on the hand-held steering wheel of the driver is detected, converted into an electric signal and transmitted to the information processing module 2 for processing, and the data and other two sensors jointly determine the alcohol content in the driver; meanwhile, the measuring result is prevented from being influenced by insufficient sweat amount when the driver detects the sweat amount. An electric heating piece and a chemical drug Pilocarpine (Pilocarpine) are arranged on an automobile steering wheel, so that sweat can be promoted to be discharged, alcohol can be volatilized, and the detection precision is enhanced;
the gas alcohol sensor 4 is a device for detecting alcohol gas around a driver seat based on the principle that the detected alcohol concentration in the gas is converted into a useful electric signal, and information related to the existence of the gas to be detected in the environment can be obtained according to the strength of the electric signals, when a driver sits on a seat and starts an ignition device, the gas alcohol sensor 4 starts to work, but not limited to an alcohol sensitive sensor, is used for checking the alcohol concentration value P around the driver in a vehicle, when a certain threshold value is reached, the information is transmitted to the information processing module 2 and then is transmitted to the center of the main control module 1 for judgment, the gas alcohol sensor 4 is installed on the left side of the driver seat, and the sensor, other sensors and cameras comprehensively judge whether the driver is drunk;
the infrared sensor 5 is designed based on an infrared detection circuit system, is arranged on the right side of a steering wheel, takes the fingers of a driver as detection objects, when the driver holds the steering wheel, the alcohol detection module 14 starts to circularly detect alcohol in the driver body within set time, an analog signal output by the infrared sensor 5 is converted into a digital signal through the A/D analog-to-digital converter 3, the digital signal is input into an AT89C51 single chip of the information processing module 2 for data analysis and processing, the alcohol content in the driver body is further judged, a signal output by the infrared sensor is transmitted to the information processing module 2, the information processing module 2 judges and processes the signal and transmits the signal to the center of the main control module 1, and the infrared sensor 5, other two sensors and the camera comprehensively judges whether the driver is drunk.
The face recognition and alcohol detection module comprises a camera, a storage unit and a processor; the camera is used for collecting head image information of a driver, the storage unit stores pre-stored registered face data information of the driver, the processor preprocesses the picture, and adopts a trained Facenet network to carry out face recognition and verify the identity of the driver; identifying facial features such as congestion of eyes, flushing or tragic white face, stiff expression, tiredness, drowsiness and the like by adopting a convolutional neural network (Nin network), evaluating whether a driver drives drunk or not, and transmitting a judgment result to a main control module; because the expressions such as distortion, confusion and the like can appear on the facial expression of the alcoholics, and the expressions such as lethargy and the like can appear at the same time, whether the driver drunk or not can be effectively evaluated through the facial expression.
The main control module adopts a circuit which takes a single chip microcomputer as main control, integrates the numerical values of concentration detection of the three alcohol sensors and the result of face image recognition through deep learning to obtain the result of alcohol detection and take corresponding measures, and is used for controlling and coordinating the work of other modules. The alcohol collection module can be controlled to detect the alcohol in the blood of the driver at regular time. When the automobile is started, after the identity authentication is successful, alcohol detection is carried out; if the detection result is drunk driving, the license plate, the position information, the alcohol concentration and the personal information of the driver are uploaded to a superior management department of a company in a network mode, meanwhile, a buzzer gives an alarm for prompting, a voice prompt unit in a communication module is started, the driver is inquired whether to make a call or search for drunk driving, convenience is provided for the drunk driver, and the call comprises a pre-stored parent call, a pre-stored drunk driving call and the like; in the driving process, the camera compares facial features of a driver in real time, simultaneously, the face recognition module, the alcohol detection module and the alcohol acquisition module perform alcohol detection, if the detection result is judged to be drunk driving, the positioning module and the communication module are started, license plates, position information, alcohol concentration and personal information of the driver are uploaded to a higher management department of a company and a traffic police head team command center in a network mode, meanwhile, the singlechip controls the acousto-optic alarm module to be started with the display module, the display screen can display warning information, the buzzer sends out harsh buzzing to prompt the driver to pay attention to own state, and the vehicle is stopped when being prompted by voice.
The communication module 8 uploads a license plate, a positioning device, alcohol concentration and driver personal information to a traffic police headquarters command center in a mobile network form when a driver is found drunk during the driving of the automobile, meanwhile, a voice prompt unit plays a result in a voice mode when the driver is detected drunk, in addition, the driver is detected to be drunk when the automobile is started, and the telephone prompt unit provides telephone service for the driver, such as inquiring whether the driver needs to make a call or search for a substitute drive, so that convenience is provided for the drunk driver, and the telephone comprises a pre-stored affiliated telephone, a substitute drive telephone and the like.
The positioning module 9 uses GPS positioning technology to receive positioning signals through a GPS antenna to determine the location of the module including longitude, latitude, time information, and the like, so as to realize the positioning of the vehicle. And when receiving the alarm signal of the main control module 1, the sound and light alarm module 15 controls the alarm lamp and the buzzer to alarm so as to warn a driver. The display module is used for displaying the judgment result of the main control module 1 so that the driver can know the state of the driver.
The foregoing are merely exemplary embodiments of the present invention, which enable those skilled in the art to understand or practice the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (9)

