Intelligent vehicle-mounted drunk driving alarm method and system

Through entropy weight method and facial recognition technology combined with multiple data sources, the problem of high misjudgment rate and cumbersome detection process in the existing technology is solved, and more efficient and accurate drunk driving testing is achieved, ensuring the reliability of the detection results. rational use of sex and resources.

CN118887784APending Publication Date: 2024-11-01HUAIYIN INSTITUTE OF TECHNOLOGY
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202410713982.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-04
Publication Date
2024-11-01

AI Technical Summary

Technical Problem

The existing drunk driving detection technology has problems such as high misjudgment rate, cumbersome testing process, resource occupancy and traffic congestion, especially the problems of fewer deep learning detection parameters and the inability to judge the driver's thinking logic.

Method used

The entropy weight method is used to combine multiple data sources (breathing test, heart rate test, logical score) for drunk driving evaluation, and the driver's identity is judged through facial recognition technology, which improves detection accuracy and efficiency, and performs preset alarm operations.

Benefits of technology

It improves the accuracy and efficiency of drunk driving testing, avoids misjudgment and replaces driver testing, and ensures the reliability of test results and the rational use of resources.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118887784A_ABST
    Figure CN118887784A_ABST
Patent Text Reader

Abstract

The invention discloses an intelligent vehicle-mounted drunk driving alarm method. The method comprises the following steps: acquiring a plurality of alcohol detection score data of a to-be-detected driver; wherein the alcohol detection score data comprises a breath test score, a heart rate test score and a logic score; according to the alcohol detection score data, calculating a drunk driving evaluation score of the to-be-detected driver through an entropy weight method, and determining a driving state of the to-be-detected driver according to the drunk driving evaluation score; when it is determined that the driving state is drunk driving, preset drunk driving alarm operation is executed. According to the method, a logic scoring unit for the concentration degree and conversation of the human brain is introduced, the drunk driving judgment accuracy is improved, in addition, the method for weighting data through the entropy weight method is an open mode, more data can be fused to improve the judgment accuracy, and the drunk driving detection accuracy and efficiency can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of drunk driving alarm, and in particular to an intelligent vehicle-mounted drunk driving alarm method and system. Background Art

[0002] Drunk driving has always been an important issue in road traffic safety in my country and even in the world.

[0003] At present, my country's basic approach to drunk driving is to set up checkpoints on the road by the traffic management department. It is a sampling test method to detect whether the driver has been drinking to investigate and punish the driver's drunk driving behavior. There are two main detection methods: alcohol sensor and blood sampling test. The former is more direct and quick, and generally the traffic management department conducts preliminary tests on the driver on the road, but there are certain false positives; the latter method measures the alcohol content more accurately, but the process is too cumbersome and time-consuming. In the case of positive alcohol sensor detection, blood sampling tests are generally required as a standard for punishment, which requires a large number of police forces and occupies tight hospital resources. In addition, the above methods often cause traffic jams. Therefore, vehicle-mounted drunk driving detection technology has been proposed as a solution to the above-mentioned problem. For example, the invention patent with publication number CN113370786A is based on a comprehensive on-vehicle drunk driving detection system for unit vehicles based on multi-source information fusion. Through the alcohol collection module and the face recognition and alcohol detection module, a deep learning training method is used to integrate multiple sensors and cameras to identify the driver's facial features to determine whether the driver has been drinking. This can improve the accuracy of alcohol concentration judgment. However, since the deep learning detection parameters are relatively small and there is no judgment on whether the driver has a clear and reasonable thinking logic, the system is more likely to make misjudgments. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides an intelligent vehicle-mounted drunk driving alarm method and system, which can improve the accuracy and efficiency of drunk driving detection.

[0005] An embodiment of the present invention provides an intelligent vehicle-mounted drunk driving alarm method, comprising the following steps:

[0006] Collecting several alcohol test score data of the driver to be tested; wherein the alcohol test score data includes a breathing test score, a heart rate test score and a logic score;

[0007] According to the alcohol test score data, the drunk driving assessment score of the driver to be tested is calculated by entropy weight method, and the driving state of the driver to be tested is determined according to the drunk driving assessment score;

[0008] When it is determined that the driving status is drunk driving, a preset drunk driving alarm operation is executed.

