Lane changing assistance method and system based on electronic rearview mirror

Abstract

The present invention belongs to the field of automobile technology, and specifically relates to a lane change assistance method and system based on an electronic rearview mirror. The present invention breaks through the inherent limitations of traditional rearview mirrors, and achieves full coverage of more than 120 degrees to the side and rear by integrating a wide-angle camera with an intelligent zoom function under the side rearview mirror of the vehicle. The image processing unit performs real-time optimization processing on the images captured by the camera. Through the application of intelligent cameras, image processing technology, automatic field of view adjustment and advanced early warning systems, the driver's perception of the side and rear environment is significantly enhanced, the risk of changing lanes is reduced, and the overall safety and comfort of driving are improved. The application of this technology not only solves the problem of field of view limitations of traditional rearview mirrors, but also opens up a new path for the development of intelligent driving assistance systems, showing the cutting-edge trend of automotive safety technology.

Description

Technical Field

[0001] The present invention belongs to the field of automobile technology, and in particular relates to a lane change assistance method and system based on an electronic rearview mirror. Background Art

[0002] With the rapid development of the automobile industry, driving safety has become one of the core issues of automobile technology innovation. In the modern traffic environment, lane change is one of the common causes of traffic accidents, especially when the vision is obstructed or driving at night. Traditional rearview mirrors are difficult to fully cover the blind spots on the side and rear of the vehicle, increasing the risk of collision when changing lanes. Therefore, how to effectively expand the driver's field of vision and accurately identify and warn of potential collision threats on the side and rear has become a technical challenge that needs to be solved urgently.

[0003] Traditional side mirrors have a narrow field of view due to physical size limitations, especially when driving at high speeds, in complex road conditions or at night. In addition, drivers often need to turn their heads frequently to check blind spots, which not only distracts driving attention but also affects driving comfort and safety. Therefore, developing a lane change assistance system that can provide a wider field of view and intelligent warning is crucial to improving driving safety and reducing driving burden. Summary of the invention

[0004] The purpose of the present invention is to provide a lane change assistance method and system based on an electronic rearview mirror, which significantly enhances the driver's perception of the side and rear environment through intelligent camera application, image processing technology, automatic field of view adjustment and advanced warning system, reduces the risk when changing lanes, and solves the problems in the prior art raised in the above background technology.

[0005] To achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a lane change assistance method based on an electronic rearview mirror, comprising the following steps: installing a camera under the side rearview mirror of the vehicle, the camera having an intelligent zoom function and an integrated night vision enhancement mode, and a viewing angle range of not less than 120 degrees; optimizing the image signal captured by the camera through an image processing unit, including but not limited to brightness adjustment, contrast enhancement and fisheye correction; transmitting the processed image signal to the vehicle-mounted display screen for real-time display, and identifying and marking the side and rear vehicles through an intelligent algorithm; The network receives the CAN bus signal of the vehicle's driving status and automatically adjusts the camera's field of view according to the vehicle speed (V), steering angle (θ) and vehicle driving direction. The calculation formula is: field of view width (W) = f(V,θ) = k1*V^2+k2*θ+k3, where k1, k2, k3 are preset coefficients; Blind spot monitoring, when a vehicle approaches to a dangerous distance (D_danger) from the side and rear, a multi-sensory lane change warning signal (A_warn) is issued. The warning signal strength is related to the speed of the approaching vehicle (V_approach) and the relative distance (D_re l). The warning formula is:

[0006] A_warn=g(D_re l,V_approach)=k4*ln(D_re l)+k5*V_approach^2, where k4 and k5 are preset constants.

[0007] Preferably, the image signal captured by the camera is optimized and processed by an image processing unit, including:

[0008] Get the original image signal output by the camera;

[0009] Using real-time data from gyroscopes and accelerometers, the vehicle's motion posture parameters in three-dimensional space are calculated, including roll angle, yaw angle, and pitch angle;

[0010] The original image is compensated according to the calculated motion posture parameters, and the inter-frame motion is corrected by translation and rotation, and the clarity and contrast of the corrected image are enhanced.

[0011] Preferably, the intelligent algorithm uses a deep learning model to perform side rear vehicle recognition, including:

[0012] Data preprocessing: Collect image samples with different lighting conditions, weather conditions, vehicle types and driving status, and perform grayscale, normalization and size unification preprocessing operations;

[0013] Training set construction: Select training sets from preprocessed samples and label them;

[0014] Build a deep learning network: Design a neural network architecture that includes convolutional layers, pooling layers, fully connected layers, and output layers. The number of network layers, the number of neurons in each layer, and the activation function are determined based on the balance between recognition accuracy and computing resources.

[0015] Network training: Using the training set data, the back propagation algorithm combined with the optimization algorithm is used to adjust the network weights and perform multiple iterative training;

[0016] Model validation and optimization: Use an independent validation set to evaluate the performance of the trained model and adjust the network parameters based on the evaluation results;

[0017] Real-time application: The optimized deep learning model is deployed in the image processing unit to perform vehicle recognition on the real-time images captured by the camera. By setting a threshold, it is determined whether there is a vehicle approaching from the side and rear, and the vehicle is marked and displayed on the vehicle display.

[0018] Preferably, the intelligent zoom method of the camera includes:

[0019] Acquire the side and rear images captured by the camera in real time, and preliminarily analyze the size and distance information of the objects in the images through the image processing unit, and set multiple focal length preset values, including but not limited to far focus, standard focus and near focus, corresponding to different detection distance ranges;

[0020] Use object recognition algorithm to identify the vehicle behind you and calculate the estimated distance between the vehicle and the vehicle.

[0021] The estimated distance is compared with the preset distance interval threshold. If the estimated distance falls within the long distance interval, the telephoto mode is activated; at a medium distance, the standard focal length is enabled; and at a close distance, the near focal length mode is switched;

[0022] The position of the lens group inside the camera is adjusted instantly through the motor drive to achieve rapid focal length conversion. After the focal length adjustment is completed, the relative distance changes between vehicles continue to be monitored and the focal length is dynamically adjusted to adapt to the dynamic distribution of vehicles in the driving environment.

[0023] Preferably, the method for issuing the multi-sensory lane change warning signal includes:

[0024] When the vehicle behind you approaches the preset danger distance, the warning algorithm is immediately triggered to calculate the warning level, which is determined by the speed and relative distance of the approaching vehicle;

[0025] Depending on the warning level, visual warnings of varying intensity, acoustic warnings or seat vibration warnings are activated.

