Surround view stitching correction method and device based on vehicle body posture and vehicle

Abstract

The present invention relates to the field of real-time observation technology for drivers using optical image capture systems, and in particular to a surround view stitching correction method, device, and vehicle based on vehicle body posture. The method of the present invention comprises the following steps: a) collecting vehicle body posture information and determining the vehicle body posture; b) stitching a surround view stitching image of the vehicle in a normal posture in a parking or straight-ahead posture; c) adjusting the position of the image stitching seam as the vehicle body posture changes in a left turn or right turn posture, a left lane change or a right lane change posture, and a reversing posture. In the method of the present invention, the position of the image stitching seam of the relevant cameras is adjusted accordingly according to the change of the vehicle body posture. By adjusting the position of the image stitching seam as the vehicle body posture changes, the blind spot of the field of view caused by the misalignment of the stitching seam can be reduced or even eliminated during the process of the vehicle body posture change. At the same time, this method of changing the position of the image stitching seam has a relatively simple processing and calculation process and a low cost of use.

Description

Technical Field

[0001] The present invention relates to the field of real-time observation technology for drivers using an optical image capture system, and in particular to a surround view stitching correction method and device based on vehicle body posture, and a vehicle. Background Art

[0002] The vehicle-mounted surround view system is a commonly used system used in the field of vehicle assisted driving. It usually includes four wide-angle cameras installed in the front, rear, left and right of the vehicle. The field of view of each camera covers a certain spatial area. The images captured by the wide-angle cameras are transformed through spatial coordinates and stitched together using a certain field of view stitching algorithm to obtain a surround view stitching map around the vehicle. This surround view stitching map can provide the driver with richer field of view information. In order for the wide-angle cameras installed on the vehicle to generate stitching images without distortion and misalignment (blind spots) through coordinate transformation and stitching algorithms, the camera's internal parameters (lens optical characteristics, image resolution, etc.) and external parameters (relative position of the camera, installation height and angle relative to the vehicle body, etc.) must be accurately calibrated.

[0003] At present, the calibration of the extrinsic parameters of surround-view cameras is mostly carried out by feature point matching. Theoretically, as long as two cameras simultaneously obtain three points in space, the relative posture between the two can be restored. When calibrating the extrinsic parameters of cameras, the usual calibration method is to place a calibration pattern with certain image features on the ground around the vehicle, calculate the relative spatial position of the two adjacent cameras by identifying the feature points of the calibration pattern, and stitch the images according to the image stitching algorithm. According to the principle of stitching calculation, this type of calibration method can only ensure that the image in the spatial area at the ground plane is complete (seamless stitching). For commercial vehicles with relatively high body heights, there is a large degree of blind spot in the stitching seam area at a certain height from the ground. Moreover, when stitching images from two adjacent surround-view cameras, the cameras' installation positions indicate that the stitching seams are usually located at the left front, right front, left rear, and right rear of the vehicle's surround view. For the driver, these four positions are prone to blind spots. When the vehicle is in different postures, especially when turning, changing lanes, and reversing, these positions will lead to even larger blind spots due to image stitching misalignment, posing a greater safety hazard.

[0004] Chinese invention patent publication number CN101763640B discloses an online calibration and processing method for a vehicle-mounted multi-camera surround view system. To eliminate the aforementioned blind spots and improve driving safety, this system provides a more accurate online calibration and image stitching system. To avoid the low accuracy of camera view stitching using fixed pose parameters due to changes in vehicle body posture, this surround view system uses a larger overlapping area for the original images of the vehicle-mounted multi-camera. The feature values ​​of the two sets of images in the overlapping area are then subtracted. When the vehicle body posture changes, the viewpoint transformation and feature value calculation are re-performed, and the images are stitched together using a table lookup to obtain the vehicle surround view.