[0009] Furthermore, the collection of several alcohol test data of the driver to be tested specifically includes:

[0010] The breath alcohol concentration of the driver to be tested is detected by the alcohol concentration sensor, and the breath test score of the driver to be tested is determined based on the comparison between the breath alcohol concentration and a preset standard range;

[0011] Detecting the blood flow rate of the driver to be tested by a pulse heart rate sensor, and obtaining the heart rate data of the driver to be tested by a PPG conversion method according to the blood flow rate, and comparing the heart rate data with a preset daily heart rate value to determine the heart rate test score of the driver to be tested;

[0012] A preset operation test broadcast is outputted through a preset voice output device. When the driver to be tested performs a corresponding operation according to the preset operation test broadcast, a logic score of the driver to be tested is calculated based on the operation completion and accuracy.

[0013] Further, the alcohol test score data is used to calculate the drunk driving assessment score of the driver to be tested by entropy weight method, and the driving state of the driver to be tested is determined according to the drunk driving assessment score, which specifically includes:

[0014] The alcohol test score data are respectively used as calculation indicators, and the weight corresponding to each calculation indicator is calculated by entropy weight method;

[0015] The drunk driving assessment score of the driver to be tested is calculated according to the weight, and when the drunk driving assessment score exceeds a preset drunk driving threshold, it is determined that the driving status of the driver to be tested is drunk driving.

[0016] Preferably, the calculating the weight corresponding to each of the calculation indicators by the entropy weight method specifically includes:

[0017] Standardizing the calculated index to obtain a standard index value;

[0018] Calculate the information entropy corresponding to the calculation indicator according to the standard indicator value;

[0019] The weight corresponding to the calculated indicator is calculated according to the information entropy.

[0020] Furthermore, when the driving state is determined to be drunk driving, a preset drunk driving alarm operation is executed, specifically including:

[0021] When it is determined that the driving status is drunk driving, the driver is reminded of the drunk driving through a preset voice broadcast, and the vehicle location is sent to a preset contact terminal, and a low level is output to the engine relay to put the engine in a power-off state.

[0022] Furthermore, the method further comprises:

[0023] Before the driver is detected, the driver's face is sampled to obtain a first sampled image;

[0024] After the driver completes the detection and starts the car, the driver's face is sampled again to obtain a second sampled image;

[0025] The first sampling image is compared with the second sampling image to obtain a driver comparison result. When the driver comparison result is abnormal, a preset abnormal alarm operation is performed.

[0026] Preferably, the step of sampling the driver's face to obtain a first sampled image specifically includes:

[0027] A plurality of facial images of the driver are obtained by a preset photographing device as characteristic face images respectively;

[0028] According to the plurality of eigenface images, an average face image is calculated by the Eigenface method and is used as the first sampling image.

[0029] Preferably, the first sampling image is compared with the second sampling image to obtain a driver comparison result, and when the driver comparison result is abnormal, a preset abnormality alarm operation is performed, which specifically includes:

[0030] Calculating the Euclidean distance between the first sampling image and the second sampling image, and determining that the driver comparison result is normal when the Euclidean distance is less than a preset distance threshold, and determining that the driver comparison result is abnormal when the Euclidean distance exceeds the preset distance threshold;

[0031] When the comparison result of the driver is abnormal, the driver is reminded of the comparison failure through a preset voice broadcast, and a low level is output to the engine relay to put the engine in a power-off state.

