Method, device and equipment for predicting age of driver and passenger, medium and vehicle

Abstract

The embodiment of the invention provides a driver and passenger age prediction method and device, equipment, a medium and a vehicle, and relates to the technical field of age prediction. The method comprises the following steps: acquiring facial features of a to-be-tested person, and determining an age prediction basic value corresponding to the to-be-tested person based on the facial features; acquiring action information of the to-be-tested person, and determining an age prediction auxiliary value corresponding to the to-be-tested person based on the action information; and determining an age prediction result corresponding to the to-be-tested person in combination with the age prediction basic value and the age prediction auxiliary value. According to the embodiment of the invention, the age prediction auxiliary information is calculated according to the action behaviors of the driver and passengers, and the final age prediction result is determined in combination with the age prediction auxiliary information and the basic age predicted based on the facial features, so that the accuracy of performing age prediction on the driver and passengers is effectively improved.

Description

Technical Field

[0001] The present application relates to the technical field of age prediction, and in particular, to a method, device, equipment, medium and vehicle for predicting the age of a driver or passenger. Background Art

[0002] In the automotive field, estimating the age of drivers and passengers is a commonly used function. The identified age information can be used to adjust vehicle parameters and perform background data analysis.

[0003] At present, the age recognition of drivers and passengers mainly relies on facial feature recognition technology. This method can obtain relatively accurate age prediction results under ideal conditions. However, in actual applications, since the collection of facial images will be affected by factors such as facial angle, occlusion, and light, the recognition of facial features often has low accuracy, resulting in the inability to accurately estimate the age information of drivers and passengers. Summary of the invention

[0004] The purpose of the embodiments of the present application is to provide a method, device, equipment, medium and vehicle for predicting the age of a driver and passenger, so as to solve the problem that the existing scheme of estimating age based on facial features has low estimation accuracy.

[0005] In a first aspect, an embodiment of the present application provides a method for predicting the age of a driver or passenger, comprising:

[0006] Acquire facial features of the person to be tested, and determine a basic value for age prediction corresponding to the person to be tested based on the facial features;

[0007] Acquiring motion information of the person to be tested, and determining an age prediction auxiliary value corresponding to the person to be tested based on the motion information;

[0008] The age prediction result corresponding to the person to be tested is determined by combining the age prediction basic value and the age prediction auxiliary value.

[0009] In an embodiment of the present application, auxiliary information is calculated based on the movements of the driver and passenger, and an age prediction result is obtained by fusing the auxiliary information with a basic age predicted based on facial features, thereby effectively improving the accuracy of age prediction for the driver and passenger.

[0010] In some possible embodiments, the acquiring the motion information of the person to be tested, and determining the age prediction auxiliary value corresponding to the person to be tested based on the motion information, includes:

[0011] Acquire action video data of the person to be tested in the process of fastening the seat belt, and acquire seat belt using action information of the person to be tested based on the action video data;

[0012] Determining an age prediction auxiliary value corresponding to the person to be tested based on the seat belt use action information;

[0013] The seat belt use action information includes at least one of hand strength information, limb angle change information, hand movement path information and action duration information.

[0014] In the embodiment of the present application, the auxiliary information is calculated by using the seat belt fastening action that drivers and passengers generally perform, thereby further improving the accuracy and applicability of age prediction for drivers and passengers.

[0015] In some possible embodiments, the method of acquiring the hand strength information includes:

[0016] Extracting key frame images of the person to be tested in the process of fastening a seat belt from the action video data;

[0017] The hand strength information is determined based on the finger bending angle of the key frame image.

[0018] In an embodiment of the present application, by extracting key frame images of the user in the process of fastening the seat belt and calculating the hand strength information, the age prediction auxiliary value is estimated based on the hand strength information, thereby further improving the accuracy of age prediction for the driver and passenger.

[0019] In some possible embodiments, the method of acquiring the limb angle change information includes:

[0020] Dividing the action video data into a plurality of action image frames, and identifying limb key points in each frame of the action image;

[0021] For each frame of the action image, determining the limb angle corresponding to the action image based on the connecting lines of adjacent limb key points;

[0022] The maximum value and the minimum value of the limb angle are selected from the limb angles of the action image in each frame, and the limb angle change information is determined based on the maximum value and the minimum value of the limb angle.

[0023] In an embodiment of the present application, by identifying the limb key points of multiple frames of action images and determining the corresponding limb angles, the maximum and minimum values ​​therein are selected and the limb angle change information is calculated, so as to estimate the age prediction auxiliary value based on the limb angle change information, thereby further improving the accuracy of age prediction for drivers and passengers.

[0024] In some possible embodiments, the method of acquiring the hand motion path information includes:

[0025] Dividing the action video data into a plurality of action image frames, identifying wrist key points in each action image frame, and obtaining coordinate positions of the wrist key points in each action image frame;

[0026] The hand motion path information is determined based on each of the coordinate positions.

