An artificial intelligence-based driver anti-drowsiness wake-up system

Abstract

The present invention provides a driver anti-drowsiness awakening system based on artificial intelligence. The system collects facial videos of the driver according to a collection frequency and performs preliminary recognition on the facial videos to obtain facial images that may be of fatigue driving, thereby acquiring and preliminarily recognizing the driver's facial condition while driving, providing a data basis for the recognition of fatigue driving, and performing facial expression and facial behavior recognition on the facial images based on artificial intelligence. Whether the driver is fatigue driving is determined based on the recognition results. The recognition accuracy is improved through artificial intelligence recognition. When it is determined that the driver is fatigue driving, the coordinate area of the driver's mouth and nose area is determined, and a refreshing liquid is sprayed into the coordinate area, thereby achieving the operation of accurately spraying the refreshing liquid on the driver and effectively making the driver awake.

Description

Technical Field

[0001] The present invention relates to the field of vehicle driving technology, and in particular to an artificial intelligence-based driver anti-drowsiness awakening system. Background Art

[0002] After long periods of continuous driving, drivers experience physiological and psychological dysfunction, which objectively leads to a decline in driving skills. Drivers also experience poor or insufficient sleep and are prone to fatigue from long hours of driving. The hazards of fatigue driving are as follows:

[0003] Decreased judgment ability: When a driver is fatigued, his judgment ability will be significantly reduced, making him prone to operational errors.

[0004] Slow reaction: The reaction time is prolonged, and it is difficult to respond quickly and correctly to emergencies.

[0005] Increased operational errors: When slightly fatigued, gear shifting is untimely and inaccurate; when moderately fatigued, operating movements are sluggish or even forgotten; when severely fatigued, subconscious operations or short periods of sleep may occur.

[0006] Existing fatigue driving recognition devices have the following technical problems: the recognition accuracy is not high, and the driver cannot be effectively awakened after fatigue driving is recognized. Summary of the Invention

[0007] The present invention provides a driver anti-drowsiness awakening system based on artificial intelligence, which is used to solve the problems raised in the background technology.

[0008] An artificial intelligence-based driver anti-drowsiness wake-up system, comprising:

[0009] A video acquisition module is used to acquire facial videos of the driver at an acquisition frequency, and to perform preliminary recognition on the facial videos to obtain facial images that may indicate fatigue driving;

[0010] An image recognition module is used to perform facial expression and facial behavior recognition on the facial image based on artificial intelligence, and determine whether the driver is driving fatigued based on the recognition results;

[0011] The servo-controlled injection module is used to determine the coordinate area of the driver's mouth and nose area after determining that the driver is driving fatigued, and to inject refreshing liquid to the coordinate area.

[0012] Preferably, the system further comprises: a trigger module, configured to trigger the servo control injection module to operate when it is determined that the driver is driving fatigued;

[0013] The trigger module includes:

[0014] a determination unit, configured to use the determination of driver fatigue as a start instruction of a trigger mechanism, and use a control instruction of an injection action as a trigger action of the trigger mechanism;

[0015] a synchronization unit, configured to establish a synchronization execution strategy between the start instruction and the trigger action, and to establish a trigger strategy based on the synchronization execution strategy and the trigger mechanism;

[0016] The execution unit realizes the immediate execution of the injection action of the servo-controlled injection module based on the trigger strategy when it is determined that the driver is driving fatigued.

[0017] Preferably, the video acquisition module includes:

[0018] a frequency determination unit, configured to set a collection frequency determination strategy based on a correspondence between a driver's historical fatigue curve and injection actions, obtain a latest fatigue curve of the driver within a preset time period from the historical fatigue curve, and determine a collection frequency based on the latest fatigue curve and the collection frequency determination strategy;

[0019] A collection unit, configured to use a camera aimed at the driving area to collect a facial video of the driver at the collection frequency;

[0020] The preliminary processing unit is used to perform preliminary recognition on the facial video based on an image comparison method, and to capture facial images that may be of fatigue driving from the facial video.

