Vehicle rearview mirror and vehicle

Abstract

This invention relates to a vehicle rearview mirror, including at least one built-in camera. The rearview mirror is disposed on the exterior of the vehicle body and has at least a first holding posture and a second holding posture relative to the vehicle body. The camera has at least a first operating state and a second operating state. In the first operating state, the camera captures images of the external environment of the vehicle; in the second operating state, the camera captures images of the vehicle's cabin. The camera is configured to be in the first operating state when the rearview mirror is in the first holding posture and in the second operating state when the rearview mirror is in the second holding posture. Thus, the switching of the rearview mirror's holding posture can be associated with the switching of the rearview mirror camera's function, allowing the rearview mirror camera to acquire information about both the external environment surrounding the vehicle and the internal environment of the vehicle's cabin, achieving functional reuse of the rearview mirror camera.

Description

Technical Field

[0001] This invention relates to a vehicle rearview mirror, and more particularly to a vehicle rearview mirror with a built-in camera. It also relates to a corresponding vehicle. Background Technology

[0002] Traditional vehicle rearview mirrors mostly use optical reflection structures, which have a limited field of view and blind spots. To improve the observation effect, cameras are usually integrated into the exterior rearview mirrors of the vehicle. These cameras can capture images of the vehicle's surroundings in real time and transmit them to the in-vehicle display screen for the vehicle occupants to view, thereby expanding the field of view, reducing blind spots, and can also be used for ADAS functions. Summary of the Invention

[0003] The purpose of this invention is to propose an improved vehicle rearview mirror that has a built-in camera, and through the functional reuse of the camera, to propose a new application scenario for the vehicle rearview mirror and the rearview mirror camera.

[0004] According to a first aspect of the present invention, a vehicle rearview mirror is provided, including at least one built-in camera, the rearview mirror being disposed outside a vehicle body and having at least a first holding posture and a second holding posture relative to the vehicle body, the camera having at least a first operating state and a second operating state, wherein in the first operating state the camera captures an image of the external environment of the vehicle, and in the second operating state the camera captures an image of the vehicle cabin, the camera being configured to be in the first operating state when the rearview mirror is in the first holding posture, and in the second operating state when the rearview mirror is in the second holding posture.

[0005] According to an optional embodiment of the present invention, the camera is a wide-angle camera or a 360-degree surround-view camera.

[0006] According to an alternative embodiment of the invention, the rearview mirror is configured to switch between a first holding posture and a second holding posture by folding relative to the vehicle body.

[0007] According to an optional embodiment of the present invention, the camera is configured to automatically initiate a switch from a first operating state to a second operating state in response to a change in the posture of the rearview mirror from a first holding posture to a second holding posture.

[0008] According to an optional embodiment of the invention, the camera is configured to automatically adjust its field of view in a second operating state based on the position of the occupants in the vehicle cabin.

[0009] According to an optional embodiment of the invention, the camera is configured to automatically recalibrate extrinsic parameters based on the vehicle's fixed structure after the rearview mirror switches between a first holding posture and a second holding posture.

[0010] According to an optional embodiment of the present invention, the extrinsic parameter matrix used for calibrating the extrinsic parameters is obtained by solving the pose of the camera after the rearview mirror switches to the second holding posture based on the EPnP algorithm and optimizing the reprojection error using the Levenberg-Marquardt algorithm.

[0011] According to an optional embodiment of the present invention, the camera is configured to automatically recalibrate extrinsic parameters based on the vehicle's fixed structure after the rearview mirror switches from a first holding posture to a second holding posture, and to automatically return to the initial extrinsic parameters after the rearview mirror switches back from the second holding posture to the first holding posture.

[0012] According to an optional embodiment of the invention, the rearview mirror is configured to switch from a first holding posture to a second holding posture in response to a parking command and / or a user command.

