Vehicle window control method and device and vehicle

Abstract

The invention relates to the technical field of vehicle window control, and particularly discloses a vehicle window control method and device and a vehicle. The method comprises the steps that in response to a first anti-pinch judgment result obtained according to Hall frequency changes, image information of a preset vehicle window area is obtained; identifying whether the image information accords with a preset anti-pinch triggering scene to obtain a second anti-pinch judgment result; and on the basis of the second anti-pinch judgment result, the vehicle window is controlled. The accuracy and the reliability of the anti-pinch function of the car window are effectively improved by combining a dual judgment mechanism of Hall frequency change and image information. According to the method, the potential anti-pinch risk can be intelligently recognized in various scenes, and the response is made in time, so that the safety and the use experience of a user are guaranteed, and the intelligent level of vehicle window control and the overall safety performance of the vehicle are improved.

Description

Technical Field

[0001] This application relates to the field of vehicle window control technology, specifically to a vehicle window control method, device, and vehicle. Background Technology

[0002] In recent years, with the improvement of automotive intelligence, the automatic anti-pinch function of power windows has become an important feature to ensure passenger safety, especially to prevent children from being pinched. Currently, most mainstream anti-pinch solutions rely on Hall effect sensors. The basic principle is: during the window's upward movement, the Hall effect frequency of the drive motor is monitored (its changes can indirectly reflect the motor speed and load) to determine whether an obstacle has been encountered.

[0003] However, existing anti-pinch schemes based on Hall sensors have limited judgment criteria, weak anti-interference capabilities, and cannot effectively distinguish between real risks and false alarms, resulting in a high false trigger rate and difficulty in guaranteeing the reliability of anti-pinch in actual complex working conditions. Summary of the Invention

[0004] In view of the above problems, this disclosure provides a method, device, and vehicle for controlling vehicle windows to overcome or at least partially solve the above problems. The technical solution is as follows: A vehicle window control method includes: in response to a first anti-pinch judgment result obtained based on a Hall frequency change, acquiring image information of a preset window area; identifying whether the image information matches a preset anti-pinch trigger scenario to obtain a second anti-pinch judgment result; and controlling the window based on the second anti-pinch judgment result.

[0005] By combining Hall effect frequency changes with image information in a dual-judgment mechanism, the accuracy and reliability of the anti-pinch window function are effectively improved. This method can intelligently identify potential anti-pinch risks in various scenarios and respond promptly, thereby ensuring user safety and user experience, and enhancing the intelligence level of window control and the overall vehicle safety performance.

[0006] In one example, controlling the window based on the second anti-pinch judgment result specifically includes: in response to the second anti-pinch judgment result being consistent with a preset anti-pinch trigger scenario, controlling the window to perform an anti-pinch rebound action; in response to the second anti-pinch judgment result not being consistent with the preset anti-pinch trigger scenario, controlling the window to perform an upward action at a preset rate.

[0007] This application further optimizes the judgment logic of the anti-pinch function of vehicle windows by introducing an image information recognition mechanism. In complex scenarios, such as rainy days, foreign object interference, or when a passenger's limbs briefly approach without posing an actual threat, relying solely on Hall frequency changes may lead to false alarms. However, by combining image recognition technology, it is possible to effectively distinguish between real risks and unnecessary triggering conditions, thereby significantly reducing the false alarm rate.

[0008] In one example, before responding to the first anti-pinch judgment result obtained based on the Hall frequency change, the method further includes: during the window raising process, acquiring the Hall frequency of the window motor and determining the Hall frequency change curve based on the Hall frequency; and when the Hall frequency change curve meets the preset first anti-pinch trigger condition, generating the first anti-pinch judgment result and pausing the window raising action.

[0009] This application dynamically assesses the window's position by monitoring the Hall effect frequency in real time during its upward movement and combining this with a preset first anti-pinch trigger condition. This method rapidly generates an initial anti-pinch assessment result upon detecting a potential obstacle, simultaneously pausing the window's upward movement and providing ample time for subsequent image recognition. This approach not only improves the anti-pinch function's response speed but also enhances the system's security.

[0010] In one example, the method further includes: recording events where the first anti-pinch judgment result conflicts with the second anti-pinch judgment result as false triggering events; when the cumulative number of false triggering events of the same window exceeds a threshold within a preset time period, adaptively increasing the sensitivity threshold of the first anti-pinch triggering condition on which the first anti-pinch judgment result depends.

[0011] This application introduces a self-learning mechanism, enabling the system to automatically adjust its sensitivity threshold when frequent false triggering events occur within a short period, thereby reducing unnecessary anti-pinch actions. This adaptive capability not only improves the system's intelligence level but also further optimizes the user experience. Through the analysis and learning of historical data, the system can dynamically adjust its judgment criteria under different environmental conditions, ensuring high reliability of the anti-pinch function in various scenarios. Furthermore, this method can effectively address sensor performance fluctuations caused by vehicle aging or changes in the external environment, extending the system's lifespan and reducing maintenance costs.

[0012] In one example, before responding to the first anti-pinch judgment result obtained based on the Hall frequency change, the method further includes: when triggering the window raising command, starting to move before the window, acquiring pre-start image information of the preset window area; identifying whether there is a pre-judged anti-pinch risk object in the pre-start image information; and in response to identifying the pre-judged anti-pinch risk object, executing a pre-start risk mitigation strategy, the pre-start risk mitigation strategy including at least one of the following: controlling the window to start rising at a slower speed than the standard rate, issuing an audible and visual warning, or prohibiting the window from starting to rise.

