Fatigue driving early warning method and system based on multiple fusion fatigue features

Abstract

The application discloses a fatigue driving early warning method and system based on multiple fusion fatigue features. The fatigue driving early warning method is applied to an intelligent automobile and comprises the following steps: model matching is performed on multiple different fatigue features to obtain matching degree results; the fatigue features include eye features, body movement features and physiological parameter features; weighted fusion processing is performed based on the matching degree results to obtain a fatigue degree quantitative value; when the fatigue degree quantitative value is higher than a preset value, a user perception early warning instruction is generated; and user early warning prompting is performed on a vehicle driver and / or a passenger in response to the user perception early warning instruction. The application detects fatigue through different fatigue features, improves the accuracy of fatigue detection, and prompts users to pay attention to safe driving through different touch feelings, thereby improving the safety of vehicle driving.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fatigue driving early warning, in particular to a fatigue driving early warning method and system based on multiple fusion fatigue features. BACKGROUND

[0002] In traffic accidents, the proportion of accidents caused by fatigue driving is relatively high. With the popularization of intelligent driving systems, fatigue driving monitoring and early warning systems are becoming more and more mature. In related technologies, the detection means of fatigue driving is relatively single, the detection accuracy is low, and false judgments of fatigue driving are easy to occur. Moreover, the existing early warning prompt effect is poor. SUMMARY

[0003] The main purpose of the present application is to provide a fatigue driving early warning method based on multiple fusion fatigue features, which aims to detect fatigue through different fatigue features, improve the accuracy of fatigue detection, and prompt users to pay attention to safe driving through different tactile sensations.

[0004] To achieve the above purpose, the present application provides a fatigue driving early warning method based on multiple fusion fatigue features, which is applied to an intelligent vehicle. The method comprises the following steps:

[0005] Model matching is performed on multiple different fatigue features to obtain matching degree results; wherein the fatigue features include eye features, body movement features, and physiological parameter features;

[0006] Based on the matching degree results, weighted fusion processing is performed to obtain fatigue degree quantization values;

[0007] When the fatigue degree quantization value is higher than a preset value, a user perception early warning instruction is generated;

[0008] In response to the user perception early warning instruction, a user early warning prompt is performed on the vehicle driver and / or passenger.

[0009] Preferably, the step of performing model matching on the eye features of the fatigue features to obtain matching degree results comprises:

[0010] An eye feature video stream of the driver is obtained;

[0011] Eye feature image frames within a preset time are extracted and processed to form an eye feature line set;

[0012] Model matching is performed on the eye feature line set to obtain eye feature matching degree results.

[0013] Preferably, the step of performing model matching on the body movement features of the fatigue features to obtain matching degree results comprises:

[0014] The hand holding pressure value and the waist leaning pressure value of the driver are obtained;

[0015] divide the hand holding pressure value and the waist abutting pressure value to form a pressure feature model curve;

[0016] model match the pressure feature model curve to obtain a body movement feature matching degree result.

[0017] Preferably, the step of model matching the physiological parameter feature of the fatigue feature to obtain a matching degree result comprises:

[0018] obtain a pulse jumping state of the driver;

[0019] carry out noise reduction processing on the pulse jumping state to form a pulse jumping model curve;

[0020] model match the pulse jumping model curve to obtain a physiological parameter feature matching degree result.

[0021] Preferably, the step of carrying out weighted fusion processing based on the matching degree result to obtain a fatigue degree quantization value comprises:

[0022] obtain fatigue influence weights corresponding to different fatigue features;

[0023] calculate a fatigue degree quantization value based on the fatigue influence weights and the eye feature matching degree result, the body movement feature matching degree result, and the physiological parameter feature matching degree result.

