A method and device for controlling a vehicle sunroof

Abstract

This invention discloses a vehicle sunroof control method and apparatus. The method includes acquiring in-vehicle image data in real time when the vehicle enters sunroof control mode. The image data includes at least driver and non-driver action image data. The image data is then input into a preset human behavior model for recognition to obtain head action recognition results for each occupant. The head action recognition results are processed to determine a sunroof control signal representing the sunroof control intention. The sunroof is then controlled to perform an action associated with the sunroof control signal. The vehicle sunroof control method and apparatus provided by this invention, through image analysis of the human behavior of occupants, obtains head action recognition results, thereby providing accurate data support for sunroof control and advancing the intelligent control process of vehicles.

Description

Technical Field

[0001] This invention relates to the field of vehicle control technology, and in particular to a method and device for controlling a vehicle sunroof. Background Technology

[0002] As people's demands for vehicle comfort increase, and as automakers strive to further enhance vehicle value, sunroofs have become a standard feature in an growing number of vehicles. When open for ventilation, sunroofs not only help quickly remove odors from the car interior, but also rapidly lower the interior temperature after being exposed to direct sunlight. Furthermore, they provide an effective way to gaze at the stars while driving at night.

[0003] Currently, vehicle sunroofs are mainly divided into two categories: manual sunroofs and electric sunroofs. Manual sunroofs require manual operation to open and close, which is inconvenient and lacks a smart experience. Although electric sunroofs can be controlled by switches or voice commands, the switch operation is inconvenient for occupants who cannot reach the sunroof switch, while voice control suffers from slow recognition speed, significant noise interference, and inaccurate recognition of regional accents.

[0004] Therefore, how to intelligently control vehicle sunroofs has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] This invention provides a method and device for controlling a vehicle sunroof. By performing image analysis on the human behavior of occupants inside the vehicle, the head movement recognition results are obtained, thereby providing accurate data support for the control of the vehicle sunroof. The vehicle sunroof opens when the head is tilted up and closes when the head is tilted down, thus advancing the intelligent control process of the vehicle.

[0006] To address the aforementioned technical problems, embodiments of the present invention provide a vehicle sunroof control method, comprising:

[0007] When the vehicle enters the sunroof control mode, real-time image data inside the vehicle is acquired, wherein the image data includes at least driver action image data and non-driver action image data;

[0008] The image data is input into a preset human behavior model for recognition to obtain the head movement recognition results of each person in the vehicle.

[0009] The head movement recognition results of each person in the vehicle are processed to determine the sunroof control signal used to represent the sunroof control intention;

[0010] Control the vehicle sunroof to perform actions associated with the sunroof control signal.

[0011] As one preferred embodiment, if the real-time luminance value inside the vehicle is detected to be greater than a preset first threshold, the vehicle is controlled to enter the sunroof control mode; or,

[0012] If the detected real-time temperature inside the vehicle exceeds a preset second threshold, the vehicle is controlled to enter the sunroof control mode; or,

[0013] If a sunroof control signal issued by the user is detected, the vehicle is controlled to enter the sunroof control mode.

[0014] As one preferred embodiment, the construction of the pre-defined human behavioral model specifically includes:

[0015] Acquire initial sample image data for various scenarios inside the vehicle, wherein the initial sample image data includes at least image data of unoccupied state, image data of full-occupancy state, image data of non-full-occupancy state, image data of driver looking up, image data of driver looking down, image data of non-driver looking up and non-driver looking down.

[0016] The initial sample image data was labeled using the LabelImg tool to obtain labels for images looking up and images looking down.

[0017] The labeled initial sample image data is divided into training and test sets;

[0018] Convolutional neural networks for human motion recognition are constructed using Faster RCNN, SSD, or Yolo V1-V7 neural network technologies.

[0019] The human motion recognition convolutional neural network is trained based on the training set, and the trained human motion recognition convolutional neural network is validated based on the test set to obtain the human behavior model.