[0032] Another embodiment of the present invention provides an intelligent vehicle-mounted drunk driving alarm system, including: a data acquisition module, a calculation module and an alarm module;

[0033] The data acquisition module is used to collect several alcohol test score data of the driver to be tested; wherein the alcohol test score data includes a breathing test score, a heart rate test score and a logic score;

[0034] The calculation module is used to calculate the drunk driving assessment score of the driver to be tested by entropy weight method according to the alcohol detection score data, and determine the driving state of the driver to be tested according to the drunk driving assessment score;

[0035] The alarm module is used to execute a preset drunk driving alarm operation when the driving status is determined to be drunk driving.

[0036] Furthermore, the system also includes: a face recognition module;

[0037] The face recognition module is used to obtain the driver's facial information before and after detection.

[0038] Compared with the prior art, the beneficial effects of the present invention are:

[0039] The present invention collects multiple data and information from multiple angles, such as human pulse and heart rate, brain concentration, conversation logic, driving status, oral alcohol concentration, etc., and uses entropy weight method to weight the information data, thereby improving the efficiency and accuracy of drunk driving detection. At the same time, face recognition and judgment technology is used to avoid the occurrence of detection instead of the driver.

[0040] Among them, the human brain concentration and conversation logic scoring unit of the present invention can output the preset operation test broadcast and the operation of the driver to be tested through a preset voice output device, and then calculate the logic score of the driver to be tested based on the operation completion and accuracy. The above unit can determine whether the driver has complete and normal thinking logic, further improving the accuracy of drunk driving detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A flowchart of an intelligent vehicle-mounted drunk driving alarm method provided by one embodiment of the present invention.

[0042] Figure 2 A schematic structural diagram of an intelligent vehicle-mounted drunk driving warning system provided in another embodiment of the present invention. DETAILED DESCRIPTION

[0043] The drawings are for illustrative purposes only and should not be construed as limiting the present patent;

[0044] It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0045] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0046] Reference Figure 1 , is a flow chart of an intelligent vehicle-mounted drunk driving alarm method provided by an embodiment of the present invention, comprising the following steps:

[0047] S1: Collecting several alcohol test score data of the driver to be tested; wherein the alcohol test score data includes a breathing test score, a heart rate test score and a logic score;

[0048] S2: Calculate the drunk driving assessment score of the driver to be tested by entropy weight method according to the alcohol test score data, and determine the driving state of the driver to be tested according to the drunk driving assessment score;

[0049] S3: When it is determined that the driving status is drunk driving, a preset drunk driving alarm operation is executed.

[0050] For step S1, specifically, the collection of several alcohol test data of the driver to be tested specifically includes:

[0051] The breath alcohol concentration of the driver to be tested is detected by the alcohol concentration sensor, and the breath test score of the driver to be tested is determined based on the comparison between the breath alcohol concentration and a preset standard range;

[0052] Detecting the blood flow rate of the driver to be tested by a pulse heart rate sensor, and obtaining the heart rate data of the driver to be tested by a PPG conversion method according to the blood flow rate, and comparing the heart rate data with a preset daily heart rate value to determine the heart rate test score of the driver to be tested;

[0053] A preset operation test broadcast is outputted through a preset voice output device. When the driver to be tested performs a corresponding operation according to the preset operation test broadcast, a logic score of the driver to be tested is calculated based on the operation completion and accuracy.

[0054] In a preferred embodiment, after starting the detection, the specific steps of collecting the detection data are as follows:

[0055] The alcohol concentration sensor is activated to detect the driver's breath alcohol concentration, and the alcohol concentration data is obtained through the analog-to-digital converter. The alcohol concentration data is compared with the preset standard interval for judgment. When the alcohol concentration data falls into a certain interval, the corresponding score of the interval is given, and then weighted processing is performed to obtain the breath test score.

[0056] The driver is prompted to put his hands on the steering wheel through the LCD, and the pulse heart rate sensor module is started to measure the blood flow rate of the blood vessels in the driver's fingers. The blood volume data is obtained through the analog-to-digital converter, and then the driver's heart rate pulse value is obtained using the PPG conversion method. The obtained value is compared with the pre-saved daily pulse heart rate value (using the k-means clustering method, if the value is within the class, it is scored as 1. If the value is outside the class, the method of taking the absolute value of the value and the class center value + 1 is used, and the percentage score is given according to the absolute value of the result.) Then weighted processing is performed to obtain the heart rate test score.

