Adjusting method and device for vehicle visor and computer equipment

Abstract

The invention relates to an adjusting method and device for a vehicle visor and computer equipment, and the method comprises the steps: determining a multi-modal target parameter based on multi-modal sensing data obtained in real time; the multi-modal target parameters comprise illumination parameters and a three-dimensional face model of the driver; determining illumination features corresponding to the illumination parameters and face posture features corresponding to the three-dimensional face model; performing correlation analysis on the illumination features and the face posture features, and determining a target angle of the to-be-adjusted visor based on a correlation analysis result and the size information of the to-be-adjusted visor; and adjusting the to-be-adjusted visor according to the target angle of the to-be-adjusted visor. By means of the method and device, the problems that real-time accurate response is difficult to achieve, and adjustment lag or adjustment angle deviation of the visor is prone to occurring are solved, adjustment of the vehicle visor is accurately controlled in real time based on multi-dimensional sensing data, and adjustment lag or adjustment angle deviation is avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automobiles, and in particular to a method and device for adjusting a sun visor of a vehicle and a computer device. BACKGROUND

[0002] During driving, direct sunlight on the driver's face can cause a blocked view, which is a long-standing safety hazard. Traditional solutions mainly rely on manual sun visors, which are folded or stretched to achieve shading. However, manual adjustment requires the driver to be distracted, increasing the risk of accidents.

[0003] To improve safety, existing automatic adjustment schemes usually introduce photosensitive sensors to detect light intensity and control the sun visor. However, there are generally problems such as insufficient adjustment accuracy, making it difficult to achieve real-time and accurate response, and prone to adjustment lag or angle deviation.

[0004] To address the problem of difficulty in achieving real-time and accurate response, and prone to adjustment lag or angle deviation of the sun visor in related technologies, there is currently no effective solution. SUMMARY

[0005] A method and device for adjusting a sun visor of a vehicle and a computer device are provided in the present embodiment to address the problem of difficulty in achieving real-time and accurate response, and prone to adjustment lag or angle deviation of the sun visor in related technologies.

[0006] In a first aspect, a method for adjusting a sun visor of a vehicle is provided in the present embodiment, comprising:

[0007] Based on the real-time acquired multi-modal perception data, a corresponding multi-modal target parameter is determined; the multi-modal target parameter includes a light parameter and a three-dimensional face model of the driver;

[0008] A light feature corresponding to the light parameter and a face posture feature corresponding to the three-dimensional face model are determined;

[0009] The light feature and the face posture feature are analyzed in association, and based on the association analysis result and the size information of the sun visor to be adjusted, a target angle of the sun visor to be adjusted is determined;

[0010] The sun visor to be adjusted is adjusted according to the target angle of the sun visor to be adjusted.

[0011] In some embodiments, the light feature and the face posture feature are analyzed in association, and based on the association analysis result and the size information of the sun visor to be adjusted, a target angle of the sun visor to be adjusted is determined, comprising:

[0012] Establishing a correlation model between the illumination feature and the facial posture feature, so as to determine a target effective shading length of the visor to be adjusted through analysis of the correlation model; wherein the illumination feature includes an incident light vector, and the facial posture feature includes a facial normal vector;

[0013] The target angle of the light shielding plate to be adjusted is determined according to the size information of the light shielding plate to be adjusted and the target effective blocking length; the size information includes the length of the light shielding plate to be adjusted.

[0014] In some embodiments, the analysis process of the association model includes:

[0015] Determining a corresponding incident angle according to the incident light vector and the facial normal vector;

[0016] Determining the eye position coordinates of the driver;

[0017] The target effective shielding length of the sun visor to be adjusted is determined according to the incident angle and the eye position coordinates of the driver.

[0018] In some embodiments, the multimodal target parameter further includes a vehicle driving parameter; and after determining the corresponding incident angle according to the incident light vector and the facial normal vector, the method further includes:

[0019] Determining a vehicle heading angle change rate, a vehicle lateral acceleration, and a light incidence change rate based on the vehicle driving parameters, wherein the vehicle driving parameters correspond to a current driving state or a predicted driving state;

[0020] Dynamic anti-drift compensation is performed on the incident angle according to the vehicle heading angle change rate, the vehicle lateral acceleration and the light incidence change rate.

[0021] In some embodiments, the lighting characteristic includes an incident light vector; and determining the lighting characteristic corresponding to the lighting parameter includes:

[0022] Acquiring real-time multimodal sensing data; the multimodal sensing data includes an interior lighting image of the vehicle;

[0023] Converting the centroid coordinates of the light spot in the interior illumination image into corresponding spherical coordinates, and determining a three-dimensional vector corresponding to the spherical coordinates;

[0024] Based on the light refraction model of the vehicle glass and the three-dimensional vector, a corresponding incident light vector is determined.

[0025] In some embodiments, determining the three-dimensional facial model of the driver includes:

[0026] acquire real-time multi-modal perception data; the multi-modal perception data includes a facial image, a facial temperature distribution information and a head depth information of the driver;

[0027] establish a corresponding head pose space coordinate system based on the head depth information of the driver;

[0028] fuse texture information in the facial image and the facial temperature distribution information to the head pose space coordinate system to obtain the three-dimensional facial model.

[0029] In some embodiments thereof, the method further comprises:

[0030] determine a driving scene corresponding to each scene vector; the scene vector includes a spatio-temporal feature, a dynamic environment perception feature and a driver feature;

[0031] based on different driving scenes, reinforcement learning is performed on an adjustment strategy of the to-be-adjusted sun visor to obtain a target adjustment strategy corresponding to each driving scene; the adjustment strategy includes an adjustment angle, an adjustment speed and an adjustment timing of the to-be-adjusted sun visor;

[0032] construct a mapping library between the driving scene and the target adjustment strategy.

[0033] In some embodiments thereof, after the mapping library between the driving scene and the target adjustment strategy is constructed, the method further comprises:

[0034] determine the target adjustment strategy corresponding to a real-time driving scene in the mapping library;

[0035] control a vehicle motor to drive the to-be-adjusted sun visor to perform angle adjustment according to the target adjustment strategy corresponding to the real-time driving scene;

[0036] based on a real-time running state of the vehicle motor, determine actual angle adjustment information of the to-be-adjusted sun visor;

[0037] based on the actual angle adjustment information, dynamically optimize the target adjustment strategy corresponding to the real-time driving scene.

[0038] In a second aspect, a vehicle sun visor adjustment device is provided in the present embodiment, comprising:

[0039] a determination module configured to determine corresponding multi-modal target parameters based on real-time acquired multi-modal perception data; the multi-modal target parameters include an illumination parameter and a three-dimensional facial model of a driver;

[0040] an extraction module configured to determine an illumination feature corresponding to the illumination parameter and a facial pose feature corresponding to the three-dimensional facial model.

[0041] analyze the illumination feature and the face posture feature, and determine a target angle of the to-be-adjusted sun visor based on a result of the correlation analysis and size information of the to-be-adjusted sun visor;

[0042] a control module, configured to adjust the to-be-adjusted sun visor according to the target angle of the to-be-adjusted sun visor.