[0020] As one preferred embodiment, the processing of the head movement recognition results of each occupant in the vehicle to determine the sunroof control signal representing the sunroof control intention specifically includes:

[0021] The head movement recognition results of each person in the vehicle are input into the preset sunroof control calculation formula to obtain the sunroof control logic result;

[0022] If the sunroof control logic result is within the first preset range, a first sunroof control signal is generated to control the sunroof to open; if the sunroof control logic result is within the second preset range, a second sunroof control signal is generated to control the sunroof to close.

[0023] As one preferred embodiment, the sunroof control calculation formula is:

[0024] S = A × s1 + B × s2 + C × s3

[0025] Wherein, S is the sunroof control logic result, A is the driver's seat weight coefficient, A = 0.37, s1 is the driver's head movement recognition logic result, B is the passenger seat weight coefficient, B = 0.32, s2 is the passenger's head movement recognition logic result, C is the rear seat weight coefficient, C = 0.31, and s3 is the rear seat occupant's head movement recognition logic result.

[0026] Another embodiment of the present invention provides a vehicle sunroof control device, comprising:

[0027] The data acquisition module is used to acquire in-vehicle image data in real time when the vehicle enters the sunroof control mode, wherein the image data includes at least driver action image data and non-driver action image data.

[0028] The motion recognition module is used to input the image data into a preset human behavior model for recognition, and obtain the head motion recognition results of each person in the vehicle.

[0029] The control signal module is used to process the head movement recognition results of each person in the vehicle to determine the sunroof control signal used to represent the sunroof control intention.

[0030] The action execution module is used to control the vehicle sunroof to perform actions associated with the sunroof control signal.

[0031] As one preferred embodiment, if the real-time luminance value inside the vehicle is detected to be greater than a preset first threshold, the vehicle is controlled to enter the sunroof control mode; or,

[0032] If the detected real-time temperature inside the vehicle exceeds a preset second threshold, the vehicle is controlled to enter the sunroof control mode; or,

[0033] If a sunroof control signal issued by the user is detected, the vehicle is controlled to enter the sunroof control mode.

[0034] As one preferred embodiment, the construction of the pre-defined human behavioral model specifically includes:

[0035] Acquire initial sample image data for various scenarios inside the vehicle, wherein the initial sample image data includes at least image data of unoccupied state, image data of full-occupancy state, image data of non-full-occupancy state, image data of driver looking up, image data of driver looking down, image data of non-driver looking up and non-driver looking down.

[0036] The initial sample image data was labeled using the LabelImg tool to obtain labels for images looking up and images looking down.

[0037] The labeled initial sample image data is divided into training and test sets;

[0038] Convolutional neural networks for human motion recognition are constructed using Faster RCNN, SSD, or Yolo V1-V7 neural network technologies.

[0039] The human motion recognition convolutional neural network is trained based on the training set, and the trained human motion recognition convolutional neural network is validated based on the test set to obtain the human behavior model.

[0040] As one preferred embodiment, the processing of the head movement recognition results of each occupant in the vehicle to determine the sunroof control signal representing the sunroof control intention specifically includes:

[0041] The head movement recognition results of each person in the vehicle are input into the preset sunroof control calculation formula to obtain the sunroof control logic result;

[0042] If the sunroof control logic result is within the first preset range, a first sunroof control signal is generated to control the sunroof to open; if the sunroof control logic result is within the second preset range, a second sunroof control signal is generated to control the sunroof to close.

[0043] As one preferred embodiment, the sunroof control calculation formula is:

[0044] S = A × s1 + B × s2 + C × s3

[0045] Wherein, S is the sunroof control logic result, A is the driver's seat weight coefficient, A = 0.37, s1 is the driver's head movement recognition logic result, B is the passenger seat weight coefficient, B = 0.32, s2 is the passenger's head movement recognition logic result, C is the rear seat weight coefficient, C = 0.31, and s3 is the rear seat occupant's head movement recognition logic result.

[0046] Compared with the prior art, the beneficial effects of the embodiments of the present invention are at least one of the following:

[0047] (1) First, when the vehicle enters the sunroof control mode, real-time image data inside the vehicle is acquired; then the image data is input into a preset human behavior model for recognition to obtain the head movement recognition results of each person in the vehicle; then the head movement recognition results of each person in the vehicle are processed to determine the sunroof control signal used to represent the sunroof control intention; finally, the vehicle sunroof is controlled to perform the action associated with the sunroof control signal.

