Vehicle seat adjusting method, system, equipment and medium

Abstract

The invention discloses a vehicle seat adjusting method, system and device and a medium. The method comprises the steps that a signal sent by an automobile remote control key is received to obtain position information of a driver; according to the position information, obtaining an outside-vehicle human body video including a driver; joint point recognition is conducted on the outside-vehicle human body video, and time sequence position data corresponding to all joint points are obtained; calculating motion data corresponding to each joint point according to the time sequence position data, and calculating a first seat adjusting parameter according to the motion data; and according to the first seat adjustment parameter, a to-be-adjusted seat of the vehicle is controlled to execute a corresponding first seat adjustment operation, so that seat adjustment is completed before a driver enters the vehicle, and the time for the driver to wait for seat adjustment is shortened.

Description

Technical Field

[0001] The present invention relates to the field of automobile technology, and in particular to a vehicle seat adjustment method, system, device and medium. Background Art

[0002] In modern society, cars have become one of the indispensable means of transportation in people's daily lives. With the continuous advancement of technology, the comfort and convenience of cars have become the focus of car research and development, and the most important of these is the design of seats.

[0003] In the prior art, the seat adjustment methods are divided into manual adjustment and automatic adjustment. The traditional manual adjustment method requires the driver to manually loosen the seat locking mechanism through the handle, and then gradually change the sitting posture to find the most suitable position to fix it. The multiple manual operations not only bring a lot of inconvenience to sitting, but also require the seat position to be manually adjusted again after each driver is changed, which cannot meet the driver's seating comfort requirements and greatly reduces the vehicle performance of seat adjustment.

[0004] Most of the current mainstream automatic adjustment solutions are to extract features of the human body and then control the adjustment of the seat based on the extracted features. However, this solution is limited to scanning after the person enters the car and sits on the seat. It does not consider whether the driver can sit in the driver's seat when the seat is too close to the steering wheel. In addition, it takes too long to adjust the seat after the driver enters the car, which cannot meet the driver's requirements for seating comfort. Summary of the invention

[0005] In order to solve the above technical problems, the present invention discloses a vehicle seat adjustment method, system, device and medium, which are used to shorten the time a driver waits for seat adjustment and meet the driver's comfort requirements.

[0006] In order to achieve the above object, in a first aspect, the present invention provides a vehicle seat adjustment method, comprising:

[0007] Receiving a signal sent by a car remote control key, and obtaining the driver's location information according to the signal;

[0008] Shooting the exterior of the vehicle according to the position information to obtain a video of the human body outside the vehicle including the driver;

[0009] Performing joint point recognition on the human body video outside the vehicle to obtain time-series position data corresponding to each joint point;

[0010] Calculating motion data corresponding to each joint point according to the time series position data, so as to calculate a first seat adjustment parameter according to the motion data;

[0011] The to-be-adjusted seat of the vehicle is controlled to perform a corresponding first seat adjustment operation according to the first seat adjustment parameter.

[0012] The present invention discloses a vehicle seat adjustment method, the method obtains the current corresponding position information of the driver according to the signal sent by the automobile remote control key, so as to accurately shoot the corresponding human body video outside the vehicle when the driver is walking outside the vehicle according to the position information, and then identify the joint points according to the human body video outside the vehicle, generate the time series position data corresponding to each joint point of the driver, and then adjust the seat according to the time series position data, thereby realizing that the seat adjustment is completed before the driver enters the vehicle, reducing the time for the driver to wait for the seat adjustment and improving the comfort of the driver.

[0013] Furthermore, in the process of adjusting the seat according to the time-series position data, the motion data corresponding to each joint point is first calculated according to the time-series position data corresponding to each joint point, so as to simulate the human body posture according to the motion data and then calculate the first seat adjustment parameter corresponding to the driver, and realize automatic adjustment of the seat according to the first seat adjustment parameter, thereby solving the technical problems caused by manual adjustment.

[0014] As a preferred example, the receiving of a signal sent by a car remote control key and obtaining the driver's location information according to the signal includes:

[0015] Receiving a signal sent by a car remote control key, and matching the signal with the car remote control key;

[0016] When the match is successful, performing spatial spectrum estimation on the signal to obtain the spatial azimuth and distance when the car remote control key sends the signal;

[0017] The spatial attitude angle and the distance are determined as position information of the driver.

[0018] The present invention uses the signal sent by the car remote control key to match the driver with the vehicle to achieve identity verification, and after the matching is successful, the signal is spatially estimated to accurately obtain the driver's current location information.

[0019] As a preferred example, the step of photographing the exterior of the vehicle according to the position information to obtain a video of a person outside the vehicle including the driver includes:

[0020] According to the spatial azimuth, determining a first camera device for shooting the video of the human body outside the vehicle; wherein the first camera device is arranged outside the vehicle;

[0021] A shooting instruction corresponding to the first camera device is generated, so as to control the first camera device to shoot a human body video outside the vehicle when the driver is walking according to the shooting instruction.

