Vehicle seat adjusting device and method, electronic equipment and storage medium

Abstract

The invention discloses a vehicle seat adjusting device and method, electronic equipment and a storage medium, and relates to the field of vehicle seat control, and the device comprises an image collection module which is used for collecting a whole body image of a person in a vehicle; the visual algorithm processing module is used for receiving the whole-body image and judging the height of a person in the vehicle through a visual algorithm, and the visual algorithm comprises the steps that the whole-body image is preprocessed; recognizing a human body region in the image through a human body detection algorithm, wherein the human body detection algorithm is a convolutional neural network model based on deep learning; the visual algorithm processing module is used for extracting feature points in a human body area, the ergonomic data storage module stores an optimal seat angle database corresponding to different heights, and the seat adjusting control module is used for calling the corresponding seat angle in the database according to height data output by the visual algorithm processing module and adjusting the seat angle according to the height data. And the electric adjusting mechanism is controlled to execute an adjusting action.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of vehicle seat control, and in particular to a vehicle seat adjusting device, method, electronic device and storage medium. BACKGROUND

[0002] In the existing automobile technology, the adjustment of the seat usually needs to be manually operated by the driver or passenger to achieve a comfortable riding experience. However, this approach has some obvious shortcomings. First, manual adjustment is tedious, and it needs to be re-adjusted every time the driver or passenger changes, wasting time and effort. Second, different people have different heights and body proportions, and manual adjustment often cannot accurately adapt to the needs of each person, which may result in uncomfortable or non-ergonomic riding. Moreover, for some people who are not familiar with the operation of the vehicle, they may not be able to correctly adjust the seat, affecting the safety and comfort of driving or riding. In addition, in some cases, the driver or passenger may forget to adjust the seat, or it is inconvenient to adjust during driving, which will bring inconvenience to the driver or passenger. With the development of technology and the increasing demand for comfort and intelligence of automobiles, the existing seat adjustment method cannot meet the needs. SUMMARY

[0003] Therefore, the purpose of the present application is to provide a vehicle seat adjusting device, method, electronic device and storage medium. The present application solves the problem of manual adjustment being tedious, inaccurate and possibly causing inconvenience and safety hazards by using a visual algorithm to determine the height of the person in the vehicle and automatically adjusting the seat angle according to ergonomic data. The present application provides the following solutions:

[0004] According to one aspect of the present application, a vehicle seat adjusting device is provided, comprising:

[0005] an image acquisition module, the image acquisition module comprising a high-resolution camera, the camera being installed at a predetermined position in the vehicle to acquire a full-body image of a person in the vehicle;

[0006] a visual algorithm processing module, the visual algorithm processing module being in communication connection with the image acquisition module, for receiving the full-body image and determining the height of the person in the vehicle through a visual algorithm, the visual algorithm comprising the following steps:

[0007] preprocessing the full-body image, including image denoising, contrast enhancement and distortion correction;

[0008] identifying the human body region in the image through a human body detection algorithm, the human body detection algorithm being a convolutional neural network model based on deep learning;

[0009] extracting feature points in the human body region, the feature points including a head point, a foot point and at least three key joint points;

[0010] According to the preset parameters of the camera and the coordinates of the feature points in the image, the actual height of the human body is calculated through perspective projection transformation;

[0011] The ergonomic data storage module stores a database of optimal seat angles corresponding to different heights, which is constructed based on ergonomic experimental data;

[0012] The seat adjustment control module is used to call the corresponding seat angle in the database according to the height data output by the vision algorithm processing module, and control the electric adjustment mechanism to perform the adjustment action.

[0013] Further, comprising:

[0014] The number of cameras of the image acquisition module is 2, which are installed on the inside of the A column and the front part of the roof, and the collection range of the two cameras covers the driver's seat and the front passenger's seat.

[0015] Further, comprising:

[0016] The convolutional neural network model in the vision algorithm processing module is a YOLO model or an SSD model, which is used to detect human regions in real time, and the detection frame rate is not less than 25fps.

