Method and system for detecting movement separation of driver of auxiliary driving system

Abstract

The invention discloses a method and system for detecting movement separation of a driver of an auxiliary driving system, and belongs to the technical field of auxiliary driving systems.The method comprises the steps that a real-time image of a steering wheel area is collected through an adjusted DMS camera and an adjusted OMS camera, and the real-time image is input into a trained visual model; performing fusion processing on the real-time image through a visual model to obtain a visual judgment result and a corresponding credibility; the auxiliary driving system receives steering wheel steering torque fed back by the EPS, processes the steering torque and compares the steering torque with a preset threshold value, and a torque judgment result whether a driver holds the steering wheel or not is obtained; based on the visual judgment result and the corresponding credibility and torque judgment result, comprehensive judgment is carried out through a judgment module, and the result of whether the driver is in the motion separation state or not is obtained. According to the invention, a steering wheel area acquisition function is added, comprehensive judgment of visual judgment and torque judgment is combined, and low-cost and high-accuracy detection is realized.

Description

Technical Field

[0001] This invention belongs to the field of driver assistance system technology, and specifically relates to a method and system for detecting driver disengagement in a driver assistance system. Background Technology

[0002] Currently, driver assistance systems generally employ two methods for driver motion detection. The first method uses a conventional steering wheel, detecting the tangential torque of the driver's hands on the steering wheel to determine if the driver has engaged in any movement. In this method, the driver assistance system determines whether the torque fed back by the EPS steering system exceeds a calibrated threshold. The second method uses a capacitive steering wheel, which determines whether the driver is holding the steering wheel based on the capacitance between the driver's hands and the steering wheel. The driver assistance system then receives the status signal from the steering wheel to determine if the driver has engaged in any movement.

[0003] While the first method currently offers the best cost-effectiveness, the driver's hands on the steering wheel don't necessarily generate tangential torque. Therefore, if the driver's hands are lightly gripping the steering wheel, the system may fail to detect this, triggering a false alarm that the driver has taken their hands off the wheel. If the driver tries to release the hands, they can only turn the steering wheel, impacting the user experience and potentially disabling the driver assistance function. The second method is more expensive but has higher accuracy. However, if the driver has a steering wheel cover, the system may fail to detect the driver's hands, leading to false alarms of hands being off the wheel. In summary, current driver assistance systems struggle to effectively detect whether the driver is touching or gripping the steering wheel, resulting in false alarms of hands being off the wheel. Further optimization is needed to accurately detect whether the driver can control the steering wheel in a timely manner. Summary of the Invention

[0004] This invention provides a method and system for detecting driver disengagement in a driver assistance system. The purpose is to solve the problems in current driver assistance systems, such as difficulty in effectively identifying whether the driver is in contact with or holding the steering wheel, the possibility of false alarms of disengagement, and the need for further optimization in the accuracy of detecting whether the driver can control the steering wheel in time.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: This invention provides a method for detecting driver disengagement in a driver assistance system, comprising the following steps: S1. Real-time images of the steering wheel area are acquired by the adjusted DMS camera and OMS camera. The real-time images are input into the trained visual model. The visual model performs fusion processing on the real-time images to obtain the visual judgment result and its corresponding trust level. S2. The driver assistance system receives the steering torque from the EPS, processes the steering torque and compares it with a preset threshold to obtain the torque judgment result of whether the driver is holding the steering wheel. S3. Based on the visual judgment results and their corresponding trust level and torque judgment results, the judgment module makes a comprehensive judgment to obtain the result of whether the driver is in a state of motion disengagement. The adjustment of the DMS and OMS cameras includes: adjusting the installation position and angle of the DMS and OMS cameras to make them compatible with image acquisition of the steering wheel area on the original basis; The training process of the trained visual model includes: collecting visual images of the steering wheel area covered by the adjusted DMS camera and OMS camera, adding the visual images of the steering wheel area to the visual training dataset, training the visual training dataset through the training model, and evaluating the preset metrics of the visual model.