[0013] This application enhances the anti-pinch function's proactiveness by introducing a pre-activation image information recognition mechanism before the window starts moving. Before the window begins to move, the system can detect potential risk objects and take corresponding mitigation measures. When a predicted anti-pinch risk object is identified, the system can choose to slowly open the window, issue an audible and visual warning, or directly prevent the window from rising. This proactive risk management strategy not only improves system safety but also significantly enhances the user experience, especially in complex environments or high-risk scenarios. In this way, the window control system can intervene in potential threats at an earlier stage, effectively avoiding unnecessary false triggers or safety accidents.

[0014] In one example, identifying whether the image information matches a preset anti-pinch trigger scenario to obtain a second anti-pinch judgment result specifically includes: using a pre-trained recognition model to identify the image information to obtain the second anti-pinch judgment result; after controlling the window based on the second anti-pinch judgment result, the method further includes: binding the image information corresponding to the current trigger event with the window motion data and marking it as positive sample data; using the positive sample data to perform periodic incremental training on the recognition model.

[0015] This application further enhances the intelligence level of the anti-pinch function of car windows by introducing an image analysis mechanism based on a pre-trained recognition model. After the window is controlled, the system binds the image information of the triggered event with the window movement data and marks it as positive sample data. This data is then used to periodically and incrementally train the recognition model, enabling the model to continuously learn new scene features and optimize its judgment ability. Through this continuous learning approach, the system can gradually improve its adaptability to complex scenarios in practical applications, ensuring that the anti-pinch function always maintains high accuracy and high reliability. At the same time, this method also effectively reduces the need for manual intervention, significantly improving the system's automation level and long-term operational stability.

[0016] In one example, identifying whether the image information conforms to a preset anti-pinch trigger scenario specifically includes: using a first model to perform target detection on the image information to obtain the target object's category information and two-dimensional bounding box information; estimating the distance information of the target object relative to the window edge based on the two-dimensional bounding box information and camera calibration parameters; and combining the target object's category information and the distance information to determine whether it conforms to the preset anti-pinch trigger scenario.

[0017] This application enhances the accuracy of the anti-pinch function for car windows by introducing a multi-dimensional analysis method. After identifying the category and distance information of the target object, the system performs a comprehensive assessment based on dynamic environmental factors to determine whether there are any potential safety hazards. Furthermore, by predicting the trajectory of the target object, the system can anticipate potential risks and take corresponding preventative measures to ensure the safety and smoothness of window operation. This multi-layered, dynamic judgment mechanism improves the system's adaptability and reliability.

[0018] In one example, the method further includes: after the second anti-pinch judgment result is consistent with the preset anti-pinch trigger scenario, calculating the estimated time for the target object to enter the preset contact risk zone based on the distance information and the real-time movement speed of the window; and in response to the estimated time being less than the preset emergency threshold, sending an alarm message while performing the window rebound action.

[0019] After calculating the estimated time for a target object to enter the preset contact risk zone, the response strategy can be dynamically adjusted based on real-time data. If the estimated time is lower than the emergency threshold, the window will not only quickly rebound but also simultaneously trigger an alarm to alert occupants inside and outside the vehicle to the potential risk. This dual-response mechanism provides more comprehensive safety protection in emergencies while avoiding the shortcomings of a single measure. Furthermore, the form of the alarm message can be customized according to specific needs, such as through sound, light, or haptic feedback, ensuring that it receives sufficient attention from users in various scenarios. This design further enhances vehicle safety performance and provides users with a more reassuring user experience.

[0020] This application also provides a vehicle window control device, comprising: an image information acquisition module, which acquires image information of a preset vehicle window area in response to a first anti-pinch judgment result obtained based on a Hall frequency change; a second anti-pinch judgment module, which identifies whether the image information conforms to a preset anti-pinch trigger scenario to obtain a second anti-pinch judgment result; and a vehicle window control module, which controls the vehicle window based on the second anti-pinch judgment result.

[0021] This application also provides a vehicle, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform a window control method as described in any of the above examples.

[0022] By employing the aforementioned technical solution, the window control method, device, and vehicle disclosed herein offer the following overall benefits: By combining a dual judgment mechanism of Hall effect frequency changes and image information, the accuracy and reliability of the window anti-pinch function are significantly improved. This not only effectively reduces the false trigger rate but also intelligently identifies potential risks in complex scenarios and takes timely response measures, thereby ensuring user safety and user experience. Furthermore, by introducing a self-learning mechanism and periodic incremental training, the system possesses dynamic adaptability, enabling it to optimize judgment criteria under different environmental conditions, extend service life, and reduce maintenance costs. Overall, this solution enhances the vehicle's intelligence level while further strengthening the safety performance of window control, providing users with a safer and more convenient driving experience.

[0023] The above description is merely an overview of the technical solution disclosed herein. In order to better understand the technical means of this disclosure and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this disclosure more apparent and understandable, specific embodiments of this disclosure are described below. Attached Figure Description

[0024] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this disclosure. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating a window control method according to an embodiment of this disclosure is shown; Figure 2 A schematic diagram of the structure of a window control device according to an embodiment of the present disclosure is shown; Figure 3 A schematic diagram of the structure of a vehicle according to an embodiment of this disclosure is shown. Detailed Implementation

[0025] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0026] The technical solutions provided by the various embodiments of this application are described in detail below with reference to the accompanying drawings.