[0026] Preferably, the blind spot monitoring further includes predicting the curvature of the road ahead through GPS and map data, including:

[0027] Receive and analyze GPS signals in real time to determine the vehicle's current location and direction of travel;

[0028] Retrieving geometric information of the road ahead from a built-in or online map database in combination with the vehicle's travel direction, including but not limited to the curve radius, curve direction, and curve length;

[0029] Analyze the characteristics of the curve and determine the appropriate time to adjust the field of view before the vehicle enters the curve. According to the characteristics of the curve, calculate the optimal field of view angle (α_opt ima l) and the field of view range adjustment amount (ΔW) that the camera should be adjusted to. The calculation formula is:

[0030] α_opt ima l=f(R_curve,V_vehicle);

[0031] ΔW = g(R_curve, V_veh icle); R_curve is the curve radius, V_veh icle is the vehicle speed;

[0032] Send instructions to the camera to adjust the camera angle and field of view. After passing the curve, it will automatically return to the standard field of view setting for straight driving.

[0033] Preferably, the method further includes: obstacle identification and BSD warning, including:

[0034] The camera continuously captures the video stream from the side and rear of the vehicle and transmits the image data to the image processing unit, which uses a deep learning algorithm to analyze the image and identify potential dangerous factors such as vehicles, pedestrians, and static obstacles in the image;

[0035] Calculate the relative distance, speed difference and potential collision time between each identified obstacle and the vehicle;

[0036] When the collision time of any obstacle is lower than the preset safety threshold, the system immediately triggers the blind spot monitoring warning. At the same time, the system dynamically adjusts the display content of the electronic rearview mirror according to the location of the obstacle, highlighting the area with collision risk to guide the driver's attention;

[0037] If the driver starts to change lanes and there is an obstacle in the blind spot, the warning level will be increased and a stronger warning signal will be used until the driver takes evasive measures or the obstacle no longer poses a threat.

[0038] Preferably, the installation position and adjustment of the camera include:

[0039] One or more preset camera installation sites are provided, and the camera is connected to the installation site using a ball joint or a universal joint seat, so that it can be adjusted in multiple dimensions in the horizontal and vertical planes;

[0040] The driver can manually or electrically adjust the specific installation angle of the camera, including the pitch angle, roll angle and yaw angle.

[0041] Preferably, the method further comprises: data recording, comprising:

[0042] connecting the data recording module to the image processing unit, the vehicle bus system and the external storage device;

[0043] When a warning event is triggered, the data recording module automatically starts recording, including but not limited to the timestamp of the warning, vehicle speed, steering angle, camera current field of view parameters, surrounding environment images and vehicle CAN bus signals;

[0044] All recorded data is compressed and encrypted and stored in high-durability storage media.

[0045] On the other hand, the present invention provides a lane change assistance system based on an electronic rearview mirror, comprising:

[0046] An image acquisition module is used to install a camera under the side rearview mirror of the vehicle, wherein the camera has an intelligent zoom function and an integrated night vision enhancement mode, and its viewing angle range is not less than 120 degrees;

[0047] An image processing module, used for optimizing the image signal captured by the camera through an image processing unit, including but not limited to brightness adjustment, contrast enhancement and fisheye correction;

[0048] The display module is used to transmit the processed image signal to the vehicle display screen for real-time display, and at the same time identify and mark the vehicles behind the side through intelligent algorithms;

[0049] The camera adjustment module is used to receive the CAN bus signal of the vehicle's driving status through the vehicle network, and automatically adjust the camera's field of view according to the vehicle speed (V), steering angle (θ) and vehicle driving direction. The calculation formula is: field of view width (W) = f(V,θ) = k1*V^2+k2*θ+k3, where k1, k2, k3 are preset coefficients;

[0050] The warning module is used for blind spot monitoring. When a vehicle approaches the dangerous distance (D_danger) from the rear, a multi-sensory lane change warning signal (A_warn) is issued. The strength of the warning signal is related to the speed (V_approach) and relative distance (D_re l) of the approaching vehicle. The warning formula is:

[0051] A_warn=g(D_re l,V_approach)=k4*ln(D_re l)+k5*V_approach^2, where k4 and k5 are preset constants.

[0052] Technical effects and advantages of the present invention: Compared with the prior art, the lane change assistance method and system based on the electronic rearview mirror proposed by the present invention has the following advantages:

[0053] The lane change assistance method of the electronic rearview mirror proposed in the present invention significantly enhances the driver's perception of the side and rear environment through the application of intelligent cameras, image processing technology, automatic field of view adjustment and advanced warning systems, reduces the risk of lane change, and improves the overall safety and comfort of driving. The application of this technology not only solves the problem of limited field of view of traditional rearview mirrors, but also opens up a new path for the development of intelligent driving assistance systems, showing the cutting-edge trend of automobile safety technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a flow chart of the lane change assistance method based on the electronic rearview mirror of the present invention;

[0055] Figure 2 It is a module diagram of the lane change assistance system based on the electronic rearview mirror of the present invention. DETAILED DESCRIPTION

[0056] The following will be combined with the accompanying drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. The specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0057] The present invention provides a lane change assistance method based on an electronic rearview mirror. The present invention breaks through the inherent limitations of traditional rearview mirrors and achieves full coverage of more than 120 degrees to the side and rear by integrating a wide-angle camera with an intelligent zoom function under the side rearview mirror of the vehicle. The camera automatically switches to night vision enhancement mode at night or in low-light environments, greatly improving the visibility of night driving. The image processing unit performs real-time optimization processing on the images captured by the camera, including automatic adjustment of brightness and contrast and correction of the fisheye effect, to ensure the clarity and realism of the displayed image, and provide the driver with more intuitive and distortion-free visual information.

[0058] The real-time monitoring capability of the vehicle's driving status is further integrated. The CAN bus receives signals such as vehicle speed and steering angle, and automatically adjusts the camera's field of view according to the vehicle's dynamics. The calculation formula ensures the flexible adaptability of the field of view under different driving conditions, effectively reducing the visual blind spot. In addition, the system's integrated blind spot monitoring system can monitor vehicles approaching from the side and rear in real time. Once a potential collision risk is detected, it immediately sends out a warning signal through a multi-sensory warning mechanism. The warning intensity is closely related to the speed and distance of the approaching vehicle, ensuring that the driver can respond quickly and effectively avoid collision accidents. The details are as follows.