[0027] In an embodiment of the present application, by identifying wrist key points in multiple frames of action images and obtaining corresponding coordinate positions, the hand motion path information is determined based on these coordinate positions, and the age prediction auxiliary value is estimated based on the hand motion path information, thereby further improving the accuracy of age prediction for drivers and passengers.

[0028] In some possible embodiments, the method of obtaining the action duration information includes:

[0029] The starting time and the ending time of the use of the safety belt by the person to be tested are determined based on the action video data, and the action duration information is determined based on the starting time and the ending time.

[0030] In an embodiment of the present application, by determining the action duration information from the start time to the end time of the user fastening the seat belt, and estimating the age prediction auxiliary value based on the action duration information, the accuracy of age prediction for the driver and passenger is further improved.

[0031] In some possible embodiments, the acquiring the motion information of the person to be tested, and determining the age prediction auxiliary value corresponding to the person to be tested based on the motion information, includes:

[0032] Obtaining a credibility corresponding to the facial feature;

[0033] In the case where it is determined that the credibility is lower than a preset credibility threshold, obtaining motion information of the person to be tested, and determining an age prediction auxiliary value corresponding to the person to be tested based on the motion information;

[0034] When it is determined that the credibility is not lower than a preset credibility threshold, an age prediction result corresponding to the person to be tested is determined based on the age prediction basic value.

[0035] In the embodiment of the present application, whether to combine the age prediction auxiliary value to estimate the age of the driver and occupant is determined according to the credibility of the facial features, thereby further improving the accuracy and flexibility of age prediction for the driver and occupant.

[0036] In some possible embodiments, obtaining the credibility corresponding to the facial feature includes:

[0037] Acquire a facial image of a person to be tested, and acquire quality assessment information corresponding to the facial image; wherein the quality assessment information includes at least one of a facial angle, a degree of occlusion, and a light intensity of the facial image;

[0038] A confidence level corresponding to the facial feature is determined based on the quality assessment information.

[0039] In an embodiment of the present application, by determining the quality assessment information of the face image based on the facial angle, occlusion degree, light intensity, etc., and determining the credibility of the facial features based on the quality assessment information of the face image, the accuracy of calculating the credibility of the facial features is further improved.

[0040] In some possible embodiments, the combining the age prediction basic value and the age prediction auxiliary value to determine the age prediction result corresponding to the person to be tested includes:

[0041] A result adjustment parameter corresponding to the age prediction auxiliary value is determined, and the age prediction basic value is adjusted based on the result adjustment parameter to obtain an age prediction result of the person to be tested.

[0042] In the embodiment of the present application, by determining the corresponding result adjustment parameter according to the age prediction auxiliary value, and adjusting the age prediction basic value according to the result adjustment parameter, the accuracy and flexibility of age prediction for the driver and passenger are further improved.

[0043] In some possible embodiments, determining a result adjustment parameter corresponding to the age prediction auxiliary value, and adjusting the age prediction basic value based on the result adjustment parameter to obtain the age prediction result of the person to be tested includes:

[0044] Comparing the age prediction auxiliary value with a preset reference benchmark value, and determining a result adjustment parameter corresponding to the age prediction auxiliary value according to the comparison result;

[0045] The age prediction basic value is adjusted based on the result adjustment parameter to obtain the age prediction result of the person to be tested.

[0046] In an embodiment of the present application, the age prediction auxiliary value is compared with a preset reference baseline value to determine a result adjustment parameter, and the age prediction basic value is adjusted according to the result adjustment parameter, thereby further improving the accuracy and flexibility of age prediction for drivers and passengers.

[0047] In some possible embodiments, the combining the age prediction basic value and the age prediction auxiliary value to determine the age prediction result corresponding to the person to be tested includes:

[0048] The age prediction basic value and the age prediction auxiliary value are weightedly summed based on a preset weight to obtain an age prediction result corresponding to the person to be tested.

[0049] In an embodiment of the present application, the age prediction basic value and the age prediction auxiliary value are weighted and summed by preset weights to obtain the final age prediction result, thereby further improving the accuracy and flexibility of age prediction for drivers and passengers.

[0050] In a second aspect, an embodiment of the present application provides a device for predicting the age of a driver or passenger, comprising:

[0051] A basic prediction module, used to obtain facial features of a person to be tested, and determine a basic age prediction value corresponding to the person to be tested based on the facial features;

[0052] an auxiliary prediction module, used to obtain the motion information of the person to be tested, and determine the age prediction auxiliary value corresponding to the person to be tested based on the motion information;

[0053] The age prediction model is used to determine the age prediction result corresponding to the person to be tested by combining the age prediction basic value and the age prediction auxiliary value.

[0054] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor can implement the method described in any embodiment of the first aspect when executing the program.

[0055] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method described in any embodiment of the first aspect can be implemented.