[0005] Although this method can continuously reassemble the surround view as the vehicle's posture changes, the pitch and horizontal angles of the vehicle constantly change as the vehicle's posture changes. The analysis and calculation of the viewpoint transformation and image stitching are large, and the system processing speed is required to be high. The entire surround view image generation process is very cumbersome and the system cost is high. Summary of the Invention

[0006] The present invention aims to provide a vehicle body posture-based surround view stitching correction method to address the cumbersome and high-cost issues of existing vehicle body posture-based surround view stitching. Furthermore, the present invention aims to provide a vehicle body posture-based surround view stitching correction device to address the cumbersome and high-cost image generation process of existing vehicle body posture-based surround view stitching devices. Furthermore, the present invention provides a vehicle including the aforementioned vehicle body posture-based surround view stitching correction device to address the cumbersome and high-cost issues of existing vehicle surround view image generation.

[0007] The vehicle body posture-based surround view stitching correction method of the present invention comprises the following steps:

[0008] a) Collect vehicle posture information and determine the vehicle posture, including stationary, straight, left turn, right turn, left lane change, right lane change, and reversing;

[0009] b) receiving the output data of the surround-view camera image acquisition module in the parking or straight-ahead posture, and stitching a surround-view mosaic image of the vehicle in normal posture based on the intrinsic and extrinsic parameters of the surround-view camera and the set stitching seam parameters;

[0010] c) In a left-turn or right-turn posture, the position of the image stitching seam between the front camera and the left camera, or the position of the image stitching seam between the front camera and the right camera, is deflected toward the left rear side or the right rear side of the vehicle surround view mosaic image by the set turning deflection angle;

[0011] d) When changing lanes left or right, the position of the stitching seam between the front camera and the left camera, or the position of the stitching seam between the front camera and the right camera, is deflected toward the left rear side or the right rear side of the vehicle surround view mosaic by the set lane change deflection angle;

[0012] e) In the reversing posture, the position of the image stitching seam between the rear camera and the left camera, and the position of the image stitching seam between the rear camera and the right camera are deflected to the left front side and the right front side of the vehicle surround view mosaic by the set reversing deflection angle respectively, or the vehicle surround view mosaic is directly replaced by the image of the reversing camera displayed separately.

[0013] The present invention proposes a new surround view stitching correction method based on vehicle body posture. In this method, the position of the image stitching seam of the relevant cameras is adjusted accordingly according to the change of the vehicle body posture. By adjusting the position of the image stitching seam as the vehicle body posture changes, it is possible to reduce or even eliminate the blind spot of the field of view caused by the misalignment of the stitching seam during the change of the vehicle body posture, thereby improving driving safety. At the same time, this method of changing the position of the image stitching seam has a relatively simple processing operation process and a low cost of use.

[0014] Furthermore, the vehicle posture information includes output signals from an onboard gyroscope and vehicle CAN messages, which include turn signal and turn angle. By combining the output signals from the onboard gyroscope and the vehicle CAN messages, the vehicle posture information can be accurately determined, providing accurate information for adjusting the position of the image stitching seam.

[0015] Furthermore, the turning deflection angle is greater than the lane change deflection angle. This allows the image stitching seam to have a larger deflection angle when the vehicle is turning, which is a larger angle than when changing lanes. This can better avoid blind spots in the field of view caused by misaligned stitching seams during turns, thereby improving driving safety.

[0016] Furthermore, the separately displayed reversing camera image is a distortion-corrected image, which can more accurately and comprehensively display the reversing camera image when reversing, thereby improving the driver's driving experience and driving safety.

[0017] Furthermore, the turning angle information output by the vehicle CAN message is used to correct the turning deflection angle, lane change deflection angle, and reverse deflection angle. Using the turning angle information output by the vehicle CAN message to further correct the turning deflection angle, lane change deflection angle, and reverse deflection angle can make the position of the seam more accurately adjusted, improving the driver's driving experience.

[0018] Furthermore, the vehicle body posture also includes an initial uphill posture and / or an initial downhill posture. In the initial uphill posture, the Y-axis pitch angle of the surround view 3D view is raised to highlight the field of view directly in front of the vehicle. In the initial downhill posture, the Y-axis pitch angle of the surround view 3D view is lowered to highlight the field of view far away from the vehicle. By dynamically adjusting the Y-axis pitch angle of the surround view 3D view, the blind spot in front of the vehicle can be highlighted, improving blind spots caused by changes in the vehicle body pitch angle, and thereby enhancing driving safety.