[0059] like Figure 1 As shown, the above lane change assistance method based on an electronic rearview mirror includes the following steps:

[0060] Step 1: Install a camera below the side rearview mirror of the vehicle, wherein the camera has an intelligent zoom function and an integrated night vision enhancement mode, and its viewing angle range is not less than 120 degrees;

[0061] Furthermore, one or more preset camera installation sites are provided on the outside of the vehicle, and the camera is connected to the installation site by using a spherical joint or a universal joint seat, and the image signal captured by the camera is optimized and processed by an image processing unit, including: obtaining the original image signal output by the camera; using the real-time data of the gyroscope and the accelerometer to calculate the vehicle's motion posture parameters in three-dimensional space, including roll angle, yaw angle and pitch angle; based on the calculated motion posture parameters, compensating the original image, correcting the inter-frame motion by translation and rotation, and performing clarity and contrast enhancement processing on the corrected image.

[0062] Specific, camera installation and configuration

[0063] First, choose to preset one or more camera installation sites under the side rearview mirror of the vehicle. These locations have been carefully calculated to ensure that the side and rear views of the vehicle can be covered to the maximum extent. The camera is fixed with a high-performance ball joint or universal joint. This design allows the camera to flexibly adjust the angle in three-dimensional space, and maintain stable image acquisition even when the vehicle is in motion. The camera itself has an intelligent zoom function, which can automatically adjust the focal length according to the vehicle's driving status and ambient light conditions. At the same time, it integrates a night vision enhancement mode and uses infrared or low-light imaging technology to ensure clear images at night or in low-light environments. The viewing angle range is at least 120 degrees, effectively widening the blind spot of vision.

[0064] Specifically, image processing and posture compensation

[0065] Raw image signal acquisition: The image signal captured by the camera is first received by the image processing unit. This unit is the "brain" of the entire system and is responsible for processing all incoming video streams. Its primary task is to obtain the raw image signal of the camera.

[0066] Attitude parameter calculation: In order to achieve accurate compensation for moving vehicles, the system has built-in gyroscopes and accelerometers. The gyroscope monitors the rotation changes of the vehicle in the horizontal and vertical planes, while the accelerometer measures the acceleration of the vehicle in three dimensions. Combining the real-time data of the two, through complex mathematical models and algorithms, the precise motion attitude parameters of the vehicle in three-dimensional space are calculated, including roll angle (vehicle rotation around the longitudinal axis), yaw angle (vehicle rotation around the vertical axis) and pitch angle (vehicle tilt angle around the lateral axis).

[0067] Image compensation processing: After obtaining the posture parameters, the image processing unit performs motion compensation on the original image. This process involves complex image processing algorithms, such as image registration, inter-frame motion estimation, etc. By implementing translation and rotation correction on each frame of the image, the blur and distortion caused by vehicle motion are eliminated to ensure that the video stream is continuous and stable. This step is crucial to maintaining the coherence of the image and reducing the discomfort of the viewer.

[0068] Image quality improvement: After completing motion compensation, the system will also perform a series of post-processing on the corrected image, including clarity enhancement and contrast adjustment. The denoising algorithm reduces noise in the image and improves detail recognition; through dynamic range adjustment, all details can be clearly presented in scenes with strong light and shadows. These processing technologies work together to ensure that the final video image presented to the driver is both clear and natural.

[0069] It uses modern sensor technology, image processing algorithms and intelligent control theory. The sensor components provide basic data of vehicle dynamic information, while the image processing unit uses this data to analyze and process the original image in real time, ensuring that the side and rear images remain stable and clear even when the vehicle moves quickly or turns. Through intelligent zoom and night vision technology, the system can provide reliable field of view information under various light and speed conditions, greatly improving driving safety and comfort.

[0070] In summary, this implementation provides the driver with an unprecedented side rear vision experience through precise camera layout, advanced posture compensation algorithm and high-quality image post-processing technology. It is a major advancement in the field of intelligent driving assistance systems and lays a solid foundation for the future development of autonomous driving technology.

[0071] Real-time feedback and control: In some applications (such as vehicle assisted driving systems), the IPU also needs to send instructions to the vehicle control system in real time based on the image analysis results, such as lane change assistance, emergency braking, etc.

[0072] Step 2: Optimizing the image signal captured by the camera through an image processing unit, including but not limited to brightness adjustment, contrast enhancement and fisheye correction;

[0073] Specifically, the Image Processing Unit (IPU) is the core component of the image processing system, responsible for receiving, processing and outputting image data. It is usually composed of two parts: hardware and software, aiming to optimize image quality and extract useful information. The following is a detailed introduction to the specific composition of the image processing unit and its working principle:

[0074] Hardware composition

[0075] Image acquisition module: includes a camera or other image sensor, which is responsible for capturing optical information of the real world and converting it into electrical signals, which are then converted into digital image data through an analog-to-digital converter (ADC).

[0076] Processor (CPU / GPU / FPGA / ASIC): Central Processing Unit (CPU): executes advanced algorithms for image processing, handles nonlinear operations and decision logic. Graphics Processing Unit (GPU): excels at parallel computing and is suitable for processing large-scale pixel-level operations of images, such as filtering and transformation. Field Programmable Gate Array (FPGA): provides highly customized hardware acceleration and is suitable for real-time image processing tasks. Application Specific Integrated Circuit (ASIC): hardware customized for specific image processing tasks that can efficiently execute specific algorithms. Storage Unit: includes RAM and flash memory for storing image data, intermediate processing results and algorithm code. Interface and Communication Module: used to exchange data with external devices (such as cameras, display screens, vehicle buses, etc.), such as the CAN bus interface for receiving vehicle status information. Front-end Subsystem: includes digitizers and arithmetic logic units (ALUs) for high-speed parallel processing of images, such as fast digitization and image averaging.

[0077] Software composition

[0078] Image processing software: includes a series of algorithms such as image enhancement, filtering, edge detection, feature extraction, object recognition, etc. These algorithms are based on mathematics, signal processing and computer vision principles. Drivers and firmware: control hardware components to ensure the correct transmission of data and smooth execution of processing flows. Operating systems and middleware: provide the operating environment, manage resource allocation, and support the efficient execution of software algorithms.

[0079] Working principle, image acquisition: images are captured through devices such as cameras and converted into digital signals through A / D converters. Preprocessing: preliminary processing of the original image, such as noise removal, grayscale conversion, histogram equalization, etc., to improve image quality. Feature extraction: key features in the image are extracted using transformation techniques (such as Fourier transform, wavelet transform) and edge detection algorithms. Image analysis and recognition: the extracted features are analyzed through algorithms to identify objects, classify or track targets. Post-processing and output: according to application requirements, the image is adjusted in the final step (such as color correction, sharpening) and output to the display or used for further decision logic.