[0013] According to an optional embodiment of the present invention, the camera is configured to automatically initiate the switch from the first operating state to the second operating state only after the rearview mirror switches from the first holding posture to the second holding posture, provided that user consent is obtained and / or the vehicle gear position is detected to be in parking gear.

[0014] According to an optional embodiment of the invention, the second operating state of the camera is disabled during vehicle operation.

[0015] According to an optional embodiment of the present invention, the camera is communicatively connected to a controller for controlling the rearview mirror to transmit captured images to the controller, the controller being configured to perform at least one of the following: storing images of the vehicle's external environment and / or vehicle cabin captured by the camera into a storage unit; displaying images of the vehicle's external environment and / or vehicle cabin captured by the camera on an in-vehicle display; and providing driving assistance based at least on the images of the vehicle's external environment captured by the camera.

[0016] According to an optional embodiment of the invention, the camera is communicatively connected to a controller for controlling the rearview mirror to transmit captured images to the controller, which is configured to acquire vehicle cabin images captured by other vehicle cameras and fuse them with the vehicle cabin images captured by the camera, particularly using lightweight 3D reconstruction technology to generate 2D or 3D fused visual information of the vehicle cabin, especially video.

[0017] According to an optional embodiment of the present invention, the controller is further configured to acquire voice information of the vehicle cabin when the camera is in a second operating state, and superimpose the voice information onto the vehicle cabin image captured by the camera to generate a video.

[0018] According to an optional embodiment of the present invention, the controller is a vehicle controller.

[0019] According to a second aspect of the invention, a vehicle is provided, including any of the vehicle rearview mirrors according to the invention.

[0020] Through certain embodiments of the present invention, the attitude switching of the vehicle's rearview mirror can be associated with the function switching of the camera, enabling the rearview mirror camera to acquire information about both the external environment surrounding the vehicle and the internal environment of the vehicle's cabin, thus achieving camera function reuse. Furthermore, by optimizing the fusion processing of images captured by the rearview mirror camera and other vehicle-mounted cameras, the quality of the in-vehicle stereoscopic video and the computational power requirements are improved.

[0021] It is worth noting that the advantages and beneficial effects of the present invention are not limited to those mentioned above. Those skilled in the art can understand other unmentioned advantages and beneficial effects of the present invention through the following specific embodiments and claims. Attached Figure Description

[0022] The invention will now be described in more detail with reference to the accompanying drawings, which will provide a better understanding of the principles, features, and advantages of the invention. In the drawings, Figure 1 A schematic diagram of an exemplary vehicle rearview mirror is shown, wherein the rearview mirror is in a first holding posture; Figure 2 Another schematic diagram of an exemplary vehicle rearview mirror is shown, in which the rearview mirror is in a second holding posture; Figure 3 A further schematic diagram of an exemplary vehicle rearview mirror is shown, in which an exemplary camera is illustrated; and Figure 4 An example of a vehicle rearview mirror's communication connection with other vehicle components is shown. Detailed Implementation

[0023] To make the technical problems to be solved, the technical solutions, and the beneficial technical effects of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and exemplary embodiments. It should be understood that the specific embodiments described herein are only for explaining the principles of the invention and are not intended to limit the scope of protection of the invention.

[0024] Figures 1 to 3 Schematic diagrams of an exemplary vehicle rearview mirror 1 are shown. In this document, a vehicle rearview mirror may also be referred to simply as a rearview mirror.

[0025] like Figures 1 to 3As shown, the vehicle rearview mirror 1 according to the present invention includes at least one built-in camera 11. The rearview mirror 1 is arranged on the exterior of the vehicle body, i.e., an exterior rearview mirror. The rearview mirror 1 can be connected to the vehicle body 2; for example, the rearview mirror 1 may include a bracket, and the posture of the rearview mirror 1 can be changed by the movement of the bracket. One rearview mirror 1 can be arranged on each side of the vehicle body. The rearview mirror 1 also includes a built-in motor for driving the movement of the bracket.