[0006] In some implementations, the visual model performs fusion processing on real-time images through a visual fusion judgment module, specifically including: the DMS camera and the OMS camera simultaneously capture real-time images of the steering wheel area within their respective fields of view; the visual fusion judgment module removes information outside the steering wheel area from the images, fuses the real-time images from the two cameras, and outputs an image around the steering wheel; and matches the coverage position of the steering wheel area in the fused image with the actual position of the driver's hands. If the scene cannot be covered, output an invalid visual judgment result; if the scene can be covered, output the visual judgment result of whether it can be released and the corresponding trust level.

[0007] In some implementations, the preset metrics include the false alarm rate and false negative rate of the visual model.

[0008] In some implementations, the steering torque feedback from the EPS is processed using a low-pass filter to remove signal interference.

[0009] In some implementations, the steering torque is processed and then compared with a preset threshold, specifically including: like And the duration is greater than If so, it is determined to be a slip of the hand; like If the duration is greater than t2, it is determined that the item has not been sold. in, The processed torque signal, For torque threshold, t1 is the first time threshold, and t2 is the second time threshold. To prevent sudden changes in speed, the following ranges apply at different vehicle speeds. Calibration needs to be performed based on the actual vehicle. , The debounce interval is set to prevent result jumps.

[0010] In some implementation methods, the comprehensive judgment is based on: When both the visual judgment result and the torque judgment result indicate that the hands have slipped, the final determination is that the hands have slipped. When the visual judgment result is "released" and the confidence level is not less than the preset value, and the torque judgment result is "not released", the final judgment is "released"; when the visual judgment result is "released" and the confidence level is less than the preset value, and the torque judgment result is "not released", the final judgment is "not released". When both the visual judgment result and the torque judgment result indicate that the hand will not be released, the final determination is that the hand will not be released; when the visual judgment result indicates that the hand will not be released, the confidence level is not less than the preset value, and the torque judgment result indicates that the hand will be released, the final determination is that the hand will not be released. When the visual judgment result is "not to be released" and the confidence level is less than the preset value, and the torque judgment result is "release", the final judgment is "release".

[0011] In some implementations, if the visual judgment result is that the scene cannot be covered and the torque judgment result is that the hand is released, then the final determination is that the hand is released; if the visual judgment result is that the scene cannot be covered and the torque judgment result is that the hand is not released, then the final determination is that the hand is not released.

[0012] In some implementations, the original basis of the DMS camera includes detecting driver fatigue and distraction, covering the driver's facial expressions and eye movements; the original basis of the OMS camera includes detecting the status of vehicle occupants, covering the driver, front passenger and rear seat areas.

[0013] In some implementations, the visual images include various scene images of whether or not the driver's hands are gripping the steering wheel. These scene images include: images of the driver's various hand gestures distributed around the steering wheel after being split, images under different lighting and angles, and negative sample images of the driver not gripping the steering wheel.

[0014] This invention also provides a detection system for driver disengagement in a driver assistance system, for implementing the aforementioned method for detecting driver disengagement in a driver assistance system, comprising a data acquisition and fusion processing module, a torque judgment module, and a comprehensive judgment module, wherein: Data acquisition and fusion processing module: used to acquire real-time images of the steering wheel area through the adjusted DMS camera and OMS camera, input the real-time images into the trained visual model, and perform fusion processing on the real-time images by the visual model to obtain the visual judgment result and its corresponding trust level. Torque determination module: Used by the driver assistance system to receive the steering torque from the EPS, process the steering torque and compare it with a preset threshold to obtain the torque determination result of whether the driver is holding the steering wheel. Comprehensive Judgment Module: Based on the visual judgment results and their corresponding trust level and torque judgment results, the judgment module performs a comprehensive judgment to determine whether the driver is in a state of motion disengagement. The adjustment of the DMS and OMS cameras includes: adjusting the installation position and angle of the DMS and OMS cameras to make them compatible with image acquisition of the steering wheel area on the original basis; The training of the visual model includes: collecting visual images of the steering wheel area covered by the adjusted DMS camera and OMS camera, adding the visual images of the steering wheel area to the visual training dataset, training the visual training dataset with the training model, and evaluating the preset metrics of the visual model.