[0024] Preferably, the intelligent automobile comprises a steering wheel and a passenger seat, and a pre-warning prompting component for user perception is arranged on the steering wheel and the passenger seat;

[0025] The step of determining that the fatigue degree quantization value is higher than a preset value and generating a user-perception pre-warning instruction comprises:

[0026] determining that the fatigue degree quantization value is higher than a preset value and accumulating a continuous duration;

[0027] determining that the accumulated duration is greater than a preset duration, and generating a first instruction for the steering wheel to carry out an electric tactile pre-warning prompt to the user.

[0028] Preferably, the first instruction further comprises:

[0029] re-obtaining a fatigue degree quantization value of the driver;

[0030] determining that the fatigue degree quantization value is higher than a preset value, and generating a second instruction for the passenger seat to carry out a physical tactile pre-warning prompt to the user; and / or,

[0031] re-obtaining a fatigue degree quantization value of the driver;

[0032] determining that the fatigue degree quantification value is higher than a preset value, obtaining a use state of the passenger seat;

[0033] determining that the passenger seat is in a passenger seated state, and generating a second instruction for the passenger seat to give a physical touch warning prompt to a user.

[0034] Preferably, before the step of performing model matching on the plurality of different fatigue features to obtain a matching degree result, the method further comprises:

[0035] obtaining a motion state of a steering wheel of the vehicle;

[0036] determining that the steering wheel of the vehicle is not turned within a preset time, and obtaining a plurality of different fatigue features;

[0037] performing model matching on the plurality of different fatigue features to obtain a matching degree result.

[0038] Preferably, the first instruction includes a high-level electrical touch warning prompt instruction and a low-level electrical touch warning prompt instruction, the second instruction includes a high-level physical touch warning prompt instruction and a low-level electrical touch warning prompt instruction, and the step of determining that the fatigue degree quantification value is higher than a preset value and generating a user perception warning instruction comprises:

[0039] determining that the fatigue degree quantification value is higher than a first preset value and lower than a second preset value, and generating a low-level electrical touch warning prompt instruction and / or a low-level electrical touch warning prompt instruction;

[0040] determining that the fatigue degree quantification value is higher than the second preset value, and generating a high-level electrical touch warning prompt instruction and / or a high-level electrical touch warning prompt instruction.

[0041] The application also provides a fatigue driving warning system, which comprises a memory for storing executable instructions and a processor for executing the executable instructions stored in the memory to implement a fatigue driving warning method based on a plurality of fusion fatigue features, the method comprising the following steps:

[0042] performing model matching on a plurality of different fatigue features to obtain a matching degree result; wherein the fatigue features include eye features, body motion features, and physiological parameter features;

[0043] performing weighted fusion processing based on the matching degree result to obtain a fatigue degree quantification value;

[0044] determining that the fatigue degree quantification value is higher than a preset value, and generating a user perception warning instruction;

[0045] in response to the user perception warning instruction, giving a user warning prompt to a driver and / or a passenger of the vehicle.

[0046] In the technical scheme of the present application, first, a plurality of different fatigue characteristics are matched to obtain a matching degree result, then the matching degree result is weighted and fused to obtain a fatigue degree quantitative value, it is determined that the fatigue degree quantitative value is higher than a preset value, a user perception warning instruction is generated, finally, in response to the user perception warning instruction, a user warning prompt is given to the vehicle driver and / or passenger; in this way, fatigue detection is performed through different fatigue characteristics, the accuracy of fatigue detection is improved, and different tactile prompts are given to users to pay attention to safe driving, thereby improving the safety of vehicle driving. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical scheme in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor based on the structures shown in the drawings.