[0026] The rearview mirror 1 has at least a first holding posture and a second holding posture relative to the vehicle body 2. The rearview mirror 1 can be configured to move at least partially relative to the vehicle body 2 while having at least the first holding posture and the second holding posture. The movement of the rearview mirror 1 relative to the vehicle body 2 can include translation, rotation, or a combination thereof.

[0027] In a specific example, the rearview mirror 1 can be configured to switch between a first holding posture and a second holding posture by folding it relative to the vehicle body 2. For example, the first holding posture of the rearview mirror 1 can correspond to the normal posture of the rearview mirror when the vehicle is in motion, that is, extending laterally outward from the vehicle body 2 and in a fully open state. The rearview mirror 1 can switch from the first holding posture to the second holding posture by at least folding and rotating it toward the vehicle body 2, that is, in the second holding posture, the outer end of the rearview mirror 1 away from the vehicle body 2 is closer to the vehicle body 2 than in the first holding posture.

[0028] In a specific example, the angle at which the rearview mirror 1 rotates toward the vehicle body 2 during the transition from a first holding posture to a second holding posture is in the range of about 20 degrees to about 30 degrees, particularly in the range of about 22 degrees to about 28 degrees.

[0029] The rearview mirror 1 can be fully retracted into the vehicle body 2 so that the surface of the rearview mirror 1 is flush with the surface of the vehicle body 2. This is known in the art and is mainly used to narrow the vehicle body width to facilitate the vehicle's passage through narrow sections of road or for parking. The second holding posture of the rearview mirror 1 may be different from this fully retracted posture.

[0030] The attitude change of the rearview mirror 1 can be manually performed by the vehicle occupants or by other means. Figure 4 The controller 3 shown is based on program command control. Here, controller 3 can be a vehicle controller (such as an ECU), in which case the rearview mirror 1 can be connected to controller 3 via the vehicle CAN bus to be controlled by controller 3. Controller 3 can also be a separate rearview mirror controller, in which case the separate rearview mirror controller can be connected to the vehicle controller via the vehicle CAN bus.

[0031] In addition, the image or video data captured by the rearview mirror camera can be transmitted to the vehicle controller via Ethernet.

[0032] The built-in camera 11 of the rearview mirror 1 can have at least a first operating state and a second operating state. In the first operating state, the camera 11 captures images of the vehicle's external environment. This means that in this first operating state, the camera 11 can be oriented to capture images of the vehicle's external environment, i.e., its field of view can cover the vehicle's external environment. In the second operating state, the camera 11 captures images of the vehicle's cabin. This means that in this second operating state, the camera 11 can be oriented to capture images of the vehicle's cabin, i.e., its field of view can cover the vehicle's cabin. This can be achieved, for example, by adjusting or controlling the holding posture of the rearview mirror 1, or, for example, by adjusting the field of view of the camera 11 individually.

[0033] The camera 11 can be configured to be in a first working state when the rearview mirror 1 is in a first holding posture, and in a second working state when the rearview mirror 1 is in a second holding posture.

[0034] In this way, the attitude switching of the rearview mirror 1 can be linked to the function switching of the camera 11, allowing the built-in camera 11 of the rearview mirror 1 to capture both the external environment around the vehicle and the internal environment of the vehicle cabin, thus achieving functional reuse of the camera 11. Compared with existing cabin video recording, the built-in camera of the vehicle rearview mirror can be reused without the need for additional or new hardware, which has a cost advantage.

[0035] The camera 11 can communicate with the controller 3 independently of the rearview mirror 1. That is, the signal / information transmission path between the controller 3 and the camera 11 can be different from the signal / information transmission path between the controller 3 and other components of the rearview mirror 1.

[0036] The camera 11 can be fixed in the rearview mirror 1, thereby changing its field of view as the rearview mirror 1 moves. For example, when the rearview mirror 1 is in the second holding position, the field of view of the camera 11 can be turned to the interior of the cabin, covering the front and rear seats of the cabin.