[0022] The present invention arranges camera devices in different directions on the vehicle to meet the needs of obtaining corresponding videos when the driver approaches the vehicle from different directions. After determining the position information of the driver, the camera device in the corresponding direction is started to avoid the collection of useless videos and improve the efficiency of shooting.

[0023] As a preferred example, the step of performing joint point recognition on the human body video outside the vehicle to obtain time series position data corresponding to each joint point includes:

[0024] Extracting video frames from the human body video outside the vehicle to obtain multiple video frame images;

[0025] Input each of the video frame images into a pre-trained human joint target detection model, and extract a plurality of joint point regions in the video frame image through the human joint target detection model;

[0026] According to the joint point area, the position coordinates of each joint point in each of the video frame images are obtained through a pre-trained human body two-dimensional key point extraction network;

[0027] According to the time axis of the human body video outside the vehicle and the position coordinates of each joint point in each video frame image, the temporal position data of each joint point is generated.

[0028] The present invention decomposes the human body video outside the vehicle into a plurality of video frame images, and then identifies the joint points in each of the images to generate the position coordinates corresponding to each joint point in each image. Then, using the time axis of the video frame and the position coordinates, the time series position data corresponding to each joint point can be generated, so that the movement of the joint point can be simulated according to the time series position data, and then the corresponding movement data can be generated to generate the human body features corresponding to the driver, thereby improving the accuracy of the human body features.

[0029] As a preferred example, the step of calculating the motion data corresponding to each joint point according to the time series position data, so as to calculate the first seat adjustment parameter according to the motion data, includes:

[0030] According to the time series position data of each joint point, the three-dimensional motion data of each joint point is calculated by a pre-trained three-dimensional human joint point detection network;

[0031] Calculate initial seat adjustment parameters based on the three-dimensional motion data of each joint point and ergonomics;

[0032] Simulation is performed based on the initial seat adjustment parameters and the three-dimensional motion data to adjust the initial seat adjustment parameters based on the simulation results to obtain first seat adjustment parameters; wherein the seat adjustment parameters include seat longitudinal position, backrest inclination, seat height, steering wheel angle and steering wheel extension.

[0033] The present invention uses the three-dimensional motion data of each joint point to estimate the accurate human posture, and then simulates the initial seat adjustment parameters corresponding to the driver sitting on the seat according to the human posture. Then, simulation is performed based on the initial seat adjustment parameters and the three-dimensional motion data, and finally the corresponding first seat adjustment parameters are output to improve the accuracy of seat adjustment.

[0034] As a preferred example, after controlling the to-be-adjusted seat of the vehicle to perform the corresponding first seat adjustment operation according to the first seat adjustment parameter, the method includes:

[0035] Acquiring a pressure tolerance value of the seat to be adjusted in real time, and calculating a second seat adjustment parameter according to the pressure tolerance value;

[0036] The seat to be adjusted that has performed the first seat adjustment operation is controlled to perform a corresponding second seat adjustment operation according to the second seat adjustment parameter.

[0037] After the seat executes the first seat adjustment operation corresponding to the first seat adjustment parameter, the present invention obtains the pressure bearing value of the seat in real time, that is, obtains the pressure caused to the seat by the driver sitting on the seat, and then performs a second seat adjustment operation on the current seat according to the pressure, so that the seat after executing the second seat adjustment operation better fits the driver's sitting posture, thereby improving the driver's comfort.

[0038] As a preferred example, the real-time acquisition of the pressure tolerance value of the seat to be adjusted, and calculation of the second seat adjustment parameter according to the pressure tolerance value, includes:

[0039] Acquire in real time first pressure data of the seat surface of the seat to be adjusted and second pressure data of the backrest of the seat to be adjusted;

[0040] detecting whether the driver is sitting on the seat to be adjusted according to the first pressure data and the second pressure data;

[0041] When it is detected that the driver is sitting on the seat to be adjusted, querying from a pre-built database according to the first pressure data and the second pressure data to obtain corresponding target seat adjustment parameters;

[0042] A difference between the target seat adjustment parameter and the first seat adjustment parameter is calculated, and the difference is determined as a second seat adjustment parameter.

[0043] The present invention obtains in real time the pressure exerted on the seat by the driver after sitting on the seat, so as to perform secondary adjustment on the seat according to the pressure, thereby ensuring that the seat after the secondary adjustment better fits the driver's sitting posture and improves the driver's comfort.

[0044] In a second aspect, the present invention provides a vehicle seat adjustment system, the system comprising a position recognition module, a data acquisition module, a joint point recognition module, a parameter calculation module and a first adjustment module;

[0045] The position identification module is used to receive a signal sent by the car remote control key and obtain the driver's position information according to the signal;

[0046] The data acquisition module is used to shoot the exterior of the vehicle according to the position information to obtain a video of the human body outside the vehicle including the driver;

[0047] The joint point recognition module is used to recognize joint points of the human body video outside the vehicle to obtain the time sequence position data corresponding to each joint point;

[0048] The parameter calculation module is used to calculate the motion data corresponding to each joint point according to the time series position data, so as to calculate the first seat adjustment parameter according to the motion data;

[0049] The first adjustment module is used to control the to-be-adjusted seat of the vehicle to perform a corresponding first seat adjustment operation according to the first seat adjustment parameter.