[0021] Preferably, the preliminary processing unit includes:

[0022] an image acquisition unit configured to acquire an image frame set from a facial video based on a preset sampling frequency, perform image comparison on adjacent image frames in the image frame set to obtain an image difference, select an image frame having an image difference greater than a preset difference as a target image frame, and obtain a target image frame set with a preset number of frames centered on the target image frame from the image frame set;

[0023] The image selection unit is used to perform a preliminary comparison between the target image frame set and a pre-acquired standard image of a driver driving fatigue to obtain image similarity, and obtain an image frame with an image similarity greater than a preset similarity as a facial image that may be of a driver driving fatigue.

[0024] Preferably, the image recognition module includes:

[0025] A model building unit is used to train an artificial intelligence model based on a large number of facial images with annotated expressions and facial features that are standard for fatigue driving, and to build a facial recognition model based on the expression and facial features training results;

[0026] an image recognition unit, configured to input the facial image into a facial recognition model, recognize each facial feature region, determine whether the expression feature of each facial feature is a fatigue feature, and determine whether the facial image is an image of fatigued driving based on the facial recognition model;

[0027] a probability determination unit, configured to determine a probability value of the driver being fatigued driving based on the fatigue characteristics and the fatigue driving image;

[0028] The fatigue judgment unit is used to determine whether the driver is driving fatigued based on the probability value in combination with vehicle state information and driving environment information.

[0029] Preferably, the probability determination unit includes:

[0030] an evaluation unit, configured to obtain a similarity between each facial feature in the fatigue driving image and a corresponding fatigue feature, determine a weight for each facial feature based on the influence of the facial features on fatigue driving, and obtain a fatigue evaluation value for the fatigue driving image based on the facial feature weight and the similarity between the facial features and the corresponding fatigue feature;

[0031] The conversion unit is used to convert the fatigue evaluation value of the fatigue driving image into a probability value of the driver being fatigued driving.

[0032] Preferably, the fatigue judgment unit includes:

[0033] An information acquisition unit, configured to acquire vehicle status information and driving environment information within a time period corresponding to the fatigue driving image;

[0034] a feature acquisition unit, configured to determine a vehicle speed change feature and a vehicle driving direction feature from the vehicle state information, and to acquire a driving road condition feature and other vehicle features from the driving environment information;

[0035] a feature analysis unit, configured to obtain a first degree of matching between the vehicle speed change feature and the driving road condition feature and other vehicle features, obtain a second degree of matching between the vehicle driving direction feature and the driving road condition feature and other vehicle features, and obtain a first change value of the vehicle speed change feature and obtain a second change value of the vehicle driving direction feature;

[0036] an awareness evaluation unit, configured to determine a comprehensive matching degree between the vehicle state and the driving environment based on the first matching degree and the second matching degree, and determine a driver awareness evaluation value based on the first change value and the second change value in combination with driving environment information;

[0037] a standard determination unit, configured to determine the driver's operational standardization level based on the comprehensive matching degree and the awareness evaluation value;

[0038] a habit analysis unit, configured to obtain historical driving data of the driver, obtain relevant driving data related to the driving environment information from the historical driving habits, determine the driver's driving habits based on the relevant driving data, and determine a degree of habit matching between the driver's current driving operation and the driving habits based on the vehicle state information;

[0039] The fatigue judgment unit is used to determine the driver's driving safety level based on the degree of operating specifications and the degree of habit matching, and to weight the probability value based on the driving safety level to obtain a target probability value. When the target probability value is greater than a preset probability value, it is determined that the driver is driving fatigued. Otherwise, it is determined that the driver may be driving fatigued and further monitoring is required.

[0040] Preferably, the servo-controlled injection module comprises:

[0041] an area determination unit, configured to determine a coordinate area of the driver's mouth and nose area based on the driver's facial image after determining that the driver is driving fatigued;

[0042] The spraying unit is configured to, upon receiving a control spray start command, determine a spray action instruction based on the coordinate area, and spray the refreshing liquid toward the coordinate area according to the spray action instruction.