[0037] exist Figure 3 The diagram schematically shows a rearview mirror 1 having a glass lens, an outer frame surrounding and supporting the glass lens, and a camera 11 disposed within the outer frame. However, it is not limited to this; the camera 11 can be disposed in any other suitable location on the rearview mirror 1. For example, the camera 11 can also be disposed in a bracket of the rearview mirror 1 that connects to the outer frame and the vehicle body 2.

[0038] Through the reflective properties of the glass lens, vehicle occupants, including the driver, can directly observe the external environment information of the vehicle, especially the road conditions behind and to the sides. However, in some other embodiments, the rearview mirror 1 may not have the glass lens shown in the figure. In this case, the external environment information captured by the camera 11 can be transmitted to the in-vehicle display screen for viewing by vehicle occupants, including the driver.

[0039] Camera 11 can be a wide-angle camera or a 360-degree surround-view camera. This allows camera 11 to have a larger field of view in its first operating state, thereby capturing a wider range of images of the vehicle's external environment and effectively reducing blind spots. For example, camera 11 can have a horizontal field of view (FOV) of at least 120 degrees.

[0040] Additionally, the rearview mirror 1 may have at least one of the aforementioned cameras 11. For example, the rearview mirror 1 may have multiple cameras 11, each of which may have at least a first operating state and a second operating state. Each camera 11 may be controlled independently of the others, for example, its field of view may be adjusted individually in the second operating state.

[0041] The rearview mirror 1 may also have another camera in addition to the camera 11 described above. This other camera may be the same as or different from the camera 11 described above, but may not be configured in accordance with the inventive concept and may have different working states associated with the attitude change of the rearview mirror 1.

[0042] The camera 11 can be configured to automatically switch from a first operating state to a second operating state in response to a change in the attitude of the rearview mirror 1, which changes from a first holding posture to a second holding posture. Thus, when the holding posture of the rearview mirror 1 switches between the first and second holding postures, the function and corresponding operating state of the camera 11 can automatically switch accordingly. This facilitates convenient function switching control of the camera 11.

[0043] The changes in the holding posture of the rearview mirror 1 can be detected and acquired in various ways. For example, a position sensor can be installed separately at the rearview mirror to directly detect the actual holding posture of the rearview mirror 1 in real time. Alternatively, the rearview mirror controller (such as an onboard controller or ECU) can detect changes in the holding posture of the rearview mirror 1 based on the rearview mirror control commands it issues or the state parameters of specific components of the rearview mirror. Here, the state parameters of specific components of the rearview mirror may include, for example, the current signal of the rearview mirror actuator motor or the rotation angle.

[0044] The two rearview mirrors 1 on both sides of the vehicle body can move synchronously to switch between holding postures, thereby causing the corresponding cameras 11 to switch their working states synchronously. In some embodiments, only one of the two rearview mirrors 1 on both sides of the vehicle body switches from a first holding posture to a second holding posture, and only its corresponding camera 11 switches from a first working state to a second working state to acquire images of the vehicle cabin.

[0045] In the first working state, the field of view of camera 11 can be kept fixed; in the second working state, the field of view of camera 11 can be finely adjusted within a preset range.

[0046] Camera 11 can be configured to automatically adjust its field of view based on the position of the occupants within the vehicle cabin in a second operating state. That is, when camera 11 switches to the second operating state, its field of view has already shifted from the external environment of the vehicle to the vehicle cabin. Then, in the second operating state, camera 11 further adjusts its field of view based on the specific position of the occupants within the cabin. This may include determining the specific position of the occupants within the cabin through appropriate means, such as image recognition or seat weight sensors, and adjusting the field of view of camera 11 based on the occupants' positions after determining their specific positions. Here, image recognition technology may rely on image data acquired by camera 11 and / or image data acquired by other onboard cameras.