[0017] Further, comprising:

[0018] The database in the ergonomic data storage module also includes associated data of seat front and rear distance, backrest inclination angle and headrest height corresponding to different heights.

[0019] Further, comprising:

[0020] The seat adjustment control module also includes a user feedback unit, which is used to receive the fine adjustment instruction of the user to the automatically adjusted seat angle, and store the angle deviation data corresponding to the fine adjustment instruction to the ergonomic data storage module, forming a personalized preference database.

[0021] Further, comprising:

[0022] The seat adjustment control module preferentially calls the data in the personalized preference database to perform the adjustment action when detecting the same user next time, and the user recognition is realized through the face features or body features collected by the image acquisition module.

[0023] According to the two aspects of the present application, a vehicle seat adjustment method is provided, which comprises the following steps:

[0024] After the vehicle is started, the full-body image of the person in the vehicle is collected by the image acquisition module;

[0025] The actual height of the person in the vehicle is calculated by analyzing the full-body image through a visual algorithm processing module, and the visual algorithm includes human body detection and feature point extraction steps based on deep learning;

[0026] The seat adjustment control module calls the matched seat angle parameters from the ergonomic data storage module according to the actual height;

[0027] The control seat electric adjustment mechanism performs angle adjustment, and the adjustment process is monitored in real time;

[0028] The feedback of the user on the adjustment result is received, and if the user performs fine adjustment, the angle data after fine adjustment is recorded and updated to the database.

[0029] Further, comprising:

[0030] The calculation of the actual height includes: combining the three-dimensional coordinates of the camera installation position and the pixel coordinates of the human body feature points in the image, and calculating the physical size through the perspective transformation formula.

[0031] Further, comprising:

[0032] Real-time monitoring includes detecting changes in the sitting posture of the person through the pressure sensor built in the seat, and if an abnormal sitting posture is detected, the adjustment process is retriggered.

[0033] According to the three aspects of the application, an electronic device is provided, comprising: a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete communication with each other through the communication bus.

[0034] The memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of a vehicle seat adjustment method.

[0035] According to the four aspects of the application, a computer readable storage medium is provided, which stores a computer program executable by an electronic device, and when the computer program runs on the electronic device, the electronic device executes the steps of a vehicle seat adjustment method.

[0036] Through the above scheme, the following beneficial technical effects are obtained:

[0037] The present application determines the height of the person in the vehicle through a visual algorithm and automatically adjusts the seat angle according to the ergonomic data for the corresponding person, without the need for the driver or passenger to manually adjust the seat angle, so that the system can automatically adjust the seat angle for the driver or passenger and provide the best riding experience.

[0038] The application can avoid discomfort or unsafe factors caused by improper manual adjustment through automatic adjustment, greatly improving the intelligent level of the vehicle and the comfort and safety of driving and riding. The improvement solves the problems of tediousness, inaccuracy and possible inconvenience and safety hazards in the prior art. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 is a structural diagram of a vehicle seat adjustment device provided by one or more embodiments of the application.

[0040] Figure 2 is a flowchart of a vehicle seat adjustment method provided by one or more embodiments of the application.

[0041] Figure 3 is a control flowchart of a vehicle seat adjustment method of one specific embodiment of the application.

[0042] Figure 4 is a structural block diagram of an electronic device for a vehicle seat adjustment method provided by one or more embodiments of the application. DETAILED DESCRIPTION

[0043] The technical solutions of the application will be described in detail below with reference to the drawings. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the application.

[0044] Figure 1 is a structural diagram of a vehicle seat adjustment device provided by one or more embodiments of the application.