[0015] Compared with the prior art, the present invention provides a method and system for detecting driver disengagement in a driver assistance system, which has the following advantages: This invention provides a method for detecting driver disengagement in an assisted driving system. By adjusting the original vehicle's DMS and OMS cameras to be compatible with the steering wheel area acquisition, it eliminates the reliance on torque signals alone. The visual model can directly identify whether the driver is holding the steering wheel, and even without tangential torque, it can visually confirm that the driver has not let go, thus improving the accuracy of driver disengagement detection. This invention reduces the implementation cost of the detection system through optimization and adjustment of existing DMS and OMS cameras and software-level visual model training. The visual model is based on image recognition logic for whether the driver is holding the steering wheel, and is unaffected by steering wheel covers, solving the false alarm problem caused by steering wheel covers in capacitive steering wheels and ensuring the reliability of detection. This invention retains the original detection functions of the adjusted DMS and OMS cameras while adding steering wheel area acquisition functionality. Through a comprehensive judgment combining visual and torque assessment, it leverages the advantage of direct visual judgment of hand grip status while retaining the auxiliary reference value of torque assessment, improving the reliability of the detection results and achieving low-cost, high-accuracy detection, thereby enhancing the reliability of the assisted driving system. Attached Figure Description

[0016] The accompanying drawings are provided to further understand the invention and constitute a part of this invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0017] Figure 1 This is a flowchart illustrating a method for detecting driver disengagement in an assisted driving system according to the present invention.

[0018] Figure 2 This is a schematic diagram of the torque determination process in the driver disengagement detection method of an assisted driving system according to the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0020] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0021] It should be noted that, in this document, 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 a 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 said element.

[0022] It should be noted that the apparatus and methods disclosed in the embodiments herein can also be implemented in other ways. The apparatus embodiments described above are merely illustrative; for example, the flowcharts and block diagrams in the accompanying drawings show the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments herein. In this regard, each block in a flowchart or block diagram may represent a module, program, or part of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system to perform the specified function or action, or can be implemented using a combination of dedicated hardware and computer instructions.

[0023] In addition, the functional modules in the various embodiments of this article can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0024] How to provide a method for detecting driver disengagement in an assisted driving system, reducing the cost of the detection system, and improving the accuracy and reliability of the detection.

[0025] Based on this, the present invention provides a method for detecting driver disengagement in a driver assistance system, comprising the following steps: S1. Real-time images of the steering wheel area are acquired by the adjusted DMS camera and OMS camera. The real-time images are input into the trained visual model. The visual model performs fusion processing on the real-time images to obtain the visual judgment result and its corresponding trust level. S2. The driver assistance system receives the steering torque from the EPS, processes the steering torque and compares it with a preset threshold to obtain the torque judgment result of whether the driver is holding the steering wheel. S3. Based on the visual judgment results and their corresponding trust level and torque judgment results, the judgment module makes a comprehensive judgment to obtain the result of whether the driver is in a state of motion disengagement. The adjustment of the DMS and OMS cameras includes: adjusting the installation position and angle of the DMS and OMS cameras to make them compatible with image acquisition of the steering wheel area on the original basis; The training of the visual model includes: collecting visual images of the steering wheel area covered by the adjusted DMS camera and OMS camera, adding the visual images of the steering wheel area to the visual training dataset, training the visual training dataset with the training model, and evaluating the preset metrics of the visual model.

[0026] The present invention provides a method for detecting driver disengagement in an assisted driving system. Compared with traditional torque steering wheels, this method directly determines whether the driver's hands are gripping the steering wheel using a visual algorithm, avoiding the false alarms caused by relying solely on the torque of the torque steering wheel. This optimizes the accuracy of the judgment and improves the experience of assisted driving functions. Compared with capacitive steering wheels, this invention optimizes and adjusts the original vehicle's camera, resulting in lower costs. It also solves the problem of false alarms caused by users installing steering wheel covers, thus improving reliability.

