Automatic control method for opening angle of vehicle door

Abstract

The invention provides a method for automatically controlling the opening angle of a vehicle door, which comprises the following steps of: detecting by combining a freespace technology arranged on a vehicle and a panoramic image system technology so as to judge the height of an obstacle near the vehicle door, and forming a first height between the vehicle door and the ground; the vehicle detects that an obstacle exists near the vehicle door, and when the height of the obstacle is larger than or equal to the first height, the vehicle limits the angle when the vehicle door is opened according to the height of the obstacle; and when the height of the obstacle is smaller than the first height, the vehicle does not limit the vehicle door opening angle.

Description

TECHNICAL FIELD

[0001] The present application relates to a kind of car door opening angle automatic control method. BACKGROUND

[0002] With the rapid development of automatic driving and intelligent driving assistance system (ADAS), the safety and intelligentization of vehicle are more and more valued by users, especially the opening of car door, due to the complexity of real road conditions, car door needs to adjust the angle in the opening process for different environment, the conventional method is through the active observation of the personnel in the vehicle, but this will appear many visual blind area, so that the car door is inconvenient to open.

[0003] The existing technology is the reflection mechanism of the electromagnetic wave signal of ultrasonic sensor, the transmitter of sensor will emit detection signal outward, these detection signals will be reflected when encountering object, the echo signal reflected back will be captured by the receiver of radar, by analyzing the characteristics of echo signal received by the receiver of radar, the outside obstacle condition is obtained.

[0004] The disadvantage of such car door obstacle avoidance technology is that when the surface of obstacle is too large relative to the incident angle of sensor, radar wave can be reflected at a large angle, instead of returning directly to sensor receiver. In this case, sensor can not receive enough echo signal to identify the existence of obstacle.

[0005] In addition, the road edge scene below the lower edge of car door can open the car door, and the car door obstacle avoidance technology based on sensor will misjudge this scene as not opening the car door. In short, the opening door obstacle avoidance technology based on radar will judge the environment that does not meet the opening of car door as opening the car door, and judge the environment that can open the car door as not opening the car door.

[0006] Therefore, the use of sensor is limited, and many common scenes cannot be accurately judged, affecting the user experience. SUMMARY

[0007] The technical problem to be solved by the present application is to provide a car door opening angle automatic control method, which adopts freespace technology and panoramic image system technology to detect and control the automatic opening angle of car door, combines the technology of panoramic image system and automatic driving field, and determines the angle of car door that can be automatically opened when parking by judging the height of surrounding obstacles.

[0008] To solve the above technical problems, the technical scheme of the present application is a car door opening angle automatic control method, and the operation process is as follows: A first step, providing a vehicle with at least one camera installed thereon to obtain a surrounding image of the vehicle, the surrounding image comprising a plurality of frames of images with a time sequence, the camera being installed on the vehicle at a height H, wherein the vehicle has at least one door, a first height between a bottom end of the door and a ground surface, the first height being 35 cm, and a projection point of the camera corresponding to the ground surface; A second step, extracting two frames of images from the images and identifying whether each frame has an obstacle, and continuously identifying if no obstacle is identified, and entering a next step if the obstacle is identified. A third step, using a free space to obtain a contour line of a bottom end point of the obstacle, obtaining the bottom end point of the obstacle from the images, obtaining a corresponding top end point from the bottom end point in a vertical direction, a height h of the obstacle between the bottom end point and the top end point, wherein the images can be bird's-eye view images, and using parallax information of the two frames of images and relative positions of the camera to perform triangulation, wherein a distance D between the camera and the top end point, and a distance Δy between the bottom end point and the projection point, and calculating the height h of the obstacle, the height h of the obstacle being calculated by h = D - Δy. ; A fourth step, determining an angle of opening the door according to the height h of the obstacle, and limiting the angle of opening the door according to the height h of the obstacle when the height h of the obstacle is greater than or equal to the first height, and not limiting the angle of opening the door when the height h of the obstacle is less than the first height.

[0009] Further, in the first step, the two frames of images are an image when parking is completed and an image when a longitudinal displacement of 20 cm before parking is completed.

[0010] Further, in the first step, a lookup table of a range of movement of each door is constructed, a sector region is defined based on a hinge point of each door and the vehicle, a length of the door, a preset maximum opening angle of the door, and a preset minimum opening angle of the door, and the sector region is used as an effective range of movement of the door, wherein the preset maximum opening angle of the door is 70°, and the preset minimum opening angle of the door is 15°.