[0027] In recent years, with the improvement of automotive intelligence, the automatic anti-pinch function of power windows has become an important feature to ensure passenger safety, especially to prevent children from being pinched. Currently, most mainstream anti-pinch solutions rely on Hall effect sensors. The basic principle is to monitor the Hall frequency of the drive motor during the window's upward movement (its changes indirectly reflect motor speed and load) to determine if an obstacle has been encountered. This single Hall sensor-based anti-pinch solution has significant limitations in practical applications, leading to insufficient system reliability and a poor user experience. Specific problems include: First, the system is highly susceptible to environmental interference and changes in vehicle status, leading to false triggering. For example, under extreme temperature conditions, the resistance characteristics of the window guide rail rubber strip may change significantly; when the vehicle travels on bumpy roads, body vibration may cause fluctuations in motor load; or as the vehicle ages, the window lifting mechanism components may age and become insufficiently lubricated, all of which may cause abnormal fluctuations in the Hall frequency. These frequency changes, not caused by obstacle clamping, may be misinterpreted by the system as meeting the "anti-pinch trigger condition," resulting in unnecessary anti-pinch rebound. This accidental triggering not only disrupts the normal window closing operation and affects usability, but frequent malfunctions may also raise concerns among users about the vehicle's quality.

[0028] Secondly, the reliance on a single Hall frequency judgment logic fails to distinguish the fundamental differences in the sources of resistance. The system can only perceive the phenomenon of "increased resistance leading to decreased speed," but it cannot determine whether the resistance originates from living beings requiring emergency protection, such as the occupant's hands or head, or from non-living interference caused by environmental or mechanical factors (such as icing or uneven friction of the rubber strips). This results in the system employing the same response strategy when facing genuine anti-pinch scenarios and pseudo-scenario scenarios caused by interference, lacking targeted intelligent judgment capabilities.

[0029] In summary, existing anti-pinch schemes based on Hall sensors rely solely on a single physical signal (Hall frequency) for decision-making, resulting in a one-sided judgment basis, weak anti-interference capability, and an inability to effectively distinguish between real risks and false alarms. This leads to a high false trigger rate and difficulty in guaranteeing anti-pinch reliability under complex working conditions. Therefore, there is an urgent need for a window control scheme that can overcome the above defects and accurately and reliably perform anti-pinch judgments even in complex environments.

[0030] Therefore, this application provides a method for controlling vehicle windows, such as... Figure 1 The diagram shown is a flowchart illustrating a window control method provided in one or more embodiments of this specification. This method can be applied to a vehicle's window control system to achieve an anti-pinch function for the windows. The process can be executed by a control device installed in the vehicle (e.g., a vehicle-mounted control module connected to the window, or a server located in the cloud). Certain input parameters or intermediate results in the process can be manually adjusted to help improve accuracy.

[0031] The analysis method involved in the embodiments of this application can be implemented by a terminal device or a server, and this application does not impose any special limitations on it. For ease of understanding and description, the following embodiments are all described in detail using an in-vehicle infotainment system as an example.

[0032] It should be noted that the server can be a single device or a system composed of multiple devices, i.e., a distributed server. This application does not make any specific limitations in this regard.

[0033] like Figure 1 As shown in the figure, this application provides a method for controlling vehicle windows, including: S101: In response to the first anti-pinch judgment result obtained based on the Hall frequency change, obtain image information of the preset window area.

[0034] When the vehicle's infotainment system detects a change in the window's Hall effect frequency, and the initial anti-pinch assessment indicates a risk of injury, requiring the window to stop rising or retract, the system immediately triggers the image acquisition module to obtain real-time image information of a preset window area. This process ensures that when a Hall effect frequency change indicates a potential risk, visual data can be used to further verify the presence of a real obstacle. The image acquisition module is typically equipped with a high-resolution camera and dynamically adjusts exposure parameters using an ambient light sensor to adapt to shooting needs under different lighting conditions, thereby ensuring stable image quality.

[0035] S102: Identify whether the image information matches the preset anti-pinch trigger scenario to obtain the second anti-pinch judgment result.

[0036] After acquiring image information of the window area, the vehicle system analyzes the image content to determine whether there is a target object that matches the preset anti-pinch trigger scenario. When a hand, head, or other soft object is detected in the window area, it is considered that there is a target object that matches the preset anti-pinch trigger scenario, and a corresponding second anti-pinch judgment result is generated. S103: Based on the second anti-pinch judgment result, control the window.

[0037] Upon receiving the second anti-pinch judgment result, the vehicle's infotainment system will immediately take corresponding control measures based on this result. These measures may include controlling the window to continue rising, stopping the window from rising, performing a rebound action, or lowering the window to a safe position, thereby controlling the window. This image-based secondary judgment mechanism effectively compensates for the shortcomings of relying solely on Hall frequency judgment, significantly improving the accuracy of the anti-pinch function and vehicle safety. In one embodiment, after obtaining the second anti-pinch judgment result, different control strategies are adopted based on this result when controlling the window to ensure the safety and reliability of window operation. For example, if the second anti-pinch judgment result confirms the presence of a real obstacle, the system will prioritize executing an emergency rebound action and adjust the window's movement speed to avoid further risks. If the second anti-pinch judgment result indicates the absence of an obstacle, the system will continue to complete the normal raising or lowering operation of the window, thereby reducing unnecessary interruptions and improving the user experience.