[0050] The present invention discloses a vehicle seat adjustment system, wherein the system obtains the current position information of the driver according to the signal sent by the automobile remote control key, so as to accurately shoot the corresponding human body video outside the vehicle when the driver is walking outside the vehicle according to the position information, and then identify the joint points according to the human body video outside the vehicle, generate the time series position data corresponding to each joint point of the driver, and then adjust the seat according to the time series position data, thereby achieving the completion of the seat adjustment before the driver enters the vehicle, reducing the time the driver waits for the seat adjustment and improving the driver's comfort.

[0051] Furthermore, in the process of adjusting the seat according to the time-series position data, the motion data corresponding to each joint point is first calculated according to the time-series position data corresponding to each joint point, so as to simulate the human body posture according to the motion data and then calculate the first seat adjustment parameter corresponding to the driver, and realize automatic adjustment of the seat according to the first seat adjustment parameter, thereby solving the technical problems caused by manual adjustment.

[0052] In a third aspect, the present invention provides a vehicle seat adjustment device, comprising:

[0053] at least one processor, and a memory communicatively coupled to at least one of the processors;

[0054] Wherein, the memory stores instructions that can be executed by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can execute a vehicle seat adjustment method as described in the first aspect.

[0055] In a fourth aspect, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, wherein the computer-executable instructions are used to enable a computer to execute a vehicle seat adjustment method as described in the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 : A schematic flow chart of an embodiment of a vehicle seat adjustment method provided by the present invention;

[0057] Figure 2 : A schematic flow chart of an embodiment of a vehicle seat adjustment system provided by the present invention;

[0058] Figure 3 : A schematic diagram of a flow chart of another embodiment of a vehicle seat adjustment system provided by the present invention;

[0059] Figure 4 : A flow chart of another embodiment of the vehicle seat adjustment method provided by the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] Embodiment 1

[0062] Please refer to Figure 1 , is a flow chart of an embodiment of the vehicle seat adjustment method provided by the present invention, which mainly includes steps 101 to 105, which are specifically as follows:

[0063] Step 101: receiving a signal sent by a car remote control key, and obtaining the driver's location information according to the signal;

[0064] In this embodiment, the step is specifically as follows: receiving a signal sent by a car remote control key, and matching the signal with the car remote control key; when the match is successful, performing spatial spectrum estimation on the signal to obtain the spatial azimuth and distance when the car remote control key sends the signal; and determining the spatial azimuth and the distance as the location information of the driver.

[0065] In this embodiment, this step uses the signal sent by the car remote control key to match the driver and the vehicle to verify the identity, and after the match is successful, the signal is spatially estimated to accurately obtain the driver's current location information.

[0066] Step 102: photograph the exterior of the vehicle according to the position information to obtain a video of the human body outside the vehicle including the driver.

[0067] In this embodiment, the step is specifically as follows: according to the spatial azimuth, determine a first camera device for shooting the video of the human body outside the vehicle; wherein the first camera device is arranged outside the vehicle; generate a shooting instruction corresponding to the first camera device, so as to control the first camera device to shoot the video of the human body outside the vehicle when the driver is walking according to the shooting instruction.

[0068] In this embodiment, the step sets up camera devices at different positions outside the vehicle to ensure that the driver can obtain corresponding videos when approaching the vehicle from different positions. After determining the position information of the driver, the camera device at the corresponding position is started to avoid the collection of useless videos and improve the efficiency of shooting.

[0069] Step 103: performing joint point recognition on the human body video outside the vehicle to obtain time-series position data corresponding to each joint point.

[0070] In this embodiment, the step is specifically as follows: extracting video frames from the human body video outside the vehicle to obtain multiple video frame images; inputting each of the video frame images into a pre-trained human joint target detection model, such as a human ROI target detection model BRP or a YOLOv8-based model ONNX, etc., which can be used for human joint detection. And extracting several joint point areas in the video frame image through the human joint point target detection model; according to the joint point area, obtaining the position coordinates of each joint point in each of the video frame images through a pre-trained human two-dimensional key point extraction network. Among them, the human two-dimensional key point extraction network can adopt a 2DKN human two-dimensional key point extraction network or a convolutional neural network or a recurrent network for 2DPose human key point detection. According to the time axis of the human body video outside the vehicle and the position coordinates of each of the joint points in each of the video frame images, the temporal position data of each of the joint points is generated.

[0071] In this embodiment, this step decomposes the human body video outside the vehicle into a plurality of video frame images, and then identifies the joint points in each of the images to generate the position coordinates corresponding to each joint point in each image. Then, using the time axis of the video frame and the position coordinates, the time series position data corresponding to each joint point can be generated, so that the movement of the joint point can be simulated according to the time series position data, and then the corresponding movement data can be generated to generate the human body features corresponding to the driver, thereby improving the accuracy of the human body features.