[0056] In a fifth aspect, an embodiment of the present application provides a computer program product, wherein the computer program product includes a computer program, wherein when the computer program is executed by a processor, the method described in any embodiment of the first aspect can be implemented.

[0057] In a sixth aspect, an embodiment of the present application provides a vehicle, comprising the electronic device described in the third aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0059] Figure 1 A schematic diagram of a process for predicting the age of a driver or passenger provided in an embodiment of the present application;

[0060] Figure 2 An overall logic flow chart of a method for predicting the age of a driver or passenger provided in an embodiment of the present application;

[0061] Figure 3 An example schematic diagram of the definition of finger key points provided in an embodiment of the present application;

[0062] Figure 4 A schematic diagram of the structure of a device for predicting the age of a driver or passenger provided in an embodiment of the present application;

[0063] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0064] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

[0065] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0066] It should be noted that the existing technology mainly predicts age based on facial features. However, in the field of driver and passenger age prediction, the recognition accuracy of facial features is often limited due to factors such as facial angle, occlusion (wearing a hat, sunglasses), and poor lighting in the actual environment. Therefore, a solution that only predicts age based on facial features is difficult to achieve satisfactory results.

[0067] In view of the problems existing in the above-mentioned prior art, an embodiment of the present application provides a method for predicting the age of a driver and passenger, which predicts the age by combining the action and behavior characteristics and facial features of the driver and passenger, thereby effectively improving the accuracy of the age prediction results.

[0068] like Figure 1As shown, the embodiment of the present application provides a method for predicting the age of a driver and passenger, which may include the steps of:

[0069] S1. Obtain facial features of the person to be tested, and determine a basic value for age prediction corresponding to the person to be tested based on the facial features.

[0070] Specifically, a camera is first installed at an appropriate position in the vehicle, and the camera can be used to capture the facial image of the person to be tested (the driver or passenger). For example, for passengers, image data can be collected when the passenger gets on the vehicle, so as to ensure that the clearest facial image can be obtained.

[0071] Based on the collected face images, facial detection and recognition algorithms (such as VGGFace, FaceNet, etc.) can be used to extract corresponding facial features, and then the extracted facial features are input into a pre-trained age prediction model to obtain the age prediction basic value corresponding to the person to be tested output by the age prediction model. The age prediction model can be trained based on pre-acquired training data, and the training data can include facial feature samples and their corresponding actual ages.

[0072] S2. Obtaining motion information of the person to be tested, and determining an age prediction auxiliary value corresponding to the person to be tested based on the motion information.

[0073] Specifically, the relevant action information of the person to be tested can be obtained by means of a camera or a sensor, and the action information can be related information about actions such as fastening a seat belt and closing a car door. For example, the video data of the relevant actions of the person to be tested can be collected based on the camera, and the corresponding action information can be identified from the video data; in addition, the action information of the person to be tested (such as the strength of fastening the seat belt, etc.) can also be determined based on the data collected by the sensor.

[0074] Exemplarily, after the motion information of the person to be tested is collected, the age prediction auxiliary value corresponding to the current motion information can be obtained based on the preset correspondence between the motion information and the age prediction auxiliary value.

[0075] In addition, the currently collected motion information can also be input into a pre-trained auxiliary age prediction model to obtain an auxiliary age prediction value output by the auxiliary age prediction model. The auxiliary age prediction model can be trained based on pre-acquired training data, and the training data can include motion information and its corresponding auxiliary age value.

[0076] S3. Determine the age prediction result corresponding to the person to be tested by combining the age prediction basic value and the age prediction auxiliary value.

[0077] Specifically, the final age prediction result can be obtained by performing weighted summation based on the age prediction basic value and the age prediction auxiliary value with preset weights; in addition, the age prediction basic value can be fine-tuned according to the age prediction auxiliary value to obtain the final age prediction result.

[0078] Based on this, by comprehensively considering the facial features and action information of the driver and passengers and calculating the corresponding age prediction basic value and age prediction auxiliary value, the age prediction basic value and the age prediction auxiliary value are fused and calculated to obtain the age prediction result. Compared with the prior art that only predicts age based on facial features, the embodiment of the present application effectively improves the accuracy of age prediction for the driver and passengers.

[0079] In some possible embodiments, step S2, obtaining motion information of the person to be tested, and determining the age prediction auxiliary value corresponding to the person to be tested based on the motion information, may include:

[0080] S201, obtaining action video data of the person to be tested in the process of fastening a seat belt, and obtaining seat belt using action information of the person to be tested based on the action video data;

[0081] S202, determining an age prediction auxiliary value corresponding to the person to be tested based on the seat belt use action information;

[0082] Among them, the seat belt use action information includes at least one of hand strength information, limb angle change information, hand movement path information and action duration information.