[0038] By combining Hall effect frequency changes with image information in a dual-judgment mechanism, the window control system can more accurately identify potential risks in complex scenarios. For example, during vehicle operation, the system can effectively distinguish between mechanical vibrations caused by road bumps and resistance changes caused by actual obstacles, thus avoiding false triggering. Simultaneously, through in-depth analysis of image information, the system can further confirm the nature of obstacles, such as whether they are human body parts or soft objects, ensuring the targeted and effective nature of anti-pinch actions. This multi-layered judgment logic not only enhances the system's intelligence level but also significantly strengthens users' sense of security and trust in actual use.

[0039] In one embodiment, when determining the first anti-pinch judgment result, the Hall frequency of the window motor can be acquired during the window's upward movement, and a Hall frequency change curve can be determined based on the Hall frequency. In response to the Hall frequency change curve meeting a preset first anti-pinch trigger condition, a first anti-pinch judgment result is generated, and the window's upward movement is paused. Specifically, when the window encounters an obstacle within the anti-pinch area, the upward resistance increases, the motor speed decreases, and the Hall frequency changes accordingly. The system determines whether to trigger an anti-pinch action (such as stopping the upward movement or reversing the downward movement) by detecting whether the frequency change curve matches a preset "anti-pinch trigger condition" (such as a specific deceleration mode). In this way, the system can monitor the Hall frequency change in real time during window operation and quickly generate a first anti-pinch judgment result by comparing it with preset conditions. This method not only improves detection efficiency but also provides a reliable triggering basis for the subsequent introduction of image information.

[0040] Furthermore, in practical applications, to further improve the accuracy of judgment, the system can also combine historical data and dynamic environmental parameters for comprehensive evaluation. For example, based on the characteristics of different vehicle models and window structures, personalized Hall frequency reference values ​​and change thresholds can be preset; or the sensitivity of the anti-pinch trigger can be dynamically adjusted according to the vehicle's current driving status (such as whether it is on a bumpy road) and external ambient temperature. This flexible adaptive mechanism enables the window control system to maintain stable performance under various complex operating conditions, effectively avoiding false triggering or missed triggering problems caused by a single fixed standard.

[0041] To enhance system robustness, multi-sensor fusion technology can be introduced. For example, data from Hall effect sensors can be combined with other types of sensors (such as pressure sensors and ultrasonic sensors) to build a more comprehensive sensing network. Specifically, pressure sensors can detect abnormal pressure distributions on window guide rails, while ultrasonic sensors can detect the presence of obstacles near the window without contact. These additional sensing methods not only further improve the accuracy of anti-pinch detection but also provide redundancy for the system; even if one sensor malfunctions or is interfered with, other sensors can still maintain normal system operation. Through cross-validation of multi-source data, the system can more accurately identify the presence and nature of obstacles, thereby further optimizing the response speed and reliability of the anti-pinch function.

[0042] In one embodiment, when the number of false alarms in the first anti-pinch judgment result is too high within a short period, a self-learning step can be added to the generation process of the first anti-pinch judgment result. This involves recording events where the first and second anti-pinch judgment results conflict as false triggering events. When the cumulative number of false triggering events for the same window exceeds a threshold within a preset time period, the sensitivity threshold of the first anti-pinch triggering condition upon which the first anti-pinch judgment result depends is adaptively increased. Specifically, the vehicle system automatically analyzes the characteristics of false triggering events, extracts relevant environmental parameters and window operating status data, and dynamically adjusts the judgment threshold of the Hall frequency change curve based on this information. For example, when false triggering occurs frequently on bumpy roads or under extreme temperature conditions, the system will appropriately relax the sensitivity to abnormal fluctuations in the Hall frequency, thereby reducing the false alarm rate. Simultaneously, this self-learning mechanism can continuously optimize model parameters through periodic incremental training, allowing the system to gradually adapt to individual vehicle characteristics and user habits during long-term use, further improving the accuracy of the judgment. Furthermore, to ensure that the adjusted sensitivity threshold does not significantly reduce the response speed of the anti-pinch function, the system performs real-time verification after each update, evaluating the performance of the new threshold in real-world scenarios and fine-tuning it based on feedback to achieve the optimal balance between safety and convenience. This intelligent learning capability not only enhances the system's adaptability but also provides users with a more stable and reliable window control experience.

[0043] In one embodiment, besides performing image recognition after obtaining the first anti-pinch judgment result, the system can also start moving before the window is triggered when the window raising command is activated, acquiring pre-start image information of the preset window area; identifying whether there is a predicted anti-pinch risk object in the pre-start image information; and executing a pre-start risk mitigation strategy in response to the identification of a predicted anti-pinch risk object. The pre-start risk mitigation strategy includes at least one of the following: controlling the window to start rising slowly at a rate lower than the standard rate, issuing an audible and visual warning, or prohibiting the window from starting to rise. Specifically, before the window raising command is triggered, the system can detect potential risk objects in advance through the acquisition and analysis of pre-start image information. For example, when a hand, head, or other object that may cause pinching is detected in the window area, the system will take corresponding mitigation measures according to preset rules.