[0027] In some embodiments, the driver disengagement detection method for the assisted driving system of the present invention is based on an architecture including a DMS camera, an OMS camera, a torque steering wheel, a vision fusion judgment module, a torque judgment module, and a final judgment module; the main steps of the method are as follows: Optimize the detection areas of the DMS (Driver Monitoring System) camera and OMS (Occupant Monitoring System) camera to be compatible with the detection of the steering wheel area; Add the scene of whether the hands are holding the steering wheel to the visual training dataset of DMS and OMS cameras; The fusion judgment module fuses the collected DMS / OMS images to determine whether the driver is holding the steering wheel. The visual fusion method fuses visual images from two cameras, and the output is a determination of whether the driver has let go of the hands. The driver assistance system determines whether the driver has taken their hands off the wheel based on the steering torque feedback from the EPS. By combining visual fusion information and torque data from torque steering wheel feedback, a comprehensive determination is made as to whether the driver is in a state of motion disengagement, thus achieving the final determination of driver motion disengagement.

[0028] This invention also provides a detection system for driver disengagement in an assisted driving system, comprising a data acquisition and fusion processing module, a torque judgment module, and a comprehensive judgment module, wherein: Data acquisition and fusion processing module: used to acquire real-time images of the steering wheel area through the adjusted DMS camera and OMS camera, input the real-time images into the trained visual model, and perform fusion processing on the real-time images by the visual model to obtain the visual judgment result and its corresponding trust level. Torque determination module: Used by the driver assistance system to receive the steering torque from the EPS, process the steering torque and compare it with a preset threshold to obtain the torque determination result of whether the driver is holding the steering wheel. Comprehensive Judgment Module: Based on the visual judgment results and their corresponding trust level and torque judgment results, the judgment module performs a comprehensive judgment to determine whether the driver is in a state of motion disengagement. The adjustment of the DMS and OMS cameras includes: adjusting the installation position and angle of the DMS and OMS cameras to make them compatible with image acquisition of the steering wheel area on the original basis; The training of the visual model includes: collecting visual images of the steering wheel area covered by the adjusted DMS camera and OMS camera, adding the visual images of the steering wheel area to the visual training dataset, training the visual training dataset with the training model, and evaluating the preset metrics of the visual model.

[0029] like Figure 1 and Figure 2 As shown below, the method and system for detecting driver disengagement in an assisted driving system according to the present invention will be further described in detail through specific embodiments.

[0030] I. This embodiment involves two cameras for observing the occupants. However, the current purpose of these two cameras is unrelated to determining whether the driver has let go of their hands. If they are to be used to determine whether the driver has let go of their hands, the arrangement and field of view of the DMS and OMS cameras need to be adjusted accordingly.

[0031] Vehicle DMS cameras are generally used to detect driver fatigue and distraction. The main coverage area is the driver's facial expressions and eye movements. Therefore, they cannot completely cover the steering wheel area involved in determining whether the driver has taken their hands off the wheel. The OMS camera is used to detect the status of the occupants in the vehicle, mainly covering the driver, passenger, and rear seat areas, and therefore cannot completely cover the steering wheel area proposed in this invention.

[0032] Using only one of the DMS and OMS cameras to cover both its functional area and the entire hand area presents significant challenges in terms of camera design and overall vehicle layout. Therefore, this invention employs both DMS and OMS cameras to provide comprehensive coverage of the steering wheel area. The main difference lies in adjusting the installation position and angle of the DMS and OMS cameras to achieve both functionality and effective steering wheel coverage.

[0033] II. Training the visual image of the steering wheel area mainly involves the following steps: The driver's hand gestures are broken down into six types: single-hand grip, double-hand grip, single-hand rest, double-hand rest, multi-finger grip, and multi-finger rest. These six types of hand gestures need to be distributed around the steering wheel. Collect images under different hand gestures, lighting conditions, and angles, and simultaneously add negative samples (images of people not holding the steering wheel) to improve the robustness of the trained model. It is important to note that the photos need to cover possible shooting angles from DMS and OMS cameras; Select a training model, train it on the collected images, and evaluate metrics such as false positive rate and false negative rate.