[0048] Figure 1 Structure diagram of an intelligent automobile in the fatigue driving warning system of the present application;

[0049] Figure 2 Structure diagram of a steering wheel of an intelligent automobile in the fatigue driving warning system of the present application;

[0050] Figure 3 Flowchart of one embodiment of the fatigue driving warning method based on a plurality of fused fatigue characteristics of the present application;

[0051] Figure 4 Flowchart of another embodiment of the fatigue driving warning method based on a plurality of fused fatigue characteristics of the present application;

[0052] Figure 5 Flowchart of still another embodiment of the fatigue driving warning method based on a plurality of fused fatigue characteristics of the present application;

[0053] Figure 6 Flowchart of still another embodiment of the fatigue driving warning method based on a plurality of fused fatigue characteristics of the present application;

[0054] Figure 7 Flowchart of still another embodiment of the fatigue driving warning method based on a plurality of fused fatigue characteristics of the present application;

[0055] Figure 8 Flowchart of still another embodiment of the fatigue driving warning method based on a plurality of fused fatigue characteristics of the present application;

[0056] Figure 9A flowchart of a process of a third embodiment of the fatigue driving early warning method based on the fusion of multiple fatigue features of the present application is shown in FIG. 6.

[0057] Figure 10 A flowchart of a process of a third embodiment of the fatigue driving early warning method based on the fusion of multiple fatigue features of the present application is shown in FIG. 6.

[0058] Figure 11 A flowchart of a process of a third embodiment of the fatigue driving early warning method based on the fusion of multiple fatigue features of the present application is shown in FIG. 6.

[0059] Figure 12 A flowchart of a process of a third embodiment of the fatigue driving early warning method based on the fusion of multiple fatigue features of the present application is shown in FIG. 6.

[0060] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0061] The technical solutions in the embodiments of the present application will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present application.

[0062] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present application are only used to explain the relative positional relationship, movement condition, etc. between components in a certain posture (as shown in the drawings), and if the certain posture changes, the directional indications also change accordingly.

[0063] In addition, the description of "first", "second", etc. in the present application is only for the purpose of description, and cannot be understood as indicating or implying the relative importance of the indicated technical features or implicitly indicating the number of the indicated technical features. Therefore, the features limited by "first", "second" can explicitly or implicitly include at least one of the features. In addition, "and / or" throughout the text includes three solutions, for example, A and / or B includes A technical solution, B technical solution, and A and B simultaneously meet the technical solution. In addition, the technical solutions of each embodiment can be combined with each other, but it must be based on the realization of those skilled in the art, when the combination of technical solutions appears contradictory or unachievable, it should be considered that the combination of technical solutions does not exist, nor within the protection scope of the present application.

[0064] Reference Figures 1 to 2The application mainly provides a fatigue driving early warning method based on multiple fusion fatigue features, and applies the fatigue driving early warning method to the driving process of an intelligent vehicle. During the driving process of the vehicle on the road, dangerous driving states such as deviation from the lane inevitably occur, and when the vehicle is in an abnormal state, the driver needs to immediately correct the state to make the vehicle return to the normal state and ensure the safety of the vehicle. Whether the driver can immediately correct the state needs to ensure that the driver has a high degree of concentration and timely discovers the abnormal state of the vehicle. When the driver is in a fatigue state such as dozing off, poor spirit and low concentration, the vehicle is prone to danger. Therefore, when the driver is tired, the driver and / or the passenger of the vehicle need to be prompted for user perception to ensure that the driver is in a sober state at all times to ensure the safe driving of the vehicle.

[0065] The intelligent vehicle has many forms, such as a traditional fuel vehicle or a pure electric new energy vehicle. The intelligent vehicle 100 includes a vehicle body, a steering wheel 110 and a safety seat arranged in the vehicle body. The safety seat includes a driver seat 120 and a passenger seat 130. In this embodiment, when it is detected that the driver has a high degree of fatigue, the user is prompted to pay attention to safe driving in different tactile sensations. The steering wheel 110 of the intelligent vehicle 100 is provided with an electric tactile prompting component for driver perception, such as a metal electrode ring 111. The metal electrode ring 111 can be controlled by a controller to output an electric stimulation current (the current size is within a safe range). The electric stimulation current includes at least two different stimulation forces. The higher the degree of fatigue of the driver, the greater the electric stimulation force. The driver is electrically stimulated by the electric stimulation current to achieve the warning effect. The passenger seat is also provided with a tactile prompting component for user perception. The form of tactile prompting has many kinds, such as an electrode type tactile sensation or a mechanical type physical tactile sensation. Taking the mechanical type physical tactile sensation as an example, the mechanical type physical tactile sensation includes at least two different pushing, hammering or knocking forces to quantitatively prompt the user perception of the passenger. The size of the force of the tactile prompting is used for quantitatively prompting the fatigue degree of the driver. That is, the greater the force of the physical tactile prompting, the higher the degree of fatigue of the driver. After the passenger is physically touched, it indicates that the driver is in a fatigue state. At this time, the passenger can directly warn the driver to keep sober or suggest the driver to stop and rest as soon as possible in the vehicle, so as to indirectly remind the driver by reminding the passenger to achieve the warning effect.