[0043] Preferably, the region determination unit includes:

[0044] a marking unit, configured to retrieve a standard driving area after determining that the driver is driving fatigued, and mark the coordinates of the standard driving area to obtain a coordinate driving area;

[0045] a matching unit, configured to obtain edge environment features from the driver's facial image, match the edge environment features with the coordinate driving area, determine the coordinate values of the edge environment in the coordinate driving area, and obtain a matching result;

[0046] A mapping unit is used to map the facial image to a coordinate driving area based on the matching result, and determine the coordinate area of the driver's mouth and nose area according to the mapping result.

[0047] Preferably, the injection unit comprises:

[0048] a position determination unit for determining the relative position of the driver's mouth and nose and the injection port based on the coordinate area and the position information of the injection port after receiving the control injection start;

[0049] The instruction generating unit is used to determine the injection direction of the injection port based on the relative position, and generate an injection action instruction in combination with a preset injection amount and injection speed.

[0050] Compared with the prior art, the present invention has achieved the following beneficial effects:

[0051] By collecting the driver's facial video at a collection frequency and performing preliminary recognition on the facial video, a facial image that may be of fatigue driving is obtained, the facial condition of the driver while driving is acquired and preliminary recognition is achieved, providing a data basis for the recognition of fatigue driving, and facial expressions and facial behaviors of the facial image are recognized based on artificial intelligence. Whether the driver is fatigued is determined based on the recognition result, and recognition accuracy is improved through artificial intelligence. When it is determined that the driver is fatigued, the coordinate area of the driver's mouth and nose area is determined, and a refreshing liquid is sprayed into the coordinate area, so as to achieve the operation of accurately spraying the refreshing liquid on the driver and effectively make the driver awake.

[0052] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purpose and other advantages of the present invention can be achieved and obtained through the structures specifically pointed out in this application document.

[0053] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0055] Figure 1 This is a structural diagram of an artificial intelligence-based driver anti-drowsiness wake-up system in an embodiment of the present invention;

[0056] Figure 2 is a structural diagram of the trigger module in an embodiment of the present invention;

[0057] Figure 3 2 is a structural diagram of the image recognition module in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0059] Example 1:

[0060] The embodiment of the present invention provides a driver anti-drowsiness awakening system based on artificial intelligence, such as Figure 1 As shown, including:

[0061] A video acquisition module is used to acquire facial videos of the driver at an acquisition frequency, and to perform preliminary recognition on the facial videos to obtain facial images that may indicate fatigue driving;

[0062] An image recognition module is used to perform facial expression and facial behavior recognition on the facial image based on artificial intelligence, and determine whether the driver is driving fatigued based on the recognition results;

[0063] The servo-controlled injection module is used to determine the coordinate area of the driver's mouth and nose area after determining that the driver is driving fatigued, and to inject refreshing liquid to the coordinate area.

[0064] In this embodiment, the preliminary recognition of the facial video is specifically performed by intercepting the video with large facial expressions or changes in head movements as facial images that may indicate fatigue driving.

[0065] In this embodiment, the image recognition module uses embedded artificial intelligence technology to realize the recognition of strolling images.

[0066] In this embodiment, facial expressions include opening the mouth, closing the eyes, etc., and facial behaviors include touching the face, rubbing the eyes, etc.

[0067] The beneficial effects of the above design scheme are: by collecting the driver's facial video according to the collection frequency, and performing preliminary identification on the facial video, a facial image that may be of fatigue driving is obtained, the facial condition of the driver while driving is obtained and preliminary identification is achieved, providing a data basis for the identification of fatigue driving, facial expression and facial behavior of the facial image are recognized based on artificial intelligence, and whether the driver is fatigued is determined according to the recognition result. The recognition accuracy is improved through artificial intelligence recognition. When it is determined that the driver is fatigued, the coordinate area of the driver's mouth and nose area is determined, and a refreshing liquid is sprayed into the coordinate area, so as to achieve the operation of accurately spraying the refreshing liquid on the driver and effectively make the driver awake.