[0057] Finally, the voice broadcast module randomly orders three MP3 files for test broadcasting. The driver performs corresponding operations according to the broadcast content. The correct and incorrect operations will be fed back to the control system for scoring. The test questions can be set according to the driving test standards, and different difficulty questions will be scored differently. After the test is completed, the system obtains the total test score and performs weighted processing to obtain the logic score. Among them, the control system can be an STM32 series single-chip microcomputer system.

[0058] For step S2, specifically, the alcohol test score data is used to calculate the drunk driving assessment score of the driver to be tested by the entropy weight method, and the driving state of the driver to be tested is determined according to the drunk driving assessment score, specifically including:

[0059] The alcohol test score data are respectively used as calculation indicators, and the weight corresponding to each calculation indicator is calculated by entropy weight method;

[0060] The drunk driving assessment score of the driver to be tested is calculated according to the weight, and when the drunk driving assessment score exceeds a preset drunk driving threshold, it is determined that the driving status of the driver to be tested is drunk driving.

[0061] Preferably, the calculating the weight corresponding to each of the calculation indicators by the entropy weight method specifically includes:

[0062] Standardizing the calculated index to obtain a standard index value;

[0063] Calculate the information entropy corresponding to the calculation indicator according to the standard indicator value;

[0064] The weight corresponding to the calculated indicator is calculated according to the information entropy.

[0065] In a preferred embodiment, after obtaining the calculated indicators, the calculated indicators must first be standardized, and in the standardization process, the positive indicators and negative indicators are processed differently. The positive and negative values ​​of the indicators are determined by the superior goal. In this preferred embodiment, the breathing test score and the heart rate test score are proportional to the final drunk driving score, and these two indicators are positive indicators. In contrast, the logic score is negatively correlated with the final drunk driving score, so the logic score is a negative indicator.

[0066] Suppose there are n indicators, m items of data are collected, and a certain calculation indicator is r i ′ j (i=1,2,...n;j=1,2,...m), then the corresponding standard index value r ij The calculation formula is as follows: i ′ j When is a positive indicator, When i ′ j When it is a negative indicator,

[0067] After obtaining the standard index value, a set of data r can be calculated according to the following formula j The corresponding information entropy H j :

[0068]

[0069] in, m is the total number of data collected for this indicator.

[0070] Finally, the weight corresponding to the indicator data can be calculated based on the information entropy

[0071] For step S3, specifically, when it is determined that the driving state is drunk driving, a preset drunk driving alarm operation is executed, specifically including:

[0072] When it is determined that the driving status is drunk driving, the driver is reminded of the drunk driving through a preset voice broadcast, and the vehicle location is sent to a preset contact terminal, and a low level is output to the engine relay to put the engine in a power-off state.

[0073] In a preferred embodiment, when it is determined that the driver is driving under the influence of alcohol, the GPS module and the GSM module are started, the GPS performs positioning and hands over the positioning to the control system, sets the variable Location, and the control system writes the Location into the text message content and sends it to the designated mobile phone number through the GSM module. At the same time, the control system inputs a low level to the relay to put the engine in a power-off state, and broadcasts through the voice broadcast module to remind the driver of driving under the influence of alcohol.

[0074] Furthermore, the method further comprises:

[0075] Before the driver is detected, the driver's face is sampled to obtain a first sampled image;

[0076] After the driver completes the detection and starts the car, the driver's face is sampled again to obtain a second sampled image;

[0077] The first sampling image is compared with the second sampling image to obtain a driver comparison result. When the driver comparison result is abnormal, a preset abnormal alarm operation is performed.

[0078] Preferably, the step of sampling the driver's face to obtain a first sampled image specifically includes:

[0079] A plurality of facial images of the driver are obtained by a preset photographing device as characteristic face images respectively;

[0080] According to the plurality of eigenface images, an average face image is calculated by the Eigenface method and is used as the first sampling image.