[0045] As Figure 1 shown, the device comprises:

[0046] An image acquisition module, the image acquisition module comprising a high-resolution camera, the camera being installed at a predetermined position in the vehicle to acquire a full-body image of a person in the vehicle;

[0047] A visual algorithm processing module, the visual algorithm processing module being in communication connection with the image acquisition module, for receiving the full-body image and determining the height of the person in the vehicle through a visual algorithm, the visual algorithm comprising the following steps:

[0048] Preprocessing the full-body image, including image denoising, contrast enhancement and distortion correction;

[0049] Identifying the human body region in the image through a human body detection algorithm, the human body detection algorithm being a convolutional neural network model based on deep learning;

[0050] Extract feature points in the human body region, including the head vertex, foot bottom point, and at least three key joint points;

[0051] According to the preset parameters of the camera and the coordinates of the feature points in the image, the actual height of the human body is calculated through perspective projection transformation;

[0052] An ergonomic data storage module stores a database of optimal seat angles corresponding to different heights, which is constructed based on ergonomic experimental data;

[0053] A seat adjustment control module is used to call the corresponding seat angle in the database according to the height data output by the visual algorithm processing module, and control the electric adjustment mechanism to perform the adjustment action.

[0054] Specifically, the height of the driver or passenger is automatically recognized by the visual algorithm, and the preset ergonomic database is directly called to realize "one-key adaptation" of the seat parameters, saving the manual operation steps, and the parameters based on scientific experimental data are more suitable for human physiological needs than subjective adjustment.

[0055] Optimize image quality by "image preprocessing (denoising, contrast enhancement, distortion correction)" to reduce environmental interference;

[0056] Use a deep learning model (convolutional neural network) to accurately locate the human body region and avoid background interference;

[0057] Extract key feature points (head, foot, joint points), combine with camera preset parameters (internal parameters, installation angle), and convert pixel coordinates to actual height through perspective projection transformation to solve the conversion problem from "two-dimensional image to three-dimensional size".

[0058] No need to manually adjust the seat: after the driver or passenger gets on the vehicle, the system automatically completes height recognition and seat adaptation, especially suitable for vehicles shared by multiple people (such as family cars, online taxis), reducing the time cost of repeated adjustments;

[0059] Lower the operation threshold: for the elderly, children, or users unfamiliar with the vehicle's functions, avoid the use obstacles caused by "not knowing how to adjust the seat".

[0060] Accurate adaptation based on ergonomics: seat angle, height, and other parameters match the height of the driver or passenger, reducing the pressure on the waist, neck, and legs (such as tall people needing more legroom, and short people needing higher seat height to ensure visibility);

[0061] Reduce driving fatigue: scientific seat parameters can maintain the natural curvature of the spine, reduce muscle tension during long-distance driving, and improve driving comfort.

[0062] Optimized field of view: Adjust the seat height to ensure that drivers of different heights can clearly observe the dashboard and the road ahead (e.g., short drivers avoid obstructing the view with the steering wheel by lowering the seat too low).

[0063] Convenient operation: After adjusting the seat angle, the driver's actions of steering the steering wheel and operating the pedals are more natural, reducing the delay caused by awkward posture (e.g., reaction speed during emergency braking).

[0064] This solution solves the technical problems of visual measurement and parameter matching in the vehicle scene, and effectively improves the convenience, comfort, and safety of users. It is a typical landing scenario that combines vehicle intelligence and humanization.

[0065] Further, comprising:

[0066] The number of cameras in the image acquisition module is 2, installed on the inside of the A-pillar and the front of the roof. The collection range of the two cameras covers the driver's seat and the front passenger's seat.

[0067] Further, comprising:

[0068] The convolutional neural network model in the visual algorithm processing module is a YOLO model or an SSD model, which is used to detect human regions in real time with a frame rate not less than 25fps.

[0069] Specifically, the input image (such as the full-body image captured by the in-vehicle camera) is scaled to a fixed size (such as YOLOv5 default 640x640 pixels), ensuring uniform input size for parallel network computation. At the same time, combined with image denoising, contrast enhancement, and distortion correction preprocessing, the interference of image quality on detection is reduced.