[0019] The surround view stitching correction device based on vehicle posture of the present invention includes a vehicle posture perception module, a surround view camera image acquisition module, an image stitching algorithm module, an image display module and a main control processing unit. The vehicle posture perception module is used to collect vehicle posture information and judge the current vehicle body movement posture. The vehicle body posture includes several postures such as stationary, straight, left turn, right turn, left lane change, right lane change, and reversing. The surround view camera image acquisition module is responsible for collecting images from the surround view camera and outputting the collected data to the image stitching algorithm module. The image stitching algorithm module receives the output data of the surround view camera image acquisition module and performs stitching calculation of the vehicle 3D surround view stitching map according to the internal and external parameters of the surround view camera and the adjustment parameters of the main control processing unit. The image display module is responsible for the display output of the vehicle 3D surround view stitching map. The main control processing unit adjusts the image stitching according to the vehicle posture obtained by the vehicle posture perception module. Parameters are used to adjust the position of the stitching seam of the surround view 3D image of the vehicle in a specific motion posture; in the left turn or right turn posture, the position of the stitching seam of the image between the front camera and the left camera or the position of the stitching seam of the image between the front camera and the right camera is deflected to the left rear side or the right rear side of the vehicle surround view mosaic image by the set turning deflection angle; in the left lane change posture or the right lane change posture, the position of the stitching seam of the image between the front camera and the left camera or the position of the stitching seam of the image between the front camera and the right camera is deflected to the left rear side or the right rear side of the vehicle surround view mosaic image by the set lane change deflection angle; in the reversing posture, the position of the stitching seam of the image between the rear camera and the left camera and the position of the stitching seam of the image between the rear camera and the right camera are deflected to the left front side and the right front side of the vehicle surround view mosaic image respectively by the set reversing deflection angle, or the vehicle surround view mosaic image is directly replaced by the image of the reversing camera displayed separately.

[0020] The main control processing unit of the surround view stitching correction device based on vehicle body posture of the present invention adjusts the position of the image stitching seam of the relevant cameras accordingly according to the change of the vehicle body posture. By adjusting the position of the image stitching seam as the vehicle body posture changes, it can reduce or even eliminate the blind spots of vision caused by the misalignment of the stitching seam during the change of the vehicle body posture, thereby improving driving safety. At the same time, this method of changing the position of the image stitching seam has a relatively simple processing operation process and a low cost of use.

[0021] Furthermore, the vehicle posture information includes output signals from an onboard gyroscope and vehicle CAN messages, which include turn signal and turn angle. By combining the output signals from the onboard gyroscope and the vehicle CAN messages, the vehicle posture information can be accurately determined, providing accurate information for adjusting the position of the image stitching seam.

[0022] Furthermore, the turning deflection angle is greater than the lane change deflection angle. This allows the image stitching seam to have a larger deflection angle when the vehicle is turning, which is a larger angle than when changing lanes. This can better avoid blind spots in the field of view caused by misaligned stitching seams during turns, thereby improving driving safety.

[0023] Furthermore, the separately displayed reversing camera image is a distortion-corrected image, which can more accurately and comprehensively display the reversing camera image when reversing, thereby improving the driver's driving experience and driving safety.

[0024] Furthermore, the turning angle information output by the vehicle CAN message is used to correct the turning deflection angle, lane change deflection angle, and reverse deflection angle. Using the turning angle information output by the vehicle CAN message to further correct the turning deflection angle, lane change deflection angle, and reverse deflection angle can make the position of the seam more accurately adjusted, improving the driver's driving experience.

[0025] Furthermore, the vehicle body posture also includes an initial uphill posture and / or an initial downhill posture. In the initial uphill posture, the Y-axis pitch angle of the surround view 3D view is raised to highlight the field of view directly in front of the vehicle. In the initial downhill posture, the Y-axis pitch angle of the surround view 3D view is lowered to highlight the field of view far away from the vehicle. By dynamically adjusting the Y-axis pitch angle of the surround view 3D view, the blind spot in front of the vehicle can be highlighted, improving blind spots caused by changes in the vehicle body pitch angle and thereby enhancing driving safety.