[0047] When the occupant's specific position inside the cabin changes, the camera 11 can also automatically adjust its field of view in the second operating state to follow the occupant's position change. This allows the camera 11 to achieve optimized dynamic viewing angle adjustment in the second operating state.

[0048] Camera 11 can be configured to automatically recalibrate its extrinsic parameters based on the vehicle's fixed structure after the rearview mirror 1 switches between a first holding posture and a second holding posture. When the holding posture of the rearview mirror 1 changes, the extrinsic parameters of camera 11 need to be recalibrated to ensure the accuracy of the images captured by camera 11. Therefore, camera 11 can be configured to automatically recalibrate its extrinsic parameters when a change in the holding posture of the rearview mirror 1 is detected. The extrinsic parameter calibration step can be completed during the interval when camera 11 switches operating states or after the switch. This eliminates the need for a manual calibration plate and adapts to dynamic changes in the viewing angle of camera 11.

[0049] The aforementioned vehicle fixed structure can refer to a vehicle structure that remains fixed relative to the vehicle body at least when the camera 11 performs external parameter calibration, including but not limited to air conditioning vents, B-pillars, interior door handles, etc.

[0050] To perform extrinsic parameter calibration, feature detection is the first step. In this step, different algorithms can be used for different vehicle fixed structures, or multiple algorithms can be used in collaboration. For example, the Harris algorithm can be used to accurately extract the corner points of the air conditioning vents. This algorithm has high sensitivity to image edges and corners, effectively capturing the key features of the air conditioning vents. The LSD algorithm can be used to extract the vertical lines of the B-pillars, thus quickly identifying the linear features of the B-pillars. Edge detection algorithms can be used to extract the contours of the interior door handles to clearly capture their shape features. Therefore, key features inside the vehicle can be accurately captured, providing a reliable basis for subsequent pose determination and extrinsic parameter calibration of camera 11.

[0051] The extrinsic parameter matrix used for calibration can be obtained by solving the pose of camera 11 after the rearview mirror 1 switches to the second holding posture based on the EPnP algorithm, and then optimizing the reprojection error using the Levenberg-Marquardt algorithm. For example, the spatial pose of camera 11 (including rotation matrix and translation vector) can be inferred using the feature information extracted in the feature detection step mentioned above, thereby optimizing the accuracy and finally obtaining the extrinsic parameter calibration matrix. In this way, projection deviation can be minimized, the accuracy of camera pose solution can be improved, and the visual positioning requirements in vehicle scenarios can be adapted.

[0052] When using the above method for external parameter calibration, the calibration time can be within 10 seconds, the calibration error can be less than 2 mm, and the relative error of the depth map accuracy can be less than 5%.

[0053] The camera 11 can be configured to automatically recalibrate its extrinsic parameters based on the vehicle's fixed structure after the rearview mirror switches from a first holding posture to a second holding posture, and automatically return to its initial extrinsic parameters after the rearview mirror switches back from the second holding posture to the first holding posture. Typically, the rearview mirror has a preset first holding posture, and correspondingly, the camera 11 can have a preset initial extrinsic parameter. Therefore, when the rearview mirror switches back to the first holding posture, it is not necessary to recalibrate the extrinsic parameters; instead, the extrinsic parameters of the camera 11 can be restored to their initial extrinsic parameters.

[0054] The rearview mirror 1 can be configured to switch from a first holding posture to a second holding posture in response to a parking command and / or a user command. That is, when the vehicle is parked, the rearview mirror 1 can automatically switch from the first holding posture to the second holding posture.

[0055] Camera 11 can be configured to automatically initiate the switch from the first operating state to the second operating state only after obtaining user consent and / or detecting that the vehicle's gear position is park. That is, when the rearview mirror changes its operating state, controller 3 can send a confirmation message to the user and only allow camera 11 to switch from the first operating state to the second operating state after obtaining user consent. Additionally or alternatively, when the rearview mirror changes its operating state, controller 3 can also detect the vehicle's gear position and only allow camera 11 to switch from the first operating state to the second operating state after determining that the current gear position is park. This allows for precise control of the camera's operating state switching and makes the control more closely match the user's actual intentions.