[0066] The specific steps of the fatigue driving early warning method based on multiple fusion fatigue features will be mainly described below.

[0067] Reference Figure 3 In the embodiment of the application, the fatigue driving early warning method based on multiple fusion fatigue features includes the following steps:

[0068] S100: Model matching is performed on multiple different fatigue features to obtain a matching degree result. The fatigue features include eye features, body movement features and physiological parameter features.

[0069] S200: performing weighted fusion processing based on the matching degree result to obtain a fatigue degree quantization value;

[0070] S300: determining that the fatigue degree quantization value is higher than a preset value, and generating a user perception early warning instruction;

[0071] S400: responding to the user perception early warning instruction, and performing a user early warning prompt on the vehicle driver and / or passenger.

[0072] In this embodiment, first, matching degree results are obtained by respectively performing model matching on a plurality of different fatigue features, then a fatigue degree quantization value is obtained by performing weighted fusion processing based on the matching degree results, then it is determined that the fatigue degree quantization value is higher than a preset value, and a user perception early warning instruction is generated, and finally, responding to the user perception early warning instruction, a user early warning prompt is performed on the vehicle driver and / or passenger; in this way, fatigue detection is performed through different fatigue features, the accuracy of fatigue detection is improved, and different touch prompts are used to remind the user to pay attention to safe driving, thereby improving the safety of vehicle driving.

[0073] Specifically, fatigue features can be various, such as facial features, mouth features, eye features, body movement features, and physiological parameter features, and the body movement features can include hand movement features, eye movement features, and leg movement features, and the physiological parameter features can be blood oxygen solubility parameter features, heartbeat parameter features, and pulse parameter features. In this embodiment, eye features, body movement features, and physiological parameter features are fused to judge fatigue degree. It can be understood that eye features, body movement features, and physiological parameter features have different influences on fatigue state and fatigue degree. For example, if the driver's eye disease causes the features to match the fatigue model, it will lead to misjudgment of fatigue state, and body movement features and physiological parameter features also have influences in special cases. Therefore, a single fatigue feature is easy to cause misjudgment of fatigue state and fatigue degree. Based on this, eye features, body movement features, and physiological parameter features are fused to judge fatigue degree, which can improve the accuracy of fatigue detection and judgment. There are various fusion processing methods for different fatigue features, such as weighted fusion processing or integral fusion processing. This embodiment takes weighted fusion processing as an example for description. For example, the influence weight of eye features is 0.4, the influence weight of body movement features is 0.4, and the influence weight of physiological parameter features is 0.2. The influence weights of different fatigue features can be configured according to the actual situation of the driver. The fatigue degree quantization value is calculated by combining the influence weights with the matching degree results obtained by model matching. When the fatigue degree quantization value is higher than a preset value, a user perception early warning instruction is generated to warn the driver to drive safely. The steps of the matching degree results obtained by model matching are described as follows:

[0074] Referring toFigure 4 In some embodiments, the step of model matching the eye feature of the fatigue feature to obtain a matching degree result comprises:

[0075] S110: Obtain a driver eye feature video stream;

[0076] S120: Extract eye feature image frames in a preset time and process to form an eye feature line set;

[0077] S130: Model match the eye feature line set to obtain an eye feature matching degree result.