[0026] The vehicle of the present invention includes a surround-view stitching correction device based on vehicle body posture, which includes a vehicle body posture perception module, a surround-view camera image acquisition module, an image stitching algorithm module, an image display module and a main control processing unit. The vehicle body posture perception module is used to collect vehicle body posture information and judge the current vehicle body movement posture. The vehicle body posture includes several postures such as stationary, straight, left turn, right turn, left lane change, right lane change, and reversing; the surround-view camera image acquisition module is responsible for collecting images from the surround-view camera and outputting the collected data to the image stitching algorithm module; the image stitching algorithm module receives the output data of the surround-view camera image acquisition module, and performs stitching calculation of the vehicle 3D surround-view stitching map according to the internal and external parameters of the surround-view camera and the adjustment parameters of the main control processing unit; the image display module is responsible for displaying and outputting the vehicle 3D surround-view stitching map; the main control processing unit adjusts the image according to the vehicle body posture obtained by the vehicle posture perception module The image stitching parameters are used to adjust the position of the stitching seam of the surround view stitching 3D image of the vehicle in a specific motion posture; in the left turn or right turn posture, the position of the image stitching seam of the front camera and the left camera or the position of the image stitching seam of the front camera and the right camera is deflected to the left rear side or the right rear side of the vehicle surround view stitching image by the set turning deflection angle; in the left lane change or right lane change posture, the position of the image stitching seam of the front camera and the left camera or the position of the image stitching seam of the front camera and the right camera is deflected to the left rear side or the right rear side of the vehicle surround view stitching image by the set lane change deflection angle; in the reversing posture, the position of the image stitching seam of the rear camera and the left camera and the position of the image stitching seam of the rear camera and the right camera are deflected to the left front side and the right front side of the vehicle surround view stitching image respectively by the set reversing deflection angle, or the vehicle surround view stitching image is directly replaced by the image of the reversing camera displayed separately.

[0027] The vehicle of the present invention improves the surround-view stitching correction device based on the vehicle body posture, so that the main control processing unit of the device adjusts the position of the image stitching seam of the relevant cameras accordingly according to the change of the vehicle body posture. By adjusting the position of the image stitching seam as the vehicle body posture changes, it is possible to reduce or even eliminate the blind spots in the field of view caused by the misalignment of the stitching seam during the change of the vehicle body posture, thereby improving driving safety. At the same time, this method of changing the position of the image stitching seam has a relatively simple processing and calculation process, thereby reducing the cost of the vehicle.

[0028] Furthermore, the vehicle posture information includes output signals from an onboard gyroscope and vehicle CAN messages, which include turn signal and turn angle. By combining the output signals from the onboard gyroscope and the vehicle CAN messages, the vehicle posture information can be accurately determined, providing accurate information for adjusting the position of the image stitching seam.

[0029] Furthermore, the turning deflection angle is greater than the lane change deflection angle. This allows the image stitching seam to have a larger deflection angle when the vehicle is turning, which is a larger angle than when changing lanes. This can better avoid blind spots in the field of view caused by misaligned stitching seams during turns, thereby improving driving safety.

[0030] Furthermore, the separately displayed reversing camera image is a distortion-corrected image, which can more accurately and comprehensively display the reversing camera image when reversing, thereby improving the driver's driving experience and driving safety.

[0031] Furthermore, the turning angle information output by the vehicle CAN message is used to correct the turning deflection angle, lane change deflection angle, and reverse deflection angle. Using the turning angle information output by the vehicle CAN message to further correct the turning deflection angle, lane change deflection angle, and reverse deflection angle can make the position of the seam more accurately adjusted, improving the driver's driving experience.

[0032] Furthermore, the vehicle body posture also includes an initial uphill posture and / or an initial downhill posture. In the initial uphill posture, the Y-axis pitch angle of the surround view 3D view is raised to highlight the field of view directly in front of the vehicle. In the initial downhill posture, the Y-axis pitch angle of the surround view 3D view is lowered to highlight the field of view far away from the vehicle. By dynamically adjusting the Y-axis pitch angle of the surround view 3D view, the blind spot in front of the vehicle can be highlighted, improving blind spots caused by changes in the vehicle body pitch angle and thereby enhancing driving safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A schematic diagram showing the installation position of a camera on a vehicle in embodiment 1 of the present invention;

[0034] Figure 2 This is a schematic diagram showing the connections of various modules of the surround view stitching and correction device based on vehicle body posture in embodiment 1 of the vehicle of the present invention;

[0035] Figure 3 This is a connection block diagram of a surround view stitching correction device based on vehicle body posture in embodiment 1 of the vehicle of the present invention;

[0036] Figure 4 This is a control flow chart of a surround view stitching correction method based on vehicle body posture in embodiment 1 of the vehicle of the present invention;

[0037] Figure 5 Schematic diagram of the seam position adjustment in the surround view stitching correction method based on vehicle body posture.