[0083] It is understandable that, considering that fastening the seat belt is a behavior that the driver and passengers must perform after getting on the car, and considering that the action of fastening the seat belt is large in amplitude and relatively long in duration, it has a more general reference value. Therefore, the embodiment of the present application uses the action information of the action of fastening the seat belt as the basis for evaluating the age prediction auxiliary value.

[0084] It should be noted that based on the video data collected by the camera, the action video data of the corresponding behavior can be identified and intercepted. For the behavior of fastening the seat belt, the time node from the time when the person to be tested picks up the seat belt to the time when the person to be tested finishes fastening the seat belt can be determined by identifying the video data using image recognition technology, so as to take the moment when the person to be tested picks up the seat belt as the starting frame (usually the moment when the person to be tested first touches the seat belt or just starts to reach for the seat belt) and the moment when the person to be tested finishes fastening the seat belt as the ending frame (usually the moment when the seat belt is buckled or the person to be tested stops the action), and intercept the action video data of the person to be tested in the process of fastening the seat belt.

[0085] Based on the intercepted action video data, it can be converted into several frames of action images, and the corresponding seat belt usage action information can be identified and calculated based on these action images, wherein the seat belt usage action information includes but is not limited to at least one of the following: hand strength information, limb angle change information, hand movement path information and action duration information.

[0086] like Figure 2 As shown, exemplarily, the corresponding basic value for age prediction can be obtained by recognition based on facial image data, and at least one item of action information can be obtained by calculation based on action video data, thereby calculating the age prediction auxiliary value corresponding to the action behavior, and finally the final age prediction result is obtained by fusing the age prediction basic value and the age prediction auxiliary value.

[0087] In some possible embodiments, the hand strength information is obtained in the following manner:

[0088] Extract key frame images of the person to be tested in the process of fastening the seat belt from the action video data;

[0089] The hand strength information is determined based on the finger bending angle of the key frame image.

[0090] It is understandable that since the greater the degree of finger bending, the greater the grip strength is generally, the degree of finger bending can be expressed according to the angle corresponding to the finger joint, and the grip strength of the person to be tested during the seat belt fastening process can be determined based on the degree of finger bending.

[0091] For example, based on the captured action video data, key frame images of the person to be tested in the process of fastening the seat belt can be extracted. For example, the image corresponding to the moment when the start frame is delayed by a preset time can be used as the key frame image; in addition, the image corresponding to the moment when the hand reaches a certain position can also be used as the key frame image.

[0092] Specifically, Figure 3 As shown, first, the positions of the finger base, finger tip and two middle finger joints can be defined as finger key points. For example, the finger key points in the key frame image can be detected using a key point detection network model such as HRNet; then these key points are connected and the bending angle values ​​of the two fingers are calculated; finally, the hand strength information is determined according to the finger bending angle. For example, the following calculation formula can be used: F force =ω 1 (θ 1 +θ 2 ), where F force Indicates the hand strength information, ω 1 It represents the conversion coefficient of the preset finger bending angle and hand strength information. The coefficient can be preset according to statistical data.

[0093] After obtaining the hand strength information, the corresponding age prediction auxiliary value can be determined according to the preset correspondence between the hand strength information and the age prediction auxiliary value, and the correspondence can be preset according to the analysis of statistical data. Generally, for the elderly and children, the smaller their corresponding hand strength information is, and the opposite is true for young people and middle-aged people.

[0094] In some possible embodiments, the method of acquiring the limb angle change information includes:

[0095] Divide the action video data into several frames of action images, and identify the key points of the limbs in each frame of action image;

[0096] For each frame of action image, the limb angle corresponding to the action image is determined based on the connecting lines of adjacent limb key points;

[0097] The maximum value and the minimum value of the limb angle are selected from the limb angles of each frame of the action image, and the limb angle change information is determined based on the maximum value and the minimum value of the limb angle.

[0098] Specifically, four key points related to the arms and body may be defined first, for example, the neck, shoulders, elbows, and wrists may be defined as limb key points.

[0099] Exemplarily, for each frame of action image included in the action video data, the corresponding four limb key points can be detected using a key point detection algorithm.

[0100] Then, for each frame of the action image, the identified limb key points are connected in pairs to obtain three line segments, and two limb angles are formed (the line between the neck and the shoulder - the line between the shoulder and the elbow constitutes the first limb angle θ 3 , the line between the shoulder and the elbow - the line between the elbow and the wrist constitutes the second limb angle θ 4 ).

[0101] Then, the limb angles corresponding to each frame of the action image are traversed, and the maximum and minimum values ​​of the corresponding first limb angles (θ 3max and θ 3min ), and the maximum and minimum values ​​of the second limb angle (θ 4max and θ 4min ).

[0102] Finally, the following formula is used to calculate the limb angle change information:

[0103] Δθ=ω 2 (θ 3max -θ 3min )+ω 3 (θ 4max -θ 4min), where ω 2 and ω 3 Represents the preset coefficient used to balance the relationship between two angles.