[0034] III. The visual fusion method fuses visual images from two cameras, and the output is a determination of whether the driver has let go of the hands. It consists of the following steps: Two cameras capture images of their respective fields of view simultaneously. The image is fed into the visual fusion judgment module. The visual fusion module removes information outside the steering wheel area and performs fusion to output an image of the steering wheel area. Since the two cameras cover different areas of the steering wheel area and at different angles, this image is a 360-degree image of the steering wheel area. Based on the fused images from the cameras, the coverage area of ​​the steering wheel is matched with the actual position of the driver's hands. Coverage can be categorized into three scenarios: no coverage, single-camera coverage, and overlapping coverage by two cameras. For scenarios where no coverage is possible, the camera output is invalid. For scenarios where coverage is possible, based on multiple training sessions with various datasets, the system outputs a result indicating whether the driver has taken their hands off the steering wheel and the level of trust. (between 0 and 1); IV. Output invalidity based on visual training results; whether to release the device; level of trust.

[0035] The driver assistance system determines whether the driver has taken their hands off the wheel based on the steering torque feedback from the EPS (Electric Power Steering) system. The steps are as follows: The driver assistance system receives steering torque feedback from the EPS (Electric Power Steering) Low-pass filtering is performed to remove signal interference; The driver assistance system uses the processed torque signal and the set threshold Compare and determine whether the item has been sold: like And the duration is greater than If so, it is determined to be a slip of the hand; like And the duration is greater than If so, it is determined that the item has not been sold; in, The processed torque signal, For torque threshold, The first time threshold, The second time threshold is, To prevent sudden changes in speed, the following ranges apply at different vehicle speeds. Calibration needs to be performed based on the actual vehicle. , The debounce interval is set to prevent result jumps; Because the impact of road surface excitation on steering torque varies at different vehicle speeds and road curvatures, therefore, at different vehicle speeds... There are discrepancies, and calibration needs to be performed based on the actual vehicle. V. The final judgment module determines whether the driver has released their hands based on the visual output and the torque judgment result from EPS feedback. The judgment strategy is shown in the table below: Table 1 Judgment Strategy

[0036] In summary, the present invention provides a method and system for detecting driver disengagement in an assisted driving system. Based on the vehicle's existing torque steering wheel and DMS (Driver Monitoring System), it uses visual fusion combined with steering torque to determine the disengagement state. This solution addresses the problem of driver disengagement detection with a low-cost approach. By adjusting the arrangement and performance of the vehicle's existing DMS and OMS cameras, and through a fusion judgment strategy, it solves the problem of inaccurate disengagement diagnosis by the torque steering wheel. Overall, it improves the detection accuracy and reliability of driver disengagement and has certain practical significance.

[0037] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Anyone skilled in the art can readily implement the present invention according to the description and above. Any modifications, alterations, or equivalent variations made using the technical content disclosed above are equivalent embodiments of the present invention. Furthermore, any modifications, alterations, or variations made to the above embodiments based on the essential technology of the present invention are still within the protection scope of the present invention.

Claims

1. A method for detecting driver disengagement in a driver assistance system, characterized in that, Includes the following steps: S1. Real-time images of the steering wheel area are acquired by the adjusted DMS camera and OMS camera. The real-time images are input into the trained visual model. The visual model performs fusion processing on the real-time images to obtain the visual judgment result and its corresponding trust level. S2. The driver assistance system receives the steering torque from the EPS, processes the steering torque and compares it with a preset threshold to obtain the torque judgment result of whether the driver is holding the steering wheel. S3. Based on the visual judgment results and their corresponding trust level and torque judgment results, the judgment module makes a comprehensive judgment to obtain the result of whether the driver is in a state of motion disengagement. The adjustment of the DMS and OMS cameras includes: adjusting the installation position and angle of the DMS and OMS cameras to make them compatible with image acquisition of the steering wheel area on the original basis; The training process of the trained visual model includes: collecting visual images of the steering wheel area covered by the adjusted DMS camera and OMS camera, adding the visual images of the steering wheel area to the visual training dataset, training the visual training dataset through the training model, and evaluating the preset metrics of the visual model.

2. The method for detecting driver disengagement in an assisted driving system according to claim 1, characterized in that, The visual model performs fusion processing on real-time images through a visual fusion judgment module, specifically including: the DMS camera and the OMS camera simultaneously capture real-time images of the steering wheel area within their respective fields of view; the visual fusion judgment module removes information outside the steering wheel area from the images, fuses the real-time images from the two cameras, and outputs an image around the steering wheel; and matches the coverage position of the steering wheel area in the fused image with the actual position of the driver's hands. If the scene cannot be covered, output an invalid visual judgment result; if the scene can be covered, output the visual judgment result of whether it can be released and the corresponding trust level.