[0078] In this embodiment, first, a driver eye feature video stream is obtained. There are various ways to obtain a driver eye feature video stream, such as installing a monitoring camera in the cockpit, aiming the camera at the driver's face, obtaining a face video stream, then processing the face video stream to obtain an eye video stream, then extracting eye feature image frames in a preset time and processing to form an eye feature line set, such as extracting eye feature image frames in 30 seconds and processing all the eye feature image frames in sequence to form an eye feature line set, and finally model matching the eye feature line set to obtain an eye feature matching degree result, which is represented by a value of 10-100, such as a matching degree of 10, 20, 30, 40, 50, 60, 70, 80, 90, and 100.

[0079] Referring to Figure 5 In some embodiments, the step of model matching the body movement feature of the fatigue feature to obtain a matching degree result comprises:

[0080] S140: Obtain a driver hand holding pressure value and a waist leaning pressure value;

[0081] S150: Divide the hand holding pressure value and the waist leaning pressure value to form a pressure feature model curve;

[0082] S160: Model match the pressure feature model curve to obtain a body movement feature matching degree result.

[0083] In this embodiment, first, the driver's hand holding pressure value and waist leaning pressure value are acquired, and the hand holding pressure value and waist leaning pressure value can be acquired by a pressure sensor, such as a pressure sensor arranged at a part of the steering wheel hand holding, and a pressure sensor arranged at the backrest of the driver's seat. Then, the hand holding pressure value and the waist leaning pressure value are divided to form a pressure characteristic model curve. With the change of time, the quotient forms a pressure characteristic model curve. It can be understood that when the driver is drowsy, he is in a low head position, at this time, the hand holding pressure value increases, the waist leaning pressure value decreases, and the quotient of the hand holding pressure value / waist leaning pressure value increases. Therefore, the quotient can be used to determine whether the driver is in a fatigue state and the degree of fatigue. Finally, the pressure characteristic model curve is matched to obtain a body movement characteristic matching degree result, and the matching degree result is represented by a value of 10-100, such as matching degrees of 10, 20, 30, 40, 50, 60, 70, 80, 90, and 100.

[0084] Referring to Figure 6 In some embodiments, the step of matching the physiological parameter characteristic of the fatigue characteristic to obtain a matching degree result comprises:

[0085] S170: acquiring a pulse jumping state of the driver;

[0086] S180: performing noise reduction processing on the pulse jumping state to form a pulse jumping model curve;

[0087] S190: matching the pulse jumping model curve to obtain a physiological parameter characteristic matching degree result.

[0088] In this embodiment, first, the pulse jumping state of the driver is acquired. The pulse jumping state can be acquired by the driver wearing a bracelet or a watch, or by an optical sensor arranged on the steering wheel. Then, the pulse jumping state is subjected to noise reduction processing to form a pulse jumping model curve. The noise reduction processing is to remove abnormal data and then continuously integrate normal data to form a pulse jumping model curve. Finally, the pulse jumping model curve is matched to obtain a physiological parameter characteristic matching degree result, and the matching degree result is represented by a value of 10-100, such as matching degrees of 10, 20, 30, 40, 50, 60, 70, 80, 90, and 100.

[0089] Referring to Figure 7 In some embodiments, the step of performing weighted fusion processing based on the matching degree result to obtain a fatigue degree quantification value comprises:

[0090] S210: acquiring a fatigue influence weight corresponding to different fatigue characteristics;

[0091] S220: calculate a fatigue degree quantization value based on the fatigue influence weight and the eye feature matching degree result, the body movement feature matching degree result, and the physiological parameter feature matching degree result.

[0092] In this embodiment, first, fatigue influence weights corresponding to different fatigue features are obtained. The fatigue influence weights corresponding to different fatigue features can be stored in a memory, and relevant data can be temporarily obtained when needed. For example, the eye feature influence weight stored in the memory is 0.4, the body movement feature influence weight is 0.4, and the physiological parameter feature influence weight is 0.2. Then, a fatigue degree quantization value is calculated based on the fatigue influence weight and the eye feature matching degree result, the body movement feature matching degree result, and the physiological parameter feature matching degree result. For example, the eye feature matching degree is 80, the body movement feature matching degree is 70, and the physiological parameter feature matching degree is 80. Then, the fatigue degree quantization value can be calculated as 80*0.4+70*0.4+80*0.2=76.