[0048] (2) Passengers do not need to control the sunroof manually, by switching on or by voice. Instead, the head movements during the process of looking at the sunroof are directly associated with the sunroof control. This allows the sunroof to be opened when looking up and closed when looking down. This avoids the problem of the sunroof being difficult to control due to the distance of the switch. In addition, the target recognition algorithm based on convolutional neural network is faster than voice recognition and is less affected by tire noise, wind noise and local accents during high-speed driving. Therefore, it is more stable and versatile. Attached Figure Description

[0049] Figure 1 This is a flowchart illustrating a vehicle sunroof control method in one embodiment of the present invention.

[0050] Figure 2 This is a schematic diagram of the structure of a vehicle sunroof control device in one embodiment of the present invention;

[0051] Figure label:

[0052] Among them, 11 is the data acquisition module; 12 is the action recognition module; 13 is the control signal module; and 14 is the action execution module. Detailed Implementation

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0054] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0055] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components. The terms "vertical," "horizontal," "left," "right," "upper," "lower," and similar expressions used herein are for illustrative purposes only and do not indicate or imply that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0056] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0057] One embodiment of the present invention provides a method for controlling a vehicle sunroof; for details, please refer to [link / reference]. Figure 1 , Figure 1 The diagram shown is a flowchart of a vehicle sunroof control method according to one embodiment of the present invention, which specifically includes steps S1 to S4:

[0058] S1. When the vehicle enters the sunroof control mode, real-time image data inside the vehicle is acquired, wherein the image data includes at least driver action image data and non-driver action image data.

[0059] S2. Input the image data into a preset human behavior model for recognition to obtain the head movement recognition results of each person in the vehicle;

[0060] S3. Process the head movement recognition results of each person in the vehicle to determine the sunroof control signal used to represent the sunroof control intention;

[0061] S4. Control the vehicle sunroof to perform actions associated with the sunroof control signal.

[0062] It should be noted that, to avoid issues such as the sunroof switch being too far away, difficult for users to reach, and unstable voice recognition leading to sunroof control failures and poor intelligent user experience, this invention associates the user's head movements with sunroof control. This allows the sunroof to open when the user looks up and close when the user looks down, thus achieving intelligent automatic control. The in-vehicle image data is acquired by in-vehicle cameras; the number, location, and specific model of these cameras are determined by the actual vehicle model and will not be elaborated upon here.

[0063] The sunroof control mode can be activated by the user or automatically by the vehicle. For example, a user can enter the sunroof control mode by pressing the sunroof control mode button on the vehicle, thereby completing steps S1 to S4 above. Alternatively, a user can generate relevant control signals via a mobile terminal such as a smartphone or tablet to control the vehicle to enter the sunroof control mode. In the scenario where the vehicle automatically enters the sunroof control mode, this could be triggered by the vehicle's light sensor detecting a real-time light intensity value inside the vehicle exceeding a preset first threshold, or by the vehicle's temperature sensor detecting a real-time temperature value inside the vehicle exceeding a preset second threshold.

[0064] In the above embodiments, the human behavioral model specifically includes:

[0065] First, a large number of in-vehicle images under different scenarios are collected using in-vehicle cameras, serving as initial sample image data for each scenario. This initial sample image data includes at least the following: images of an unoccupied vehicle, images of a fully occupied vehicle, images of a partially occupied vehicle, images of the driver looking up, images of the driver looking down, images of non-drivers looking up, and images of non-drivers looking down.

[0066] Then, using LabelImg or other target recognition and annotation tools, the head area of ​​the person looking up in the in-vehicle camera image is annotated to obtain label data.

[0067] Next, the labeled initial sample image data is divided into a training set and a test set. In this embodiment, preferably, 80% of the sample data is used as the training set and 20% as the validation set, thereby completing the creation of the head tilting and head lowering action recognition dataset.