[0038] In the figure: 1, vehicle body; 11, front camera; 12, rear camera; 13, left camera; 14, right camera. DETAILED DESCRIPTION

[0039] The features and performance of the present invention are further described in detail below with reference to the embodiments.

[0040] Embodiment 1 of the vehicle of the present invention:

[0041] This embodiment is mainly described by taking a commercial vehicle as an example. Of course, in other embodiments, a passenger vehicle may also be used.

[0042] like Figure 1-5 As shown, specifically, the commercial vehicle of this embodiment is equipped with a 360° surround view host, and the surround view host is equipped with a surround view stitching correction device based on the vehicle body posture, which includes a vehicle body posture perception module, a surround view camera image acquisition module, an image stitching algorithm module, an image display module and a main control processing unit.

[0043] Among them, the vehicle body posture perception module is used to collect vehicle body posture information and judge the current vehicle body motion posture, specifically including the gyroscope configured on the vehicle body and the output signal of the vehicle CAN message. The vehicle body posture includes several postures such as stationary, straight, left turn, right turn, left lane change, right lane change, and reversing; the front camera 11, rear camera 12, left camera 13 and right camera 14 are arranged on the front, rear, left and right directions respectively on the vehicle body 1. The surround view camera image acquisition module is mainly responsible for acquiring the image of the surround view camera and outputting the acquired data to the image stitching algorithm module; the image stitching algorithm The module receives the output data of the surround-view camera image acquisition module, and performs the stitching calculation of the vehicle's 3D surround-view mosaic image based on the internal and external parameters of the surround-view camera and the adjustment parameters of the main control processing unit; the image display module is responsible for the display and output of the vehicle's 3D surround-view mosaic image and fault information; the main control processing unit adjusts the image stitching parameters and the position of the stitching seams of the surround-view mosaic 3D image under specific motion postures based on the vehicle body posture obtained by the vehicle body posture perception module. It is also responsible for detecting the fault status of each module and outputting specific information to the image display module for output display according to the definition of the fault mode.

[0044] More critically, the main control processing unit adopts the following surround view stitching correction method based on the vehicle body posture to adjust the position of the surround view stitching seam relative to the vehicle body in real time, thereby changing the position of the surround view stitching seam blind spot and reducing driving safety hazards.

[0045] The method specifically comprises the following steps:

[0046] S1, the main control processing unit determines the valid status of the surround view video data by detecting the status parameters of the surround view camera image acquisition module. If the surround view video data status is normal, indicating that the camera image data can be received, then execute S2; otherwise, execute S5;

[0047] S2. The vehicle body posture perception module calculates the current posture of the vehicle body through the output information of the gyroscope and the message information on the vehicle CAN bus (optional signals, such as turn signal, turning angle, etc.), including static posture, straight posture, left turn posture, right turn posture, left lane change posture, right lane change posture, reversing posture, etc.

[0048] S3. The main control processing unit calculates the stitching algorithm adjustment parameters required by the image stitching algorithm module based on the vehicle posture information output by the vehicle posture perception module:

[0049] In static and straight-ahead postures, there is no need to adjust the stitching algorithm. The image stitching algorithm module outputs a normal surround stitching image according to the calibration parameters of the camera's intrinsic and extrinsic parameters.