[0068] Example 2:

[0069] Based on Example 1, the embodiment of the present invention provides a driver anti-drowsiness wake-up system based on artificial intelligence, such as Figure 2 As shown, it also includes: a trigger module, which is used to trigger the servo control injection module to work when it is determined that the driver is driving fatigued;

[0070] The trigger module includes:

[0071] a determination unit, configured to use the determination of driver fatigue as a start instruction of a trigger mechanism, and use a control instruction of an injection action as a trigger action of the trigger mechanism;

[0072] a synchronization unit, configured to establish a synchronization execution strategy between the start instruction and the trigger action, and to establish a trigger strategy based on the synchronization execution strategy and the trigger mechanism;

[0073] The execution unit realizes the immediate execution of the injection action of the servo-controlled injection module based on the trigger strategy when it is determined that the driver is driving fatigued.

[0074] The beneficial effects of the above design scheme are: by using the determination of driver fatigue driving as the starting instruction of the trigger mechanism, and the control instruction of the injection action as the trigger action of the trigger mechanism, a synchronous execution strategy between the starting instruction and the trigger action is established, and a trigger strategy is established based on the synchronous execution strategy and the trigger mechanism. Based on the trigger strategy, the injection action of the servo-controlled injection module is immediately executed after it is determined that the driver is driving fatigued, thereby ensuring real-time synchronization between the injection action and the driver's fatigue state, and ensuring that the driver can wake up quickly.

[0075] Example 3:

[0076] Based on Example 1, this embodiment of the present invention provides an artificial intelligence-based driver anti-drowsiness wake-up system, wherein the video acquisition module includes:

[0077] a frequency determination unit, configured to set a collection frequency determination strategy based on a correspondence between a driver's historical fatigue curve and injection actions, obtain a latest fatigue curve of the driver within a preset time period from the historical fatigue curve, and determine a collection frequency based on the latest fatigue curve and the collection frequency determination strategy;

[0078] A collection unit, configured to use a camera aimed at the driving area to collect a facial video of the driver at the collection frequency;

[0079] The preliminary processing unit is used to perform preliminary recognition on the facial video based on an image comparison method, and to capture facial images that may be of fatigue driving from the facial video.

[0080] In this embodiment, the acquisition frequency determination strategy is that the more curve segments with ejection actions in the historical fatigue curve, the higher the corresponding acquisition frequency. The specific setting value can be reasonably set according to the actual situation.

[0081] In this embodiment, the latest fatigue curve is the latest one at the current time.

[0082] In this embodiment, the image comparison method specifically compares the image of the facial video with the preset image of the driver when fatigued, and makes a judgment based on the similarity.

[0083] The beneficial effects of the above design scheme are: by setting the acquisition frequency determination strategy based on the correspondence between the driver's historical fatigue curve and the injection action, the latest fatigue curve of the driver within a preset time period is obtained from the historical fatigue curve, and the acquisition frequency is determined based on the latest fatigue curve in combination with the acquisition frequency determination strategy; the camera aimed at the driving area is used to collect the driver's facial video at the acquisition frequency; based on the image comparison method, the facial video is preliminarily identified, and facial images that may be fatigue driving are intercepted from the facial video, providing a data basis for the identification of fatigue driving.

[0084] Example 4:

[0085] Based on Example 3, an embodiment of the present invention provides an artificial intelligence-based driver anti-drowsiness awakening system, wherein the preliminary processing unit includes:

[0086] an image acquisition unit configured to acquire an image frame set from a facial video based on a preset sampling frequency, perform image comparison on adjacent image frames in the image frame set to obtain an image difference, select an image frame having an image difference greater than a preset difference as a target image frame, and obtain a target image frame set with a preset number of frames centered on the target image frame from the image frame set;

[0087] The image selection unit is used to perform a preliminary comparison between the target image frame set and a pre-acquired standard image of a driver driving fatigue to obtain image similarity, and obtain an image frame with an image similarity greater than a preset similarity as a facial image that may be of a driver driving fatigue.

[0088] In this embodiment, the standard image of driver fatigue driving is acquired in advance from historical data.

[0089] The beneficial effects of the above design scheme are: by collecting a set of image frames in a facial video based on a preset sampling frequency, performing image comparison on adjacent image frames in the image frame set to obtain image difference, and taking image frames with image difference greater than the preset difference as target image frames, obtaining a target image frame set with a preset number of frames centered on the target image frame from the image frame set, performing a preliminary comparison between the target image frame set and a pre-acquired standard image of the driver driving fatigue to obtain image similarity, and obtaining image frames with image similarity greater than the preset similarity as facial images that may be of fatigue driving, thereby ensuring the correctness of screening of facial images that may be of fatigue driving and providing a data basis for the identification of fatigue driving.