1. The utility model provides a vehicle-mounted wine of unit car drives comprehensive testing system based on multisource information fusion which characterized in that: the system mainly comprises a main control module (1), a display module (12), a communication module (8), a positioning module (9), an audible and visual alarm (15), a face recognition and alcohol detection module (7) and an alcohol acquisition module (14) which are arranged in an automobile; the main control module (1) is provided with a first processor; the positioning module (9) is provided with a GPS positioning system; the audible and visual alarm (15) is provided with a buzzer and an alarm lamp; the display module (12) is provided with a display screen (10); the face recognition and alcohol detection module (7) comprises a camera (71), a storage unit and a second processor; the camera (71) and the storage unit are respectively electrically connected with the second processor; the communication module (8) comprises an information transmission unit, a voice prompt unit and a telephone prompt unit;
the information transmission unit, the voice prompt unit, the telephone prompt unit, the second processor, the GPS, the buzzer, the alarm lamp and the display screen (10) are respectively and electrically connected with the first processor; the first processor is connected with an ignition control module of the automobile through a relay; the first processor is also electrically connected with a lamp control module of the automobile;
the alcohol collection module (14) comprises a storage battery, a sweat alcohol sensor (6), a gas alcohol sensor (4) and an infrared sensor (5); the storage battery, the sweat alcohol sensor (6), the gas alcohol sensor (4) and the infrared sensor (5) are respectively electrically connected with the first processor.
2. The unit vehicle-mounted drunk driving comprehensive detection system based on multi-source information fusion of claim 1, characterized in that: the sweat alcohol sensor (6) is arranged on the left side of the automobile steering wheel; the infrared sensor (5) is arranged on the right side of the automobile steering wheel; the gas alcohol sensor (4) is arranged on the left side of the driver seat of the automobile.
3. The unit vehicle-mounted drunk driving comprehensive detection system based on multi-source information fusion of claim 1, characterized in that: the camera (71) is arranged on one side in front of the copilot of the automobile, and the camera (71) faces to the driver seat of the automobile and monitors the head state of the driver.
4. The unit vehicle-mounted drunk driving comprehensive detection system based on multi-source information fusion of claim 3, characterized in that: the automobile steering wheel is further provided with an electric heating piece and a chemical drug pilocarpine, and the electric heating piece is electrically connected with the first processor.
5. The unit vehicle-mounted drunk driving comprehensive detection system based on multi-source information fusion of claim 1, characterized in that: the first processor and the second processor are both of AT89C51 single-chip microcomputer.
6. The unit vehicle-mounted drunk driving comprehensive detection system based on multi-source information fusion of claim 1, characterized in that: the positioning module (9) is arranged at the position of an A column of the automobile.
7. The use method of the unit vehicle-mounted drunk driving comprehensive detection system based on multi-source information fusion as claimed in any one of claims 1 to 6, characterized in that: the alcohol detection method comprises the steps of automobile starting and running alcohol detection, is suitable for unit vehicles such as official vehicles, tourism charters, transport fleets and the like, and comprises the following steps:
s1: the driver starts the automobile engine through a key, the camera (71) is opened at the moment, the facial image of the driver is collected, the second processor adopts the Facenet network to verify the identity of the driver, and meanwhile the obtained facial features of the driver are uploaded to a cloud end to be managed to call;
s2: after the identity authentication is successful, the sweat alcohol sensor (6), the gas alcohol sensor (4) and the infrared sensor (5) are powered on to start alcohol detection of the driver, and signals are transmitted to the main control module (1) for processing; meanwhile, the camera (71) acquires a facial image of the driver and transmits the facial image to the second processor, the second processor adopts a Nin network to identify facial features such as congestion of eyes, flushing or tragic white face, stiff expression, tiredness, sleepiness and the like, whether the driver drives drunk or not is evaluated, and the result is transmitted to the main control module (1);
s3: the method comprises the steps that a main control module (1) performs data fusion by adopting a prediction function obtained by deep learning training, a sweat alcohol sensor (6), a gas alcohol sensor (4), an infrared sensor (5) and an image recognition result jointly judge whether a driver drives drunk or not, the data content and the image recognition result detected by the three sensors are input into the prediction function to obtain a prediction real value, and then the prediction real value is converted into 0 or 1 through a Sigmoid function, wherein 0 represents drunk driving behavior, and 1 represents drunk driving behavior, so that whether drunk driving behavior occurs or not is judged;