[0044] This predictive mechanism not only effectively reduces the risk of pinching injuries before the window is opened, but also provides users with additional safety. Furthermore, by combining multimodal alerts such as audible and visual warnings, the vehicle's infotainment system can further enhance the user's perception of potential hazards, thereby preventing accidents caused by negligence. This forward-thinking design significantly improves the active safety of the window control system, while also demonstrating the optimization and innovation of intelligent technology in its details.

[0045] During the pre-start image information recognition process, the same recognition method as the second anti-pinch judgment result generation process can be adopted. However, the confidence level required for the pre-start image information recognition process can be appropriately reduced. That is, in the pre-start phase, the system's requirements for image information recognition accuracy can be relatively relaxed to improve detection efficiency and reduce computational resource consumption. For example, in the pre-start image information recognition process, the system can speed up processing by simplifying the algorithm or reducing the number of analysis frames, while ensuring that obvious risk objects can be captured. This strategy can ensure safety while avoiding delay problems caused by excessive precision. In addition, since the car window has not yet started to move during the pre-start phase, even if a slight misjudgment occurs, it will not have an actual impact on the user. Therefore, appropriately reducing the confidence threshold is reasonable and feasible.

[0046] Furthermore, to enhance the reliability of pre-start image information recognition, the system can be optimized by incorporating historical data. For example, by recording and analyzing past user operating habits and risk distribution patterns in specific scenarios, the system can more accurately predict potential safety hazards and dynamically adjust the risk assessment model during the pre-start phase. For instance, if a user frequently places items in the rear seats, the system can strengthen monitoring of that area; or, in situations where children frequently ride in the car, priority can be given to monitoring potential physical activities near the windows. This intelligent optimization based on personalized data not only improves the system's usability but also provides more targeted safety protection for different user groups.

[0047] Furthermore, when implementing pre-activation risk mitigation strategies, the system can flexibly select from various combinations of measures based on the actual situation. For example, when a high-probability risk object is identified, in addition to controlling the window to open slowly, an audible and visual warning can be triggered simultaneously to alert passengers to potential danger. For extremely low-probability but high-risk situations (such as detecting a head near the window edge), the window can be directly prevented from opening and rising until safety is confirmed. This tiered response mechanism balances user experience with maximizing passenger safety.

[0048] Furthermore, to meet the personalized needs of different users, this application also provides a configurable anti-pinch strategy management mechanism. Users can customize parameters such as the sensitivity, response mode, and warning method of the anti-pinch function through the vehicle system interface or a mobile application. For example, families with children can choose higher anti-pinch sensitivity and stricter judgment conditions; while users who prioritize ease of operation can appropriately reduce sensitivity to minimize unnecessary interruptions. This flexible configuration capability not only enhances the system's applicability but also allows users to adjust the window control behavior according to their own needs, thereby obtaining an experience more closely aligned with actual usage scenarios. Simultaneously, the system also supports a learning function based on users' historical operation data, automatically recommending the optimal anti-pinch strategy configuration, further simplifying the user's setup process and improving overall satisfaction.

[0049] In one embodiment, when generating the second anti-pinch judgment result, a pre-trained recognition model can be used to recognize the image information to obtain the second anti-pinch judgment result. When using the pre-trained recognition model, the system can quickly locate and classify target objects in the image using deep learning algorithms. For example, the YOLO model can be used to quickly locate and classify target objects in the image. The YOLO model is a deep learning object detection algorithm that can analyze the image of the car window area in a short time and output a judgment result indicating whether a target object exists that matches the preset anti-pinch trigger scenario. To further improve recognition accuracy, the model can also be fine-tuned based on the specific characteristics of the car window area, such as optimizing model parameters for different lighting conditions, car window materials, or interior decoration styles. Furthermore, the system can periodically update model weights to adapt to new scenario requirements or user feedback, thereby ensuring that the recognition capability remains at a high level.

[0050] Furthermore, during the generation of the second anti-pinch judgment result, the system can also incorporate a multi-frame image temporal analysis mechanism. By monitoring the dynamic changes of consecutive multi-frame images, the system can more accurately determine the movement trend and potential risks of the target object. For example, when an object is detected gradually approaching the edge of the car window, even if the current frame does not fully meet the anti-pinch trigger conditions, the system can issue an early warning and take corresponding preventive measures. This temporal information-based analysis method not only enhances the foresight of the judgment but also effectively reduces erroneous actions caused by misjudgments of single-frame images.

[0051] In certain special scenarios, such as low-light environments at night or under strong backlight conditions, relying solely on image information may not fully meet the requirements for anti-pinch detection. Therefore, the system can combine other auxiliary data sources for comprehensive decision-making. For example, by fusing data from infrared sensors or depth cameras, thermal imaging information or the three-dimensional spatial position of the target object can be obtained, thus compensating for the limitations of visible light images under specific conditions. This multimodal data fusion approach not only improves the system's robustness but also provides more reliable protection for anti-pinch functionality in complex environments.

[0052] Furthermore, based on the second anti-pinch judgment result, after controlling the car window, the image information corresponding to this triggering event can be bound with the car window motion data and marked as positive sample data; and the positive sample data can be used to perform periodic incremental training on the recognition model.