[0056] While the vehicle is in motion, the second operating state of camera 11 can be disabled. This ensures that even if the rearview mirror 1 is accidentally or unintentionally switched from the first holding posture to the second holding posture while the vehicle is in motion, camera 11 will not switch from the first operating state to the second operating state. This ensures that camera 11 can always remain in the first operating state to monitor the external environment of the vehicle in real time while the vehicle is in motion, thereby ensuring driving safety.

[0057] Therefore, the current driving status of the vehicle, such as the vehicle speed, can be detected before camera 11 switches to the second operating state. When the vehicle speed is not zero, the second operating state of camera 11 can be disabled, thereby preventing camera 11 from switching to the second operating state.

[0058] Camera 11 can communicate with controller 3, which controls rearview mirror 1, to transmit captured images to controller 3. Controller 3 can be a vehicle controller, such as an ECU. Figure 4 As shown, the camera 11 can communicate with the controller 3 via the vehicle's CAN bus to be controlled by the controller 3 and transmit information / commands, etc.

[0059] The controller 3 can be configured to store images of the vehicle's external environment and / or the vehicle's cabin captured by the camera 11 into a storage unit. Here, the storage unit can be a storage unit such as that of a vehicle controller or a cloud storage unit.

[0060] The controller 3 can be configured to display images of the vehicle's external environment and / or the vehicle's cabin captured by the camera 11 on an in-vehicle display. This in-vehicle display can be a center console display, an instrument cluster auxiliary display, a rear-seat entertainment display, or a head-up display (HUD), etc. The images of the vehicle's external environment and the vehicle's cabin can be displayed on the same or different in-vehicle displays.

[0061] The controller 3 can be configured to provide driving assistance based at least on images of the vehicle's external environment captured by the camera 11. This is known in the art and will not be elaborated upon here.

[0062] Controller 3 can be configured to acquire data from other vehicle cameras 4 (such as...). Figure 4 The system captures images of the vehicle cabin (as shown) and fuses them with images of the vehicle cabin captured by camera 11 to generate 2D or 3D fused visual information of the vehicle cabin. This fused visual information can be video.

[0063] Lightweight 3D reconstruction technology can be used during video generation. This balances the computational constraints and imaging accuracy requirements of in-vehicle scenarios, enabling real-time and efficient generation of stereoscopic visual information.

[0064] This lightweight 3D reconstruction technology can integrate an improved stereo matching algorithm with an onboard optimized NeRF (Neural Radiation Field) model.

[0065] The improved stereo matching algorithm optimizes feature extraction, matching cost calculation, and discrete optimization for in-vehicle scenarios. This can effectively improve the accuracy and speed of feature matching in complex in-vehicle environments (such as strong light, low light, rain, and occlusion scenarios) and reduce computing power consumption to adapt to the limited computing power resources of in-vehicle terminals.

[0066] The vehicle-optimized NeRF model is a lightweight improvement based on the standard NeRF model. By simplifying the network structure, optimizing the sampling strategy, and introducing prior knowledge of vehicle scenarios, it reduces parameters and computational load while retaining the advantages of high-precision stereo reconstruction. It can quickly convert two-dimensional images into spatial depth stereo data.

[0067] By deeply integrating the two optimization technologies mentioned above, lightweight 3D reconstruction technology can give full play to the advantages of both technologies. It can process vehicle cabin images captured by rearview mirror cameras and other optional vehicle cameras in real time with low computing power consumption, generate continuous and accurate spatial stereoscopic videos, and clearly present the spatial positional relationships and detailed features of the cabin.