[0043] In a third aspect, a computer device is provided in the embodiment, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the vehicle sun visor adjustment method in the first aspect when executing the computer program.

[0044] Compared with the related art, the vehicle sun visor adjustment method, device, and computer device provided in the embodiment determine corresponding multi-modal target parameters based on real-time acquired multi-modal perception data, the multi-modal target parameters including an illumination parameter and a three-dimensional face model of a driver, determine an illumination feature corresponding to the illumination parameter and a face posture feature corresponding to the three-dimensional face model, perform correlation analysis on the illumination feature and the face posture feature, and determine a target angle of a to-be-adjusted sun visor based on a result of the correlation analysis and size information of the to-be-adjusted sun visor, and adjust the to-be-adjusted sun visor according to the target angle of the to-be-adjusted sun visor, thereby solving the problem that it is difficult to achieve real-time accurate response and the to-be-adjusted sun visor is prone to adjustment lag or adjustment angle deviation, achieving real-time accurate control of the adjustment of the vehicle sun visor based on multi-dimensional perception data, avoiding adjustment lag or adjustment angle deviation, and improving driving comfort and safety and effectively preventing sunlight from directly shining on the driver and interfering with the driver's vision.

[0045] Details of one or more embodiments of the present application are presented in the following drawings and description to make other features, objects, and advantages of the present application more apparent. BRIEF DESCRIPTION OF DRAWINGS

[0046] The accompanying drawings illustrated herein are used to provide further understanding of the present application, constitute a part of the present application, and the illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute improper limitations on the present application. In the drawings:

[0047] Figure 1 is a hardware structure block diagram of a terminal device of the vehicle sun visor adjustment method provided in an embodiment of the present application;

[0048] Figure 2 is a flowchart of the vehicle sun visor adjustment method provided in an embodiment of the present application;

[0049] Figure 3is a schematic diagram of a principle of image distortion correction provided by an embodiment of the present application;

[0050] Figure 4 is a schematic diagram of a flow of incident angle dynamic compensation provided by an embodiment of the present application;

[0051] Figure 5 is a schematic diagram of a method for constructing a three-dimensional face model provided by an embodiment of the present application;

[0052] Figure 6 is a schematic diagram of a flow of a method for adjusting a vehicle sun visor provided by an embodiment of the present application;

[0053] Figure 7 is a structural block diagram of an adjusting device for a vehicle sun visor provided by an embodiment of the present application.

[0054] In the figure: 102, a processor; 104, a memory; 106, a transmission device; 108, an input / output device; 10, a determination module; 20, an extraction module; 30, an analysis module; 40, a control module. DETAILED DESCRIPTION

[0055] In order to more clearly understand the objectives, technical solutions and advantages of the present application, the present application is described and explained below in conjunction with the accompanying drawings and embodiments.

[0056] Unless otherwise defined, technical terms or scientific terms used in the present application shall have the general meaning understood by a person with ordinary skill in the art to which the present application belongs. In the present application, "one", "a", "an", "the", "these" and similar words do not represent a quantitative limitation, and they can be singular or plural. In the present application, the terms "include", "contain", "have" and any variants thereof have the purpose of covering non-exclusive inclusion; for example, a process, method and system, product or device containing a series of steps or modules (units) are not limited to the listed steps or modules (units), but can include steps or modules (units) not listed, or can include other steps or modules (units) inherent to the process, method, product or device. In the present application, the words "connected", "connected", "coupled" and similar words do not limit to physical or mechanical connection, but can include electrical connection, whether direct or indirect. In the present application, "multiple" means two or more. The association between the associated objects is described by "and / or", which means that there can be three relationships, for example, "A and / or B" can mean that A exists alone, A and B exist together, and B exists alone. In general, the character " / " represents an "or" relationship between the associated objects. In the present application, the terms "first", "second", "third" and the like are only used to distinguish similar objects, and do not represent a specific order of the objects.

[0057] The method embodiments provided in the present embodiment can be executed in a terminal, a computer or a similar computing device. For example, the method embodiments are executed on a terminal, Figure 1 is a hardware structure block diagram of a terminal of the adjusting method of a vehicle sun visor of the present embodiment. As shown in Figure 1 , the terminal can include one or more (only one is shown in Figure 1 ) processors 102 and a memory 104 for storing data, wherein the processor 102 can include but not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA. The terminal can further include a transmission device 106 for communication function and an input / output device 108. Those skilled in the art can understand that Figure 1 the structure shown is only schematic, which does not limit the structure of the terminal. For example, the terminal can further include more or less components than those shown in Figure 1 , or have different configurations from those shown in Figure 1 .

[0058] The memory 104 can be used to store computer programs, for example, software programs of application software and modules, such as the computer program corresponding to the adjusting method of a vehicle sun visor in the present embodiment. The processor 102 executes various functional applications and data processing by running the computer programs stored in the memory 104, that is, implements the method described above. The memory 104 can include a high-speed random access memory, and can further include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 104 can further include a memory remotely arranged with respect to the processor 102, which can be connected to the terminal through a network. Examples of the network include but not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0059] The transmission device 106 is used to receive or send data via a network. The network includes a wireless network provided by a communication provider of the terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC for short), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (Radio Frequency, RF for short) module, which is used to communicate with the Internet in a wireless manner.

[0060] An adjusting method of a vehicle sun visor is provided in the present embodiment, Figure 2 is a flowchart of the adjusting method of a vehicle sun visor of the present embodiment, as shown in Figure 2 , the flowchart includes the following steps:

[0061] Step S210: determining corresponding multimodal target parameters based on the multimodal perception data acquired in real time; the multimodal target parameters include illumination parameters and a three-dimensional facial model of the driver;

[0062] Step S220, determining illumination features corresponding to the illumination parameters and facial posture features corresponding to the three-dimensional facial model;

[0063] Step S230, performing correlation analysis on the illumination features and the facial posture features, and determining a target angle of the visor to be adjusted based on the correlation analysis result and the size information of the visor to be adjusted;

[0064] Step S240 , adjusting the shading plate to be adjusted according to the target angle of the shading plate to be adjusted.

[0065] Specifically, in the vehicle sun visor adjustment scenario, multimodal perception data is collected in real time. This multimodal perception data includes illumination and facial perception data. The illumination detection module, comprised of an onboard camera, a multi-angle light intensity sensor, a spectral sensor, and other devices, accurately captures illumination perception data such as the intensity distribution, color temperature parameters, and spectral characteristics (e.g., visible light and infrared intensity) from different directions inside and outside the vehicle. The facial perception module, equipped with a high-resolution camera, an infrared thermal imager, and a depth perception camera, acquires facial perception data such as the driver's facial image, facial temperature distribution, and head depth information in real time. It should be noted that traditional camera solutions have limitations in extreme lighting environments (e.g., strong backlight, low light, and dynamic strobe light). This embodiment utilizes a combination of an infrared thermal imager, a depth perception camera, and a spectral sensor during multimodal perception data collection. This enables multimodal perception fusion, avoiding overexposure in strong light and noise in low light that can affect subsequent data processing. This multi-sensor fusion overcomes the limitations of a single sensor and expands scenario usability.