[0053] In this way, the system can continuously optimize the performance of the recognition model, gradually adapting it to diverse scenario requirements in real-world applications. For example, over long-term use, the system can accumulate a large amount of positive sample data containing different lighting conditions, object types, and motion patterns, and use this data to fine-tune the model. This incremental training mechanism based on real-world scene data not only improves the model's generalization ability but also significantly enhances its judgment accuracy in complex environments. Furthermore, to ensure the efficiency and stability of the training process, the system can also adopt a batch update strategy, using only a portion of the newly added data for model optimization each time, avoiding wasted computational resources or overfitting problems caused by excessive data volume. Simultaneously, the system periodically evaluates the model's performance after updates and adjusts training parameters or re-selects training data based on actual results to ensure the model remains in optimal condition. This continuous learning capability enables the window control system to maintain a high level of reliability and intelligence in dynamically changing usage environments, providing users with a safer and more convenient experience.

[0054] In one embodiment, in addition to using a recognition model, the distance between the target object and the window area can be determined to determine whether the image information meets the preset anti-pinch trigger scenario. The first model can be used to perform target detection on the image information to obtain the target object's category information and two-dimensional bounding box information; based on the two-dimensional bounding box information and camera calibration parameters, the distance information of the target object relative to the window edge is estimated; combining the target object's category information and the distance information, it is determined whether the preset anti-pinch trigger scenario is met.

[0055] Specifically, by calibrating parameters using cameras, the system can map two-dimensional information from images to real three-dimensional space, thereby estimating the actual distance between the target object and the car window. For example, when an object is detected as a human body part and its distance from the car window is less than a preset threshold, the system determines that it meets the preset anti-pinch trigger scenario and immediately takes corresponding control measures. This distance-based judgment method not only improves the accuracy of recognition but also effectively avoids the misjudgment problems that may arise from relying solely on image classification. Furthermore, to further improve estimation accuracy, the system can also combine data from multi-view cameras and use methods such as triangulation to more accurately locate the spatial position of the target object. This method is particularly advantageous in complex scenarios, such as when the target object is partially occluded or in a non-standard posture, it still maintains high judgment reliability.

[0056] Meanwhile, to address the differences in vehicle models and window structures, the system supports personalized calibration of the distance estimation model. For example, calibration parameters and distance thresholds are adjusted based on the specific vehicle's window shape, installation angle, and camera position to ensure the applicability and consistency of the judgment results. This flexible adaptability allows the anti-pinch function to perform excellently on various vehicle models, further enhancing the system's versatility. Furthermore, the system regularly optimizes calibration parameters, dynamically adjusting them based on data accumulated during actual use, thereby continuously improving the accuracy and stability of distance estimation.

[0057] In certain special circumstances, such as when the camera lens is damaged or there are drastic changes in ambient light, relying solely on image information may lead to inaccurate distance estimations. To address this, the system can introduce a redundancy mechanism by fusing data from other sensors (such as ultrasonic ranging or LiDAR) to verify the reliability of the estimation results. For example, when there is a significant difference between the target distance derived from image analysis and the measurement value from the ultrasonic sensor, the system will prioritize the more reliable data source and trigger anomaly handling procedures, such as pausing window operation or issuing a warning. This multi-sensor collaborative approach not only improves the system's robustness but also provides users with more comprehensive safety assurance.

[0058] Furthermore, to enhance user experience, the system can dynamically adjust the anti-pinch trigger strategy based on the category of the target object. For example, for high-risk objects (such as a child's hand or head), the system sets a stricter distance threshold and a faster response speed; while for low-risk objects (such as interior decorations or fixed obstacles), the judgment conditions can be appropriately relaxed to reduce unnecessary interference. This tiered processing mechanism not only balances safety and convenience but also reflects the meticulous design of intelligent technology in the details. By continuously optimizing algorithms and integrating multi-source data, the window control system can achieve efficient and accurate anti-pinch functionality in various complex scenarios, providing users with a more reassuring user experience.

[0059] In one embodiment, when the object at risk is too close to the window and the window's initial upward speed is too fast, an alarm message can be sent to prevent the object from approaching further and thus avoid potential pinching accidents. Specifically, after the second anti-pinch judgment result meets the preset anti-pinch trigger scenario, the estimated time for the target object to enter the preset contact risk zone is calculated based on the distance information and the window's real-time movement speed. If the estimated time is less than a preset emergency threshold, an alarm message is sent while the window is retracting.

[0060] Specifically, the system dynamically assesses the risk level based on the distance to the target object and the current operating status of the window, triggering an alarm mechanism when necessary. For example, if a child's hand is detected to be less than a safety threshold from the window edge and the window is rapidly rising, the system will immediately issue an audible and visual warning to alert occupants. Simultaneously, the window's rising motion may be temporarily slowed or paused to allow sufficient time for the user to take appropriate action. This real-time risk intervention strategy not only effectively prevents accidents but also significantly enhances the user's perceived sense of security. Furthermore, to ensure efficient alarm information delivery, the system can optimize the prompting method based on the characteristics of the in-vehicle environment. For instance, in noisy environments, the system may prioritize high-brightness visual warnings, while in quiet scenarios, it tends to use softer audio prompts to avoid unnecessary disturbance to the user. This flexible alarm mechanism further enhances the system's adaptability and user-friendly design.