[0080] Specifically, in the lane change assistance system based on the electronic rearview mirror, the optimization processing of the image signal captured by the camera is a key link, which directly affects the accuracy and real-time nature of the information obtained by the driver. This link involves the hardware configuration and algorithm application of the image processing unit (IPU), and the specific implementation method is as follows:

[0081] Hardware Configuration

[0082] 1. HD camera (C1)

[0083] Select models such as Sony IMX477, which has up to 12 million pixels and supports a 120-degree ultra-wide-angle lens, ensuring that clear and wide side and rear views can be captured in all lighting conditions. Its high dynamic range (HDR) function helps to retain more details in environments with strong contrast between light and dark.

[0084] 2. Image Processing Unit (IPU)

[0085] The core processor recommended is NVIDIA Jetson Xavier NX, which is a high-performance embedded system designed for edge computing, with powerful AI computing capabilities and efficient image processing capabilities. Jetson Xavier NX supports multi-channel camera input, can process high-resolution video streams in real time, and integrates deep learning models to achieve complex image analysis.

[0086] 3. Storage and transmission

[0087] Considering the high data throughput during image processing, DDR4 RAM (such as 16GB) is used as cache and eMMC5.1 (such as 32GB) is used as storage medium to ensure smooth data processing. In terms of transmission interface, USB3.0 or higher is used to ensure high-speed data transmission.

[0088] Image processing principles and algorithms

[0089] Brightness adjustment, using the histogram equalization algorithm to automatically adjust the image brightness. This algorithm analyzes the distribution of image pixels and dynamically adjusts the brightness value of each pixel to make the overall image brightness uniform, improve dark details, and avoid overexposing bright areas.

[0090] Contrast enhancement uses a local contrast enhancement algorithm, such as CLAHE (Contrast Limited Adaptive Histogram Equatorialion), which can improve the contrast of local areas of the image while maintaining the global contrast without distortion, making the image clearer and more detailed.

[0091] Fisheye correction, the fisheye correction algorithm is based on polynomial or inverse perspective transformation. By calculating the camera distortion center and distortion parameters, the image is reversely deformed to restore the straight line to a straight line, eliminating or reducing the fisheye effect. The specific algorithm can use the cv::und i stort() function in the OpenCV library, which supports custom camera intrinsic matrix and distortion coefficients to achieve accurate image correction.

[0092] Implementation steps:

[0093] Data acquisition: The original image signal captured by the camera C1 is transmitted to the image processing unit IPU at high speed through the M IPI interface (Mobile Industry Processor Interface).

[0094] Preprocessing: In the IPU, basic noise filtering, such as median filtering or bilateral filtering, is first performed to remove random noise and spots in the image.

[0095] Brightness and contrast adjustment: Apply histogram equalization and CLAHE algorithm to process brightness and contrast respectively, so as to significantly improve the image clarity.

[0096] Fisheye correction: Call the OpenCV library to perform inverse perspective transformation, input preset or previously calibrated distortion parameters, and output the corrected image.

[0097] Intelligent algorithm application: Use deep learning models, such as YOLOv5 or MobileNet-SSD, to identify and mark vehicles and assist driving decisions.

[0098] Output display: The processed high-quality image signal is transmitted to the vehicle display screen through the HDMI or LVDS interface, and the optimized side and rear view is displayed in real time.

[0099] Through the above hardware configuration and algorithm implementation, the system can significantly improve the driver's ability to perceive the side and rear environment, reduce visual blind spots, and enhance driving safety. The system design fully considers real-time, robustness and intelligence to ensure that reliable and effective auxiliary information can be provided in various driving environments.

[0100] Step 3: The processed image signal is transmitted to the vehicle display screen for real-time display, and the vehicle behind is identified and marked through intelligent algorithms;

[0101] Specifically, the intelligent algorithm uses a deep learning model to identify side and rear vehicles, including:

[0102] Data preprocessing: Collect image samples with different lighting conditions, weather conditions, vehicle types and driving status, and perform grayscale, normalization and size unification preprocessing operations;

[0103] Training set construction: Select training sets from preprocessed samples and label them;

[0104] Build a deep learning network: Design a neural network architecture that includes convolutional layers, pooling layers, fully connected layers, and output layers. The number of network layers, the number of neurons in each layer, and the activation function are determined based on the balance between recognition accuracy and computing resources.

[0105] Network training: Using the training set data, the back propagation algorithm combined with the optimization algorithm is used to adjust the network weights and perform multiple iterative training;

[0106] Model validation and optimization: Use an independent validation set to evaluate the performance of the trained model and adjust the network parameters based on the evaluation results;

[0107] Real-time application: The optimized deep learning model is deployed in the image processing unit to perform vehicle recognition on the real-time images captured by the camera. By setting a threshold, it is determined whether there is a vehicle approaching from the side and rear, and the vehicle is marked and displayed on the vehicle display.

[0108] Step 4: Receive the CAN bus signal of the vehicle's driving status through the vehicle network, and automatically adjust the camera's field of view according to the vehicle speed (V), steering angle (θ) and vehicle driving direction. The calculation formula is: field of view width (W) = f(V,θ) = k1*V^2+k2*θ+k3, where k1, k2, k3 are preset coefficients;

[0109] In the lane change assistance system based on electronic rearview mirror, automatically adjusting the camera's field of view is a core technology that aims to provide the best side and rear view according to the real-time vehicle driving status to ensure driving safety. The realization of this function depends on the precise coordination of vehicle network communication, data processing algorithms and hardware. The following will introduce the hardware selection, principles and specific application steps of this process in detail.

[0110] Hardware selection and system architecture

[0111] CAN bus interface module: TI's TCAN1042 is used, which is a high-performance CAN transceiver that supports the ISO11898-2 standard and can reliably receive key information such as vehicle speed and steering angle from the vehicle's CAN bus network. TCAN1042 has good electromagnetic compatibility and anti-interference capabilities, ensuring the accuracy and stability of data transmission.

[0112] Microprocessor: NXP's S32K144 series microcontroller is selected. This high-performance MCU is designed for automotive applications and integrates a powerful ARMCortex-M4F core, which can efficiently handle complex algorithm calculations, such as the logic operation of dynamically adjusting the camera's field of view according to vehicle speed and steering angle. The S32K144 also has a built-in CAN-FD interface, which perfectly matches the TCAN1042 to improve data transmission rate and efficiency.

[0113] Camera module: OmniVision's OV9712 is a CMOS image sensor with high dynamic range and excellent performance under low light conditions. It supports multiple image processing functions such as automatic exposure control and digital wide dynamic range (DWDR). By connecting to the I2C or SPI interface of the microprocessor, the camera's zoom control can be achieved to adapt to the calculated field of view width.