[0066] Based on the real-time acquisition of multi-modal perception data, the corresponding multi-modal target parameters are analyzed and determined, including the illumination parameter and the three-dimensional face model of the driver. Among them, the illumination perception data is analyzed to obtain the main source direction of the current illumination (usually determined by the illumination vector), the light hardness (such as distinguishing direct light and diffuse reflection light) and other illumination parameters; based on the face image, face temperature distribution information and head depth information and other face perception data of the driver, multi-modal fusion modeling is carried out, and the key points of the face (such as eyes, nose and mouth) are recognized through deep learning algorithm, the face orientation, blinking frequency and whether wearing sunglasses and other key features are judged, so as to construct the three-dimensional face model of the driver. The three-dimensional face model can be used to determine the line of sight direction of the driver, predict the head posture position of the driver according to the head posture motion track, and at the same time detect whether there is an article on the head of the driver. For example, when the user wears sunglasses, the visual data, eye area depth, heat distribution and texture are different compared with the state of not wearing, at this time there is no need to adjust the light shield; when the user wears a hat on the head, the direct eye angle is different compared with the state of not wearing, at this time it is judged whether the hat brim angle interferes with the light incidence angle, if it is shielded, there is no need to adjust the light shield.

[0067] Further, the multi-modal target parameters are preprocessed, and the preprocessing operations include filtering processing, noise reduction processing and time synchronization, etc., so as to ensure that the multi-source data is accurate, reliable and synchronous, and to lay a foundation for subsequent fusion analysis. Key features are extracted from the preprocessed multi-modal target parameters through feature extraction algorithm, including illumination features (such as incident light vector) corresponding to illumination parameters, face posture features (such as face normal vector, face key point coordinate sequence) corresponding to three-dimensional face model, etc.

[0068] The correlation analysis is performed on the light feature and the facial pose feature, a correlation model between the light feature and the facial pose feature is established, and the target effective shielding length of the to-be-adjusted sun visor is determined through the correlation model analysis, that is, the length of the sun visor that needs to cover the face is calculated. In the correlation model, the corresponding incident angle is determined according to the incident light vector and the facial normal vector obtained based on the three-dimensional facial model, and the eye position coordinates of the driver are determined, and the target effective shielding length of the to-be-adjusted sun visor is calculated according to the incident angle and the eye position coordinates of the driver. Then, the target angle of the to-be-adjusted sun visor is determined according to the size information of the to-be-adjusted sun visor and the target effective shielding length. The size information of the to-be-adjusted sun visor includes the length of the to-be-adjusted sun visor, which is the size of the side of the sun visor that is not fixed (i.e. the free end), commonly the short side of a rectangular sun visor. In other embodiments, the correlation feature between the light feature and the facial pose feature can be learned by a neural network, and the target angle can be output in combination with the size parameters of the to-be-adjusted sun visor. Specifically, based on the light feature (including light intensity values in each direction) and the facial pose feature (including pitch angle, yaw angle and facial centroid coordinates), a two-dimensional coordinate mapping of the shielding area is determined, that is, the to-be-shielded area of the driver's face is projected onto a specific two-dimensional plane (such as the rotation plane of the sun visor itself or other virtual planes set based on actual needs), and a corresponding two-dimensional region representation is generated on the plane, so as to accurately represent the region to be shielded on the face by two-dimensional projection, and output the target angle of the to-be-adjusted sun visor based on the shielding area, the length of the sun visor and the current installation height, so as to automatically learn the shielding rule under complex light and adapt to changes in facial pose.

[0069] Finally, the target angle of the to-be-adjusted sun visor is controlled to drive the vehicle motor to adjust the angle of the sun visor, so as to achieve effective and accurate shielding, prevent direct sunlight from interfering with the driver's line of sight, and improve driving comfort and safety. It can be understood that the above sun visor adjustment method is also applicable to other passengers in the vehicle, for example, by adjusting the sun visor angle of the co-driver seat, the comfort of the corresponding passenger is improved.

[0070] It should be noted that on the in-vehicle central control screen or dedicated display screen, intuitive graphical interfaces can be used to display key information such as current lighting conditions, sun visor adjustment status, facial pose of in-vehicle personnel, etc., so that in-vehicle personnel can timely obtain system operation status and enhance system transparency and user trust. At the same time, in-vehicle personnel can adjust the sun visor angle through various interaction methods such as touch, voice, gesture, etc. according to their actual needs, and record the personalized preference settings of in-vehicle personnel (such as the preferred sun visor angle range, adjustment speed, etc. under different lighting scenarios), and the system will automatically refer to the historical preference data in the subsequent operation process to optimize the sun visor adjustment strategy, realizing personalized and intelligent sun visor adjustment service.

[0071] To improve safety, a photosensitive sensor is usually introduced in the existing automatic adjustment scheme to realize sun visor control by detecting light intensity, but there are problems such as response delay and insufficient adjustment accuracy, which makes it difficult to realize real-time and accurate response and easy to cause adjustment angle deviation.

[0072] Compared with the prior art, the present application determines the corresponding multi-modal target parameters based on the real-time acquired multi-modal perception data; the multi-modal target parameters include the illumination parameter and the three-dimensional face model of the driver; the illumination feature corresponding to the illumination parameter and the face posture feature corresponding to the three-dimensional face model are determined; the illumination feature and the face posture feature are associated and analyzed, and based on the association analysis result and the size information of the to-be-adjusted sun visor, the target angle of the to-be-adjusted sun visor is determined; and the to-be-adjusted sun visor is adjusted according to the target angle of the to-be-adjusted sun visor. Based on this, through the association analysis between the illumination feature and the face posture feature, the influence degree of the illumination change on the visual experience of the vehicle occupant is determined, so that the target overturning angle of the sun visor can be calculated according to the association analysis result, the problem of difficult real-time and accurate response and easy to cause sun visor adjustment lag or adjustment angle deviation is solved, real-time and accurate control of the adjustment of the sun visor of the vehicle based on multi-dimensional perception data is realized, adjustment lag or adjustment angle deviation is avoided to improve driving comfort and safety, and interference of direct sunlight on the driver's vision is effectively prevented.

[0073] In some embodiments, the association analysis of the illumination feature and the face posture feature in step S230, and the determination of the target angle of the to-be-adjusted sun visor based on the association analysis result and the size information of the to-be-adjusted sun visor, includes the following steps:

[0074] Step S231, an association model between the illumination feature and the face posture feature is established to determine the target effective shielding length of the to-be-adjusted sun visor through the association model analysis; wherein the illumination feature includes an incident light vector, and the face posture feature includes a face normal vector;

[0075] Step S232, the target angle of the to-be-adjusted sun visor is determined according to the size information of the to-be-adjusted sun visor and the target effective shielding length; the size information includes the length of the to-be-adjusted sun visor.