[0068] Finally, a human motion recognition convolutional neural network is constructed based on Faster RCNN, SSD, or YOLO V1-V7 neural network technology (other related neural network technologies can also be selected, which will not be elaborated here); the human motion recognition convolutional neural network is trained based on the training set, and the trained human motion recognition convolutional neural network is verified based on the test set to obtain the human behavior model.

[0069] The construction of the aforementioned human behavior model can be completed autonomously by the vehicle. For example, during normal vehicle use, the image data collected by the in-vehicle camera can be transmitted to the vehicle's domain controller, such as the Vehicle Dynamics Control System (VDCM), via Ethernet or USB interface. The relevant processor will then process and analyze the data to obtain a model that can be used in real-world scenarios.

[0070] Furthermore, in the above embodiments, the processing of the head movement recognition results of each occupant in the vehicle to determine the sunroof control signal representing the sunroof control intention specifically includes:

[0071] The head movement recognition results of each person in the vehicle are input into the preset sunroof control calculation formula to obtain the sunroof control logic result;

[0072] If the sunroof control logic result falls within a first preset range, a first sunroof control signal is generated to control the sunroof opening; if the sunroof control logic result falls within a second preset range, a second sunroof control signal is generated to control the sunroof closing. By quantifying the processing results, it is easier to accurately analyze the user's sunroof control intentions, thereby enabling precise control of the vehicle's sunroof. Of course, the first and second preset ranges need to be set according to the actual vehicle model and user needs, and will not be elaborated further here.

[0073] The sunroof control calculation formula in the above embodiment is:

[0074] S = A × s1 + B × s2 + C × s3

[0075] Wherein, S represents the sunroof control logic result, A represents the driver's seat weight coefficient (A = 0.37), s1 represents the driver's head movement recognition logic result, B represents the front passenger seat weight coefficient (B = 0.32), s2 represents the front passenger's head movement recognition logic result, C represents the rear seat weight coefficient (C = 0.31), and s3 represents the rear passenger's head movement recognition logic result. It should be noted that each head movement recognition logic result can be determined by setting logic values; for example, the logic result for a head-tilt state is 1, and the logic result for a head-down state is 2. Furthermore, the weight coefficients need to be set according to the actual scenario requirements. In this embodiment, the driver's seat weight coefficient is set to a higher 0.37, while the non-driver's seat weight coefficient is lower. For other scenarios, such as picking up VIPs, the corresponding front passenger seat or rear seat weight coefficient can be increased, and the driver's seat weight coefficient can be decreased, depending on the actual situation. Further details are omitted here.

[0076] Another embodiment of the present invention provides a vehicle sunroof control device; for details, please refer to [link to details]. Figure 2 , Figure 2 The diagram shown illustrates the structure of a vehicle sunroof control device according to one embodiment of the present invention, which includes:

[0077] The data acquisition module 11 is used to acquire in-vehicle image data in real time when the vehicle enters the sunroof control mode, wherein the image data includes at least driver action image data and non-driver action image data.

[0078] The motion recognition module 12 is used to input the image data into a preset human behavior model for recognition, and obtain the head motion recognition results of each person in the vehicle.

[0079] The control signal module 13 is used to process the head movement recognition results of each person in the vehicle and determine the sunroof control signal used to represent the sunroof control intention.

[0080] The action execution module 14 is used to control the vehicle sunroof to perform actions associated with the sunroof control signal.

[0081] Furthermore, in the above embodiments, if the real-time luminance value inside the vehicle is detected to be greater than a preset first threshold, the vehicle is controlled to enter the sunroof control mode; or,

[0082] If the detected real-time temperature inside the vehicle exceeds a preset second threshold, the vehicle is controlled to enter the sunroof control mode; or,

[0083] If a sunroof control signal issued by the user is detected, the vehicle is controlled to enter the sunroof control mode.

[0084] Furthermore, in the above embodiments, the construction of the preset human behavioral model specifically includes:

[0085] Acquire initial sample image data for various scenarios inside the vehicle, wherein the initial sample image data includes at least image data of unoccupied state, image data of full-occupancy state, image data of non-full-occupancy state, image data of driver looking up, image data of driver looking down, image data of non-driver looking up and non-driver looking down.