[0050] In the left-turn posture, the main control processing unit adjusts the stitching algorithm so that the position of the image stitching seam between the front camera and the left camera is deflected to the left rear side of the vehicle surround view stitching image by R L1 Degrees, and preferably, the turning angle information output by the vehicle CAN message is combined with the R L1 That is, the value of the turning deflection angle is corrected;

[0051] In the right-turn posture, the main control processing unit adjusts the stitching algorithm so that the position of the image stitching seam between the front camera and the right camera is deflected to the right rear side of the vehicle surround view stitching image by R R1 Degrees, and preferably, the turning angle information output by the vehicle CAN message is combined with the R R1 That is, the value of the turning deflection angle is corrected;

[0052] In the left lane change posture, the main control processing unit adjusts the stitching algorithm so that the position of the image stitching seam between the front camera and the left camera is deflected to the left rear side of the vehicle surround view stitching image by R L2 Degrees, and preferably, the turning angle information output by the vehicle CAN message is combined with the R L2 That is, the value of the lane change deflection angle is corrected;

[0053] Among them, R L2 <R L1 ;

[0054] In the right lane change posture, the main control processing unit adjusts the stitching algorithm so that the position of the image stitching seam between the front camera and the right camera is deflected to the right rear side of the vehicle surround view stitching image by R R2 , and preferably, the turning angle information output by the vehicle CAN message is combined with the R R2 That is, the value of the lane change deflection angle is corrected;

[0055] Among them, R R2 <RR1 ;

[0056] In the reverse posture, the main control processing unit adjusts the stitching algorithm so that the position of the image stitching seam between the rear camera and the left camera turns to the left front side of the vehicle surround view stitching image. L3 degrees, so that the position of the image stitching seam between the rear camera and the right camera is turned to the right front side of the vehicle surround view stitching image. R3 Degrees, and preferably, the turning angle information output by the vehicle CAN message is combined with the R L3 、R R3 That is, the value of the reverse deflection angle is corrected.

[0057] S4. The image stitching algorithm module determines the working status of the surround view camera image acquisition module according to the status indication information of the main control processing unit. When the working status is normal, the image stitching algorithm module performs stitching calculations of the surround view stitching map according to the adjustment parameters output by the main control processing unit in step S3 and the internal and external parameters of the camera, and outputs the stitched image to the image display module.

[0058] S5. The image display module receives the image display data from the image stitching algorithm module or the fault display information from the main control processing unit, and drives the hardware to perform display output.

[0059] Taking into account that commercial vehicles also have other postures during actual driving, such as the initial uphill posture (part of the vehicle body is on the uphill slope, part of the vehicle body is on the plane, and the vehicle body and the uphill slope are at a certain angle) and the initial downhill posture (part of the vehicle body is on the downhill slope, part of the vehicle body is on the plane, and the vehicle body and the downhill slope are at a certain angle), it is more preferred but not necessary. In this embodiment, in S2, the vehicle body posture perception module calculates the current posture of the vehicle body through the output information of the gyroscope and the message information on the vehicle CAN bus, which also includes the initial uphill posture and the initial downhill posture. Correspondingly, in S3, when the main control processing unit calculates the stitching algorithm adjustment parameters required by the image stitching algorithm module based on the vehicle body posture information output by the vehicle body posture perception module, in the initial uphill posture, the main control processing unit raises the pitch angle Y in the Y-axis direction of the surround view 3D view. e degrees, so that the field of view directly in front of the vehicle is highlighted; in the initial downhill posture, the main control processing unit reduces the pitch angle Y of the surround 3D view in the Y-axis direction e1 By dynamically adjusting the pitch angle of the surround 3D view's Y axis, the blind spot in front of the vehicle is highlighted, improving the blind spot problem caused by changes in the vehicle's pitch angle.

[0060] As can be seen from the above description, the commercial vehicle of this embodiment, with the aid of the vehicle posture-based surround view stitching correction device described above, can calculate the vehicle posture information based on the output signal of the gyroscope in the vehicle posture perception module and the signal output by the vehicle CAN message. It can dynamically adjust the parameters of the image stitching algorithm module in real time when the vehicle turns left, turns right, changes lanes left, changes lanes right, reverses, and performs initial uphill and initial downhill movements, thereby changing the position of the camera image stitching seams of the vehicle surround view stitching image. This can reduce the blind spots caused by the misalignment of the stitching seams under different vehicle postures and improve driving safety. Moreover, the surround view stitching correction method adopted by this surround view stitching correction device has a simple computational processing process, generates images quickly through surround view stitching correction, has low hardware requirements, and reduces vehicle costs.