[0068] In a specific example, the original vehicle cabin image captured by camera 11 can first be preprocessed. The Bouguet algorithm is used for epipolar correction, and the CLAHE algorithm is used for illumination equalization, laying a stable foundation for subsequent depth calculations. Subsequently, an improved SGBM algorithm can be used for stereo matching and depth estimation. The matching cost is calculated using Census transform, and edge-aware dynamic programming is combined to optimize disparity solving, resulting in high-precision cabin depth information. In the real-time 3D rendering stage, a lightweight NeRF model can be used, introducing Instant-NGP hash encoding to improve feature encoding efficiency. The MLP network is compressed to four layers and combined with Tensor RT to accelerate inference, ensuring real-time rendering under low computational constraints. For dynamic scenes such as passenger movement, Deformable NeRF can be used to model human deformation. Furthermore, the RAFT optical flow method can be used to complete the texture and depth of occluded areas, eliminating reconstruction errors and information loss caused by dynamic targets.

[0069] Through the above processing flow, cockpit images can be processed in real time with low computing power consumption, ultimately generating continuous and accurate spatial stereoscopic video that clearly restores the spatial positional relationships and object details inside the cockpit.

[0070] Compared with traditional 360-degree surround view systems, this solution can support 3D interactive perspectives and improve the accuracy of passenger motion reproduction by about 40%.

[0071] Compared to commercial NeRF solutions, by utilizing the aforementioned automotive-grade lightweight NeRF, the model size can be significantly reduced (e.g., reduced to 10MB, a reduction of approximately 90%), rendering latency can be reduced (e.g., less than 20ms at 1080p resolution), and power consumption can be reduced (by approximately 70%).

[0072] The controller 3 can also be configured to acquire voice information from the vehicle cabin when the camera 11 is in its second operating state, and overlay the voice information onto the vehicle cabin image captured by the camera 11 to generate video. To this end, the controller 3 can be configured to automatically activate or invoke the in-cabin radio to acquire voice information from the vehicle cabin after the camera 11 switches to its second operating state.

[0073] In a specific example, after the vehicle stops, the rearview mirror 1 automatically flips up from a first holding posture to a second holding posture. Correspondingly, the camera 11 automatically switches from a first operating state to a second operating state, thereby capturing images of the vehicle's cabin. These images can be fused with cabin images captured by other vehicle cameras to generate a 3D video. This allows for capturing scenes such as parent-child interactions and meetings / office work within the vehicle cabin. The controller 3 can control the automatic flipping of the rearview mirror 1 by, for example, receiving a parking command, detecting zero speed, or detecting that the vehicle is in parking gear.

[0074] The generated videos can be saved in the storage unit for later playback. Images acquired by camera 11 and the generated videos can also be categorized and grouped according to preset tags to create in-car photo albums or video collections. Preset tags can include shooting time, location, people, etc. Categorization and grouping rules can be based on time, location, people, and / or scenes. Therefore, the rearview mirror camera 11 can be integrated into the in-vehicle entertainment system to provide personalized services such as "in-car time memories."

[0075] The video format can be either Side-by-Side 3D (SBS 3D) or MV-HEVC multi-view encoding. Both formats require no additional plugins, support Oculus Quest controller interaction, and are compatible with mainstream in-vehicle screens, offering optimized stability and support for voice and touch interaction to meet dual-scenario needs. The SBS 3D format, through frame-by-frame arrangement of left and right views and the use of VR device lenses for refraction and parallax processing, presents an immersive 3D effect. It can be directly recognized by Oculus Quest and supports basic interactions such as view switching and pausing / playing, making it simple and easy to understand. The MV-HEVC format, based on High Efficiency Video Coding (HEVC, or H.265), is a multi-view video encoding standard that offers high compression efficiency and clear, detailed images. It flexibly adapts to the display needs of different devices, meeting the multi-view immersive playback requirements of VR devices and adapting to the display ratio and resolution of in-vehicle screens, avoiding image stretching and blurring, and ensuring viewing clarity in in-vehicle scenarios.

[0076] Some embodiments of the present invention also relate to a vehicle that includes a vehicle rearview mirror according to any of the present invention.