[0070] The standardized image is divided into S*S grids (such as YOLOv5 uses a 19x19 grid, adjusted with the model version), and each grid is responsible for detecting human targets whose centers fall within the grid. For example, if the center point of the human body (the geometric center of the human body region in the image) is located in the 5th row and 8th column grid, the grid will predict the bounding box of the human body.

[0071] Each grid predicts multiple candidate bounding boxes, each containing 4 types of key information that collectively describe the position and reliability of the human body region:

[0072] Including: center coordinates (x, y), offset of bounding box center relative to the top left corner of the grid (normalized to 0-1), finally converted to absolute image coordinates through grid coordinate conversion;

[0073] width and height of the bounding box (normalized value relative to the whole image, e.g. w=0.3 means the width is 30% of the image width);

[0074] confidence, which comprehensively reflects the probability that the box contains a human body and the positioning accuracy of the box.

[0075] Taking the in-vehicle scene as an example, if a certain grid predicts a candidate box, its (x, y) corresponds to the driver's waist position, (w, h) covers the range from the top of the head to the bottom of the feet, and the confidence is 0.95, it means that the box is probably a reliable human body area.

[0076] Since multiple grids may predict multiple candidate boxes for the same human body (there is overlap), the non-maximum suppression (NMS) is needed to filter out the optimal result, and the steps are as follows:

[0077] Sort all candidate boxes according to the product of "confidence + human class probability", and keep the highest score box as the initial optimal box;

[0078] Calculate the IoU (intersection over union) of the remaining boxes and the optimal box, if IoU>threshold (such as 0.5), it is considered as a duplicate box and deleted;

[0079] Repeat the above steps until all non-overlapping human body boxes are filtered out, and the final output box is the detected human body area.

[0080] In the in-vehicle scene, NMS can effectively remove redundant boxes caused by "seat back similar to human body contour" and "multiple grid repeated prediction", ensuring that only one most accurate area box is kept for each human body.

[0081] Among them, there is a threshold for the smooth perception of dynamic pictures by human eyes. In general, when the frame rate of the picture reaches 24fps and above, the human eye will perceive the continuous frame sequence as a smooth dynamic scene, rather than a discrete picture switch. The detection frame rate is not less than 25fps, which can ensure that the detection result of the human body area by the algorithm is synchronized with the dynamic change of the real scene, so that the detection information (such as human body position, motion state, etc.) output by the system meets the perception expectation of the human eye for "real-time", avoiding the perception of picture lag and delay.

[0082] For dynamic targets such as human bodies, their motion state (such as walking, running, turning, etc.) is in continuous change. Higher detection frame rate (≥25fps) can obtain more target frames in unit time, thereby more densely recording the position and shape change of the human body, reducing the "jump" or "loss" of the target caused by insufficient frame rate.

[0083] The detection frame rate is not less than 25 fps, which is a key technical index balancing real-time performance, dynamic target capture capability, decision response speed and system performance, and has a basic guarantee function for application scenarios (such as monitoring, interaction, automatic driving, etc.) relying on real-time human region detection.

[0084] Further, comprising:

[0085] The database in the ergonomics data storage module also includes associated data of seat front and rear distance, backrest inclination angle and headrest height corresponding to different heights.

[0086] Further, comprising:

[0087] The seat adjustment control module also includes a user feedback unit, which is used to receive user fine adjustment instructions for the automatically adjusted seat angle, and store the angle deviation data corresponding to the fine adjustment instructions to the ergonomics data storage module to form a personalized preference database.

[0088] Further, comprising:

[0089] The seat adjustment control module preferentially calls the data in the personalized preference database to perform the adjustment action when detecting the same user next time, and the user recognition is realized by the facial features or body features collected by the image collection module.