[0061] Of course, the commercial vehicle embodiment described above is only a preferred embodiment of the present invention, and other modified embodiments based on the design concept of the present invention are also provided below.

[0062] For example, in other embodiments, unlike the method described above in Example 1, in which the main control processing unit deflects the positions of the image stitching seams between the rear camera and the left camera, and the image stitching seams between the rear camera and the right camera, respectively, toward the left and right front sides of the vehicle surround view mosaic image in the reverse posture during S3 of the vehicle body posture-based surround view stitching correction method, the surround view mosaic image is directly replaced with a separately displayed image from the reverse camera, i.e., the rear camera image is separately displayed for the driver to observe in the reverse posture. Based on this, in other embodiments, it is more preferred to replace the surround view mosaic image with a separately displayed image from the reverse camera after distortion correction. The distortion correction method is conventional and will not be described in detail herein.

[0063] For example, in other embodiments, the difference from the embodiment 1 introduced above in which the main control processing unit corrects the turning deflection angle, lane change deflection angle and reversing deflection angle through the turning angle information output by the vehicle CAN message in S3 of executing the surround view stitching correction method based on the vehicle body posture is that when the accuracy of the turning deflection angle, lane change deflection angle and reversing deflection angle meets the requirements, the above deflection angles are no longer corrected through the turning angle information output by the vehicle CAN message.

[0064] For example, in other embodiments, unlike the above-described embodiment 1 in which the vehicle body posture information is determined by using the information output by the gyroscope and the vehicle CAN message, a posture sensor including motion sensors such as a three-axis gyroscope, a three-axis accelerometer, and a three-axis electronic compass is configured on the vehicle body to determine the vehicle body posture information.

[0065] For example, in other embodiments, unlike the setting method in which the main control processing unit sets the turning deflection angle to be greater than the lane change deflection angle in S3 of the surround view stitching correction method based on the vehicle body posture in the first embodiment introduced above, the turning deflection angle can be equal to the lane change deflection angle.

[0066] An embodiment of a surround view stitching and correction device based on vehicle body posture of the present invention: its specific structure is consistent with the surround view stitching and correction device introduced in the above vehicle embodiment, and will not be repeated here.

[0067] An embodiment of a surround view stitching correction method based on vehicle body posture of the present invention is consistent with the surround view stitching correction method performed by the surround view stitching correction device based on vehicle body posture in the above vehicle embodiment, and will not be repeated herein.

[0068] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. The scope of patent protection of the present invention shall be based on the claims. Any equivalent structural changes made using the description and drawings of the present invention shall be included in the scope of protection of the present invention.

Claims

1. A surround view stitching correction method based on vehicle body posture, characterized in that: The method comprises the following steps: a) Collect vehicle posture information and determine the vehicle posture, including stationary, straight, left turn, right turn, left lane change, right lane change, and reversing; b) receiving the output data of the surround-view camera image acquisition module in the parking or straight-ahead posture, and stitching a surround-view mosaic image of the vehicle in normal posture based on the intrinsic and extrinsic parameters of the surround-view camera and the set stitching seam parameters; c) In a left-turn or right-turn posture, the position of the image stitching seam between the front camera and the left camera, or the position of the image stitching seam between the front camera and the right camera, is deflected toward the left rear side or the right rear side of the vehicle surround view mosaic image by the set turning deflection angle; d) When changing lanes left or right, the position of the stitching seam between the front camera and the left camera, or the position of the stitching seam between the front camera and the right camera, is deflected toward the left rear side or the right rear side of the vehicle surround view mosaic by the set lane change deflection angle; e) In the reversing posture, the position of the image stitching seam between the rear camera and the left camera, and the position of the image stitching seam between the rear camera and the right camera are deflected to the left front side and the right front side of the vehicle surround view mosaic by the set reversing deflection angle respectively, or the vehicle surround view mosaic is directly replaced by the image of the reversing camera displayed separately.

2. The surround stitching correction method according to claim 1, characterized in that: The vehicle body posture information includes output signals of the vehicle-mounted gyroscope and the vehicle CAN message, and the output signals of the vehicle CAN message include turning light signals and turning angles.