[0090] Example 5:

[0091] Based on Example 1, the embodiment of the present invention provides a driver anti-drowsiness wake-up system based on artificial intelligence, such as Figure 3 As shown, the image recognition module includes:

[0092] A model building unit is used to train an artificial intelligence model based on a large number of facial images with annotated expressions and facial features that are standard for fatigue driving, and to build a facial recognition model based on the expression and facial features training results;

[0093] an image recognition unit, configured to input the facial image into a facial recognition model, recognize each facial feature region, determine whether the expression feature of each facial feature is a fatigue feature, and determine whether the facial image is an image of fatigued driving based on the facial recognition model;

[0094] a probability determination unit, configured to determine a probability value of the driver being fatigued driving based on the fatigue characteristics and the fatigue driving image;

[0095] The fatigue judgment unit is used to determine whether the driver is driving fatigued based on the probability value in combination with vehicle state information and driving environment information.

[0096] In this embodiment, facial expression features include eye features, mouth features, eyebrow features, ear features and nose features.

[0097] In this embodiment, the facial recognition model is used for recognition of each facial feature and comprehensive recognition of fatigue driving.

[0098] The beneficial effects of the above design scheme are: by training the artificial intelligence model based on a large number of facial images with annotated expressions, and training the artificial intelligence model based on a large number of facial images with standard facial features during fatigue driving, a facial recognition model is established according to the expression and facial features training results, the facial recognition model is established, the facial image is input into the facial recognition model, each facial feature area is identified, and it is determined whether the expression feature of each facial feature is a fatigue feature, and whether the facial image is a fatigue driving image is determined based on the facial recognition model, and based on the fatigue feature and the fatigue driving image, the probability value of the driver is fatigue driving is determined, and the recognition of the image from the local to the whole based on artificial intelligence is realized, so as to ensure the accuracy of the probability value of the driver being fatigue driving, and based on the probability value, combined with the vehicle status information and the driving environment information, determine whether the driver is fatigue driving, and make further judgments based on the actual driving situation to ensure the accuracy of the determination of whether the driver is fatigue driving.

[0099] Example 6:

[0100] Based on Example 5, an embodiment of the present invention provides an artificial intelligence-based driver anti-drowsiness wakeup system, wherein the probability determination unit includes:

[0101] an evaluation unit, configured to obtain a similarity between each facial feature in the fatigue driving image and a corresponding fatigue feature, determine a weight for each facial feature based on the influence of the facial features on fatigue driving, and obtain a fatigue evaluation value for the fatigue driving image based on the facial feature weight and the similarity between the facial features and the corresponding fatigue feature;

[0102] The conversion unit is used to convert the fatigue evaluation value of the fatigue driving image into a probability value of the driver being fatigued driving.

[0103] In this embodiment, the larger the fatigue evaluation value, the greater the probability that the driver is driving while fatigued.

[0104] In this embodiment, the weight of each facial feature is determined based on the influence of the facial features on fatigue driving, for example, the weight of the mouth is 0.9, the weight of the eyes is 0.9, and the weight of the ears is 0.3.

[0105] The beneficial effects of the above design scheme are: by obtaining the similarity between each facial feature in the fatigue driving image and the corresponding fatigue feature, and determining the weight of each facial feature based on the influence of the facial features on fatigue driving, a fatigue evaluation value for the fatigue driving image is obtained based on the facial feature weight of the facial features and the similarity with the corresponding fatigue feature, and the fatigue evaluation value of the fatigue driving image is converted into a probability value of the driver being fatigued driving, thereby realizing image recognition from local to overall based on artificial intelligence, and ensuring the accuracy of the probability value of the driver being fatigued driving.