[0076] Specifically, the illumination feature and the face posture feature are associated and modeled to obtain an association model between the illumination feature and the face posture feature. Wherein the illumination feature includes an incident light vector, and the face posture feature includes a face normal vector, the association model performs eye shielding analysis based on the spatial relationship between the illumination direction and the face posture to determine the target effective shielding length of the to-be-adjusted sun visor, so as to realize the analysis of the influence degree of the illumination change on the visual experience of the driver through the association model.

[0077] It should be noted that the above correlation model can be combined with information such as light intensity distribution, spectral characteristics, light hardness and light source type to assist in occlusion analysis, optimize the occlusion analysis result, and improve the calculation accuracy of the target effective occlusion length.

[0078] Further, according to the length of the to-be-adjusted sun visor and the target effective occlusion length, the target angle of the to-be-adjusted sun visor is calculated, which is the to-be-flipped angle of the to-be-adjusted sun visor compared with the initial state, and the specific calculation formula is as follows:

[0079] (1)

[0080] In formula (1), d represents the target angle of the to-be-adjusted sun visor; L represents the target effective occlusion length of the to-be-adjusted sun visor; and A represents the length of the to-be-adjusted sun visor. In other embodiments, the target angle of the to-be-adjusted sun visor can be optimized in combination with information such as the real-time relative distance between the to-be-adjusted sun visor and the driver's face and the predicted head position of the driver, which helps to achieve accurate occlusion in a dynamic scene.

[0081] Through the embodiment, a correlation model between the light feature and the face posture feature is established to determine the target effective occlusion length of the to-be-adjusted sun visor through the correlation model analysis, the light feature includes an incident light vector, and the face posture feature includes a face normal vector. Then, according to the size information of the to-be-adjusted sun visor and the target effective occlusion length, the target angle of the to-be-adjusted sun visor is determined, the size information includes the length of the to-be-adjusted sun visor, so as to analyze the influence degree of light change on the visual experience of the vehicle passenger, determine the priority and target angle range of sun visor adjustment, and realize accurate calculation of the sun visor adjustment angle.

[0082] In some of the embodiments, the analysis process of the above correlation model includes the following steps:

[0083] According to the incident light vector and the face normal vector, the corresponding incident angle is determined;

[0084] The eye position coordinates of the driver are determined;

[0085] According to the incident angle and the eye position coordinates of the driver, the target effective occlusion length of the to-be-adjusted sun visor is determined.

[0086] Specifically, in the correlation analysis, according to the incident light vector I and the face normal vector N, the incident angle of the incident light is calculated , and the eye position coordinates of the driver are determined. Wherein, taking the nose tip or other approximate position (such as eyebrow center, etc.) of the driver as the origin, a two-dimensional coordinate system is established in the vertical plane, the X axis extends horizontally along the line connecting the two eyes, and the Y axis is vertically upward. The eye position coordinates are quantitatively calibrated through the coordinate system.

[0087] Then, according to the incident angle and the eye position coordinate of the driver, the target effective shielding length of the sun visor to be adjusted is calculated, and the specific calculation formula is as follows:

[0088] (2)

[0089] In formula (2), L represents the target effective shielding length of the sun visor to be adjusted, represents the incident angle; (h, s) represents the eye position coordinate of the driver.

[0090] Through the embodiment, the corresponding incident angle is determined according to the incident light vector and the face normal vector, the eye position coordinate of the driver is determined, and then the target effective shielding length of the sun visor to be adjusted is determined according to the incident angle and the eye position coordinate of the driver, so that the precise calculation of the target effective shielding length is realized.

[0091] In some embodiments, the multi-modal target parameter further includes a vehicle driving parameter; after the corresponding incident angle is determined according to the incident light vector and the face normal vector, the following steps are further included:

[0092] Based on the vehicle driving parameter, the vehicle heading angle change rate, the vehicle lateral acceleration and the light incidence change rate are determined; the vehicle driving parameter corresponds to the current driving state or the predicted driving state;

[0093] According to the vehicle heading angle change rate, the vehicle lateral acceleration and the light incidence change rate, the incident angle is dynamically anti-offset compensated.

[0094] In the embodiment, the multi-modal perception parameter collected in real time includes environmental perception data. The environmental perception module integrates an in-vehicle temperature sensor, a humidity sensor and a vehicle speed sensor, etc., and obtains environmental perception data such as in-vehicle environmental parameters (such as temperature, humidity) and vehicle driving states (such as speed, acceleration) through the environmental perception module. Among them, the in-vehicle environmental parameters are used to assist in analyzing the comfort demand of the driver, and the vehicle driving state is used to assist in analyzing the light change trend, so as to dynamically anti-offset compensate the incident angle calculated in real time. Dynamic anti-offset compensation refers to the influence of the body posture change during driving or in the future on the light incident angle, and the dynamic anti-offset compensation adopts a torque compensation algorithm based on the vehicle acceleration signal to suppress the bump interference.

[0095] Specifically, the vehicle driving state is analyzed to obtain vehicle driving parameters corresponding to the current driving state or the predicted driving state, and the vehicle heading angle change rate, the vehicle lateral acceleration and the light incidence change rate are calculated according to the vehicle driving parameters. Wherein, the current position and speed information of the vehicle are obtained through the Global Positioning System (GPS), and the vehicle acceleration is obtained through the Inertial Measurement Unit (IMU) sensor, the vehicle body posture after n seconds is predicted according to the measurement data of the GPS and the IMU sensor, and the corresponding sun visor adjustment operation is performed.

[0096] Then, the incident angle is dynamically anti-offset compensated according to the vehicle heading angle change rate, the vehicle lateral acceleration and the light incidence change rate, and the specific compensation algorithm is as follows:

[0097] (3)

[0098] In formula (3), represents the dynamic anti-offset compensation calculation result; represents the vehicle heading angle change rate; represents the vehicle lateral acceleration, which is used to calculate the steering compensation; represents the light incidence change rate; , and are calibration parameters. For example, when the lateral acceleration >0.3g is detected, the torque compensation algorithm is triggered to correct the angle error =1.2°.

[0099] As shown in Figure 4 , in the scenario that the vehicle is about to turn, the current position and speed information of the vehicle are obtained through the Global Positioning System (GPS), and the vehicle acceleration and vehicle yaw rate are obtained through the Inertial Measurement Unit (IMU) sensor. In the processing unit, the light incidence angle change amount after 3 seconds is predicted according to the measurement data of the GPS and the IMU sensor, the sun visor adjustment operation corresponding to the light incidence angle change amount is determined, and is sent to the actuator to complete the corresponding adjustment operation through the actuator, such as the light incidence angle change amount after 3 seconds is predicted to be 25°, the sun visor adjustment operation is to control the sun visor to be adjusted downward by 28°, so as to ensure that the light is just blocked when actually turning.

[0100] By this embodiment, based on the vehicle driving parameters corresponding to the current driving state or the predicted driving state, the vehicle heading angle change rate, the vehicle lateral acceleration and the light incidence change rate are determined, and the dynamic anti-offset compensation of the incident angle is performed according to the vehicle heading angle change rate, the vehicle lateral acceleration and the light incidence change rate, so as to avoid the angle offset of the electric sun visor in the bumpy road condition, and significantly improve the accuracy of the angle adjustment of the sun visor in the vehicle.