s4: in the starting process, if the automobile is detected to be qualified, the automobile normally runs, if drunk driving is detected, the engine is shut down, and meanwhile, a voice prompt unit in the communication module (8) inquires whether a driver needs to make a call or search for a designated driver, wherein the call comprises a pre-stored parent call and a pre-stored designated driver call; in the driving process of the automobile, the camera (71) compares facial features of a driver in real time to ensure that the information of the driver can be correctly matched, if the information of the driver cannot be successfully matched, the audible and visual alarm (15) is started, the automobile starts alarming, double-flash is started to prompt parking, and meanwhile, the license plate, the positioning, the alcohol concentration and the personal information of the driver are uploaded to a superior management department of a company; in addition, drunk driving detection is carried out in real time in the driving process, if the detection result is drunk driving, the main control module (1) receives the position information of the positioning module (9), the license plate, the positioning, the alcohol concentration and the personal information of a driver are uploaded to a higher management department of a company and a traffic police command center through the communication module (8), and meanwhile, the car gives an alarm and opens double-flash prompt to stop.
8. The use method of the unit vehicle-mounted drunk driving comprehensive detection system based on multi-source information fusion as claimed in claim 7, characterized in that: the prediction function is obtained by a deep learning training method and comprises the following steps: arranging relevant drivers to perform experiments, measuring the alcohol content in the body of the driver by using a blood alcohol content testing method, then arranging the drivers to get on the vehicle to measure three groups of sensor data, finding out the relation between the three groups of sensor data, picture recognition drunk driving results and data obtained by using the blood alcohol content testing by using a deep learning training method to obtain a prediction function, wherein the measured values of the three sensors are respectively marked as x1,x2,x3And the image recognition result is marked as x4；
The specific algorithm is as follows: a set of data x obtained from the experiment1,x2,x3,x4Is marked as
Figure FDA0003111179440000021
The data measured by the blood alcohol content test method is recorded as yi(ii) a Multiple groups of data are obtained through multiple experiments and are recorded as a sample set
Figure FDA0003111179440000022
Figure FDA0003111179440000023
Obtaining a prediction function by linear regression
Figure FDA0003111179440000024
So that
Figure FDA0003111179440000025
Get the optimal solution
Figure FDA0003111179440000026
b, predicting the linear regression
Figure FDA0003111179440000027
Is recorded as z, then
Figure FDA0003111179440000028
And z is the predicted real value, now to convert z to a y value close to 0 or 1, a Sigmoid function is introduced
Figure FDA0003111179440000031
The conversion of the measured values is completed, after which a conversion value of 0 or 1, i.e. y, is obtainediE (0,1), wherein 0 represents the drunk driving behavior, and 1 represents the drunk driving behavior, thereby judging whether the drunk driving behavior occurs.
9. The use method of the unit vehicle-mounted drunk driving comprehensive detection system based on multi-source information fusion as claimed in claim 7, characterized in that: the identity verification model and the image drunk driving recognition model are recognized and trained in a cloud, the recognition training comprises the construction and training of Facenet and Nin network and deep learning, and the steps are as follows: the method comprises the steps of finishing tasks of face detection and face alignment by adopting MTCNN (convolutional neural network), adopting Facenet to carry out face recognition, mapping a face to a feature vector of an Euclidean space by CNN (convolutional neural network), calculating distances of face features of different pictures, and training a network by using a priori knowledge that the distance of the same individual face is always smaller than that of the different individual faces; drunk driving identification is carried out through a Nin network model, firstly, a large number of samples are collected by a drunk driving detection system, and the drunk driving detection system is divided into drunk driving and non-drunk driving according to whether a driver is drunk or not, wherein most of the drunk driving detection system is used as a training sample, and the rest of the drunk driving detection system is used as a testing sample; acquiring face region images through a camera to obtain training samples, and training a Nin network model by using the face region images; the purpose of image recognition is whether the driver drives drunk or not, and only 2 types of drunk driving and drunk-free driving are required; the trained Nin network model can identify whether the driver is drunk driving in the image.
CN202110649485.7A 2021-06-10 2021-06-10 Vehicle-mounted drunk driving comprehensive detection system for unit vehicle based on multi-source information fusion Active CN113370786B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110649485.7A CN113370786B (en) 2021-06-10 2021-06-10 Vehicle-mounted drunk driving comprehensive detection system for unit vehicle based on multi-source information fusion