[0061] In terms of user experience, this application's design fully considers ease of operation and timely feedback. When the system detects a potential risk and takes corresponding measures, it will convey relevant information to the user in an intuitive way. For example, the system can display concise prompts on the in-vehicle display screen or inform the user of the current operating status via voice broadcast. This transparent interactive design not only enhances the user's understanding of the system but also effectively reduces confusion caused by misoperation or information asymmetry, thereby improving overall user satisfaction.

[0062] By employing the aforementioned technical solution, the window control method, device, and vehicle disclosed herein offer the following overall benefits: By combining a dual judgment mechanism of Hall effect frequency changes and image information, the accuracy and reliability of the window anti-pinch function are significantly improved. This not only effectively reduces the false trigger rate but also intelligently identifies potential risks in complex scenarios and takes timely response measures, thereby ensuring user safety and user experience. Furthermore, by introducing a self-learning mechanism and periodic incremental training, the system possesses dynamic adaptability, enabling it to optimize judgment criteria under different environmental conditions, extend service life, and reduce maintenance costs. Overall, this solution enhances the vehicle's intelligence level while further strengthening the safety performance of window control, providing users with a safer and more convenient driving experience.

[0063] In addition, such as Figure 2 As shown, Figure 2 This is a schematic diagram of the structure of the window control device provided in the embodiment of this application. The device includes: The image information acquisition module 201 acquires image information of a preset window area in response to the first anti-pinch judgment result obtained based on the Hall frequency change.

[0064] The second anti-pinch judgment module 202 identifies whether the image information matches a preset anti-pinch trigger scenario to obtain a second anti-pinch judgment result.

[0065] The window control module 203 controls the window based on the second anti-pinch judgment result.

[0066] In one specific embodiment, the image information acquisition module 203 includes: in response to the second anti-pinch judgment result being consistent with a preset anti-pinch triggering scenario, controlling the window to perform an anti-pinch rebound action; and in response to the second anti-pinch judgment result being inconsistent with the preset anti-pinch triggering scenario, controlling the window to perform an upward action at a preset rate.

[0067] In one specific embodiment, the image information acquisition module 201 includes: acquiring the Hall frequency of the window motor during the window raising process, and determining the Hall frequency change curve based on the Hall frequency; and generating the first anti-pinch judgment result and pausing the window raising action when the Hall frequency change curve meets the preset first anti-pinch trigger condition.

[0068] In one specific embodiment, the image information acquisition module 201 includes: recording events where the first anti-pinch judgment result and the second anti-pinch judgment result conflict as false triggering events; when the cumulative number of false triggering events of the same window exceeds a threshold within a preset time period, adaptively increasing the sensitivity threshold of the first anti-pinch triggering condition on which the first anti-pinch judgment result depends.

[0069] In one specific embodiment, the second anti-pinch judgment module 201 includes: when a window raising command is triggered, starting to move before the window, acquiring pre-start image information of the preset window area; identifying whether there is a pre-judged anti-pinch risk object in the pre-start image information; and in response to identifying the pre-judged anti-pinch risk object, executing a pre-start risk mitigation strategy, wherein the pre-start risk mitigation strategy includes at least one of the following: controlling the window to start rising at a slower speed than the standard rate, issuing an audible and visual warning, or prohibiting the window from starting to rise.

[0070] In one specific embodiment, the second anti-pinch judgment module 202 includes: using a pre-trained recognition model to recognize the image information to obtain the second anti-pinch judgment result; after controlling the window based on the second anti-pinch judgment result, binding the image information corresponding to the current triggering event with the window motion data and marking it as positive sample data; and using the positive sample data to perform periodic incremental training on the recognition model.

[0071] In one specific embodiment, the second anti-pinch judgment module 202 includes: using a first model to perform target detection on the image information to obtain the category information and two-dimensional bounding box information of the target object; estimating the distance information of the target object relative to the edge of the car window based on the two-dimensional bounding box information and camera calibration parameters; and combining the category information of the target object and the distance information to determine whether it meets the preset anti-pinch triggering scenario.

[0072] In one specific embodiment, the second anti-pinch judgment module 202 includes: after the second anti-pinch judgment result is consistent with the preset anti-pinch trigger scenario, calculating the estimated time for the target object to enter the preset contact risk zone based on the distance information and the real-time movement speed of the window; and in response to the estimated time being less than the preset emergency threshold, sending an alarm message while performing the window rebound action.

[0073] Regarding the apparatus in the above embodiments, the specific manner in which each unit performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0074] Figure 3 This is a schematic diagram of the structure of a vehicle provided in an embodiment of this application.

[0075] For example, such as Figure 3 As shown, the vehicle includes a memory 301 and a processor 302. The memory 301 stores executable program code 3011, and the processor 302 is used to call and execute the executable program code 3011 to perform a window control method.

[0076] This embodiment can divide the vehicle into functional modules according to the above method example. For example, each function can be assigned to a separate module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware. It should be noted that the module division in this embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0077] When each functional module is divided according to its corresponding function, the vehicle may include: The image information acquisition module, in response to the first anti-pinch judgment result obtained based on the Hall frequency change, acquires image information of a preset window area; The second anti-pinch judgment module identifies whether the image information matches a preset anti-pinch trigger scenario to obtain a second anti-pinch judgment result; The window control module controls the windows based on the second anti-pinch judgment result.

[0078] It should be noted that all relevant content of each step involved in the above method embodiments can be referenced from the functional description of the corresponding functional module, and will not be repeated here.