[0101] In some embodiments, the light feature includes an incident light vector; the determining, in step S220, the light feature corresponding to the light parameter includes the following steps:

[0102] In step S221, real-time multi-modal perception data is obtained; the multi-modal perception data includes an in-vehicle light image;

[0103] In step S222, the centroid coordinates of the light spot in the in-vehicle light image are converted into corresponding spherical coordinates, and a three-dimensional vector corresponding to the spherical coordinates is determined;

[0104] In step S223, based on the light refraction model of the vehicle glass and the three-dimensional vector, the corresponding incident light vector is determined.

[0105] Specifically, real-time multi-modal perception data is obtained, and the multi-modal perception data includes an in-vehicle light image. The centroid coordinates of the light spot in the in-vehicle light image are analyzed and determined, the centroid coordinates of the light spot are converted into corresponding spherical coordinates, and a three-dimensional vector corresponding to the spherical coordinates is calculated. Further, considering the existence of the car window glass refraction, the three-dimensional vector is operated based on the light refraction model of the vehicle glass, and finally the corresponding incident light vector is obtained.

[0106] Illustratively, through a vehicle-mounted fisheye lens capable of realizing ultra-wide angle coverage of omnidirectional light in the cabin (such as a field of view angle of 240° and a focal length f=1.4mm), a RAW format in-vehicle light image is obtained. The in-vehicle light image is pre-filtered, preferably multi-scale Gaussian filter denoising, and the filtered image is adaptively threshold segmented. Based on the threshold segmentation result, the light spot profile in the image is extracted, and the centroid coordinates of the light spot are calculated, so as to obtain the light center coordinates (u, v).

[0107] Since the image collected by the fisheye lens is a distorted image, the distortion coefficient is determined through camera calibration, the original image coordinates (u, v) are normalized to obtain the normalized coordinate system (x, y), and the distance between the pixel point and the light center is calculated . Distortion correction is performed on the image, that is, the position of the initial coordinates (x, y) when it is not distorted is calculated, and the specific calculation formula is as follows:

[0108] (4)

[0109] (5)

[0110] In formula (4) and formula (5), denotes a radial distortion coefficient; , denotes a tangential distortion coefficient; , is a corrected coordinate. It can be understood that, as shown in the above distortion correction process, the pixel (original point P) in the image deviating from the ideal position (distance r) due to distortion is corrected to the ideal position (distance r`≈r, modified point P`) without distortion according to the lens distortion law. Figure 3

[0111] Then, the corrected optical center coordinates are converted into corresponding spherical coordinates (x, y, z), and a three-dimensional vector corresponding to the spherical coordinates is calculated, and the specific calculation formula is as follows:

[0112] (6)

[0113] In formula (6), denotes a three-dimensional vector converted, the X axis points to the vehicle head, the Y axis points to the co-driver, and the coordinate system Z axis points to the vehicle roof.

[0114] Further, considering the existence of the vehicle window glass refraction, the three-dimensional vector is operated based on the light refraction model of the vehicle glass to obtain a corresponding incident light vector, and the specific calculation formula is as follows:

[0115] (7)

[0116] In formula (7), denotes the incident light vector finally calculated; denotes the refractive index of the vehicle window glass, which is measured according to the actual vehicle window glass; denotes the vehicle window normal vector.

[0117] Through the embodiment, real-time multi-modal perception data including an in-vehicle illumination image is obtained, the centroid coordinates of a light spot in the in-vehicle illumination image are converted into corresponding spherical coordinates, a three-dimensional vector corresponding to the spherical coordinates is determined, and the corresponding incident light vector is determined based on the light refraction model of the vehicle glass and the three-dimensional vector, so as to realize accurate calculation of the incident light vector and help to improve the accuracy of correlation analysis.

[0118] In some embodiments, the determining of the three-dimensional face model of the driver in step S220 includes the following steps:

[0119] ​Step S224: acquiring real-time multimodal sensing data; the multimodal sensing data includes the driver's facial image, facial temperature distribution information, and head depth information;

[0120] Step S225: establishing a corresponding head pose space coordinate system based on the driver's head depth information;

[0121] Step S226: Fusing the texture information and facial temperature distribution information in the facial image into the head pose space coordinate system to obtain a three-dimensional facial model.

[0122] In this embodiment, the facial perception module is equipped with a high-resolution camera, an infrared thermal imager, and a depth perception camera. The main controller controls the facial perception module to acquire facial perception data in real time. This includes capturing the driver's facial image with the high-resolution camera, acquiring facial temperature distribution information (i.e., thermal imaging data) with the infrared thermal imager, and acquiring head depth information (i.e., head depth image) with the depth perception camera.

[0123] Further, such as Figure 5 As shown in the figure, a corresponding head pose space coordinate system is established based on the driver's head depth information. For example, a head pose space coordinate system is constructed with the head's center of mass as the origin, the line connecting the two eyes as the X-axis, the vertical upward direction as the Y-axis, and the Z-axis pointing in the driving direction as the Z-axis. A facial key point algorithm is used to extract key feature points such as the eyes, nose, and lips from the two-dimensional image. Facial key feature points are then registered, and texture information (such as skin texture) in the facial image and the facial temperature distribution from infrared thermal imaging are mapped to the head pose space coordinate system. Ultimately, a three-dimensional facial model is generated that integrates geometric structure, visual texture, and thermal features.

[0124] Through this embodiment, real-time multimodal perception data is obtained, and the multimodal perception data includes the driver's facial image, facial temperature distribution information and head depth information. Based on the driver's head depth information, a corresponding head pose space coordinate system is established, and the texture information and facial temperature distribution information in the facial image are fused into the head pose space coordinate system to obtain a three-dimensional facial model, thereby achieving accurate head pose modeling, so that facial pose features can be subsequently extracted based on the three-dimensional facial model.

[0125] In some embodiments, the vehicle sun visor adjustment method further includes the following steps:

[0126] Determine a driving scenario corresponding to each scenario vector; the scenario vector includes spatiotemporal features, dynamic environment perception features, and driver features;

[0127] The adjustment strategy of the to-be-adjusted sun visor is reinforced learning based on different driving scenes, and a target adjustment strategy corresponding to each driving scene is obtained; the adjustment strategy includes an adjustment angle, an adjustment speed, and an adjustment time of the to-be-adjusted sun visor.

[0128] A mapping library between the driving scene and the target adjustment strategy is constructed.

[0129] Specifically, in order to adapt to the sun visor adjustment in different driving scenes, different driving scenes are defined in advance according to scene vectors, and the scene vectors include space-time features, dynamic environment perception features, and driver features. The space-time features include latitude, longitude, time, and road type (which can be obtained based on high-precision map information), such as setting the space-time feature S_base = [lat, lon, t, road_type], where lat represents latitude, lon represents longitude, t represents time, and road_type represents road type. The actually obtained space-time feature can be S_base = [34.0522 ° N, 118.2437 ° W, 12:00, highway]. The dynamic environment perception features include vehicle speed, vehicle acceleration, light direction, and light intensity, such as setting the dynamic environment perception feature S_dynamic = [v, a, θ_light, brightness], where v represents speed, a represents acceleration, θ_light represents light direction, and brightness represents light intensity. The actually obtained dynamic environment perception feature can be S_base = [100 km / h, 0.3 g, 15 °, 10 lux]. The driver features include face orientation, head turning angle, head status (such as whether wearing sunglasses and a hat), and historical preference, such as the driver feature S_user = [pitch, yaw, status, pref_sensitivity], where pitch represents face orientation, yaw represents head turning angle, status represents head status, and pref_sensitivity represents historical preference.