[0081] Preferably, the first sampling image is compared with the second sampling image to obtain a driver comparison result, and when the driver comparison result is abnormal, a preset abnormality alarm operation is performed, which specifically includes:

[0082] Calculating the Euclidean distance between the first sampling image and the second sampling image, and determining that the driver comparison result is normal when the Euclidean distance is less than a preset distance threshold, and determining that the driver comparison result is abnormal when the Euclidean distance exceeds the preset distance threshold;

[0083] When the comparison result of the driver is abnormal, the driver is reminded of the comparison failure through a preset voice broadcast, and a low level is output to the engine relay to put the engine in a power-off state.

[0084] In a preferred embodiment, before the driver is tested, a preset shooting device is used to first capture a photo of the driver (i.e., the characteristic face image) Γ1, Γ2, ...Γ M , (where M can be set by yourself, and in this preferred embodiment, the number is 2. The more M you get from shooting, the more accurate the face recognition is, but it cannot exceed the memory limit), put these M vectors in a set, and get S = {Γ1,Γ2,...Γ M}.

[0085] Then, we use the traversal method to accumulate the vectors in the set S, using: Formula to calculate the average. Create an M-dimensional vector ψ according to the number of photos we need to compare M. Convert ψ back to the original image format, and the system can obtain the average photo of the driver's previous and subsequent photos, that is, the average face image.

[0086] Finally, after the driver completes the test items, the face image is collected again to obtain the second sampled image Γ, which is represented by the eigenface as For the kth eigenface μ k , its weight can be calculated by the following formula: Where k = 1, 2, ...M.

[0087] M weights can form a vector, namely: Ω T =[ω1,ω2,ω3,……,ω k ]. Let Ω = Finally, use εk =||Ω-Ω k || 2 Perform face recognition. Its essence is to find Ω and Ω k When the distance is less than the threshold, it means that the two people are the same person, and when it exceeds the threshold, it means that they are not the same person. Since the training amount involved in this preferred embodiment is very small, the threshold can be set relatively small.

[0088] Reference Figure 2 , is a schematic diagram of the structure of an intelligent vehicle-mounted drunk driving alarm system provided by another embodiment of the present invention, comprising: a data acquisition module 101, a calculation module 102 and an alarm module 103;

[0089] The data acquisition module 101 is used to collect several alcohol test score data of the driver to be tested; wherein the alcohol test score data includes a breathing test score, a heart rate test score and a logic score;

[0090] The calculation module 102 is used to calculate the drunk driving assessment score of the driver to be tested by entropy weight method according to the alcohol detection score data, and determine the driving state of the driver to be tested according to the drunk driving assessment score;

[0091] The alarm module 103 is used to execute a preset drunk driving alarm operation when the driving status is determined to be drunk driving.

[0092] Furthermore, the system further includes: a face recognition module 104;

[0093] The face recognition module 104 is used to obtain the driver's facial information before and after detection.

[0094] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. An intelligent vehicle-mounted drunk driving alarm method, characterized in that: The steps include: Collecting several alcohol test score data of the driver to be tested; wherein the alcohol test score data includes a breathing test score, a heart rate test score and a logic score; According to the alcohol test score data, the drunk driving assessment score of the driver to be tested is calculated by entropy weight method, and the driving state of the driver to be tested is determined according to the drunk driving assessment score; When it is determined that the driving status is drunk driving, a preset drunk driving alarm operation is executed.

2. The intelligent vehicle-mounted drunk driving alarm method according to claim 1, characterized in that: The collection of several alcohol test data of the driver to be tested specifically includes: The breath alcohol concentration of the driver to be tested is detected by the alcohol concentration sensor, and the breath test score of the driver to be tested is determined based on the comparison between the breath alcohol concentration and a preset standard range; Detecting the blood flow rate of the driver to be tested by a pulse heart rate sensor, and obtaining the heart rate data of the driver to be tested by a PPG conversion method according to the blood flow rate, and comparing the heart rate data with a preset daily heart rate value to determine the heart rate test score of the driver to be tested; A preset operation test broadcast is outputted through a preset voice output device. When the driver to be tested performs a corresponding operation according to the preset operation test broadcast, a logic score of the driver to be tested is calculated based on the operation completion and accuracy.