[0114] Formula principle and calculation method:

[0115] The field of view width adjustment formula (W) = f(V,θ) = k1*V^2+k2*θ+k3 is based on driving physics and ergonomics principles and is designed to dynamically adapt to changes in driving conditions.

[0116] k1*V^2: As the vehicle speed V increases, the field of view width W needs to be expanded accordingly. This is because the faster the vehicle speed, the longer the vehicle braking distance, and the observation distance of the surrounding environment also needs to be increased to identify dangers in advance. k1 is a coefficient proportional to the square of the vehicle speed, reflecting the nonlinear growth of the vehicle speed on the field of view.

[0117] k2*θ: The addition of steering angle θ takes into account the change in field of vision requirements when the vehicle turns. During the steering process, especially sharp turns, the driver needs a wider side and rear field of vision to avoid the risks brought by blind spots. k2 represents the direct contribution of the steering angle to the width of the field of vision, helping the system to quickly adjust the field of vision when turning.

[0118] k3: This is a constant term, representing the basic field of view width requirement, which is not directly affected by vehicle speed and steering angle, and reflects the basic field of view requirement when the vehicle is stationary or moving straight normally.

[0119] Specific implementation steps

[0120] Data reading: Receive the vehicle speed V and steering angle θ data sent by the vehicle CAN bus through TCAN1042 to ensure the real-time and accuracy of the data.

[0121] Parameter calculation: The S32K144 microprocessor uses the above formula to calculate the current required field of view width W according to the received vehicle speed and steering angle. Adaptive adjustment to different driving conditions is achieved through the pre-calibrated coefficients 1k1, k2, and k3.

[0122] Camera control: Send the calculated field of view width command to the camera module (such as OV9712) to change its actual field of view by adjusting the camera's optical or digital zoom. This step involves lens movement or image processing algorithm adjustment to achieve dynamic changes in the field of view.

[0123] Feedback and optimization: The system continuously monitors the vehicle status and driver operation, evaluates the effect of the adjusted field of view, and makes fine adjustments when necessary to ensure the best assistance effect. At the same time, the system records relevant data for continuous optimization of the algorithm and performance evaluation.

[0124] Through the introduction of the above hardware selection, formula principles and specific implementation steps, it can be seen that the lane change assistance system based on the electronic rearview mirror achieves intelligent adaptation to the driving environment through precise data collection, efficient computing and processing, and flexible hardware control, significantly improving driving safety and driving comfort.

[0125] It is worth noting that the intelligent zoom method of the camera includes:

[0126] Acquire the side and rear images captured by the camera in real time, and preliminarily analyze the size and distance information of the objects in the images through the image processing unit, and set multiple focal length preset values, including but not limited to far focus, standard focus and near focus, corresponding to different detection distance ranges;

[0127] Use object recognition algorithm to identify the vehicle behind you and calculate the estimated distance between the vehicle and the vehicle.

[0128] According to the comparison between the estimated distance and the preset distance interval threshold, if the estimated distance falls within the long distance interval, the telephoto mode is activated; when it is at a medium distance, the standard focal length is enabled; and the close distance is switched to the close focus mode; specifically, according to the estimated distance between vehicles, the system sets multiple levels of focal length preset values, generally including telephoto (suitable for observing distant objects), standard focal length (moderate distance) and close focus (close distance). By comparing with the preset distance interval threshold, the system dynamically selects the most appropriate focal length mode: Long distance: When the estimated distance exceeds a certain threshold, the telephoto mode is activated, and the lens is configured to the minimum focal length to expand the field of view and capture details farther away. Medium distance: If the estimated distance falls within the medium range, it switches to the standard focal length. At this time, the camera provides a balanced field of view and detail resolution, which is suitable for most daily driving scenarios. Close distance: When the rear vehicle is found to be very close, the system quickly adjusts to the close focus mode, increases the magnification of the lens, ensures that the details of the close vehicle can be clearly identified, and provides the driver with immediate risk avoidance information.

[0129] The position of the lens group inside the camera is adjusted instantly through the motor drive to achieve rapid focal length conversion. After the focal length adjustment is completed, the relative distance changes between vehicles continue to be monitored, and the focal length is dynamically adjusted to adapt to the dynamic distribution of vehicles in the driving environment. Specifically, the motor drive module adjusts the position of the lens group inside the lens instantly through a precise mechanical structure according to the instructions of the image processing unit to achieve rapid focal length conversion. This process must be smooth and rapid to reduce image jitter and improve driver comfort and safety. At the same time, the system continuously monitors the relative distance changes between vehicles and dynamically adjusts the focal length through a closed-loop feedback mechanism to ensure that the best visual information support is provided even in complex and changing driving environments.

[0130] In addition, the installation position and adjustment of the camera include:

[0131] One or more preset camera installation sites are provided, and the camera is connected to the installation site using a ball joint or a universal joint seat, so that it can be adjusted in multiple dimensions in the horizontal and vertical planes;

[0132] Furthermore, the preset camera installation position needs to comprehensively consider the structural characteristics of the vehicle, the driving vision requirements and the road regulations. Generally speaking, the ideal installation position of the side rear view camera is located at the rear corner of the vehicle, slightly higher than the waistline of the vehicle body, which can not only reduce the blind spot, but also effectively monitor the traffic conditions on the side and rear without interfering with the driver's direct line of sight. When designing the installation position, it is also necessary to ensure that the camera can avoid components that block the line of sight, such as wipers and door handles, and take into account the changes in the vehicle's viewing angle under different driving conditions (such as turning, uphill and downhill) to maximize the widening of the field of view. In order to enable the camera to be freely adjusted in three-dimensional space, a ball joint or a universal joint is usually used as a connector. The ball joint is embedded between two supporting parts through a central sphere, allowing the camera to rotate in any direction to achieve full adjustment of the pitch angle (up and down tilt), roll angle (left and right tilt) and yaw angle (left and right rotation). The universal joint also achieves the purpose of multi-dimensional adjustment through a more complex joint design, which is particularly suitable for installation environments that need to withstand large vibrations or torque.

[0133] The driver adjusts the specific installation angle of the camera manually or electrically, including pitch, roll and yaw. The electric adjustment mechanism usually includes three parts: sensor, actuator and control logic. The sensor (such as angle encoder) monitors the current angle position of the camera in real time, and the information is transmitted to the vehicle's ECU (electronic control unit). The driver inputs the adjustment command through the button or touch screen on the control panel in the car. The ECU calculates the target angle according to the command and drives the motor to rotate the corresponding angle until the camera reaches the ideal position. The whole process is precisely controlled by the software algorithm to ensure smooth and impact-free adjustment effect. In summary, through the carefully designed preset installation site, flexible spherical joint or universal joint connection, combined with manual and electric adjustment mechanism, the camera can achieve precise positioning and efficient adjustment, provide high-quality visual information input for the intelligent driving assistance system, and is an important cornerstone for achieving a safe and intelligent driving experience. With the continuous evolution of technology, the future camera adjustment system will be more intelligent and adaptive, further improving the safety and convenience of driving.