[0090] It is worth noting that although the device only discloses the image collection module, the visual algorithm processing module, the ergonomics data storage module and the seat adjustment control module, it does not mean that the device is limited to the above basic functional modules. On the contrary, the meaning expressed by the present application is that on the basis of the above basic functional modules, those skilled in the art can add one or more functional modules to form infinite embodiments or technical solutions in combination with the prior art. That is to say, the system is open rather than closed, and it cannot be considered that the protection scope of the present application is limited to the above disclosed basic functional modules because the present embodiment only discloses individual basic functional modules.

[0091] Figure 2 is a flowchart of a vehicle seat adjustment method provided by one or more embodiments of the present application.

[0092] As Figure 2 shown, a vehicle seat adjustment method includes the following steps:

[0093] Step S1, after the vehicle starts, the whole body image of the person in the vehicle is collected by the image collection module;

[0094] Step S2, the whole body image is analyzed by the visual algorithm processing module to calculate the actual height of the person in the vehicle, and the visual algorithm includes human detection and feature point extraction steps based on deep learning.

[0095] Step S3, the seat adjustment control module calls the matching seat angle parameters from the ergonomic data storage module according to the actual height;

[0096] Step S4, control the seat electric adjustment mechanism to execute angle adjustment and monitor the adjustment process in real time;

[0097] Step S5, receive user feedback on the adjustment result, if the user performs fine adjustment, record the fine-adjusted angle data and update to the database.

[0098] Further, comprising:

[0099] Calculating the actual height includes: combining the three-dimensional coordinates of the camera installation position and the pixel coordinates of the human feature points in the image, and calculating the physical size through the perspective transformation formula.

[0100] Further, comprising:

[0101] Real-time monitoring includes detecting changes in the user's sitting posture through the seat's built-in pressure sensor, and if an abnormal sitting posture is detected, the adjustment process is triggered again.

[0102] Specifically, by combining the three-dimensional coordinates of the camera + the pixel coordinates of the feature points, the actual height is calculated through perspective transformation, and the mapping relationship between pixels and physical size is established using geometric principles (such as camera installation height 1.5m, pitch angle 30°, combined with head / foot pixel difference, directly converted height through perspective formula).

[0103] Combined with the precise human region positioning of the YOLO model, the reliability of feature point (head, foot) extraction is ensured, providing high-quality input for height calculation.

[0104] And through the optimization of the YOLO model, such as dynamic confidence threshold, small target detection enhancement, to meet the human recognition needs of children, curled sitting posture and other special scenes.

[0105] Solve the problem of "insufficient matching degree between automatic adjustment and user's actual needs".

[0106] Through the seat pressure sensor to monitor the change of sitting posture, real-time judge whether there is an abnormality in the adjustment process, if there is an abnormality, trigger the re-adjustment process, avoid the adjustment deviation or security risks caused by the change of posture (such as the user suddenly leans forward when the seat reclines, the system pauses and recalculates the adjustment parameters).

[0107] Through the above technical features, the problems of insufficient precision, lack of individualization, and safety hazards of traditional automatic adjustment are solved.

[0108] Figure 4is an electronic device structure block diagram of a vehicle seat adjustment method provided by one or more embodiments of the present application.

[0109] As shown in Figure 4 The present application provides an electronic device, comprising: a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete mutual communication through the communication bus.

[0110] The memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of a vehicle seat adjustment method.

[0111] The present application also provides a computer readable storage medium storing a computer program executable by an electronic device, and when the computer program runs on the electronic device, the electronic device executes the steps of a vehicle seat adjustment method.

[0112] For the method embodiments, in order to simply describe, they are all expressed as a series of action combinations, but those skilled in the art should know that the embodiments of the present application are not limited to the order of the described actions, because according to the embodiments of the present application, certain steps can be performed in other order or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions involved are not necessarily the necessary of the embodiments of the present application.

[0113] From the above description of the embodiments, those skilled in the art can clearly understand that the present application can be implemented by means of software plus necessary universal hardware platforms. Based on such understanding, the technical solutions of the present application can be embodied in the form of software products, which can be stored in storage media, such as ROM / RAM, magnetic disks, optical disks, etc., and include a number of instructions to make a computer device (which can be a personal computer, a server or a network device, etc.) execute the methods described in the various embodiments or some parts of the embodiments of the present application.