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110649485.7A CN113370786B (en) 2021-06-10 2021-06-10 Vehicle-mounted drunk driving comprehensive detection system for unit vehicle based on multi-source information fusion

Publications (2)

Publication Number Publication Date
CN113370786A true CN113370786A (en) 2021-09-10
CN113370786B CN113370786B (en) 2023-04-28

Family

ID=77573718

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110649485.7A Active CN113370786B (en) 2021-06-10 2021-06-10 Vehicle-mounted drunk driving comprehensive detection system for unit vehicle based on multi-source information fusion

Country Status (1)

Country Link
CN (1) CN113370786B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114216941A (en) * 2021-12-13 2022-03-22 合肥艾鑫智能科技有限公司 Drunken driving prevention alarm system based on composite array sensor
CN116434029A (en) * 2023-06-15 2023-07-14 西南石油大学 Drinking detection method
GB2618313A (en) * 2022-04-22 2023-11-08 Continental Automotive Tech Gmbh A method and system for detecting a state of abnormality within a cabin

Citations (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH10181380A (en) * 1996-12-24 1998-07-07 Toyota Motor Corp Alarm system for vehicle
US20060202842A1 (en) * 2005-02-11 2006-09-14 Stephanie Sofer Car alcohol monitoring system
CN101470951A (en) * 2008-01-08 2009-07-01 徐建荣 Vehicle security drive monitoring system
JP2009202655A (en) * 2008-02-26 2009-09-10 Nissan Motor Co Ltd Drunk driving preventive device and drunk driving preventive method
US20110091079A1 (en) * 2009-10-21 2011-04-21 Automotive Research & Testing Center Facial image recognition system for a driver of a vehicle
WO2011120353A1 (en) * 2010-04-02 2011-10-06 深圳市赛格导航科技股份有限公司 On-vehicle navigation equipment, on-vehicle drunken driving detecting device and method, safe driving service system and method for vehicle
US8224608B1 (en) * 2011-10-24 2012-07-17 AK Global Tech Corp. Calibrating breathalyzer
CN102991353A (en) * 2012-12-06 2013-03-27 安科智慧城市技术（中国）有限公司 Vehicular alcohol detection system, method and automobile
CN203020082U (en) * 2012-08-31 2013-06-26 浙江吉利汽车研究院有限公司杭州分公司 Vehicle-mounted alcohol monitoring device
CN103434400A (en) * 2013-08-09 2013-12-11 浙江吉利汽车研究院有限公司 Drunk driving and drowsiness prevention system and drunk driving and drowsiness prevention method
US20140365142A1 (en) * 2013-06-11 2014-12-11 Antimatter Research, Inc. Wearable Blood Alcohol Measuring Device
CN104401229A (en) * 2014-12-26 2015-03-11 南京信息工程大学 Vehicle-mounted drunk-driving-preventing on-line measurement and control system based on sensor array
CN104434066A (en) * 2014-12-05 2015-03-25 上海电机学院 Physiologic signal monitoring system and method of driver
CN104691332A (en) * 2015-02-16 2015-06-10 桂林电子科技大学 Anti-drunk driving system with alcohol inspecting, monitoring and positioning functions and control method of system
CN104742802A (en) * 2014-12-19 2015-07-01 北京联合大学 Drunken driving detection system and method based on multiple sensors and video recognition technology
CN105667312A (en) * 2016-04-15 2016-06-15 吉林大学 Intelligent drunk driving prevention and control system based on vehicle
CN205344773U (en) * 2016-01-04 2016-06-29 安徽理工大学 Novel prevent wine door of driving
CN108382299A (en) * 2017-12-01 2018-08-10 上海电机学院 A kind of vehicle-mounted drunk-driving prevention system based on infrared ray Yu sweat alcohol detection