[0072] Step 104: Calculate motion data corresponding to each joint point according to the time-series position data, so as to calculate a first seat adjustment parameter according to the motion data.

[0073] In this embodiment, the step is specifically as follows: according to the time series position data of each joint point, the three-dimensional motion data of each joint point is calculated through a pre-trained three-dimensional human joint point detection network. Wherein, the three-dimensional human joint point detection network uses the time series data of the two-dimensional joint point to obtain the motion data of the human walking joint point. Specifically, a network that can output the 3D coordinate data of the human joint point, such as a trained TRN time series three-dimensional human joint point detection network and Mediapipe, can be used. The initial seat adjustment parameters are calculated according to the three-dimensional motion data of each joint point and ergonomics; simulation is performed according to the initial seat adjustment parameters and the three-dimensional motion data to adjust the initial seat adjustment parameters according to the simulation results to obtain the first seat adjustment parameters; wherein, the seat adjustment parameters include the longitudinal position of the seat, the degree of inclination of the backrest, the seat height, the steering wheel angle and the steering wheel extension.

[0074] In this embodiment, this step uses the three-dimensional motion data of each joint point to estimate the accurate human posture, and then simulates the initial seat adjustment parameters corresponding to the driver sitting on the seat according to the human posture. Then, simulation is performed based on the initial seat adjustment parameters and the three-dimensional motion data, and finally the corresponding first seat adjustment parameters are output to improve the accuracy of seat adjustment.

[0075] Step 105: Controlling the to-be-adjusted seat of the vehicle to perform a corresponding first seat adjustment operation according to the first seat adjustment parameter.

[0076] Furthermore, in this embodiment, in order to improve the comfort of the seat, the seat can be adjusted for a second time after the user enters the vehicle and sits on the seat. Specifically, the process of adjusting the seat for a second time is as follows:

[0077] Step S1: acquiring the pressure tolerance value of the seat to be adjusted in real time, and calculating the second seat adjustment parameter according to the pressure tolerance value.

[0078] Step S2: controlling the seat to be adjusted that has performed the first seat adjustment operation to perform a corresponding second seat adjustment operation according to the second seat adjustment parameter.

[0079] In this embodiment, the step S1 also includes: acquiring first pressure data of the seat surface of the seat to be adjusted and second pressure data of the backrest of the seat to be adjusted in real time; detecting whether the driver is sitting on the seat to be adjusted according to the first pressure data and the second pressure data; when it is detected that the driver is sitting on the seat to be adjusted, querying from a pre-constructed database according to the first pressure data and the second pressure data to obtain corresponding target seat adjustment parameters; calculating the difference between the target seat adjustment parameters and the first seat adjustment parameters, and determining the difference as the second seat adjustment parameters.

[0080] In this embodiment, step S1 obtains in real time the pressure exerted by the driver on the seat after sitting on the seat, so as to perform secondary adjustment on the seat according to the pressure, thereby ensuring that the secondary-adjusted seat better fits the driver's sitting posture and improves the driver's comfort.

[0081] In this embodiment, please refer to Figure 2 , which is a structural schematic diagram of an embodiment of the vehicle seat adjustment system provided by the present invention, specifically, includes a position recognition module 201, a data acquisition module 202, a joint point recognition module 203, a parameter calculation module 204 and a first adjustment module 205.

[0082] The position identification module 201 is used to receive a signal sent by a car remote control key and obtain the driver's position information according to the signal.

[0083] The data acquisition module 202 is used to shoot the exterior of the vehicle according to the position information to obtain a video of the human body outside the vehicle including the driver.

[0084] The joint point recognition module 203 is used to perform joint point recognition on the human body video outside the vehicle to obtain the time sequence position data corresponding to each joint point.

[0085] The parameter calculation module 204 is used to calculate the motion data corresponding to each joint point according to the time series position data, so as to calculate the first seat adjustment parameter according to the motion data.

[0086] The first adjustment module 205 is used to control the to-be-adjusted seat of the vehicle to perform a corresponding first seat adjustment operation according to the first seat adjustment parameter.

[0087] In this embodiment, the location identification module 201 includes a matching unit and a signal processing unit.

[0088] The matching unit is used to receive a signal sent by the car remote control key and match the car remote control key according to the signal.

[0089] The signal processing unit is used to perform spatial spectrum estimation on the signal when the match is successful, obtain the spatial azimuth and distance when the car remote control key sends the signal; and determine the spatial azimuth and distance as the location information of the driver.

[0090] In this embodiment, the data acquisition module 202 includes a positioning unit and a collection unit.

[0091] The positioning unit is used to determine a first camera device for shooting the video of the human body outside the vehicle according to the spatial azimuth; wherein the first camera device is arranged on the vehicle.