[0130] Further, the adjustment strategy of the to-be-adjusted sun visor is reinforced learning based on different driving scenes, and a target adjustment strategy corresponding to each driving scene is obtained. It can be understood that the adjustment strategy includes an adjustment angle, an adjustment speed, and an adjustment time, and the angle adjustment strategy of the to-be-adjusted sun visor can correspond to the sun visor adjustment method in any one of the above embodiments. For example, in the highway cruising scene, the system predicts the change of the sun azimuth angle in the next 5 minutes, and at this time, the sun visor is adjusted gently at a speed of 0.5 ° / s according to the prediction result, and the advance is calculated based on the vehicle speed of 100 km / h.

[0131] Afterwards, a mapping library between the driving scene and the target adjustment strategy is constructed, and different driving scenes in the mapping library can be defined based on one or more combinations of the spatiotemporal features, the dynamic environment perception features and the driver features. For example, the high-speed strong backlight scene is defined based on the spatiotemporal features and the dynamic environment perception features, the spatiotemporal features of which include the noon time and the highway, and the dynamic environment perception features of which include the high-speed driving (such as v>100km / h) and the backlight driving (such as the light incident angle θ<20°), and the corresponding target adjustment strategy is to lower the sun visor by 28° to avoid the strong backlight directly shining into the eyes; the city tree shadow driving scene is defined based on the spatiotemporal features and the dynamic environment perception features, the spatiotemporal features of which include the urban road and the afternoon time, and the dynamic environment perception features of which include the large acceleration change and the large light intensity fluctuation frequency, and the corresponding target adjustment strategy is to delay the dimming response by more than 1 second to suppress the flicker caused by repeated adjustment; the night opposite light driving scene is defined based on the spatiotemporal features, the dynamic environment perception features and the driver features, the spatiotemporal features of which include the night driving, the dynamic environment perception features of which include the backlight driving and the instantaneous increase of the light intensity, and the driver features of which include not wearing sunglasses and being used to complete blocking, and the corresponding target adjustment strategy is to respond quickly and adopt full blocking to effectively prevent glare.

[0132] Through the embodiment, the driving scene corresponding to each scene vector is determined, the scene vector includes the spatiotemporal features, the dynamic environment perception features and the driver features, the adjustment strategy of the to-be-adjusted sun visor is reinforced learning based on different driving scenes, the target adjustment strategy corresponding to each driving scene is obtained, wherein the adjustment strategy includes the adjustment angle, the adjustment speed and the adjustment timing of the to-be-adjusted sun visor, and then the mapping library between the driving scene and the target adjustment strategy is constructed, so as to combine the light change law and the driver visual demand characteristics in different driving scenes (such as the urban road, the highway and the rural road), use the intelligent decision algorithm based on the reinforcement learning to formulate the sun visor adjustment strategy, reduce the misjudgment rate in the complex light scene, ensure the smooth, accurate and efficient sun visor adjustment process, and minimize the interference to the driver's line of sight. At the same time, by constructing the mapping library, the appropriate adjustment strategy can be retrieved in time, and the efficiency of the adjustment decision is improved.

[0133] In some embodiments, after the mapping library between the driving scene and the target adjustment strategy is constructed, the above adjustment method of the sun visor of the vehicle further includes the following steps:

[0134] determining the target adjustment strategy corresponding to the real-time driving scene in the mapping library;

[0135] controlling the vehicle motor to drive the to-be-adjusted sun visor to adjust the angle according to the target adjustment strategy corresponding to the real-time driving scene;

[0136] determining the actual angle adjustment information of the to-be-adjusted sun visor based on the real-time running state of the vehicle motor.

[0137] Based on the actual angle adjustment information, the target adjustment strategy corresponding to the real-time driving scene is dynamically optimized.

[0138] Specifically, based on the real-time driving scene, the corresponding target adjustment strategy is called from the mapping library, and the vehicle motor is controlled to drive the to-be-adjusted sun visor to adjust the angle according to the target adjustment strategy corresponding to the real-time driving scene. At the same time, the running state of the vehicle motor is monitored in real time, and the real-time running state of the vehicle motor includes speed, torque, position feedback, etc.

[0139] Further, according to the real-time running state of the vehicle motor, the actual angle adjustment information of the to-be-adjusted sun visor is analyzed and determined, and the actual angle adjustment information of the to-be-adjusted sun visor is fed back to the adjustment strategy planning module. Based on the actual angle adjustment information, the target adjustment strategy corresponding to the real-time driving scene is dynamically optimized, so as to form a closed loop control, realize accurate control and dynamic compensation of the sun visor angle, and ensure that the sun visor is always in the best shading position.

[0140] Through the embodiment, the target adjustment strategy corresponding to the real-time driving scene in the mapping library is determined, the vehicle motor is controlled to drive the to-be-adjusted sun visor to adjust the angle according to the target adjustment strategy corresponding to the real-time driving scene, the actual angle adjustment information of the to-be-adjusted sun visor is determined based on the real-time running state of the vehicle motor, and the target adjustment strategy corresponding to the real-time driving scene is dynamically optimized based on the actual angle adjustment information. In this way, the dynamic optimization of the adjustment strategy is realized, and the accuracy of the adjustment strategy is improved.

[0141] The following will be described and explained in combination with Figure 6 , through specific embodiments.

[0142] In the vehicle sun visor adjustment scene, multi-modal perception data is collected in real time, the multi-modal perception data includes illumination perception data, face perception data and environment perception data, etc., the illumination perception data, the face perception data and the environment perception data are data-synchronized, and feature extraction is performed on the multi-modal perception data after data synchronization. Among them, the illumination perception data includes an in-vehicle illumination image collected through a vehicle-mounted camera, the centroid coordinates of a light spot in the in-vehicle illumination image are analyzed and determined, the centroid coordinates of the light spot are converted into corresponding spherical coordinates, a three-dimensional vector corresponding to the spherical coordinates is calculated, and a corresponding incident light vector is obtained by operating the three-dimensional vector based on a light refraction model of the vehicle glass; the face image, the face temperature distribution information and the head depth information of the driver are obtained in real time through the face perception module, and a three-dimensional face model of the driver is constructed by performing multi-modal fusion modeling based on the face image, the face temperature distribution information and the head depth information.