[0114] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A vehicle seat adjustment device, characterized in that, include: An image acquisition module, comprising a high-resolution camera, which is installed at a preset position inside the vehicle to acquire full-body images of the occupants. A visual algorithm processing module, communicatively connected to the image acquisition module, is used to receive the full-body image and determine the height of the occupants using a visual algorithm. The visual algorithm includes the following steps: The full-body image is preprocessed, including image denoising, contrast enhancement, and distortion correction; The human body region in the image is identified by a human body detection algorithm, which is a convolutional neural network model based on deep learning. Extract feature points within the human body region, including the top of the head, the bottom of the feet, and at least three key joint points; Based on the camera's preset parameters and the coordinates of the feature points in the image, the actual height of the human body is calculated through perspective projection transformation. An ergonomic data storage module stores a database of optimal seat angles for different heights, and the database is constructed based on ergonomic experimental data. The seat adjustment control module is used to retrieve the corresponding seat angle from the database based on the height data output by the vision algorithm processing module, and control the electric adjustment mechanism to perform the adjustment action.

2. The vehicle seat adjustment device according to claim 1, characterized in that, The image acquisition module has two cameras, which are installed on the inside of the A-pillar and the front of the roof, respectively. The acquisition range of the two cameras covers the driver's seat and the passenger seat.

3. The vehicle seat adjustment device according to claim 1, characterized in that, The convolutional neural network model in the visual algorithm processing module is either the YOLO model or the SSD model, used for real-time detection of human body regions, with a detection frame rate of no less than 25fps.

4. A vehicle seat adjustment device according to claim 1, characterized in that, The database in the ergonomic data storage module also includes associated data on the seat's fore-and-aft distance, backrest tilt angle, and headrest height for different heights.

5. A vehicle seat adjustment device according to claim 1, characterized in that, The seat adjustment control module also includes a user feedback unit; The user feedback unit is used to receive the user's fine-tuning instruction for the automatically adjusted seat angle; and to store the angle deviation data corresponding to the fine-tuning instruction in the ergonomic data storage module to form a personalized preference database.

6. A vehicle seat adjustment device according to claim 5, characterized in that, When the same user is detected again, the seat adjustment control module will prioritize calling the data in the personalized preference database to perform the adjustment action; the user recognition is achieved by facial features or body features collected by the image acquisition module.

7. A method for adjusting a vehicle seat, characterized in that, Includes the following steps: After the vehicle starts, the image acquisition module captures full-body images of the occupants inside the vehicle. The full-body image is analyzed by a visual algorithm processing module to calculate the actual height of the people inside the vehicle. The visual algorithm includes human body detection and feature point extraction steps based on deep learning. The seat adjustment control module retrieves the matching seat angle parameters from the ergonomic data storage module based on the actual height. Control the electric adjustment mechanism of the seat to perform angle adjustment and monitor the adjustment process in real time; Receive user feedback on the adjustment results. If the user performs a fine-tuning operation, record the fine-tuned angle data and update it to the database.

8. A vehicle seat adjustment method according to claim 7, characterized in that, The real-time monitoring includes detecting changes in the person's sitting posture using pressure sensors built into the seat, and re-triggering the adjustment process if an abnormal sitting posture is detected.

9. An electronic device, characterized in that, include: The processor, communication interface, memory, and communication bus are connected, with the processor, communication interface, and memory communicating with each other via the communication bus. The memory stores a computer program that, when executed by the processor, causes the processor to perform the steps of a vehicle seat adjustment method according to any one of claims 7-8.

10. A computer-readable storage medium, characterized in that, It stores a computer program executable by an electronic device, which, when run on the electronic device, causes the electronic device to perform the steps of a vehicle seat adjustment method as described in any one of claims 7-8.

Images