[0092] The acquisition unit is used to generate a shooting instruction corresponding to the first camera device, so as to control the first camera device to shoot a human body video outside the vehicle when the driver is walking according to the shooting instruction.

[0093] In this embodiment, the joint point recognition module 203 includes a region detection unit and a data processing unit.

[0094] The area detection unit is used to extract video frames from the human body video outside the vehicle to obtain multiple video frame images; each of the video frame images is input into a pre-trained human joint point target detection model, and a number of joint point areas in the video frame image are extracted through the human joint point target detection model.

[0095] The data processing unit is used to obtain the position coordinates of each joint point in each of the video frame images according to the joint point area through a pre-trained human body two-dimensional key point extraction network; and generate the temporal position data of each joint point according to the time axis of the external human body video and the position coordinates of each joint point in each of the video frame images.

[0096] In this embodiment, the parameter calculation module 204 includes a motion unit and a simulation unit.

[0097] The motion unit is used to calculate the three-dimensional motion data of each joint point according to the time-series position data of each joint point through a pre-trained three-dimensional human joint point detection network.

[0098] The simulation unit is used to calculate initial seat adjustment parameters based on the three-dimensional motion data of each joint point and ergonomics; perform simulation based on the initial seat adjustment parameters and the three-dimensional motion data to adjust the initial seat adjustment parameters according to the simulation results to obtain first seat adjustment parameters; wherein the seat adjustment parameters include seat longitudinal position, backrest inclination, seat height, steering wheel angle and steering wheel extension.

[0099] Optionally, in this embodiment, in order to improve the comfort of the seat, a parameter adjustment module and a second adjustment module may be newly added to the vehicle seat adjustment system, so that after the user sits on the seat, the parameter adjustment module and the second adjustment module can perform a secondary adjustment on the seat according to the pressure exerted on the seat by the user, thereby ensuring that the seat after the secondary adjustment better fits the driver's sitting posture and improves the driver's comfort.

[0100] Specifically, the parameter adjustment module is used to obtain the pressure tolerance value of the seat to be adjusted in real time, and calculate the second seat adjustment parameter according to the pressure tolerance value.

[0101] The second adjustment module is used to control the seat to be adjusted that has performed the first seat adjustment operation to perform a corresponding second seat adjustment operation according to the second seat adjustment parameter.

[0102] Furthermore, in this embodiment, the parameter adjustment module also includes a pressure detection unit and an adjustment unit.

[0103] The pressure detection unit is used to obtain first pressure data of the seat surface of the seat to be adjusted and second pressure data of the backrest of the seat to be adjusted in real time; and detect whether the driver is sitting on the seat to be adjusted based on the first pressure data and the second pressure data.

[0104] The adjustment unit is used to query from a pre-constructed database according to the first pressure data and the second pressure data to obtain corresponding target seat adjustment parameters when it is detected that the driver is sitting on the seat to be adjusted; calculate the difference between the target seat adjustment parameter and the first seat adjustment parameter, and determine the difference as the second seat adjustment parameter.

[0105] In this embodiment, in addition to the method and system, a vehicle seat adjustment device and a computer-readable storage medium are also provided; the vehicle seat adjustment device includes at least one processor and a memory connected to the at least one processor; wherein the memory stores instructions executable by at least one processor, and the instructions are executed by at least one processor so that at least one processor can execute a vehicle seat adjustment method as described in this embodiment. The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute a vehicle seat adjustment method as described in this embodiment.

[0106] The vehicle seat adjustment method, system, device and medium provided in the present embodiment obtain the driver's current corresponding position information according to the signal sent by the car remote control key, so as to accurately capture the corresponding video of the human body outside the vehicle when the driver is walking outside the vehicle according to the position information, and then identify the joint points according to the video of the human body outside the vehicle, and generate time-series position data corresponding to each joint point of the driver, and then adjust the seat according to the time-series position data, thereby completing the seat adjustment before the driver enters the vehicle, reducing the time the driver waits for the seat adjustment and improving the driver's comfort.

[0107] Furthermore, in the process of adjusting the seat according to the time-series position data, the motion data corresponding to each joint point is first calculated according to the time-series position data corresponding to each joint point, so as to simulate the human body posture according to the motion data and then calculate the first seat adjustment parameter corresponding to the driver, and the seat is automatically adjusted according to the first seat adjustment parameter, thereby solving the technical problems caused by manual adjustment. After the seat performs the first seat adjustment operation corresponding to the first seat adjustment parameter, the pressure bearing value of the seat is obtained in real time, that is, the pressure caused to the seat by the driver after sitting on the seat is obtained, and then the second seat adjustment operation is performed on the current seat according to the pressure, so that the seat after the second seat adjustment operation is performed is more in line with the driver's sitting posture, thereby improving the driver's comfort.

[0108] Embodiment 2

[0109] Please refer to Figure 3 , Figure 3 The module diagram of a vehicle seat adjustment system provided by an embodiment of the present invention is as follows. In this embodiment, the adjustment system mainly includes a human body acquisition module 301 , an image processing module 302 , a first parameter calculation module 303 and a seat adjustment module 304 .