[0104] After obtaining the limb angle change information, the corresponding age prediction auxiliary value can be determined according to the preset correspondence between the limb angle change information and the age prediction auxiliary value, and the correspondence can be preset according to the analysis of statistical data. Generally, for the elderly and children, the corresponding limb angle change information is smaller, and the opposite is true for young people and middle-aged people.

[0105] In some possible embodiments, the method of obtaining the hand motion path information includes:

[0106] Divide the action video data into several frames of action images, identify the wrist key points in each frame of action image, and obtain the coordinate position of the wrist key points in each frame of action image;

[0107] Hand motion path information is determined based on each coordinate position.

[0108] It is understandable that passengers of different ages may show significant differences in the hand motion paths when fastening their seat belts. For example, the motion paths of young passengers may be shorter, smoother, and have a lower rate of change, while the motion paths of older passengers may be longer, more complex, and have a higher rate of change. Based on this, by detecting and tracking the hand motion paths of passengers in the process of fastening their seat belts and calculating the relevant path features, these features can be used as a basis for evaluating the smoothness of the seat belt fastening movements of the person to be tested, and then as a basis for assisting in predicting age.

[0109] Furthermore, the hand motion path information may include one or more of path length, path change rate, and path smoothness.

[0110] Specifically, firstly, the wrist key points in several frames of action images included in the action video data are respectively identified, and the coordinate position of the wrist key point in the image is determined, which can be expressed as W i =(x i ,y i ). Where W i Indicates the coordinate position (or corresponding vector) of the wrist key point in the i-th frame.

[0111] The path length refers to the total length of the wrist keypoint motion path, which can be calculated by accumulating the Euclidean distance between adjacent frames:

[0112]

[0113] The path change rate (denoted as R) can be measured by calculating the change in direction of the path in each frame, where the change in direction can be expressed by calculating the change in angle in each frame:

[0114]

[0115] Among them, W i Represents the vector corresponding to the coordinate position of the wrist key point in the i-th frame.

[0116] The path change rate R can be expressed as the sum of all angle changes:

[0117] Path smoothness (denoted as S) can be measured by calculating the acceleration change of the path.

[0118] Among them, the speed of the wrist key point of each frame action image is expressed as: Here, represents the time interval between two adjacent frames of action images.

[0119] The acceleration of the wrist key point can be expressed as:

[0120] The path smoothness S can be expressed as the sum of all acceleration changes:

[0121] For example, after obtaining parameters such as path length, path change rate, and path smoothness, the corresponding hand motion path information (expressed as Road) can be calculated based on the following formula: Road = ω 4 L+ω 5 R+ω 6 S, where ω 4 ,ω 5 and ω 6 Represents a preset coefficient used to balance the impact of various parameters on the hand motion path information.

[0122] After the hand motion path information is acquired, the corresponding age prediction auxiliary value may be determined according to a preset correspondence between the hand motion path information and the age prediction auxiliary value, and the correspondence may be preset according to analysis of statistical data.

[0123] In some possible embodiments, the method of obtaining the action duration information includes:

[0124] The starting time and the ending time of the use of the safety belt by the person to be tested are determined based on the action video data, and the action duration information is determined based on the starting time and the ending time.

[0125] It should be noted that, based on the start frame and the end frame when the action video data is intercepted as the start time and the end time of the test person using the seat belt, the specific action of the test person in the process of fastening the seat belt can also be identified to determine the start time and the end time. Then, the corresponding action duration information is calculated based on the duration between the start time and the end time.

[0126] After obtaining the action duration information, the corresponding age prediction auxiliary value can be determined according to the correspondence between the preset action duration information and the age prediction auxiliary value, and the correspondence can be preset according to the analysis of statistical data. Generally, the elderly and children wear seat belts for a longer time, while young people and middle-aged people wear them for a shorter time.

[0127] It should be noted that after obtaining one or more seat belt usage action information such as hand strength information, limb angle change information, hand movement path information and action duration information, these information items can be weighted summed based on preset weights to obtain the final seat belt usage action information, and then the corresponding age prediction auxiliary value can be determined based on the correspondence between the preset seat belt usage action information and the age prediction auxiliary value. The correspondence can be pre-set based on the analysis of statistical data.

[0128] In addition, hand strength information, limb angle change information, hand movement path information and action duration information can be respectively used as four types of feature information, and model training is performed by using the four types of feature information and their corresponding preset age prediction auxiliary values ​​as training data, so as to obtain a trained auxiliary value prediction model, so as to be able to identify the correspondence between the seat belt usage action information and the preset age prediction auxiliary value, and then by inputting the acquired seat belt usage action information into the trained auxiliary value prediction model, the age prediction auxiliary value output by the auxiliary value prediction model is obtained.