[0093] Referring to Figure 1 , 2 and Figure 8 , in some embodiments, the intelligent vehicle includes a steering wheel and a passenger seat, and a pre-warning prompt component for user perception is arranged on the steering wheel and the passenger seat.

[0094] The step of determining that the fatigue degree quantization value is higher than a preset value and generating a user-perception pre-warning instruction includes:

[0095] S310: determining that the fatigue degree quantization value is higher than a preset value and accumulating a continuous duration.

[0096] S320: determining that the accumulated duration is greater than a preset duration, and generating a first instruction for the steering wheel to perform an electric tactile pre-warning prompt to the user.

[0097] To give the driver a certain fatigue self-response and warning time, in this embodiment, first, it is determined that the fatigue degree quantization value is higher than a preset value and an accumulated continuous duration is accumulated. Then, it is determined that the accumulated duration is greater than a preset duration, and a first instruction for the steering wheel to perform an electric tactile pre-warning prompt to the user is generated. For example, if the fatigue degree quantization value is higher than a preset value for 10 seconds, it indicates that the driver is in a fatigue state and is unaware of his fatigue. The controller controls the metal electrode ring 111 of the steering wheel 110 to output a first instruction of an electric stimulation current to produce an electric stimulation warning to the driver and to have an electric stimulation refreshing effect.

[0098] It can be understood that the electric stimulation warning may not have an effective effect on every driver. For example, for a driver with a particularly high fatigue degree, the electric stimulation warning may not have a sufficient warning effect. Therefore, the passenger on the vehicle needs to directly wake up the driver to have an effective warning effect.

[0099] Referring to Figure 9 and 10 In some embodiments, the first instruction triggered electric touch early warning prompt further comprises:

[0100] S330: reacquire the quantified value of the driver's fatigue degree;

[0101] S340: determine that the quantified value of the fatigue degree is higher than the preset value, and generate a second instruction for the passenger seat to give the user a physical touch early warning prompt;

[0102] Specifically, in the embodiment, after the electric touch early warning prompt triggered by the first instruction ends, the quantified value of the driver's fatigue degree is reacquired first. When the quantified value of the driver's fatigue degree is still higher than the preset value, it indicates that the electric stimulation warning does not have enough warning effect. At this time, a second instruction for the passenger seat to give the user a physical touch early warning prompt is generated, such as generating a hammering force on the passenger seat. When the passenger perceives the hammering force, it indicates that the driver is in a fatigue state. At this time, the passenger can directly wake up the driver in person. And / or,

[0103] S330: reacquire the quantified value of the driver's fatigue degree;

[0104] S350: determine that the quantified value of the fatigue degree is higher than the preset value, and acquire the use state of the passenger seat;

[0105] S360: determine that the passenger seat is in a passenger seated state, and generate a second instruction for the passenger seat to give the user a physical touch early warning prompt.

[0106] Specifically, in the embodiment, after the electric touch early warning prompt triggered by the first instruction ends, the quantified value of the driver's fatigue degree is reacquired first. When the quantified value of the driver's fatigue degree is still higher than the preset value, it indicates that the electric stimulation warning does not have enough warning effect. At this time, the use state of the passenger seat is acquired. When the user is detected to use the passenger seat, a second instruction for the passenger seat to give the user a physical touch early warning prompt is generated, such as generating a hammering force on the passenger seat. When the passenger perceives the hammering force, it indicates that the driver is in a fatigue state. At this time, the passenger can directly wake up the driver in person.

[0107] Therefore, the technical scheme can prompt the user to pay attention to safe driving by different touch (driver electric touch and passenger physical touch), thereby improving the safety of vehicle driving.