[0110] In certain implementations of this embodiment, in order to improve the comfort of the seat, a sensor module 305 may be added to the vehicle seat adjustment system to detect the pressure exerted on the seat by the user after the user sits on the seat through the sensor module 305, and then subsequently perform a secondary adjustment on the seat based on the pressure to ensure that the seat after the secondary adjustment better fits the driver's sitting posture.

[0111] like Figure 3 As shown, the human body acquisition module 301 is used to obtain the driver's position information according to the signal sent by the car remote control key when the driver approaches the vehicle, and call the first shooting device preset on the vehicle to shoot the driver according to the position information to obtain the human body video outside the vehicle.

[0112] The image processing module 302 is used to process the human body video outside the vehicle, extract the human body features corresponding to the driver, and then estimate the human body posture corresponding to the driver according to the human body features.

[0113] The first parameter calculation module 303 is used to calculate the first adjustment parameter information of the seat according to the human body posture; wherein the adjustment parameter information includes information such as the longitudinal position of the seat, the degree of backrest inclination, the seat height, the steering wheel angle and the steering wheel extension.

[0114] The seat adjustment module 304 is used to control the seat to perform corresponding adjustment operations according to the received first adjustment parameter information and second adjustment parameter information.

[0115] The sensor module 305 is used to collect pressure data of the seat after the driver sits on the seat that has been adjusted according to the first adjustment parameter information, and calculate the second adjustment parameter information of the seat according to the pressure data.

[0116] In this embodiment, before the image processing module 302 is used to estimate the driver's body posture, the image processing module 302 is trained to ensure the accuracy of the image processing module 302 .

[0117] The training process of the image processing module 302 is as follows:

[0118] 1) Introducing the TWGE algorithm into the image processing module 302, and estimating the human body posture by the TWGE algorithm;

[0119] 2) Build a walking motion experimental scene, collect 2D motion video data of human motion, and then annotate it to form a walking motion dataset containing human motion video data, human region annotation, and 2D and 3D key point coordinate information of human body;

[0120] 3) Testing the TWGE algorithm for estimating human posture based on the walking motion data set. Specifically, the process of testing the TWGE algorithm is as follows: inputting each of the two-dimensional motion video data in the walking motion data set into the TWGE algorithm to obtain the two-dimensional and three-dimensional key point coordinates of the TWGE algorithm for estimating human posture from the two-dimensional motion video data, and comparing the two-dimensional and three-dimensional key point coordinates output by the TWGE algorithm with the two-dimensional and three-dimensional key point coordinate information of the human body in the annotation of each two-dimensional motion video data in the walking motion data set, so as to update the parameters in the TWGE algorithm according to the comparison result, until the error between the two-dimensional and three-dimensional key point coordinates output by the TWGE algorithm after the parameter update and the two-dimensional and three-dimensional key point coordinate information in the annotation of the two-dimensional motion video data is within the preset error range, and the test of the TWGE algorithm is completed.

[0121] In this embodiment, before using the first parameter calculation module 303 to calculate the seat adjustment parameters corresponding to the current human body posture, the first parameter calculation module 303 is trained so that the first parameter calculation module 303 can output the seat adjustment parameter information corresponding to the human body posture. The training process of the first parameter calculation module 303 is:

[0122] 1) Construct an initial data set; a certain number of images of a human body standing frontally and sideways are taken by a camera. In order to better simulate the driving state of the driver, image information of the side-up, side-sitting, front-sitting and front-side image information in the car are also prepared, and the seat parameter information corresponding to the optimal driving position of the 3D model is recorded, including the longitudinal position, the tilt degree and height of the backrest, etc.;

[0123] 2) Develop a seat adjustment program; process the collected human body 3D model data information, use ergonomics to calculate the seat longitudinal position, backrest tilt, seat height, distance between the seat and the steering wheel, and tilt angle, simulate the 3D model after it is seated using simulation software, and substitute it into the initial data set to obtain the seat longitudinal position, backrest tilt, seat height, distance between the seat and the steering wheel, and tilt angle parameter information that can be adjusted according to the current driver's needs;

[0124] 3) Migrate the developed seat adjustment program to the first parameter calculation module 303 .

[0125] In this embodiment, the construction process of the sensor module 305 is as follows:

[0126] 1) Build a database; collect initial pressure data when people of different height and weight ranges sit on the seat, and then adjust the longitudinal position of the seat, backrest inclination, seat height, steering wheel angle, steering wheel extension and other adjustment parameter information according to the requirements of sitting posture, field of view and operable space during driving, and record the test data.

[0127] 2) Data collection: When the driver sits down, the pressure sensors distributed on the seat cushion, backrest, waist and head first collect the pressure data received by each part to obtain the corresponding pressure tolerance value;

[0128] 3) The pressure, contact area, deformation and other parameters of different areas of the human body when sitting in a stable state are obtained through sensors. The comfort calculation results obtained from the currently public dummy test are used as the standard range, and compared with the previously obtained test data to optimize and obtain the final database.