[0134] Step 5: Blind spot monitoring. When a vehicle approaches the dangerous distance (D_danger) from the rear, a multi-sensory lane change warning signal (A_warn) is issued. The strength of the warning signal is related to the speed (V_approach) and relative distance (D_re l) of the approaching vehicle. The warning formula is:

[0135] A_warn=g(D_re l,V_approach)=k4*ln(D_re l)+k5*V_approach^2, where k4 and k5 are preset constants.

[0136] Formula structure analysis

[0137] Distance term (D_re l): The natural logarithm ln(D_re l) is used to reflect the influence of relative distance. As D_re l decreases (i.e., the closer the vehicle is), the logarithmic function value increases negatively, which means that the smaller the distance, the greater the contribution to the warning signal. The choice of logarithmic function makes the warning intensity grow faster at close distances and slower at long distances, which is in line with the psychological characteristics of humans being more sensitive to close-range threats. At the same time, it also ensures a smooth transition of the warning signal strength and avoids abrupt changes.

[0138] Speed ​​term (V_approach): Using the square of speed V_approach^2, means that the impact of approach speed on the intensity of warning increases nonlinearly. This is because the faster the vehicle approaches, the less time the driver has to react, and the risk of collision rises sharply. The square term strengthens this sense of urgency, ensuring that vehicles approaching at high speeds can cause stronger warnings, prompting the driver to take immediate action.

[0139] Coefficients k4 and k5: These are preset constants based on safety standards, ergonomics research and measured data, used to adjust the contribution of distance and speed factors to the warning signal strength. k4 determines the baseline effect of distance changes on the warning strength at a given speed; k5 controls the gain amplitude of speed changes on the warning strength. The reasonable setting of these two coefficients is crucial to ensure that the warning is neither frequently false nor missed, and requires careful experimental verification and adjustment.

[0140] Principles of early warning signal design

[0141] Multi-sensory warning: Although not directly reflected in the formula itself, the mention of "multi-sensory lane change warning signal" means that the system reminds the driver simultaneously through multiple means such as visual (such as dashboard warning lights), auditory (warning sound) and even tactile (seat vibration) to ensure that attention can be quickly attracted in any driving environment.

[0142] Dynamic adaptability: By closely combining the intensity of the warning signal with the actual driving situation, the formula achieves the dynamic adaptability of the warning system. It can automatically adjust the intensity and urgency of the warning according to the actual approaching speed and distance of the vehicle, thereby providing the most appropriate warning information in different situations and helping the driver make timely and reasonable decisions.

[0143] Furthermore, the method for issuing the multi-sensory lane change warning signal includes: when the vehicle behind the side approaches to a preset danger distance, immediately triggering a warning algorithm to calculate a warning level, which is determined based on the speed and relative distance of the approaching vehicle; according to the warning level, activating visual warnings, sound warnings or seat vibration warnings of different intensities.

[0144] The blind spot monitoring also includes predicting the curvature of the road ahead through GPS and map data, including: receiving and analyzing GPS signals in real time to determine the current position and driving direction of the vehicle; retrieving geometric information of the road ahead from a built-in or online map database in combination with the driving direction of the vehicle, including but not limited to the radius, direction and length of the curve; analyzing the curve characteristics to determine the appropriate field of view adjustment time before the vehicle is about to enter the curve, and calculating the optimal field of view angle (α_opt ima l) and field of view range adjustment amount (ΔW) to which the camera should be adjusted according to the curve characteristics, and the calculation formula is:

[0145] α_opt ima l=f(R_curve,V_vehicle);

[0146] ΔW=g(R_curve,V_veh ic le); R_curve is the curve radius, V_veh ic le is the vehicle speed;

[0147] Send instructions to the camera to adjust the camera angle and field of view. After passing the curve, it will automatically return to the standard field of view setting for straight driving.

[0148] Specifically, the curve field of view angle adjustment (α_opt ima l): formula: α_opt ima l = f(R_curve, V_veh ic le), this formula reflects the influence of the curve radius (R_curve) and the vehicle speed (V_veh icle) on the optimal field of view angle (α_opt ima l) required by the camera. The smaller the curve radius, the sharper the curve. In order to ensure sufficient field of view coverage, the camera needs to be adjusted to a larger tilt angle to capture more scenes in the curve. At the same time, the vehicle speed will also affect the required field of view: when the speed is faster, the driver needs to see the situation deep in the curve earlier in order to react in time, so the camera angle is adjusted to a larger angle; conversely, if the speed is slower, a smaller tilt angle is sufficient.

[0149] Field of view adjustment (ΔW): Formula: ΔW = g(R_curve, V_veh ic le), Field of view adjustment (ΔW) refers to the change in the width of the camera's field of view, which is also affected by the curve radius and vehicle speed. The sharper the curve, the wider the field of view the vehicle needs to identify obstacles or other vehicles in the curve in advance, so the smaller the curve radius, the greater the field of view adjustment. When driving at a high speed, increasing the width of the field of view helps to detect and respond to emergencies in advance, so ΔW increases with increasing speed. This adjustment is intended to ensure that the driver can promptly detect possible blind spots in the curve.

[0150] The method also includes: obstacle recognition and BSD warning, including: the camera continuously captures the video stream behind the side of the vehicle and transmits the image data to the image processing unit, the image processing unit uses a deep learning algorithm to analyze the image, and identifies the potential dangerous factors of vehicles, pedestrians, and static obstacles in the image; calculates the relative distance, speed difference and potential collision time between each identified obstacle and the vehicle; when the collision time of any obstacle is lower than a preset safety threshold, the system immediately triggers a blind spot monitoring warning. At the same time, the system dynamically adjusts the display content of the electronic rearview mirror according to the position of the obstacle, highlights the area with collision risk, and guides the driver's attention; if the driver starts to perform a lane change operation and there is an obstacle in the blind spot, the warning level will be strengthened and a stronger warning signal will be used until the driver takes risk avoidance measures or the obstacle no longer poses a threat.