[0086] The initial sample image data was labeled using the LabelImg tool to obtain labels for images looking up and images looking down.

[0087] The labeled initial sample image data is divided into training and test sets;

[0088] Convolutional neural networks for human motion recognition are constructed using Faster RCNN, SSD, or Yolo V1-V7 neural network technologies.

[0089] The human motion recognition convolutional neural network is trained based on the training set, and the trained human motion recognition convolutional neural network is validated based on the test set to obtain the human behavior model.

[0090] Furthermore, in the above embodiments, the processing of the head movement recognition results of each occupant in the vehicle to determine the sunroof control signal representing the sunroof control intention specifically includes:

[0091] The head movement recognition results of each person in the vehicle are input into the preset sunroof control calculation formula to obtain the sunroof control logic result;

[0092] If the sunroof control logic result is within a first preset range, a first sunroof control signal is generated to control the sunroof to open; if the sunroof control logic result is within a second preset range, a second sunroof control signal is generated to control the sunroof to close. By quantifying the processing results, it is convenient to accurately analyze the user's sunroof control intentions, thereby enabling precise control of the vehicle sunroof.

[0093] Furthermore, in the above embodiments, the sunroof control calculation formula is:

[0094] S = A × s1 + B × s2 + C × s3

[0095] Wherein, S is the sunroof control logic result, A is the driver's seat weight coefficient, A = 0.37, s1 is the driver's head movement recognition logic result, B is the passenger seat weight coefficient, B = 0.32, s2 is the passenger's head movement recognition logic result, C is the rear seat weight coefficient, C = 0.31, and s3 is the rear seat occupant's head movement recognition logic result.

[0096] For the vehicle sunroof control method provided in the embodiments of the present invention, those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0097] Accordingly, embodiments of the present invention provide a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform steps in the vehicle sunroof control method of the above embodiments, for example... Figure 1 Steps S1 to S4 as described above.

[0098] The vehicle sunroof control method and device provided in this invention have the following advantages:

[0099] (1) First, when the vehicle enters the sunroof control mode, real-time image data inside the vehicle is acquired; then the image data is input into a preset human behavior model for recognition to obtain the head movement recognition results of each person in the vehicle; then the head movement recognition results of each person in the vehicle are processed to determine the sunroof control signal used to represent the sunroof control intention; finally, the vehicle sunroof is controlled to perform the action associated with the sunroof control signal.

[0100] (2) Passengers do not need to control the sunroof manually, by switching on or by voice. Instead, the head movements during the process of looking at the sunroof are directly associated with the sunroof control. This allows the sunroof to be opened when looking up and closed when looking down. This avoids the problem of the sunroof being difficult to control due to the distance of the switch. In addition, the target recognition algorithm based on convolutional neural network is faster than voice recognition and is less affected by tire noise, wind noise and local accents during high-speed driving. Therefore, it is more stable and versatile.

[0101] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for controlling a vehicle sunroof, characterized in that, include: When the vehicle enters the sunroof control mode, real-time image data inside the vehicle is acquired, wherein the image data includes at least driver action image data and non-driver action image data; The image data is input into a preset human behavior model for recognition to obtain the head movement recognition results of each person in the vehicle. The head movement recognition results of each person in the vehicle are processed to determine the sunroof control signal used to represent the sunroof control intention; Control the vehicle sunroof to perform actions associated with the sunroof control signal; The construction of the pre-defined human behavioral model specifically includes: Acquire initial sample image data for various scenarios inside the vehicle, wherein the initial sample image data includes at least image data of unoccupied state, image data of full-occupancy state, image data of non-full-occupancy state, image data of driver looking up, image data of driver looking down, image data of non-driver looking up and non-driver looking down. The initial sample image data was labeled using the LabelImg tool to obtain labels for images looking up and images looking down. The labeled initial sample image data is divided into training and test sets; Convolutional neural networks for human motion recognition can be constructed using Faster RCNN, SSD, or Yolo V1~V7 neural network technologies. The human motion recognition convolutional neural network is trained based on the training set, and the trained human motion recognition convolutional neural network is validated based on the test set to obtain the human behavior model. The process of processing the head movement recognition results of each occupant in the vehicle to determine the sunroof control signal used to represent the sunroof control intention specifically includes: The head movement recognition results of each person in the vehicle are input into the preset sunroof control calculation formula to obtain the sunroof control logic result; If the sunroof control logic result is within the first preset range, a first sunroof control signal for controlling the sunroof to open is generated; if the sunroof control logic result is within the second preset range, a second sunroof control signal for controlling the sunroof to close is generated. The formula for sunroof control is: S = A × s1 + B × s2 + C × s3 Wherein, S is the sunroof control logic result, A is the driver's seat weight coefficient, A=0.37, s1 is the driver's head action recognition logic result, B is the passenger seat weight coefficient, B=0.32, s2 is the passenger's head action recognition logic result, C is the rear seat weight coefficient, C=0.31, and s3 is the rear seat occupant's head action recognition logic result.