[0108] Referring to Figure 11 In some embodiments, the step of matching a plurality of different fatigue characteristics to obtain a matching degree result further comprises:

[0109] S000: Obtain a motion state of a steering wheel of the vehicle;

[0110] S010: Determine that the steering wheel of the vehicle is not turned within a preset time, and obtain a plurality of different fatigue features:

[0111] S100: Respectively perform model matching on the plurality of different fatigue features to obtain a matching degree result.

[0112] In this embodiment, first, the motion state of the steering wheel of the vehicle is obtained. When the steering wheel of the vehicle is not turned within a preset time, it is indicated that the driver is highly likely not to operate the vehicle, and it is further indicated that the driver is highly likely in a fatigue state. At this time, a plurality of different fatigue features need to be obtained to determine whether the driver is fatigued. After the plurality of different fatigue features are obtained, model matching is performed on the plurality of different fatigue features to obtain a matching degree result.

[0113] Reference Figure 12 In some embodiments, the first instruction includes a high-level electrical tactile early warning instruction and a low-level electrical tactile early warning instruction, the second instruction includes a high-level physical tactile early warning instruction and a low-level electrical tactile early warning instruction, and the step of determining that the fatigue degree quantization value is higher than a preset value and generating a user-perceived early warning instruction includes:

[0114] S370: Determine that the fatigue degree quantization value is higher than a first preset value and lower than a second preset value, and generate a low-level electrical tactile early warning instruction and / or a low-level electrical tactile early warning instruction.

[0115] S380: Determine that the fatigue degree quantization value is higher than the second preset value, and generate a high-level electrical tactile early warning instruction and / or a high-level electrical tactile early warning instruction.

[0116] Specifically, in this embodiment, when the fatigue degree quantization value is higher than the first preset value and lower than the second preset value, such as 50 < fatigue degree quantization value < 70, it is indicated that the driver is in a fatigue state, but the fatigue degree is low. At this time, a low-level electrical tactile early warning instruction and / or a low-level electrical tactile early warning instruction can be generated. The low-level electrical tactile early warning instruction and the low-level electrical tactile early warning instruction correspond to small-intensity electrical stimulation or physical stimulation. When 70 < fatigue degree quantization value, it is indicated that the driver is in a fatigue state and the fatigue degree is high. At this time, a high-level electrical tactile early warning instruction and / or a high-level electrical tactile early warning instruction are generated. The high-level electrical tactile early warning instruction and the high-level electrical tactile early warning instruction correspond to large-intensity electrical stimulation or physical stimulation, which warns the driver or the passenger through the large-intensity electrical stimulation or physical stimulation.

[0117] The application further provides a fatigue driving early warning system, comprising a memory for storing executable instructions; and a processor for executing the executable instructions stored in the memory to implement a fatigue driving early warning method based on multiple fusion fatigue features, which comprises the following steps:

[0118] S100: model matching is performed on multiple different fatigue features respectively to obtain matching degree results; wherein the fatigue features include eye features, body movement features and physiological parameter features;

[0119] S200: weighted fusion processing is performed based on the matching degree results to obtain a fatigue degree quantization value;

[0120] S300: when the fatigue degree quantization value is higher than a preset value, a user perception early warning instruction is generated;

[0121] S400: in response to the user perception early warning instruction, a user early warning prompt is performed on the driver and / or passenger of the vehicle.

[0122] The above description is merely preferred embodiments of the application, and does not limit the patent scope of the application, and any equivalent structural transformation made according to the application concept, or direct / indirect application in other related technical fields, is included in the patent protection scope of the application.