[0151] In some other embodiments, the above-mentioned lane change assistance method based on an electronic rearview mirror further includes: data recording, specifically including:

[0152] connecting the data recording module to the image processing unit, the vehicle bus system and the external storage device;

[0153] When a warning event is triggered, the data recording module automatically starts recording, including but not limited to the timestamp of the warning, vehicle speed, steering angle, camera current field of view parameters, surrounding environment images and vehicle CAN bus signals; all recorded data are compressed and encrypted and stored in a high-durability storage medium.

[0154] The data recording function realizes the automatic collection, processing and storage of key data before and after the warning event through highly integrated hardware and software design, providing valuable data resources for post-analysis, troubleshooting and driving behavior research. The specific implementation methods are as follows:

[0155] Integration and connection of data logging modules

[0156] Hardware connection: The data recording module (usually called a black box or event data recorder EDR) is connected to the image processing unit, the vehicle bus system, and an external storage device (such as an SD card, solid-state drive SSD) through the vehicle's internal communication network (such as the CAN bus). Such a connection ensures that the data recording module can access the vehicle status, image processing results, and other sensor information in real time.

[0157] Software Integration: At the software level, the control logic of the data logging module is integrated into the vehicle’s main control system, so that when a warning event is triggered, the data logging process can be automatically started without human intervention.

[0158] Data recording trigger conditions and contents

[0159] Warning event trigger: When the lane change assist system detects a lane change hazard and triggers a warning, it is considered a data recording event. In addition, other trigger conditions can also be set according to the system design, such as manual emergency braking, collision sensing, etc.

[0160] Record data content:

[0161] Timestamp: records the exact time when the warning occurs. Vehicle speed: current vehicle speed information. Steering angle: real-time angle data of the vehicle's steering system, reflecting the driver's operating intention. Camera current field of view parameters: including camera focal length, field of view angle, exposure settings, etc., to evaluate image quality. Surrounding environment image: video or image of the surrounding environment captured by the electronic rearview mirror before and after the danger occurs. Vehicle CAN bus signal: including but not limited to engine status, brake status, accelerator pedal position, etc., providing a comprehensive vehicle operating status.

[0162] Data processing and storage

[0163] Data compression and encryption: All collected data will be efficiently compressed before storage to save storage space. At the same time, in order to protect privacy and data security, the data will be encrypted to ensure that only authorized personnel or systems can access and decrypt.

[0164] High-durability storage media: In order to preserve data for a long time and withstand extreme environments, data is stored in storage media with high durability and vibration resistance. These media are often designed with fault tolerance, such as solid-state drives in RAID configurations, or SD cards with special protection measures to ensure that data is traceable throughout the vehicle's life cycle.

[0165] On the other hand, the present invention proposes a lane change assistance system based on an electronic rearview mirror, such as Figure 2 As shown, it includes: an image acquisition module, an image processing module, a display module, a camera adjustment module and an early warning module.

[0166] Specifically, the image acquisition module is used to install a camera under the side rearview mirror of the vehicle, the camera has an intelligent zoom function and an integrated night vision enhancement mode, and its viewing angle range is not less than 120 degrees;

[0167] Specifically, the image processing module is used to optimize the image signal captured by the camera through the image processing unit, including but not limited to brightness adjustment, contrast enhancement and fisheye correction;

[0168] Specifically, the display module is used to transmit the processed image signal to the vehicle display screen for real-time display, and at the same time identify and mark the side and rear vehicles through intelligent algorithms;

[0169] Specifically, the camera adjustment module is used to receive the CAN bus signal of the vehicle's driving status through the vehicle network, and automatically adjust the camera's field of view according to the vehicle speed (V), steering angle (θ) and vehicle driving direction. The calculation formula is: field of view width (W) = f(V,θ) = k1*V^2+k2*θ+k3, where k1, k2, k3 are preset coefficients;

[0170] Specifically, the warning module is used for blind spot monitoring. When a vehicle approaches to a dangerous distance (D_danger) from the side and rear, a multi-sensory lane change warning signal (A_warn) is issued. The strength of the warning signal is related to the speed (V_approach) and relative distance (D_re l) of the approaching vehicle. The warning formula is:

[0171] A_warn=g(D_re l,V_approach)=k4*ln(D_re l)+k5*V_approach^2, where k4 and k5 are preset constants.

[0172] In addition, when the above-mentioned image acquisition module, image processing module, display module, camera adjustment module and warning module are executed, they are also used to realize other functions of the above-mentioned lane change assistance method based on electronic rearview mirror, which will not be described one by one here.

[0173] In summary, the lane change assistance method of the electronic rearview mirror proposed in the present invention significantly enhances the driver's perception of the side and rear environment through the application of intelligent cameras, image processing technology, automatic field of view adjustment and advanced warning systems, reduces the risk of lane change, and improves the overall safety and comfort of driving. The application of this technology not only solves the problem of limited field of view of traditional rearview mirrors, but also opens up a new path for the development of intelligent driving assistance systems, showing the cutting-edge trend of automobile safety technology.

[0174] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A lane change assistance method based on an electronic rearview mirror, characterized in that: The following steps are involved: A camera is installed below the side rearview mirror of the vehicle, the camera having an intelligent zoom function and an integrated night vision enhancement mode, and its viewing angle range is not less than 120 degrees; The image signal captured by the camera is optimized and processed by an image processing unit, including brightness adjustment, contrast enhancement and fisheye correction; The processed image signal is transmitted to the on-board display screen for real-time display, and the vehicles behind are identified and marked through intelligent algorithms; The CAN bus signal of the vehicle's driving status is received through the vehicle network, and the camera's field of view is automatically adjusted according to the vehicle speed V, steering angle θ and vehicle driving direction. The calculation formula is: field of view width W=f(V,θ)=k1*V^2+k2*θ+k3, where k1, k2, k3 are preset coefficients; Blind spot monitoring: When a vehicle approaches the dangerous distance D_danger from the rear, a multi-sensory lane change warning signal A_warn is issued. The strength of the warning signal is related to the speed V_approach and relative distance D_rel of the approaching vehicle. The warning formula is: A_warn=g(D_rel,V_approach)=k4*ln(D_rel)+k5*V_approach^2, where k4 and k5 are preset constants.

2. The lane change assistance method based on an electronic rearview mirror according to claim 1, characterized in that: The image signal captured by the camera is optimized and processed by an image processing unit, including: Get the original image signal output by the camera; Using real-time data from gyroscopes and accelerometers, the vehicle's motion posture parameters in three-dimensional space are calculated, including roll angle, yaw angle, and pitch angle. The original image is compensated according to the calculated motion posture parameters, and the inter-frame motion is corrected by translation and rotation, and the clarity and contrast of the corrected image are enhanced.