[0129] In a preferred application embodiment, based on Figure 3 The vehicle seat adjustment system shown executes a vehicle seat adjustment method. Specifically, the method includes steps 401 to 405, which are as follows:

[0130] Step 401: receiving a signal sent by a key in real time according to a human body acquisition module, and locating the position information of a driver according to the signal, so as to acquire two-dimensional video image information corresponding to the driver according to the position information.

[0131] Specifically, when the driver gradually approaches the vehicle and the distance from the door handle is approximately enough for the PEPS system in the body controller to match it according to the signal sent by the key, the spatial azimuth and distance of the high-frequency signal sent by the key are obtained by performing spatial spectrum estimation on the high-frequency signal sent by the key, thereby obtaining the driver's position information.

[0132] The RGB cameras installed on the door handles and rearview mirrors obtain the driver's position information, determine the direction from which the driver approaches, and transmit the information to the central control system to issue a command for the camera to start shooting video. The RGB camera in the corresponding direction is then controlled to start shooting a video of the driver walking from a certain direction. During the shooting process, the human body is ensured to occupy the main body and central position of the picture, so as to obtain two-dimensional video image information of human movement.

[0133] Step 402: The two-dimensional video image information is transmitted to an image processing module, and the human body posture data corresponding to the driver is extracted by using a TWGE algorithm provided in the image processing module.

[0134] Specifically, the collected two-dimensional video image information is transmitted to the image processing module and processed by the TWGE algorithm preset in the image processing module.

[0135] Optionally, a human ROI target detection model BRP is designed in the image processing module to extract the ROI position of the main person in the image. Then, according to the ROI position, the position coordinates of each human joint point are obtained through a 2DKN human two-dimensional key point extraction network.

[0136] According to the position coordinates of each human joint point in each image and the time data in the two-dimensional video image information, time series data corresponding to each joint point is formed, wherein the time series data includes the position coordinates and the time corresponding to the position coordinates.

[0137] The TRN time-series 3D human joint detection network uses the 2D joint time-series data (i.e., the time-series data corresponding to each joint) to obtain the motion data of the human walking joints and output the 3D motion data of each joint during the human action. Among them, the 3D human skeleton joints are mainly the nose, left and right eyes, left and right ears, left and right shoulders, left and right elbows, left and right wrists, left and right hips, left and right knees, and left and right ankles.

[0138] Step 403: Transmitting the human body posture data to a first parameter calculation module, and obtaining corresponding first adjustment parameter information through a seat adjustment program preset in the first parameter calculation module.

[0139] Specifically, after obtaining the three-dimensional motion data of each joint during the human body's movement, it is transmitted to the first parameter calculation module, and the first adjustment parameter information such as the seat longitudinal position, backrest inclination, seat height, steering wheel angle and steering wheel extension are obtained according to the seat adjustment program developed in the construction of the first parameter calculation module.

[0140] Step 404: Control the seat of the vehicle to perform a corresponding first adjustment operation according to the first adjustment parameter information.

[0141] Specifically, the first adjustment parameter information is transmitted to the vehicle's seat adjustment system, which includes a stepper motor, a sensor and an ECU control system, with each component connected with a central processing unit as the center. After receiving the data information, the corresponding motor operation adjusts the seat to the optimal position suitable for human driving.

[0142] Step 405: The pressure data on the seat caused by the driver sitting on the seat is obtained in real time through the sensor module, and corresponding second adjustment parameter information is obtained based on the pressure data, and the vehicle seat is controlled to perform a corresponding second adjustment operation based on the second adjustment parameter information.

[0143] Specifically, when the driver opens the car door and sits on the driver's seat, the pressure sensor located on the seat detects the pressure data of the seat surface and the backrest, and sends this data to the seat adjustment module for small adjustments.

[0144] After the sensor module receives the pressure data detected by the pressure sensor, it finds the adjustment parameter information corresponding to the pressure data according to the database constructed during the training process of the system, compares the adjustment parameter information with the first adjustment parameter information, and automatically adjusts the headrest, backrest and seat cushion according to the comparison result to make the seat more suitable for the passenger's body shape. When the driver has different sitting postures, the sensor determines the sitting posture through the detected pressure distribution, automatically adjusts according to the data, and drives the motor to make small adjustments to the seat to provide the most comfortable fit support for the passenger.

[0145] The present embodiment provides a vehicle seat adjustment system and method, which uses an AI algorithm to scan the driver before the driver enters the vehicle to obtain the driver's body data information, and can adjust the seat to the optimal position before the driver gets in the vehicle, greatly improving the time problem of waiting for seat adjustment and improving the driver's comfort and convenience.