[0129] In some possible embodiments, step S2, obtaining motion information of the person to be tested, and determining the age prediction auxiliary value corresponding to the person to be tested based on the motion information, may include:

[0130] S211, obtaining the credibility corresponding to the facial features;

[0131] S212: when it is determined that the credibility is lower than a preset credibility threshold, obtaining motion information of the person to be tested, and determining an age prediction auxiliary value corresponding to the person to be tested based on the motion information;

[0132] S213: When it is determined that the credibility is not lower than a preset credibility threshold, determine an age prediction result corresponding to the person to be tested based on the age prediction basic value.

[0133] It should be noted that before obtaining the motion information of the person to be tested and calculating the corresponding age prediction auxiliary value, it can be first determined whether the credibility of the facial features is lower than a preset credibility threshold;

[0134] If so, it means that the basic value of age prediction obtained based on the current facial features may be less accurate, and it is necessary to further obtain auxiliary values ​​for age prediction and comprehensively determine the final age prediction result; based on this, when the facial features are unreliable, further auxiliary information can be obtained to comprehensively calculate the age prediction result, effectively improving the accuracy of age prediction.

[0135] If not, it means that the age prediction base value calculated based on the current facial features may be accurate enough, and the age prediction base value can be directly used as the age prediction result. Based on this, the age prediction result can be directly output when the facial features are highly credible, saving unnecessary calculation expenses and improving the efficiency of age prediction.

[0136] In some possible embodiments, step S211, obtaining the credibility corresponding to the facial feature, may include:

[0137] S2111, obtaining a facial image of the person to be tested, and obtaining quality assessment information corresponding to the facial image; wherein the quality assessment information includes at least one of a facial angle, an occlusion degree, and a light intensity of the facial image;

[0138] S2112. Determine the credibility corresponding to the facial features based on the quality assessment information.

[0139] Specifically, for the collected facial image of the person to be tested, relevant features in the facial image can be identified as quality assessment information, which may include but is not limited to one or more of facial angle, occlusion degree and light intensity.

[0140] For example, the more frontal the recognized facial angle is, the higher the credibility of the corresponding facial features is, and the more lateral the facial angle is, the lower the credibility of the corresponding facial features is.

[0141] The degree of occlusion can be measured by calculating the ratio of the area of ​​the face that is blocked to the area of ​​the entire face. The higher the degree of occlusion, the lower the credibility of the facial features, and the lower the degree of occlusion, the higher the credibility of the corresponding facial features.

[0142] The light intensity is measured by calculating the image brightness of the current face image. The closer the light intensity is to the preset standard light, the higher the credibility of the corresponding facial features. The farther the light intensity is from the preset standard light (too high or too low), the lower the credibility of the corresponding facial features.

[0143] In some possible embodiments, step S3, combining the age prediction basic value and the age prediction auxiliary value to determine the age prediction result corresponding to the person to be tested, may include:

[0144] S301. Determine a result adjustment parameter corresponding to the auxiliary value for age prediction, and adjust the basic value for age prediction based on the result adjustment parameter to obtain an age prediction result for the person to be tested.

[0145] It should be noted that after obtaining the age prediction basic value and the age prediction auxiliary value, the result adjustment parameter corresponding to the age prediction auxiliary value can be determined according to the preset corresponding relationship, and the age prediction basic value can be adjusted based on the result adjustment parameter to obtain the age prediction result of the person to be tested.

[0146] Among them, the correspondence between the age prediction auxiliary value and the result adjustment parameter can be set according to the analysis of statistical data. For example, the result adjustment parameter corresponding to the age prediction auxiliary value of 50 can be set to 5, the result adjustment parameter corresponding to the age prediction auxiliary value of 40 can be set to 4, the result adjustment parameter corresponding to the age prediction auxiliary value of 20 can be set to -1, etc., then after obtaining the result adjustment parameter corresponding to the current age prediction auxiliary value, the corresponding result adjustment parameter can be added or subtracted from the currently obtained age prediction basic value to obtain the final age prediction result.

[0147] In some possible embodiments, step S301, determining a result adjustment parameter corresponding to the age prediction auxiliary value, and adjusting the age prediction basic value based on the result adjustment parameter to obtain the age prediction result of the person to be tested, may include:

[0148] S3011, comparing the age prediction auxiliary value with a preset reference benchmark value, and determining a result adjustment parameter corresponding to the age prediction auxiliary value according to the comparison result;

[0149] S3012. Adjust the age prediction basic value based on the result adjustment parameter to obtain the age prediction result of the person to be tested.

[0150] Exemplarily, the preset reference benchmark value can be a preset fixed value, such as 40; the result adjustment parameter can also be a fixed value, such as 2 and -2; by comparing the age prediction auxiliary value with the preset reference benchmark value, if the current age prediction auxiliary value is greater than the reference benchmark value, the corresponding result adjustment parameter is determined to be 2, otherwise the corresponding result adjustment parameter is determined to be -2 (if the age prediction auxiliary value is equal to the preset reference benchmark value, the result adjustment parameter is 0).