[0079] The vehicle provided in this embodiment is used to execute the above-described window control method, and therefore can achieve the same effect as the above-described implementation method.

[0080] When using integrated units, the vehicle may include a processing module and a storage module. The processing module is used to control and manage the vehicle's actions. The storage module supports the vehicle in executing program code and data.

[0081] The processing module may be a processor or a controller, which can implement or execute various exemplary logic blocks, modules, and circuits as disclosed in this application. The processor may also be a combination of computing functions, such as a combination of one or more microprocessors, a combination of digital signal processing (DSP) and a microprocessor, etc., and the storage module may be a memory.

[0082] This embodiment also provides a computer-readable storage medium (including but not limited to disk storage, CD-ROM, optical storage, etc.) storing computer program code. When the computer program code is run on a computer, the computer executes the above-mentioned related method steps to implement the window control method provided in the above embodiment.

[0083] This embodiment also provides a computer program product that, when run on a computer, causes the computer to perform the aforementioned steps to implement the window control method provided in the above embodiment.

[0084] The beneficial effects of the above embodiments can be referred to the beneficial effects of the corresponding methods provided above, and will not be repeated here.

[0085] Through the above description of the embodiments, those skilled in the art will understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0086] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0087] In the description of this disclosure, it should be understood that if the terms "upper", "lower", "front", "rear", "left" and "right" are used to indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the position or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this disclosure.

[0088] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0089] The above are merely embodiments of this disclosure and are not intended to limit the scope of this disclosure. Various modifications and variations can be made to this disclosure by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the scope of the claims of this disclosure.

Claims

1. A method for controlling vehicle windows, characterized in that, include: In response to the first anti-pinch judgment result obtained based on the Hall frequency change, image information of the preset window area is acquired; The image information is identified to determine whether it matches a preset anti-pinch trigger scenario, in order to obtain a second anti-pinch judgment result. Based on the second anti-pinch judgment result, the car window is controlled.

2. The method according to claim 1, characterized in that, The step of controlling the vehicle window based on the second anti-pinch judgment result specifically includes: If the second anti-pinch judgment result is consistent with the preset anti-pinch trigger scenario, then the window is controlled to perform an anti-pinch rebound action; If the second anti-pinch judgment result does not meet the preset anti-pinch trigger scenario, the window is controlled to perform an upward action at a preset rate.

3. The method according to claim 1, characterized in that, Before responding to the first anti-pinch judgment result obtained based on the Hall frequency change, the method further includes: During the window raising process, the Hall frequency of the window motor is acquired, and the Hall frequency change curve is determined based on the Hall frequency. When the Hall frequency change curve meets the preset first anti-pinch trigger condition, the first anti-pinch judgment result is generated, and the window raising action is paused.

4. The method according to claim 3, characterized in that, The method further includes: The event in which the first anti-pinch judgment result conflicts with the second anti-pinch judgment result is recorded as a false trigger event; When the cumulative number of false trigger events of the same car window within a preset time exceeds a threshold, the sensitivity threshold of the first anti-pinch trigger condition on which the first anti-pinch judgment result depends is adaptively increased.

5. The method according to claim 1, characterized in that, Before responding to the first anti-pinch judgment result obtained based on the Hall frequency change, the method further includes: When the command to raise the window is triggered, the window begins to move before the window itself, and pre-start image information of the preset window area is obtained; Identify whether the pre-startup image information contains any objects that are pre-judged as having a risk of being pinched; In response to the identification of a pre-judged anti-pinch risk object, a pre-start risk mitigation strategy is executed, which includes at least one of the following: controlling the window to start rising slowly at a rate lower than the standard rate, issuing an audible and visual warning, or prohibiting the window from starting to rise.

6. The method according to claim 1, characterized in that, The step of identifying whether the image information matches a preset anti-pinch trigger scenario to obtain a second anti-pinch judgment result specifically includes: The image information is identified using a pre-trained recognition model to obtain the second anti-pinch judgment result; After controlling the vehicle window based on the second anti-pinch judgment result, the method further includes: The image information corresponding to this triggered event is bound to the window motion data and marked as positive sample data; The recognition model is periodically incrementally trained using the positive sample data.

7. The method according to claim 1, characterized in that, The step of identifying whether the image information matches a preset anti-pinch trigger scenario specifically includes: The first model is used to perform target detection on the image information to obtain the category information and two-dimensional bounding box information of the target object; Based on the two-dimensional bounding box information and camera calibration parameters, the distance information of the target object relative to the edge of the car window is estimated; By combining the category information of the target object and the distance information, it is determined whether the preset anti-pinch trigger scenario is met.

8. The method according to claim 7, characterized in that, The method further includes: After the second anti-pinch judgment result is that it meets the preset anti-pinch trigger scenario, the estimated time for the target object to enter the preset contact risk zone is calculated based on the distance information and the real-time movement speed of the window. If the estimated time is less than a preset emergency threshold, an alarm message is sent while the window is being retracted.

9. A vehicle window control device, characterized in that, include: The image information acquisition module, in response to the first anti-pinch judgment result obtained based on the Hall frequency change, acquires image information of a preset window area; The second anti-pinch judgment module identifies whether the image information matches a preset anti-pinch trigger scenario to obtain a second anti-pinch judgment result; The window control module controls the windows based on the second anti-pinch judgment result.

10. A vehicle, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the window control method as described in any one of claims 1 to 8.

Images