[0077] Although specific embodiments of the invention have been described in detail herein, they are given for illustrative purposes only and should not be construed as limiting the scope of the invention. Various substitutions, alterations, and modifications can be conceived without departing from the spirit and scope of the invention.

Claims

1. A vehicle rearview mirror, comprising at least one built-in camera, The rearview mirror is arranged outside the vehicle body and has at least a first holding posture and a second holding posture relative to the vehicle body. The camera has at least a first operating state and a second operating state. In the first operating state, the camera captures images of the vehicle's external environment, and in the second operating state, the camera captures images of the vehicle's cabin. The camera is configured to be in a first working state when the rearview mirror is in a first holding posture, and in a second working state when the rearview mirror is in a second holding posture.

2. The vehicle rearview mirror according to claim 1, characterized in that, The camera is a wide-angle camera or a 360-degree surround-view camera; and / or The rearview mirror is configured to switch between a first holding posture and a second holding posture by folding it relative to the vehicle body.

3. The vehicle rearview mirror according to claim 1 or 2, characterized in that, The camera is configured to automatically initiate a switch from a first operating state to a second operating state in response to a change in the rearview mirror's posture from a first holding posture to a second holding posture; and / or The camera is configured to automatically adjust its field of view based on the position of the occupants in the vehicle cabin during a second operating state.

4. The vehicle rearview mirror according to any one of claims 1-3, characterized in that, The camera is configured to automatically recalibrate extrinsic parameters based on the vehicle's fixed structure after the rearview mirror switches between a first holding posture and a second holding posture.

5. The vehicle rearview mirror according to claim 4, characterized in that, The extrinsic parameter matrix used for calibration is obtained by solving the camera pose after the rearview mirror switches to the second hold posture based on the EPnP algorithm and optimizing the reprojection error using the Levenberg-Marquardt algorithm; and / or The camera is configured to automatically recalibrate its extrinsic parameters based on the vehicle's fixed structure after the rearview mirror switches from a first holding posture to a second holding posture, and to automatically return to its initial extrinsic parameters after the rearview mirror switches back from the second holding posture to the first holding posture.

6. The vehicle rearview mirror according to any one of claims 1-5, characterized in that, The rearview mirror is configured to switch from a first holding posture to a second holding posture in response to a parking command and / or a user command; and / or The camera is configured to automatically initiate the switch from the first working state to the second working state only after obtaining user consent and / or detecting that the vehicle is in the parking gear position after the rearview mirror switches from the first holding posture to the second holding posture. and / or The second operating state of the camera is disabled while the vehicle is in motion.

7. The vehicle rearview mirror according to any one of claims 1-6, characterized in that, The camera is communicatively connected to a controller for controlling the rearview mirror to transmit captured images to the controller, which is configured to perform at least one of the following: The images of the vehicle's external environment and / or the vehicle's cabin captured by the camera are stored in the storage unit; The images of the vehicle's external environment and / or the vehicle's cabin captured by the camera are displayed on the in-vehicle display. and Driving assistance is provided based at least on images of the vehicle's external environment captured by the camera.

8. The vehicle rearview mirror according to any one of claims 1-7, characterized in that, The camera is communicatively connected to a controller for controlling the rearview mirror to transmit captured images to the controller. The controller is configured to acquire vehicle cabin images captured by other vehicle cameras and fuse them with the vehicle cabin images captured by the camera, particularly using lightweight 3D reconstruction technology to generate 2D or 3D fused visual information of the vehicle cabin, especially video.

9. The vehicle rearview mirror according to claim 8, characterized in that, The controller is also configured to acquire voice information from the vehicle cabin when the camera is in a second operating state, and to overlay the voice information onto the vehicle cabin image captured by the camera to generate a video. and / or The controller is a vehicle controller.

10. A vehicle comprising a vehicle rearview mirror according to any one of claims 1-9.

Images