[0151] Based on this, the result adjustment parameter corresponding to the age prediction auxiliary value is obtained by presetting the reference benchmark value, so that the age prediction result after adjusting the age prediction basic value is closer to the actual age of the person to be tested, effectively improving the accuracy of age prediction.

[0152] In some possible embodiments, step S3, combining the age prediction basic value and the age prediction auxiliary value to determine the age prediction result corresponding to the person to be tested, may include:

[0153] S311 . Perform weighted summation of the age prediction basic value and the age prediction auxiliary value based on preset weights to obtain an age prediction result corresponding to the person to be tested.

[0154] It should be noted that in addition to obtaining the result adjustment parameters corresponding to the age prediction auxiliary value to adjust the age prediction basic value, the fusion weight can also be set in advance, and the age prediction basic value and the age prediction auxiliary value can be weighted and summed based on the fusion weight to obtain the final age prediction result.

[0155] For example, the weight α corresponding to the basic value of age prediction can be set to 0.8, and the weight corresponding to the auxiliary value of age prediction can be set to 0.2. Then, when the basic value of age prediction is calculated to be 30 and the auxiliary value of age prediction is 50, the fused age prediction result is 30×0.8+50×0.2=34 years old.

[0156] Based on this, the age prediction basic value and the age prediction auxiliary value are weighted and summed by preset weights to obtain the final age prediction result, thereby further improving the accuracy and flexibility of age prediction for drivers and passengers.

[0157] Please refer to Figure 4 , Figure 4 FIG. 1 shows a block diagram of a driver and passenger age prediction device provided by some embodiments of the present application. It should be understood that the driver and passenger age prediction device is similar to the above-mentioned Figure 1 Corresponding to the method embodiment, it is able to execute each step involved in the above method embodiment. The specific functions of the driver and passenger age prediction device can be found in the description above. To avoid repetition, the detailed description is appropriately omitted here.

[0158] Figure 4 The age prediction device for a driver and passenger comprises at least one software function module that can be stored in a memory in the form of software or firmware or fixed in the age prediction device for a driver and passenger, and the age prediction device for a driver and passenger comprises:

[0159] A basic prediction module 410 is used to obtain facial features of the person to be tested, and determine a basic age prediction value corresponding to the person to be tested based on the facial features;

[0160] The auxiliary prediction module 420 is used to obtain the motion information of the person to be tested, and determine the age prediction auxiliary value corresponding to the person to be tested based on the motion information;

[0161] The age prediction model 430 is used to determine the age prediction result corresponding to the person to be tested by combining the age prediction basic value and the age prediction auxiliary value.

[0162] It can be understood that the above-mentioned device item embodiment corresponds to the method item embodiment of the present invention. The age prediction device for drivers and passengers provided by the embodiment of the present invention can implement the age prediction method for drivers and passengers provided by any method item embodiment of the present invention.

[0163] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the device described above can refer to the corresponding process in the aforementioned method, and will not be described in detail here.

[0164] like Figure 5 As shown, some embodiments of the present application provide an electronic device 500, which includes: a memory 510, a processor 520, and a computer program stored in the memory 510 and executable on the processor 520, wherein the processor 520 reads the program from the memory 510 through a bus 530 and executes the program to implement a method of any embodiment of the driver and passenger age prediction method as described above.

[0165] Processor 520 can process digital signals and can include various computing structures, such as complex instruction set computer structure, reduced instruction set computer structure, or a structure that implements a combination of multiple instruction sets. In some examples, processor 520 can be a microprocessor.

[0166] The memory 510 may be used to store instructions executed by the processor 520 or data related to the execution of instructions. These instructions and / or data may include codes for implementing some or all functions of one or more modules described in the embodiments of the present application. The processor 520 of the disclosed embodiment may be used to execute instructions in the memory 510 to implement the method shown above. The memory 510 includes a dynamic random access memory, a static random access memory, a flash memory, an optical memory, or other memory known to those skilled in the art.

[0167] Some embodiments of the present application also provide a vehicle, comprising the electronic device 500 described above.

[0168] Some embodiments of the present application further provide a computer-readable storage medium having a computer program stored thereon. The computer program is executed by a processor to execute the method described in the method embodiment.

[0169] Some embodiments of the present application further provide a computer program product, which, when executed on a computer, enables the computer to execute the method described in the method embodiment.

[0170] It should be noted that each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the embodiments can be referred to each other. For the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0171] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0172] In addition, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0173] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0174] The above description is only an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings.

[0175] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0176] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

Claims

1. A method for predicting the age of a driver and passenger, characterized in that: include: Acquire facial features of the person to be tested, and determine a basic value for age prediction corresponding to the person to be tested based on the facial features; Acquiring motion information of the person to be tested, and determining an age prediction auxiliary value corresponding to the person to be tested based on the motion information; The age prediction result corresponding to the person to be tested is determined by combining the age prediction basic value and the age prediction auxiliary value.