[0143] Further, based on the three-dimensional face model, the face normal vector of the driver is determined, the incident light vector and the face normal vector are analyzed, the correlation model between the illumination feature and the face posture feature is established, and the target effective shielding length of the sun visor to be adjusted is determined through the correlation model analysis, that is, the length of the sun visor covering the face is calculated. In the correlation model, the incident angle of the light is calculated according to the incident light vector and the face normal vector, at this time, the dynamic anti-offset compensation of the incident angle is performed according to the vehicle heading angle change rate, the vehicle lateral acceleration and the illumination incident change rate, and the target effective shielding length of the sun visor to be adjusted is calculated according to the dynamic compensation incident angle and the eye position coordinates of the driver.

[0144] Finally, the target angle of the sun visor to be adjusted is calculated according to the length of the sun visor to be adjusted and the target effective shielding length, the sun visor decision is completed, and the angle adjustment of the sun visor to be adjusted is controlled according to the target angle of the sun visor to be adjusted, so as to realize precise sun visor adjustment.

[0145] It should be noted that the steps shown in the above process or the flowchart of the accompanying drawings can be executed in a computer system such as a group of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0146] An adjusting device of a vehicle sun visor is also provided in the embodiment, which is used to realize the above-mentioned embodiments and preferred embodiments, and will not be described again. The terms "module", "unit", "sub-unit" and the like used below can be a combination of software and / or hardware that realizes a predetermined function. Although the device described in the following embodiments is preferably realized in software, realization of hardware, or a combination of software and hardware, is also possible and conceived.

[0147] Figure 7 is a structural block diagram of the adjusting device of the vehicle sun visor of the embodiment, as Figure 7 shown, the device comprises:

[0148] A determination module 10 is configured to determine corresponding multi-modal target parameters based on real-time acquired multi-modal perception data; the multi-modal target parameters comprise a light parameter and a three-dimensional face model of a driver;

[0149] An extraction module 20 is configured to determine a light feature corresponding to the light parameter and a face posture feature corresponding to the three-dimensional face model;

[0150] An analysis module 30 is configured to perform correlation analysis on the light feature and the face posture feature, and determine a target angle of a sun visor to be adjusted based on a correlation analysis result and size information of the sun visor to be adjusted;

[0151] A control module 40 is configured to adjust the sun visor to be adjusted according to the target angle of the sun visor to be adjusted.

[0152] Through the device provided in the embodiment, by determining corresponding multi-modal target parameters based on real-time acquired multi-modal perception data; the multi-modal target parameters comprise a light parameter and a three-dimensional face model of a driver; determining a light feature corresponding to the light parameter and a face posture feature corresponding to the three-dimensional face model; performing correlation analysis on the light feature and the face posture feature, and determining a target angle of a sun visor to be adjusted based on a correlation analysis result and size information of the sun visor to be adjusted; adjusting the sun visor to be adjusted according to the target angle of the sun visor to be adjusted, the problem that it is difficult to realize real-time accurate response and the sun visor is prone to adjustment lag or adjustment angle deviation is solved, real-time accurate control of adjustment of the vehicle sun visor based on multi-dimensional perception data is realized, adjustment lag or adjustment angle deviation is avoided to improve driving comfort and safety, and interference of direct sunlight on the driver's line of sight is effectively prevented.

[0153] In some embodiments, the analysis module 30 is further configured to establish a correlation model between the light feature and the face posture feature, to determine a target effective blocking length of the sun visor to be adjusted by correlation model analysis; wherein the light feature comprises an incident light vector, and the face posture feature comprises a face normal vector; determine a target angle of the sun visor to be adjusted according to the size information of the sun visor to be adjusted and the target effective blocking length; the size information comprises the length of the sun visor to be adjusted.

[0154] In some embodiments, the analysis module 30 is further configured to determine a corresponding incident angle according to the incident light vector and the face normal vector; determine the eye position coordinates of the driver; determine the target effective blocking length of the sun visor to be adjusted according to the incident angle and the eye position coordinates of the driver.

[0155] In some embodiments, the analysis module 30 is further configured to determine the vehicle heading angle change rate, the vehicle lateral acceleration and the light incidence change rate based on the vehicle driving parameters; the vehicle driving parameters correspond to the current driving state or the predicted driving state; and dynamically anti-offset compensate the incident angle according to the vehicle heading angle change rate, the vehicle lateral acceleration and the light incidence change rate.

[0156] In some embodiments, the extraction module 20 is further configured to obtain real-time multi-modal perception data; the multi-modal perception data comprises an in-vehicle light image; convert the centroid coordinates of the light spot in the in-vehicle light image into corresponding spherical coordinates, and determine a three-dimensional vector corresponding to the spherical coordinates; determine a corresponding incident light vector based on a light refraction model of the vehicle glass and the three-dimensional vector.

[0157] In some embodiments, the extraction module 20 is further configured to obtain real-time multi-modal perception data; the multi-modal perception data comprises a driver's face image, face temperature distribution information and head depth information; establish a corresponding head pose space coordinate system based on the driver's head depth information; fuse the texture information in the face image and the face temperature distribution information into the head pose space coordinate system to obtain a three-dimensional face model.

[0158] In some embodiments, on the basis of Figure 7 The device further comprises an adaptive module for determining a driving scene corresponding to each scene vector; the scene vector comprises a space-time feature, a dynamic environment perception feature and a driver feature; based on different driving scenes, reinforcement learning is performed on the adjustment strategy of the sun visor to be adjusted to obtain a target adjustment strategy corresponding to each driving scene; the adjustment strategy comprises an adjustment angle, an adjustment speed and an adjustment timing of the sun visor to be adjusted; a mapping library between the driving scene and the target adjustment strategy is constructed.

[0159] In some embodiments, on the basis of Figure 7On the basis of the above, the device further comprises a dynamic optimization module, configured to determine a target adjustment strategy corresponding to the real-time driving scene in the mapping library; control the vehicle motor to drive the sun visor to be adjusted to adjust the angle according to the target adjustment strategy corresponding to the real-time driving scene; determine actual angle adjustment information of the sun visor to be adjusted based on the real-time running state of the vehicle motor; and dynamically optimize the target adjustment strategy corresponding to the real-time driving scene based on the actual angle adjustment information.

[0160] It should be noted that each of the above modules can be a functional module or a program module, and can be implemented by software or hardware. For the modules implemented by hardware, each of the above modules can be located in the same processor; or each of the above modules can also be located in different processors in any combination.

[0161] In the embodiment, a vehicle is also provided, comprising a sun visor and the sun visor adjustment device in any of the above device embodiments.

[0162] In the embodiment, a computer device is also provided, comprising a memory and a processor, the memory storing a computer program, and the processor being configured to execute the computer program to perform the steps in any of the above method embodiments.

[0163] Optionally, the computer device can further comprise a transmission device and an input and output device, wherein the transmission device is connected with the processor, and the input and output device is connected with the processor.

[0164] Optionally, in the embodiment, the processor can be configured to execute the following steps through the computer program:

[0165] S1, determining corresponding multi-modal target parameters based on the real-time acquired multi-modal perception data; the multi-modal target parameters comprise illumination parameters and a three-dimensional face model of the driver;

[0166] S2, determining illumination features corresponding to the illumination parameters and face posture features corresponding to the three-dimensional face model;

[0167] S3, performing correlation analysis on the illumination features and the face posture features, and determining a target angle of the sun visor to be adjusted based on the correlation analysis result and size information of the sun visor to be adjusted;

[0168] S4, adjusting the sun visor to be adjusted according to the target angle of the sun visor to be adjusted.