[0106] Example 7:

[0107] Based on Example 5, this embodiment of the present invention provides an artificial intelligence-based driver anti-drowsiness wake-up system, wherein the fatigue judgment unit includes:

[0108] An information acquisition unit, configured to acquire vehicle status information and driving environment information within a time period corresponding to the fatigue driving image;

[0109] a feature acquisition unit, configured to determine a vehicle speed change feature and a vehicle driving direction feature from the vehicle state information, and to acquire a driving road condition feature and other vehicle features from the driving environment information;

[0110] a feature analysis unit, configured to obtain a first degree of matching between the vehicle speed change feature and the driving road condition feature and other vehicle features, obtain a second degree of matching between the vehicle driving direction feature and the driving road condition feature and other vehicle features, and obtain a first change value of the vehicle speed change feature and obtain a second change value of the vehicle driving direction feature;

[0111] an awareness evaluation unit, configured to determine a comprehensive matching degree between the vehicle state and the driving environment based on the first matching degree and the second matching degree, and determine a driver awareness evaluation value based on the first change value and the second change value in combination with driving environment information;

[0112] a standard determination unit, configured to determine the driver's operational standardization level based on the comprehensive matching degree and the awareness evaluation value;

[0113] a habit analysis unit, configured to obtain historical driving data of the driver, obtain relevant driving data related to the driving environment information from the historical driving habits, determine the driver's driving habits based on the relevant driving data, and determine a degree of habit matching between the driver's current driving operation and the driving habits based on the vehicle state information;

[0114] The fatigue judgment unit is used to determine the driver's driving safety level based on the degree of operating specifications and the degree of habit matching, and to weight the probability value based on the driving safety level to obtain a target probability value. When the target probability value is greater than a preset probability value, it is determined that the driver is driving fatigued. Otherwise, it is determined that the driver may be driving fatigued and further monitoring is required.

[0115] In this embodiment, the higher the driving safety level, the smaller the weighting process is applied to the probability value.

[0116] In this embodiment, based on the first change value and the second change value, combined with the driving environment information, the driver's awareness evaluation value is determined. Specifically, when the driving environment information needs to change more, the more the first change value and the second change value match the situation where the driving environment information needs to change more, the corresponding awareness evaluation value is higher.

[0117] The beneficial effect of the above design scheme is: based on the probability value, combined with the vehicle status information and driving environment information, it is determined whether the driver is driving fatigued, and further judgment is made in combination with the actual driving situation to ensure the accuracy of the determination of whether the driver is driving fatigued.

[0118] Example 8:

[0119] Based on Example 1, an embodiment of the present invention provides an artificial intelligence-based driver anti-drowsiness wake-up system, wherein the servo-controlled injection module includes:

[0120] an area determination unit, configured to determine a coordinate area of the driver's mouth and nose area based on the driver's facial image after determining that the driver is driving fatigued;

[0121] The spraying unit is configured to, upon receiving a control spray start command, determine a spray action instruction based on the coordinate area, and spray the refreshing liquid toward the coordinate area according to the spray action instruction.

[0122] The beneficial effects of the above design scheme are: when it is determined that the driver is driving fatigued, the coordinate area of the driver's mouth and nose area is determined based on the driver's facial image; when the control injection start is received, the injection action instruction is determined based on the coordinate area, and refreshing liquid is sprayed to the coordinate area according to the injection action instruction; when it is determined that the driver is driving fatigued, the coordinate area of the driver's mouth and nose area is determined, and refreshing liquid is sprayed to the coordinate area, thereby realizing the operation of accurately spraying refreshing liquid on the driver and effectively making the driver awake.

[0123] Example 9:

[0124] Based on Example 8, an embodiment of the present invention provides an artificial intelligence-based driver anti-drowsiness wakeup system, wherein the area determination unit includes:

[0125] a marking unit, configured to retrieve a standard driving area after determining that the driver is driving fatigued, and mark the coordinates of the standard driving area to obtain a coordinate driving area;

[0126] a matching unit, configured to obtain edge environment features from the driver's facial image, match the edge environment features with the coordinate driving area, determine the coordinate values of the edge environment in the coordinate driving area, and obtain a matching result;

[0127] A mapping unit is used to map the facial image to a coordinate driving area based on the matching result, and determine the coordinate area of the driver's mouth and nose area according to the mapping result.