[0011] Further, in the third step, a pixel coordinate of the bottom end of the obstacle is (X p , Y p ), and a pixel coordinate of the camera on the ground surface is (X o , Yo ), the scaling ratio of the image is K, and the image is a bird's eye view, then the calculation formula of the distance △y is .

[0012] Compared with the prior art, the vehicle door opening angle automatic control method provided by the application can store the data of the opening of the vehicle door in two dimensions through panoramic image technology, analyze and pre-calculate the surrounding environment of the vehicle, and store the two-dimensional coordinates of the opening angle of the vehicle door obtained by pre-calculation, so that the position of the vehicle door can be directly corresponded when the vehicle door is opened, and the real-time echo of sensors is no longer needed, the visual blind area and the sensor blind area are solved, and the accuracy of judgment and the convenience of use are improved. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figure 1 The operation flowchart of the application is shown in the figure.

[0014] Figure 2 The coordinate system diagram of the application is shown in the figure.

[0015] Figure 3 The relationship diagram between the feature matching point pairs and the vertical point pairs between adjacent frames of the bird's eye view is shown in the figure.

[0016] Figure 4 The obstacle height calculation diagram of the application is shown in the figure. DETAILED DESCRIPTION

[0017] In one embodiment, as shown in Figure 1 , a vehicle door opening angle automatic control method comprises the following steps: A first step, providing a vehicle equipped with at least one camera to obtain a surrounding image of the vehicle, the surrounding image comprising a plurality of frames of images with a time sequence, the camera being installed on the vehicle at a height H, wherein the vehicle has at least one door, the distance between the bottom end of the door and the ground has a first height, and the camera has a projection point corresponding to the ground; In addition, in the first step, a vehicle door lookup table is also established, comprising the following steps: S1. Define a global coordinate system, the origin of the coordinate system being set at the position of the ground directly below the center of the rear axle of the vehicle; the positive direction being directed to the front of the vehicle, representing the longitudinal distance X axis; the positive direction being directed to the left side of the vehicle, representing the lateral distance Y axis; and the positive direction of the Z axis being vertically upward.

[0018] S2. Define a vehicle door local coordinate system, the origin of the coordinate system being set at each door hinge, the X axis being directed to the front of the vehicle, and the Y axis being directed to the left side of the vehicle. S3. Set the sector of the door, where the door hinge point (x0, y0) is the center, the door length is the radius, the central angle is 70° (representing the maximum angle of the door from closed to fully open), and the double door length is the side length of the rectangular region with the radius of the coordinate system. Traverse the rectangular region with the hinge (x0, y0) as the center and 2R as the side length. Calculate the distance of each point in the rectangle to the center distance and the rotation angle angle in the current coordinate system. The specific calculation method is: dx = x - x0, dy = y - y0, distance = sqrt(dx2 + dy2), angle = atan2(dy, dx). Define the points with distance <= R and angle satisfying 0° <= angle <= 70° as the effective range of the door movement, and the corresponding value of the points in the effective range is the opening angle, while the remaining part of the rectangle is defined as the invalid range.

[0019] The second step is to capture two frames of images in the image and identify whether there is an obstacle respectively, and if the obstacle is not identified, the identification continues; when the obstacle is identified, the next step is entered; The third step is to use freespace to obtain the contour line of the bottom end point of the obstacle. The bottom end point of the obstacle is obtained from the image, and a corresponding top end point is obtained along the vertical direction from the bottom end point. The distance from the bottom end point to the top end point has an obstacle height h. Triangulation is performed using the parallax information of the two frames of images and the relative position of the camera, wherein the camera has a distance D to the top end point, and the bottom end point has a distance Ay to the projection point. When calculating the obstacle height h, ; The fourth step is to determine the angle of opening the door according to the height of the obstacle. When the height of the obstacle is greater than or equal to the first height, the vehicle limits the angle of opening the door according to the height of the obstacle; when the height of the obstacle is less than the first height, the vehicle does not limit the angle of opening the door.