[0169] It should be noted that the specific examples in the embodiment can refer to the examples described in the above embodiments and optional implementation manners, which will not be described herein again.

[0170] In addition, in combination with the adjusting method of the vehicle sun visor provided in the above-mentioned embodiments, a storage medium can also be provided in this embodiment to implement the adjusting method of the vehicle sun visor. The storage medium has a computer program stored thereon; and the computer program is executed by a processor to implement any one of the adjusting methods of the vehicle sun visor in the above-mentioned embodiments.

[0171] In this embodiment, a computer program product is also provided, which includes a computer program. The computer program is executed by a processor to implement the steps in the above-mentioned method embodiments.

[0172] It should be understood that the specific embodiments described herein are merely exemplary and are not intended to limit the application. Based on the embodiments provided in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of the present application.

[0173] Obviously, the drawings are only some examples or embodiments of the present application, and can be applied to other similar situations without creative efforts for those of ordinary skill in the art. In addition, it can be understood that although the work done in the development process can be complex and long, certain design, manufacture or production changes made by those of ordinary skill in the art based on the technical content disclosed in the present application are only routine technical means and should not be regarded as insufficient disclosure of the present application.

[0174] The term "embodiment" in the present application means that the specific features, structures or characteristics described in connection with the embodiment can be included in at least one embodiment of the present application. The presence of this phrase in various places in the specification does not necessarily mean the same embodiment, nor does it mean independence or alternatives to other embodiments. It can be clearly or implicitly understood by those of ordinary skill in the art that the embodiments described in the present application can be combined without conflict.

[0175] The above-described embodiments only express several implementation manners of the present application, which are described in detail and specifically, but should not be understood as limitations on the scope of patent protection. It should be noted that for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, which are all within the scope of protection of the present application. Therefore, the scope of protection of the present application should be subject to the appended claims.

Claims

1. A method for adjusting a vehicle sun visor, characterized in that: include: Determine the corresponding multimodal target parameters based on the multimodal perception data acquired in real time; The multimodal target parameters include illumination parameters and a three-dimensional facial model of the driver; determining lighting features corresponding to the lighting parameters and facial posture features corresponding to the three-dimensional facial model; performing a correlation analysis on the illumination feature and the facial posture feature, and determining a target angle of the light shield to be adjusted based on the correlation analysis result and the size information of the light shield to be adjusted; The shading plate to be adjusted is adjusted according to the target angle of the shading plate to be adjusted.

2. The method for adjusting a vehicle sun visor according to claim 1, wherein: The performing correlation analysis on the illumination feature and the facial posture feature, and determining a target angle of the visor to be adjusted based on the correlation analysis result and the size information of the visor to be adjusted, includes: Establishing a correlation model between the illumination feature and the facial posture feature, so as to determine a target effective shading length of the visor to be adjusted through analysis of the correlation model; wherein the illumination feature includes an incident light vector, and the facial posture feature includes a facial normal vector; The target angle of the light shielding plate to be adjusted is determined according to the size information of the light shielding plate to be adjusted and the target effective blocking length; the size information includes the length of the light shielding plate to be adjusted.

3. The method for adjusting a vehicle sun visor according to claim 2, wherein: The analysis process of the association model includes: Determining a corresponding incident angle according to the incident light vector and the facial normal vector; Determining the eye position coordinates of the driver; The target effective shielding length of the sun visor to be adjusted is determined according to the incident angle and the eye position coordinates of the driver.

4. The method for adjusting a vehicle sun visor according to claim 3, wherein: The multimodal target parameters also include vehicle driving parameters; after determining the corresponding incident angle according to the incident light vector and the facial normal vector, the method further includes: Determining a vehicle heading angle change rate, a vehicle lateral acceleration, and a light incidence change rate based on the vehicle driving parameters, wherein the vehicle driving parameters correspond to a current driving state or a predicted driving state; Dynamic anti-drift compensation is performed on the incident angle according to the vehicle heading angle change rate, the vehicle lateral acceleration and the light incidence change rate.

5. The method for adjusting a vehicle sun visor according to claim 1, wherein: The illumination feature includes an incident light vector; and determining the illumination feature corresponding to the illumination parameter includes: Acquiring real-time multimodal sensing data; the multimodal sensing data includes an interior lighting image of the vehicle; Converting the centroid coordinates of the light spot in the interior illumination image into corresponding spherical coordinates, and determining a three-dimensional vector corresponding to the spherical coordinates; Based on the light refraction model of the vehicle glass and the three-dimensional vector, a corresponding incident light vector is determined.

6. The method for adjusting a vehicle sun visor according to claim 1, wherein: Determining a three-dimensional facial model of the driver, comprising: Acquiring real-time multimodal perception data; the multimodal perception data includes the driver's facial image, facial temperature distribution information, and head depth information; Based on the head depth information of the driver, establishing a corresponding head pose space coordinate system; The texture information in the facial image and the facial temperature distribution information are fused into the head pose space coordinate system to obtain the three-dimensional facial model.

7. The method for adjusting a vehicle sun visor according to claim 1, wherein: The method further comprises: Determining a driving scenario corresponding to each scenario vector; the scenario vector comprising spatiotemporal features, dynamic environment perception features, and driver features; Based on different driving scenarios, reinforcement learning is performed on the adjustment strategy of the sun visor to be adjusted to obtain a target adjustment strategy corresponding to each driving scenario; the adjustment strategy includes the adjustment angle, adjustment speed and adjustment timing of the sun visor to be adjusted; A mapping library between the driving scenario and the target adjustment strategy is constructed.

8. The method for adjusting a vehicle sun visor according to claim 7, wherein: After constructing the mapping library between the driving scenario and the target adjustment strategy, the method further includes: determining the target adjustment strategy corresponding to the real-time driving scenario in the mapping library; According to the target adjustment strategy corresponding to the real-time driving scenario, controlling the vehicle motor to drive the sunshade to be adjusted to adjust the angle; Determining actual angle adjustment information of the sunshade to be adjusted based on the real-time operating status of the vehicle motor; Based on the actual angle adjustment information, the target adjustment strategy corresponding to the real-time driving scenario is dynamically optimized.

9. An adjusting device for a vehicle sun visor, characterized in that: include: A determination module, configured to determine corresponding multimodal target parameters based on multimodal perception data acquired in real time; The multimodal target parameters include illumination parameters and a three-dimensional facial model of the driver; an extraction module for determining illumination features corresponding to the illumination parameters and facial posture features corresponding to the three-dimensional facial model; an analysis module, configured to perform correlation analysis on the illumination feature and the facial posture feature, and determine a target angle of the visor to be adjusted based on the correlation analysis result and the size information of the visor to be adjusted; A control module is used to adjust the shading plate to be adjusted according to the target angle of the shading plate to be adjusted.

10. A computer device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to perform the steps of the vehicle sun visor adjustment method according to any one of claims 1 to 8.

Images