3. The surround stitching correction method according to claim 1, characterized in that: The turning deflection angle is greater than the lane changing deflection angle.

4. The surround stitching correction method according to claim 1, characterized in that: The image of the reversing camera displayed separately is the image after distortion correction.

5. The surround stitching correction method according to claim 2, characterized in that: The turning angle information output by the vehicle CAN message is used to correct the turning deflection angle, lane change deflection angle and reverse deflection angle.

6. The surround stitching correction method according to any one of claims 1 to 5, characterized in that: The vehicle body posture also includes an initial uphill posture and / or an initial downhill posture. In the initial uphill posture, the pitch angle of the surround view 3D view in the Y-axis direction is raised so that the field of view area directly in front of the vehicle is highlighted. In the initial downhill posture, the pitch angle of the surround view 3D view in the Y-axis direction is lowered so that the field of view area in the distance of the vehicle is highlighted.

7. A surround view stitching correction device based on vehicle body posture, comprising a vehicle body posture perception module, a surround view camera image acquisition module, an image stitching algorithm module, an image display module and a main control processing unit, characterized in that: The vehicle posture perception module is used to collect vehicle posture information and determine the current vehicle motion posture, including stationary, straight, left turn, right turn, left lane change, right lane change, and reversing. The surround-view camera image acquisition module is responsible for collecting images from the surround-view camera and outputting the collected data to the image stitching algorithm module; The image stitching algorithm module receives the output data of the surround-view camera image acquisition module, and performs stitching calculations of the vehicle's 3D surround-view stitching map based on the internal and external parameters of the surround-view camera and the adjustment parameters of the main control processing unit; the image display module is responsible for the display output of the vehicle's 3D surround-view stitching map; the main control processing unit adjusts the image stitching parameters according to the vehicle body posture obtained by the vehicle body posture perception module, and adjusts the position of the stitching seam of the surround-view stitching 3D map of the vehicle under a specific motion posture; in the left-turn or right-turn posture, the position of the image stitching seam of the front camera and the left camera or the position of the image stitching seam of the front camera and the right camera is correspondingly deflected to the left rear side or the right rear side of the vehicle's surround-view stitching map by the set turning deflection angle; in the left-change lane or right-change lane posture, the position of the image stitching seam of the front camera and the left camera or the position of the image stitching seam of the front camera and the right camera is correspondingly deflected to the left rear side or the right rear side of the vehicle's surround-view stitching map by the set lane change deflection angle; In the reversing posture, the position of the image stitching seam between the rear camera and the left camera, and the position of the image stitching seam between the rear camera and the right camera are deflected to the left front side and the right front side of the vehicle surround view mosaic image by the set reversing deflection angle respectively, or the vehicle surround view mosaic image is directly replaced by the image of the reversing camera displayed separately.

8. The surround view stitching correction device based on vehicle body posture according to claim 7 is characterized in that: The vehicle body posture information includes output signals of the vehicle-mounted gyroscope and the vehicle CAN message, and the output signals of the vehicle CAN message include turning light signals and turning angles.

9. The surround view stitching correction device based on vehicle body posture according to claim 7, characterized in that: The turning deflection angle is greater than the lane changing deflection angle.

10. The surround view stitching correction device based on vehicle body posture according to claim 7, characterized in that: The image of the reversing camera displayed separately is the image after distortion correction.

11. The surround view stitching correction device based on vehicle body posture according to claim 8, characterized in that: The turning angle information output by the vehicle CAN message is used to correct the turning deflection angle, lane change deflection angle and reverse deflection angle.

12. The surround view stitching correction device based on vehicle body posture according to any one of claims 7 to 11, characterized in that: The vehicle body posture also includes an initial uphill posture and / or an initial downhill posture. In the initial uphill posture, the pitch angle of the surround view 3D view in the Y-axis direction is raised so that the field of view area directly in front of the vehicle is highlighted. In the initial downhill posture, the pitch angle of the surround view 3D view in the Y-axis direction is lowered so that the field of view area in the distance of the vehicle is highlighted.

13. A vehicle, characterized in that: The vehicle includes the surround view stitching correction device based on vehicle body posture as described in any one of claims 7 to 12.

Images