2. The method for predicting the age of a driver and passenger according to claim 1, characterized in that: The acquiring the motion information of the person to be tested, and determining the age prediction auxiliary value corresponding to the person to be tested based on the motion information, comprises: Acquire action video data of the person to be tested in the process of fastening the seat belt, and acquire seat belt using action information of the person to be tested based on the action video data; Determining an age prediction auxiliary value corresponding to the person to be tested based on the seat belt use action information; The seat belt use action information includes at least one of hand strength information, limb angle change information, hand movement path information and action duration information.

3. The method for predicting the age of a driver and passenger according to claim 2, characterized in that: The method for obtaining the hand strength information includes: Extracting key frame images of the person to be tested in the process of fastening a seat belt from the action video data; The hand strength information is determined based on the finger bending angle of the key frame image.

4. The method for predicting the age of a driver and passenger according to claim 2, characterized in that: The method for acquiring the limb angle change information includes: Dividing the action video data into a plurality of action image frames, and identifying limb key points in each frame of the action image; For each frame of the action image, determining the limb angle corresponding to the action image based on the connecting lines of adjacent limb key points; The maximum value and the minimum value of the limb angle are selected from the limb angles of the action image in each frame, and the limb angle change information is determined based on the maximum value and the minimum value of the limb angle.

5. The method for predicting the age of a driver and passenger according to claim 2, characterized in that: The method for acquiring the hand motion path information includes: Dividing the action video data into a plurality of action image frames, identifying wrist key points in each action image frame, and obtaining coordinate positions of the wrist key points in each action image frame; The hand motion path information is determined based on each of the coordinate positions.

6. The method for predicting the age of a driver and passenger according to claim 2, characterized in that: The method for obtaining the action duration information includes: The starting time and the ending time of the use of the safety belt by the person to be tested are determined based on the action video data, and the action duration information is determined based on the starting time and the ending time.

7. The method for predicting the age of a driver and passenger according to claim 1, characterized in that: The acquiring the motion information of the person to be tested, and determining the age prediction auxiliary value corresponding to the person to be tested based on the motion information, comprises: Obtaining a credibility corresponding to the facial feature; In the case where it is determined that the credibility is lower than a preset credibility threshold, obtaining motion information of the person to be tested, and determining an age prediction auxiliary value corresponding to the person to be tested based on the motion information; When it is determined that the credibility is not lower than a preset credibility threshold, an age prediction result corresponding to the person to be tested is determined based on the age prediction basic value.

8. The method for predicting the age of a driver and passenger according to claim 7, characterized in that: The obtaining of the credibility corresponding to the facial feature comprises: Acquire a facial image of a person to be tested, and acquire quality assessment information corresponding to the facial image; wherein the quality assessment information includes at least one of a facial angle, a degree of occlusion, and a light intensity of the facial image; A confidence level corresponding to the facial feature is determined based on the quality assessment information.

9. The method for predicting the age of a driver and passenger according to claim 1, characterized in that: The step of combining the age prediction basic value and the age prediction auxiliary value to determine the age prediction result corresponding to the person to be tested includes: A result adjustment parameter corresponding to the age prediction auxiliary value is determined, and the age prediction basic value is adjusted based on the result adjustment parameter to obtain an age prediction result of the person to be tested.

10. The method for predicting the age of a driver and passenger according to claim 9, characterized in that: The determining of the result adjustment parameter corresponding to the age prediction auxiliary value, and adjusting the age prediction basic value based on the result adjustment parameter to obtain the age prediction result of the person to be tested, includes: Comparing the age prediction auxiliary value with a preset reference benchmark value, and determining a result adjustment parameter corresponding to the age prediction auxiliary value according to the comparison result; The age prediction basic value is adjusted based on the result adjustment parameter to obtain the age prediction result of the person to be tested.

11. The method for predicting the age of a driver and passenger according to claim 1, characterized in that: The step of combining the age prediction basic value and the age prediction auxiliary value to determine the age prediction result corresponding to the person to be tested includes: The age prediction basic value and the age prediction auxiliary value are weightedly summed based on a preset weight to obtain an age prediction result corresponding to the person to be tested.

12. A device for predicting the age of a driver or passenger, characterized in that: include: A basic prediction module, used to obtain facial features of a person to be tested, and determine a basic age prediction value corresponding to the person to be tested based on the facial features; an auxiliary prediction module, used to obtain the motion information of the person to be tested, and determine the age prediction auxiliary value corresponding to the person to be tested based on the motion information; The age prediction model is used to determine the age prediction result corresponding to the person to be tested by combining the age prediction basic value and the age prediction auxiliary value.

13. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor can implement the driver and passenger age prediction method as described in any one of claims 1-11 when executing the program.

14. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for predicting the age of a driver and passenger as described in any one of claims 1-11 is executed.

15. A vehicle, characterized in that: Comprising the electronic device as claimed in claim 13.

Images