3. The method for detecting driver disengagement in an assisted driving system according to claim 1, characterized in that, The preset metrics include the false alarm rate and false negative rate of the visual model.

4. The method for detecting driver disengagement in an assisted driving system according to claim 1, characterized in that, The steering torque feedback from the EPS is processed using a low-pass filter to remove signal interference.

5. The method for detecting driver disengagement in an assisted driving system according to claim 1, characterized in that, The steering torque is processed and then compared with a preset threshold, specifically including: like And the duration is greater than If so, it is determined to be a slip of the hand; like If the duration is greater than t2, it is determined that the item has not been sold. in, The processed torque signal, For torque threshold, t1 is the first time threshold, and t2 is the second time threshold. To prevent sudden changes in speed, the following ranges apply at different vehicle speeds. Calibration needs to be performed based on the actual vehicle. , The debounce interval is set to prevent result jumps.

6. The method for detecting driver disengagement in an assisted driving system according to claim 1, characterized in that, The comprehensive judgment adopts: When both the visual judgment result and the torque judgment result indicate that the hands have slipped, the final determination is that the hands have slipped. When the visual judgment result is "released" and the confidence level is not less than the preset value, and the torque judgment result is "not released", the final judgment is "released"; when the visual judgment result is "released" and the confidence level is less than the preset value, and the torque judgment result is "not released", the final judgment is "not released". When both the visual judgment result and the torque judgment result indicate that the hand will not be released, the final determination is that the hand will not be released; when the visual judgment result indicates that the hand will not be released, the confidence level is not less than the preset value, and the torque judgment result indicates that the hand will be released, the final determination is that the hand will not be released. When the visual judgment result is "not to be released" and the confidence level is less than the preset value, and the torque judgment result is "release", the final judgment is "release".

7. The method for detecting driver disengagement in an assisted driving system according to claim 1, characterized in that, If the visual judgment result is that the scene cannot be covered, and the torque judgment result is that the hand is released, then the final judgment is that the hand is released; if the visual judgment result is that the scene cannot be covered, and the torque judgment result is that the hand is not released, then the final judgment is that the hand is not released.

8. The method for detecting driver disengagement in an assisted driving system according to claim 1, characterized in that, The original DMS camera includes detecting driver fatigue and distraction, covering the driver's facial expressions and eye movements; the original OMS camera includes detecting the status of vehicle occupants, covering the driver, front passenger, and rear seat areas.

9. The method for detecting driver disengagement in an assisted driving system according to claim 1, characterized in that, The visual images include various scene images with or without the driver's hands gripping the steering wheel. These scene images include: images of the driver's various hand gestures distributed around the steering wheel after being split, images under different lighting and angles, and negative sample images of the driver not gripping the steering wheel.

10. A detection system for driver disengagement in a driver assistance system, used to implement the driver disengagement detection method of any one of claims 1-9, characterized in that, It includes a data acquisition and fusion processing module, a torque judgment module, and a comprehensive judgment module, among which: Data acquisition and fusion processing module: used to acquire real-time images of the steering wheel area through the adjusted DMS camera and OMS camera, input the real-time images into the trained visual model, and perform fusion processing on the real-time images by the visual model to obtain the visual judgment result and its corresponding trust level. Torque determination module: Used by the driver assistance system to receive the steering torque from the EPS, process the steering torque and compare it with a preset threshold to obtain the torque determination result of whether the driver is holding the steering wheel. Comprehensive Judgment Module: Based on the visual judgment results and their corresponding trust level and torque judgment results, the judgment module performs a comprehensive judgment to determine whether the driver is in a state of motion disengagement. The adjustment of the DMS and OMS cameras includes: adjusting the installation position and angle of the DMS and OMS cameras to make them compatible with image acquisition of the steering wheel area on the original basis; The training of the visual model includes: collecting visual images of the steering wheel area covered by the adjusted DMS camera and OMS camera, adding the visual images of the steering wheel area to the visual training dataset, training the visual training dataset with the training model, and evaluating the preset metrics of the visual model.

Images