[0146] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A vehicle seat adjustment method, characterized in that: include: Receiving a signal sent by a car remote control key, and obtaining the driver's location information according to the signal; Shooting the exterior of the vehicle according to the position information to obtain a video of the human body outside the vehicle including the driver; Performing joint point recognition on the human body video outside the vehicle to obtain time-series position data corresponding to each joint point; Calculating motion data corresponding to each joint point according to the time series position data, so as to calculate a first seat adjustment parameter according to the motion data; The to-be-adjusted seat of the vehicle is controlled to perform a corresponding first seat adjustment operation according to the first seat adjustment parameter.

2. A vehicle seat adjustment method according to claim 1, characterized in that: The receiving of the signal sent by the automobile remote control key and obtaining the driver's location information according to the signal includes: Receiving a signal sent by a car remote control key, and matching the signal with the car remote control key; When the match is successful, performing spatial spectrum estimation on the signal to obtain the spatial azimuth and distance when the car remote control key sends the signal; The spatial attitude angle and the distance are determined as position information of the driver.

3. A vehicle seat adjustment method according to claim 2, characterized in that: The step of photographing the exterior of the vehicle according to the position information to obtain a video of a person outside the vehicle including the driver includes: According to the spatial azimuth, determining a first camera device for shooting the video of the human body outside the vehicle; wherein the first camera device is arranged outside the vehicle; A shooting instruction corresponding to the first camera device is generated, so as to control the first camera device to shoot a human body video outside the vehicle when the driver is walking according to the shooting instruction.

4. A vehicle seat adjustment method according to claim 1, characterized in that: The step of performing joint point recognition on the human body video outside the vehicle to obtain time series position data corresponding to each joint point includes: Extracting video frames from the human body video outside the vehicle to obtain multiple video frame images; Input each of the video frame images into a pre-trained human joint target detection model, and extract a plurality of joint point regions in the video frame image through the human joint target detection model; According to the joint point area, the position coordinates of each joint point in each of the video frame images are obtained through a pre-trained human body two-dimensional key point extraction network; According to the time axis of the human body video outside the vehicle and the position coordinates of each joint point in each video frame image, the temporal position data of each joint point is generated.

5. A vehicle seat adjustment method according to claim 4, characterized in that: The step of calculating the motion data corresponding to each joint point according to the time series position data, so as to calculate the first seat adjustment parameter according to the motion data, includes: According to the time series position data of each joint point, the three-dimensional motion data of each joint point is calculated by a pre-trained three-dimensional human joint point detection network; Calculate initial seat adjustment parameters based on the three-dimensional motion data of each joint point and ergonomics; Simulation is performed based on the initial seat adjustment parameters and the three-dimensional motion data to adjust the initial seat adjustment parameters based on the simulation results to obtain first seat adjustment parameters; wherein the seat adjustment parameters include seat longitudinal position, backrest inclination, seat height, steering wheel angle and steering wheel extension.

6. The vehicle seat adjustment method according to claim 1, characterized in that: After controlling the to-be-adjusted seat of the vehicle to perform the corresponding first seat adjustment operation according to the first seat adjustment parameter, the method includes: Acquiring a pressure tolerance value of the seat to be adjusted in real time, and calculating a second seat adjustment parameter according to the pressure tolerance value; The seat to be adjusted that has performed the first seat adjustment operation is controlled to perform a corresponding second seat adjustment operation according to the second seat adjustment parameter.

7. A vehicle seat adjustment method according to claim 6, characterized in that: The step of acquiring the pressure tolerance value of the seat to be adjusted in real time and calculating the second seat adjustment parameter according to the pressure tolerance value includes: Acquire in real time first pressure data of a seat surface of the seat to be adjusted and second pressure data of a backrest of the seat to be adjusted; detecting whether the driver is sitting on the seat to be adjusted according to the first pressure data and the second pressure data; When it is detected that the driver is sitting on the seat to be adjusted, querying from a pre-built database according to the first pressure data and the second pressure data to obtain corresponding target seat adjustment parameters; A difference between the target seat adjustment parameter and the first seat adjustment parameter is calculated, and the difference is determined as a second seat adjustment parameter.

8. A vehicle seat adjustment system, characterized in that: The system includes a position recognition module, a data acquisition module, a joint point recognition module, a parameter calculation module and a first adjustment module; The position identification module is used to receive a signal sent by the car remote control key and obtain the driver's position information according to the signal; The data acquisition module is used to shoot the exterior of the vehicle according to the position information to obtain a video of the human body outside the vehicle including the driver; The joint point recognition module is used to recognize joint points of the human body video outside the vehicle to obtain the time sequence position data corresponding to each joint point; The parameter calculation module is used to calculate the motion data corresponding to each joint point according to the time series position data, so as to calculate the first seat adjustment parameter according to the motion data; The first adjustment module is used to control the to-be-adjusted seat of the vehicle to perform a corresponding first seat adjustment operation according to the first seat adjustment parameter.

9. A vehicle seat adjustment device, characterized in that: include: at least one processor, and a memory communicatively coupled to at least one of the processors; The memory stores instructions executable by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can execute a vehicle seat adjustment method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute a vehicle seat adjustment method as described in any one of claims 1 to 7.

Images