CN109334452A (en) * 2018-12-03 2019-02-15 华东交通大学 A kind of vehicle-mounted drunk driving test macro of Multi-sensor Fusion
CN110070078A (en) * 2019-05-27 2019-07-30 浙江科技学院 A kind of drunk driving detection method and system based on sensor and machine vision
US20190248237A1 (en) * 2018-02-09 2019-08-15 Sadeq Albakri Blood alcohol level sensing system for a vehicle
CN112406881A (en) * 2019-08-21 2021-02-26 美光科技公司 Fatigue driving detection for vehicle control
US20210113153A1 (en) * 2018-03-22 2021-04-22 KHN Solutions, Inc. Method and system for transdermal alcohol monitoring

Patent Citations (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH10181380A (en) * 1996-12-24 1998-07-07 Toyota Motor Corp Alarm system for vehicle
US20060202842A1 (en) * 2005-02-11 2006-09-14 Stephanie Sofer Car alcohol monitoring system
CN101470951A (en) * 2008-01-08 2009-07-01 徐建荣 Vehicle security drive monitoring system
JP2009202655A (en) * 2008-02-26 2009-09-10 Nissan Motor Co Ltd Drunk driving preventive device and drunk driving preventive method
US20110091079A1 (en) * 2009-10-21 2011-04-21 Automotive Research & Testing Center Facial image recognition system for a driver of a vehicle
WO2011120353A1 (en) * 2010-04-02 2011-10-06 深圳市赛格导航科技股份有限公司 On-vehicle navigation equipment, on-vehicle drunken driving detecting device and method, safe driving service system and method for vehicle
US8224608B1 (en) * 2011-10-24 2012-07-17 AK Global Tech Corp. Calibrating breathalyzer
CN203020082U (en) * 2012-08-31 2013-06-26 浙江吉利汽车研究院有限公司杭州分公司 Vehicle-mounted alcohol monitoring device
CN102991353A (en) * 2012-12-06 2013-03-27 安科智慧城市技术（中国）有限公司 Vehicular alcohol detection system, method and automobile
US20140365142A1 (en) * 2013-06-11 2014-12-11 Antimatter Research, Inc. Wearable Blood Alcohol Measuring Device
CN103434400A (en) * 2013-08-09 2013-12-11 浙江吉利汽车研究院有限公司 Drunk driving and drowsiness prevention system and drunk driving and drowsiness prevention method
CN104434066A (en) * 2014-12-05 2015-03-25 上海电机学院 Physiologic signal monitoring system and method of driver
CN104742802A (en) * 2014-12-19 2015-07-01 北京联合大学 Drunken driving detection system and method based on multiple sensors and video recognition technology
CN104401229A (en) * 2014-12-26 2015-03-11 南京信息工程大学 Vehicle-mounted drunk-driving-preventing on-line measurement and control system based on sensor array
CN104691332A (en) * 2015-02-16 2015-06-10 桂林电子科技大学 Anti-drunk driving system with alcohol inspecting, monitoring and positioning functions and control method of system
CN205344773U (en) * 2016-01-04 2016-06-29 安徽理工大学 Novel prevent wine door of driving
CN105667312A (en) * 2016-04-15 2016-06-15 吉林大学 Intelligent drunk driving prevention and control system based on vehicle
CN108382299A (en) * 2017-12-01 2018-08-10 上海电机学院 A kind of vehicle-mounted drunk-driving prevention system based on infrared ray Yu sweat alcohol detection
US20190248237A1 (en) * 2018-02-09 2019-08-15 Sadeq Albakri Blood alcohol level sensing system for a vehicle
US20210113153A1 (en) * 2018-03-22 2021-04-22 KHN Solutions, Inc. Method and system for transdermal alcohol monitoring
CN109334452A (en) * 2018-12-03 2019-02-15 华东交通大学 A kind of vehicle-mounted drunk driving test macro of Multi-sensor Fusion
CN110070078A (en) * 2019-05-27 2019-07-30 浙江科技学院 A kind of drunk driving detection method and system based on sensor and machine vision
CN112406881A (en) * 2019-08-21 2021-02-26 美光科技公司 Fatigue driving detection for vehicle control