3. The intelligent vehicle-mounted drunk driving alarm method according to claim 1, characterized in that: The step of calculating the drunk driving assessment score of the driver to be tested by an entropy weight method according to the alcohol test score data, and determining the driving state of the driver to be tested according to the drunk driving assessment score, specifically includes: The alcohol test score data are respectively used as calculation indicators, and the weight corresponding to each calculation indicator is calculated by entropy weight method; The drunk driving assessment score of the driver to be tested is calculated according to the weight, and when the drunk driving assessment score exceeds a preset drunk driving threshold, it is determined that the driving status of the driver to be tested is drunk driving.

4. The intelligent vehicle-mounted drunk driving alarm method according to claim 3, characterized in that: Calculating the weight corresponding to each of the calculation indicators by the entropy weight method specifically includes: Standardizing the calculated index to obtain a standard index value; Calculate the information entropy corresponding to the calculation indicator according to the standard indicator value; The weight corresponding to the calculated indicator is calculated according to the information entropy.

5. The intelligent vehicle-mounted drunk driving alarm method according to claim 1, characterized in that: When the driving state is determined to be drunk driving, a preset drunk driving alarm operation is executed, specifically including: When it is determined that the driving status is drunk driving, the driver is reminded of the drunk driving through a preset voice broadcast, and the vehicle location is sent to a preset contact terminal, and a low level is output to the engine relay to put the engine in a power-off state.

6. The intelligent vehicle-mounted drunk driving alarm method according to claim 1, characterized in that: The method further comprises: Before the driver is detected, the driver's face is sampled to obtain a first sampled image; After the driver completes the detection and starts the car, the driver's face is sampled again to obtain a second sampled image; The first sampling image is compared with the second sampling image to obtain a driver comparison result. When the driver comparison result is abnormal, a preset abnormal alarm operation is performed.

7. The intelligent vehicle-mounted drunk driving alarm method according to claim 6, characterized in that: The step of sampling the driver's face to obtain a first sampled image specifically includes: A plurality of facial images of the driver are obtained by a preset photographing device as characteristic face images respectively; According to the plurality of eigenface images, an average face image is calculated by the Eigenface method and is used as the first sampling image.

8. The intelligent vehicle-mounted drunk driving alarm method according to claim 6, characterized in that: The first sampling image is compared with the second sampling image to obtain a driver comparison result, and when the driver comparison result is abnormal, a preset abnormality alarm operation is performed, specifically including: Calculating the Euclidean distance between the first sampling image and the second sampling image, and determining that the driver comparison result is normal when the Euclidean distance is less than a preset distance threshold, and determining that the driver comparison result is abnormal when the Euclidean distance exceeds the preset distance threshold; When the comparison result of the driver is abnormal, the driver is reminded of the comparison failure through a preset voice broadcast, and a low level is output to the engine relay to put the engine in a power-off state.

9. An intelligent vehicle-mounted drunk driving alarm system, characterized in that: include: Data acquisition module, calculation module and alarm module; The data acquisition module is used to collect several alcohol test score data of the driver to be tested; wherein the alcohol test score data includes a breathing test score, a heart rate test score and a logic score; The calculation module is used to calculate the drunk driving assessment score of the driver to be tested by entropy weight method according to the alcohol detection score data, and determine the driving state of the driver to be tested according to the drunk driving assessment score; The alarm module is used to execute a preset drunk driving alarm operation when the driving status is determined to be drunk driving.

10. The intelligent vehicle-mounted drunk driving warning system as claimed in claim 9, characterized in that: The system also includes: a face recognition module; The face recognition module is used to obtain the driver's facial information before and after detection.

Citation Information

Patent Citations

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

    CN113370786A