[0128] The beneficial effects of the above design scheme are: when it is determined that the driver is driving fatigued, the coordinate area of the driver's mouth and nose area is determined based on the driver's facial image; when the control injection start is received, the injection action instruction is determined based on the coordinate area, and refreshing liquid is sprayed to the coordinate area according to the injection action instruction; when it is determined that the driver is driving fatigued, the coordinate area of the driver's mouth and nose area is determined, and refreshing liquid is sprayed to the coordinate area, thereby realizing the operation of accurately spraying refreshing liquid on the driver and effectively making the driver awake.

[0129] Example 10:

[0130] Based on Example 8, an embodiment of the present invention provides an artificial intelligence-based driver anti-drowsiness awakening system, wherein the injection unit includes:

[0131] a position determination unit for determining the relative position of the driver's mouth and nose and the injection port based on the coordinate area and the position information of the injection port after receiving the control injection start;

[0132] The instruction generating unit is used to determine the injection direction of the injection port based on the relative position, and generate an injection action instruction in combination with a preset injection amount and injection speed.

[0133] The beneficial effects of the above design scheme are: when it is determined that the driver is driving fatigued, the coordinate area of the driver's mouth and nose area is determined based on the driver's facial image; when the control injection start is received, the injection action instruction is determined based on the coordinate area, and refreshing liquid is sprayed to the coordinate area according to the injection action instruction; when it is determined that the driver is driving fatigued, the coordinate area of the driver's mouth and nose area is determined, and refreshing liquid is sprayed to the coordinate area, thereby realizing the operation of accurately spraying refreshing liquid on the driver and effectively making the driver awake.

[0134] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of this application document and its equivalents, the present invention is intended to include these modifications and variations.