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
F.J. ACEVEDO: "Probabilistic support vector machines for multi-class alcohol identification", 《ELSEVIER》 *
刘艳红: "基于多传感器信息融合的汽车酒驾测控系统研究", 《中国优秀硕士学位论文全文数据库 (信息科技辑)》 *
孙同辉;邵开丽;: "基于车联网的健康监测系统的设计" *
毛??,初秀民,严新平,吴超仲: "汽车驾驶员驾驶疲劳监测技术研究进展" *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114216941A (en) * 2021-12-13 2022-03-22 合肥艾鑫智能科技有限公司 Drunken driving prevention alarm system based on composite array sensor
CN114216941B (en) * 2021-12-13 2023-12-29 合肥艾鑫智能科技有限公司 Drunk driving prevention alarm system based on composite array sensor
GB2618313A (en) * 2022-04-22 2023-11-08 Continental Automotive Tech Gmbh A method and system for detecting a state of abnormality within a cabin
CN116434029A (en) * 2023-06-15 2023-07-14 西南石油大学 Drinking detection method
CN116434029B (en) * 2023-06-15 2023-08-18 西南石油大学 Drinking detection method

Also Published As

Publication number Publication date
CN113370786B (en) 2023-04-28

Similar Documents

Publication Publication Date Title
CN113370786A (en) Vehicle-mounted drunk driving comprehensive detection system for unit vehicle based on multi-source information fusion
Chen et al. D 3: Abnormal driving behaviors detection and identification using smartphone sensors
CN108791299B (en) Driving fatigue detection and early warning system and method based on vision
US11465634B1 (en) Automobile detection system
WO2018058958A1 (en) Road vehicle traffic alarm system and method therefor
US20200001892A1 (en) Passenger assisting apparatus, method, and program
CN103824420B (en) Fatigue driving identification system based on heart rate variability non-contact measurement
CN107813828A (en) Passenger verification system and method
CN107817714A (en) Passenger's monitoring system and method
CN107813829A (en) Passenger's tracing system and method
CN107161104B (en) A kind of vehicle drive assist system
US11858419B2 (en) Fatigue driving monitoring, reminding and early-warning method and system based on computer vision
CN107481521B (en) The traffic events intelligence means of proof and system based on car networking
CN111353369B (en) Application method and system of urban road side parking high-level video in auxiliary criminal investigation
CN108268849A (en) Public transport vehicle-mounted Visual intelligent sensory perceptual system based on AI technologies
CN107933471A (en) The method and vehicle-mounted automatic help system of accident active call rescue
CN103700220A (en) Fatigue driving monitoring device
CN111008566A (en) Deep learning-based school bus student getting-off omission detection device and method
CN112532928B (en) Bus-mounted system based on 5G and face recognition and use method
CN106004439A (en) Remote control intelligent on-board host computer for bus
KR20160028542A (en) an emergency management and crime prevention system for cars and the method thereof
KR101437406B1 (en) an emergency management and crime prevention system for cars and the method thereof
CN112519787A (en) Vehicle control system and vehicle
CN109770922A (en) Embedded fatigue detecting system and method
CN115909651A (en) Method, device, equipment and storage medium for protecting personal safety in vehicle

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