Claims

1. A fatigue driving early warning method based on multiple fusion fatigue features, applied to an intelligent automobile, characterized in that, The method comprises: respectively performing model matching on a plurality of different fatigue features to obtain matching degree results; wherein the fatigue features include eye features, body movement features, and physiological parameter features; performing weighted fusion processing based on the matching degree results to obtain a fatigue degree quantification value; determining that the fatigue degree quantification value is higher than a preset value, and generating a user-perception warning instruction; in response to the user-perception warning instruction, performing a user warning prompt on the vehicle driver and / or passenger; the step of performing model matching on the eye features of the fatigue features to obtain matching degree results comprises: obtaining a driver eye feature video stream; extracting eye feature image frames within a preset time and processing to form an eye feature line set; performing model matching on the eye feature line set to obtain an eye feature matching degree result; the step of performing model matching on the body movement features of the fatigue features to obtain matching degree results comprises: obtaining driver hand holding pressure values and waist leaning pressure values; performing division calculation on the hand holding pressure values and waist leaning pressure values to form a pressure feature model curve; performing model matching on the pressure feature model curve to obtain a body movement feature matching degree result.

2. The fatigue driving early warning method of claim 1, wherein, the step of performing model matching on the physiological parameter features of the fatigue features to obtain matching degree results comprises: obtaining a driver pulse jumping state; performing noise reduction processing on the pulse jumping state to form a pulse jumping model curve; performing model matching on the pulse jumping model curve to obtain a physiological parameter feature matching degree result.

3. The fatigue driving early warning method of claim 2, wherein, the step of performing weighted fusion processing based on the matching degree results to obtain a fatigue degree quantification value comprises: obtaining fatigue influence weights corresponding to different fatigue features; based on the fatigue influence weights and the eye feature matching degree result, the body movement feature matching degree result, and the physiological parameter feature matching degree result, calculating a fatigue degree quantification value.

4. The fatigue driving early warning method of claim 3, wherein, The intelligent automobile comprises a steering wheel and a passenger seat, and the steering wheel and the passenger seat are provided with a warning prompt component for user perception; the step of determining that the fatigue degree quantification value is higher than a preset value, and generating a user-perception warning instruction comprises: determining that the fatigue degree quantification value is higher than a preset value, and accumulating a continuous duration; determining that the continuous duration is greater than a preset duration, and generating a first instruction for the steering wheel to perform an electric tactile warning prompt on the user.

5. The fatigue driving early warning method of claim 4, wherein, after the electric tactile warning prompt triggered by the first instruction ends, it further comprises: re-obtaining the fatigue degree quantification value of the driver; determining that the fatigue degree quantification value is higher than a preset value, and generating a second instruction for the passenger seat to perform a physical tactile warning prompt on the user; and / or, re-obtaining the fatigue degree quantification value of the driver; determining that the fatigue degree quantification value is higher than a preset value, and obtaining a use state of the passenger seat; determining that the passenger seat is in a passenger seating state, and generating a second instruction for the passenger seat to perform a physical tactile warning prompt on the user.

6. The fatigue driving early warning method of claim 5, wherein, before the step of respectively performing model matching on a plurality of different fatigue features to obtain matching degree results, it further comprises: obtaining a motion state of the vehicle steering wheel; determining that the vehicle steering wheel has not been turned within a preset time, and obtaining a plurality of different fatigue features; respectively performing model matching on a plurality of different fatigue features to obtain matching degree results.

7. The fatigue driving early warning method of claim 6, wherein, The first instruction includes high-level electric touch early warning instruction and low-level electric touch early warning instruction, the second instruction includes high-level physical touch early warning instruction and low-level electric touch early warning instruction, the step of determining that the fatigue degree quantitative value is higher than a preset value and generating user-perceived early warning instruction includes: determining that the fatigue degree quantitative value is higher than a first preset value and lower than a second preset value, and generating low-level electric touch early warning instruction and / or low-level electric touch early warning instruction; determining that the fatigue degree quantitative value is higher than the second preset value, and generating high-level electric touch early warning instruction and / or high-level electric touch early warning instruction.

8. A fatigue driving warning system characterized by, comprise: a memory for storing executable instructions; a processor for executing the executable instructions stored in the memory to implement the method of any one of claims 1 to 7.

Images