Claims

1. An artificial intelligence-based driver anti-drowsiness wake-up system, characterized in that: include: The video acquisition module is used to collect the driver's facial video at the acquisition frequency and perform preliminary recognition on the facial video to obtain facial images that may indicate fatigue driving, including: a frequency determination unit, configured to set a collection frequency determination strategy based on a correspondence between a driver's historical fatigue curve and injection actions, obtain a latest fatigue curve of the driver within a preset time period from the historical fatigue curve, and determine a collection frequency based on the latest fatigue curve and the collection frequency determination strategy; A collection unit, configured to use a camera aimed at the driving area to collect a facial video of the driver at the collection frequency; A preliminary processing unit is configured to perform preliminary recognition on the facial video based on an image comparison method, and to extract facial images that may indicate fatigue driving from the facial video, including: an image acquisition unit configured to acquire an image frame set from a facial video based on a preset sampling frequency, perform image comparison on adjacent image frames in the image frame set to obtain an image difference, select an image frame having an image difference greater than a preset difference as a target image frame, and obtain a target image frame set with a preset number of frames centered on the target image frame from the image frame set; an image selection unit, configured to perform a preliminary comparison between the target image frame set and a pre-acquired standard image of a driver experiencing fatigue driving to obtain image similarity, and to obtain an image frame having an image similarity greater than a preset similarity as a facial image that may be of a driver experiencing fatigue driving; An image recognition module is used to perform facial expression and facial behavior recognition on the facial image based on artificial intelligence, and determine whether the driver is driving fatigued based on the recognition results, including: A model building unit is used to train an artificial intelligence model based on a large number of facial images with annotated expressions and facial features that are standard for fatigue driving, and to build a facial recognition model based on the expression and facial features training results; an image recognition unit, configured to input the facial image into a facial recognition model, recognize each facial feature region, determine whether the expression feature of each facial feature is a fatigue feature, and determine whether the facial image is an image of fatigued driving based on the facial recognition model; A probability determination unit is used to determine a probability value of the driver being fatigued driving based on the fatigue characteristics and the fatigue driving image, including: an evaluation unit, configured to obtain a similarity between each facial feature in the fatigue driving image and a corresponding fatigue feature, determine a weight for each facial feature based on the influence of the facial features on fatigue driving, and obtain a fatigue evaluation value for the fatigue driving image based on the facial feature weight and the similarity between the facial features and the corresponding fatigue feature; a conversion unit, configured to convert the fatigue evaluation value of the fatigue driving image into a probability value of the driver being fatigued; A fatigue judgment unit is used to determine whether the driver is driving fatigued based on the probability value in combination with vehicle state information and driving environment information, including: An information acquisition unit, configured to acquire vehicle status information and driving environment information within a time period corresponding to the fatigue driving image; a feature acquisition unit, configured to determine a vehicle speed change feature and a vehicle driving direction feature from the vehicle state information, and to acquire a driving road condition feature and other vehicle features from the driving environment information; a feature analysis unit, configured to obtain a first degree of matching between the vehicle speed change feature and the driving road condition feature and other vehicle features, obtain a second degree of matching between the vehicle driving direction feature and the driving road condition feature and other vehicle features, and obtain a first change value of the vehicle speed change feature and obtain a second change value of the vehicle driving direction feature; an awareness evaluation unit, configured to determine a comprehensive matching degree between the vehicle state and the driving environment based on the first matching degree and the second matching degree, and determine a driver awareness evaluation value based on the first change value and the second change value in combination with driving environment information; a standard determination unit, configured to determine the driver's operational standardization level based on the comprehensive matching degree and the awareness evaluation value; a habit analysis unit, configured to obtain historical driving data of the driver, obtain relevant driving data related to the driving environment information from the historical driving data, determine the driver's driving habits based on the relevant driving data, and determine a degree of habit matching between the driver's current driving operation and the driving habits based on the vehicle state information; a fatigue judgment unit, configured to determine the driver's driving safety level based on the degree of operational standardization and the degree of matching of the operating habits, and to weight the probability value based on the driving safety level to obtain a target probability value; when the target probability value is greater than a preset probability value, determine that the driver is driving fatigued; otherwise, determine that the driver may be driving fatigued and requires further monitoring; The servo-controlled injection module is used to determine the coordinate area of the driver's mouth and nose area after determining that the driver is driving fatigued, and to inject refreshing liquid into the coordinate area, including: an area determination unit, configured to determine a coordinate area of the driver's mouth and nose area based on the driver's facial image after determining that the driver is driving fatigued; The spraying unit is configured to, upon receiving a control spray start command, determine a spray action instruction based on the coordinate area, and spray the refreshing liquid toward the coordinate area according to the spray action instruction.

2. The artificial intelligence-based driver anti-drowsiness awakening system according to claim 1, characterized in that: Also includes: A trigger module is used to trigger the servo control injection module to work when it is determined that the driver is driving fatigued; The trigger module includes: a determination unit, configured to use the determination of driver fatigue as a start instruction of a trigger mechanism, and use a control instruction of an injection action as a trigger action of the trigger mechanism; a synchronization unit, configured to establish a synchronization execution strategy between the start instruction and the trigger action, and to establish a trigger strategy based on the synchronization execution strategy and the trigger mechanism; The execution unit realizes the immediate execution of the injection action of the servo-controlled injection module based on the trigger strategy when it is determined that the driver is driving fatigued.

3. The artificial intelligence-based driver anti-drowsiness awakening system according to claim 1, characterized in that: The area determination unit includes: a marking unit, configured to retrieve a standard driving area after determining that the driver is driving fatigued, and mark the coordinates of the standard driving area to obtain a coordinate driving area; a matching unit, configured to obtain edge environment features from the driver's facial image, match the edge environment features with the coordinate driving area, determine the coordinate values of the edge environment in the coordinate driving area, and obtain a matching result; A mapping unit is used to map the facial image to a coordinate driving area based on the matching result, and determine the coordinate area of the driver's mouth and nose area according to the mapping result.

4. The artificial intelligence-based driver anti-drowsiness awakening system according to claim 1, characterized in that: The injection unit comprises: a position determination unit for determining the relative position of the driver's mouth and nose and the injection port based on the coordinate area and the position information of the injection port after receiving the control injection start; The instruction generating unit is used to determine the injection direction of the injection port based on the relative position, and generate an injection action instruction in combination with a preset injection amount and injection speed.

Images