2. The vehicle sunroof control method as described in claim 1, characterized in that, If the real-time luminance value inside the vehicle is detected to be greater than a preset first threshold, the vehicle is controlled to enter the sunroof control mode. or, If the real-time temperature inside the vehicle is detected to be greater than the preset second threshold, the vehicle will be controlled to enter the sunroof control mode. or, If a sunroof control signal issued by the user is detected, the vehicle is controlled to enter the sunroof control mode.

3. A vehicle sunroof control device, characterized in that, include: The data acquisition module is used to acquire in-vehicle image data in real time when the vehicle enters the sunroof control mode, wherein the image data includes at least driver action image data and non-driver action image data. The motion recognition module is used to input the image data into a preset human behavior model for recognition, and obtain the head motion recognition results of each person in the vehicle. The control signal module is used to process the head movement recognition results of each person in the vehicle to determine the sunroof control signal used to represent the sunroof control intention. An action execution module is used to control the vehicle sunroof to perform actions associated with the sunroof control signal; The construction of the pre-defined human behavioral model specifically includes: Acquire initial sample image data for various scenarios inside the vehicle, wherein the initial sample image data includes at least image data of unoccupied state, image data of full-occupancy state, image data of non-full-occupancy state, image data of driver looking up, image data of driver looking down, image data of non-driver looking up and non-driver looking down. The initial sample image data was labeled using the LabelImg tool to obtain labels for images looking up and images looking down. The labeled initial sample image data is divided into training and test sets; Convolutional neural networks for human motion recognition can be constructed using Faster RCNN, SSD, or Yolo V1~V7 neural network technologies. The human motion recognition convolutional neural network is trained based on the training set, and the trained human motion recognition convolutional neural network is validated based on the test set to obtain the human behavior model. The process of processing the head movement recognition results of each occupant in the vehicle to determine the sunroof control signal used to represent the sunroof control intention specifically includes: The head movement recognition results of each person in the vehicle are input into the preset sunroof control calculation formula to obtain the sunroof control logic result; If the sunroof control logic result is within the first preset range, a first sunroof control signal for controlling the sunroof to open is generated; if the sunroof control logic result is within the second preset range, a second sunroof control signal for controlling the sunroof to close is generated. The formula for sunroof control is: S = A × s1 + B × s2 + C × s3 Wherein, S is the sunroof control logic result, A is the driver's seat weight coefficient, A=0.37, s1 is the driver's head action recognition logic result, B is the passenger seat weight coefficient, B=0.32, s2 is the passenger's head action recognition logic result, C is the rear seat weight coefficient, C=0.31, and s3 is the rear seat occupant's head action recognition logic result.

4. The vehicle sunroof control device as described in claim 3, characterized in that, If the real-time luminance value inside the vehicle is detected to be greater than a preset first threshold, the vehicle is controlled to enter the sunroof control mode. or, If the real-time temperature inside the vehicle is detected to be greater than the preset second threshold, the vehicle will be controlled to enter the sunroof control mode. or, If a sunroof control signal issued by the user is detected, the vehicle is controlled to enter the sunroof control mode.

Images