3. The lane change assistance method based on an electronic rearview mirror according to claim 1, characterized in that: The intelligent algorithm uses a deep learning model to perform side and rear vehicle recognition, including: Data preprocessing: Collect image samples with different lighting conditions, weather conditions, vehicle types and driving status, and perform grayscale, normalization and size unification preprocessing operations; Training set construction: Select training sets from preprocessed samples and label them; Build a deep learning network: Design a neural network architecture that includes convolutional layers, pooling layers, fully connected layers, and output layers. The number of network layers, the number of neurons in each layer, and the activation function are determined based on the balance between recognition accuracy and computing resources. Network training: Using the training set data, the back propagation algorithm combined with the optimization algorithm is used to adjust the network weights and perform multiple iterative training; Model validation and optimization: Use an independent validation set to evaluate the performance of the trained model and adjust the network parameters based on the evaluation results; Real-time application: The optimized deep learning model is deployed in the image processing unit to perform vehicle recognition on the real-time images captured by the camera. By setting a threshold, it is determined whether there is a vehicle approaching from the side and rear, and the vehicle is marked and displayed on the vehicle display.

4. The lane change assistance method based on an electronic rearview mirror according to claim 1, characterized in that: The intelligent zoom method of the camera comprises: The side and rear images captured by the camera are acquired in real time, and the size and distance information of the objects in the images are preliminarily analyzed by the image processing unit, and multi-level focal length preset values ​​are set, including far focus, standard focus and near focus, corresponding to different detection distance ranges; Use object recognition algorithm to identify the vehicle behind you and calculate the estimated distance between the vehicle and the vehicle. The estimated distance is compared with the preset distance interval threshold. If the estimated distance falls within the long distance interval, the telephoto mode is activated; at a medium distance, the standard focal length is enabled; and at a close distance, the near focal length mode is switched; The position of the lens group inside the camera is adjusted instantly through the motor drive to achieve rapid focal length conversion. After the focal length adjustment is completed, the relative distance changes between vehicles continue to be monitored and the focal length is dynamically adjusted to adapt to the dynamic distribution of vehicles in the driving environment.

5. The lane change assistance method based on an electronic rearview mirror according to claim 1, characterized in that: The method for issuing the multi-sensory lane change warning signal includes: When the vehicle behind you approaches the preset danger distance, the warning algorithm is immediately triggered to calculate the warning level, which is determined by the speed and relative distance of the approaching vehicle; Depending on the warning level, visual warnings of varying intensity, acoustic warnings or seat vibration warnings are activated.

6. The lane change assistance method based on an electronic rearview mirror according to claim 1, characterized in that: The blind spot monitoring also includes predicting the curvature of the road ahead through GPS and map data, including: Receive and analyze GPS signals in real time to determine the vehicle's current location and direction of travel; Retrieve geometric information of the road ahead from a built-in or online map database in combination with the vehicle's driving direction, including curve radius, curve direction, and curve length; Analyze the characteristics of the curve, determine the appropriate time to adjust the field of view before the vehicle enters the curve, and calculate the optimal field of view angle α_optimal and the field of view range adjustment amount ΔW to which the camera should be adjusted according to the characteristics of the curve. The calculation formula is: α_optimal=f(R_curve,V_vehicle); ΔW=g(R_curve,V_vehicle); R_curve is the curve radius, V_vehicle is the vehicle speed; Send instructions to the camera to adjust the camera angle and field of view. After passing the curve, it will automatically return to the standard field of view setting for straight driving.

7. The lane change assistance method based on an electronic rearview mirror according to claim 1, characterized in that: The method further includes: obstacle identification and BSD warning, including: The camera continuously captures the video stream from the side and rear of the vehicle and transmits the image data to the image processing unit, which uses a deep learning algorithm to analyze the image and identify potential dangerous factors such as vehicles, pedestrians, and static obstacles in the image; Calculate the relative distance, speed difference and potential collision time between each identified obstacle and the vehicle; When the collision time of any obstacle is lower than the preset safety threshold, the system immediately triggers the blind spot monitoring warning. At the same time, the system dynamically adjusts the display content of the electronic rearview mirror according to the location of the obstacle, highlighting the area with collision risk to guide the driver's attention; If the driver starts to change lanes and there is an obstacle in the blind spot, the warning level will be increased and a stronger warning signal will be used until the driver takes evasive measures or the obstacle no longer poses a threat.

8. The lane change assistance method based on an electronic rearview mirror according to claim 1, characterized in that: The installation position and adjustment of the camera include: One or more preset camera installation sites are provided, and the camera is connected to the installation site using a ball joint or a universal joint seat, so that it can be adjusted in multiple dimensions in the horizontal and vertical planes; The driver can manually or electrically adjust the specific installation angle of the camera, including the pitch angle, roll angle and yaw angle.

9. The lane change assistance method based on an electronic rearview mirror according to claim 1, characterized in that: The method further comprises: data recording, comprising: connecting the data recording module to the image processing unit, the vehicle bus system and the external storage device; When a warning event is triggered, the data recording module automatically starts recording, including the timestamp of the warning, vehicle speed, steering angle, camera current field of view parameters, surrounding environment images and vehicle CAN bus signals; All recorded data is compressed and encrypted and stored in high-durability storage media.

10. A lane change assistance system based on an electronic rearview mirror, characterized in that: include: An image acquisition module is used to install a camera under the side rearview mirror of the vehicle, wherein the camera has an intelligent zoom function and an integrated night vision enhancement mode, and its viewing angle range is not less than 120 degrees; An image processing module, used for optimizing the image signal captured by the camera through an image processing unit, including brightness adjustment, contrast enhancement and fisheye correction; The display module is used to transmit the processed image signal to the vehicle display screen for real-time display, and at the same time identify and mark the vehicles behind the side through intelligent algorithms; The camera adjustment module is used to receive the CAN bus signal of the vehicle's driving status through the vehicle network, and automatically adjust the camera's field of view according to the vehicle speed V, the steering angle θ and the vehicle's driving direction. The calculation formula is: field of view width W=f(V,θ)=k1*V^2+k2*θ+k3, where k1, k2, k3 are preset coefficients; The warning module is used for blind spot monitoring. When a vehicle approaches the dangerous distance D_danger from the side and rear, a multi-sensory lane change warning signal A_warn is issued. The strength of the warning signal is related to the speed V_approach and relative distance D_rel of the approaching vehicle. The warning formula is: A_warn=g(D_rel,V_approach)=k4*ln(D_rel)+k5*V_approach^2, where k4 and k5 are preset constants.

Images