[0020] As Figures 2-4 shown, in one embodiment, in the state of judging the obstacle, the method for calculating the height of the obstacle is as follows: When the obstacle enters the door movement range, two frames of images are selected to generate obstacle feature matching point pairs, and the height of the obstacle is calculated. As Figure 3As shown, the image C2 when parking is completed and the historical image C1 when the vehicle still has at least 20 cm longitudinal displacement before parking is completed are selected. The displacement ensures that there is sufficient parallax between the two images to meet the triangulation condition. At the same time, in order to ensure that the ROI regions identified in the two images are the same region in the real world, the ROI region identification range is defined within one door width laterally and from the front door hinge to the rear door tail longitudinally. The ROI region displacement on the obstacle bottom end point Pb is tracked by using the consecutive frame images to lock the obstacle part in the field of view, combined with the camera ROI, and converted into the real world distance (i.e. the camera moving distance). There is a fixed point P' on the obstacle bottom end point, which forms an image point P'11 on the overhead view C1 at T1 time and an image point P'21 on the overhead view C2 at T2 time. If P'11 and P'21 are successfully detected, P'11 and P'21 constitute a pair of feature matching points.

[0021] The freesppace boundary line is used as the obstacle bottom end contour line to extract the obstacle bottom end point Pb from the overhead view, and the obstacle bottom end point Pb in the overhead view is mapped to the original image through the transformation matrix; the corresponding obstacle top end point Pt is found by searching upward along the vertical direction from the obstacle bottom end contour line; after the obstacle top end point Pt is obtained, the Euclidean distance from the camera O to the obstacle top end point Pt is .

[0022] In the coordinate system, the pixel coordinates (xb, yb) of the road along bottom end point Pb, the pixel coordinates (xo, yo) of the projection point O' of the camera O on the ground, and the scaling ratio K of the overhead view are known, and the physical distance from the bottom end point Pb to the projection point O' of the camera O on the ground is obtained as . The actual height H of the camera O is known, and combined with the parameters obtained in the above steps, the obstacle height is obtained, wherein h is the obstacle height, H is the camera height, D is the distance from the obstacle top end point Pt to the camera O, and Δy is the distance between the obstacle bottom end point Pb and the projection point O' of the camera O.

[0023] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting technical solutions. Those of ordinary skill in the art should understand that modifications or equivalent replacements to the technical solutions of the present application without departing from the purpose and scope of the technical solutions should be covered in the scope of the claims of the present application.

Claims

1. A method of automatically controlling an opening angle of a vehicle door, characterized by, The steps are as follows: A first step is to provide a vehicle with at least one camera installed to obtain a surrounding image of the vehicle, the surrounding image including a plurality of frames of images with a front-back time sequence, the camera being installed on the vehicle at a height H, wherein the vehicle has at least one door, a distance between a bottom end of the door and the ground being a first height, and the camera has a projection point corresponding to the ground; A second step is to capture two frames of images in the images and identify whether each frame has an obstacle, and if no obstacle is identified, the identification is continued; if an obstacle is identified, a next step is entered; In the third step, the bottom end point of the obstacle is obtained from the image, a corresponding top end point is obtained along the vertical direction from the bottom end point, and the distance between the bottom end point and the top end point is the height h of the obstacle; the parallax information of the two images and the relative position of the camera are used to perform triangulation, wherein the camera has a distance D to the top end point, and the bottom end point has a distance △y to the projection point; when the height h of the obstacle is calculated, the height of the obstacle is . A fourth step is to determine an opening angle of the door based on a height of the obstacle, and when the height of the obstacle is greater than or equal to the first height, the vehicle limits the opening angle of the door based on the height of the obstacle; and when the height of the obstacle is less than the first height, the vehicle does not limit the opening angle of the door.

2. The method of claim 1, wherein In the first step, the two frames of images are an image when parking is completed and an image when a longitudinal displacement of 20 cm before parking is completed.

3. The method of claim 1, wherein In the first step, a lookup table of a range of movement of each door is also constructed, a sector region is defined based on a hinge point of each door and the vehicle, a length of the door, a preset maximum opening angle of the door, and a preset minimum opening angle of the door, and the sector region is used as an effective range of movement of the door.

4. The method of claim 3, wherein The preset maximum opening angle of the door is 70°.

5. The method of claim 3, wherein The preset minimum opening angle of the door is 15°.

6. The method of claim 1, wherein In the third step, the pixel coordinate of the bottom end of the obstacle is (X p , Y p ), the pixel coordinate of the camera on the ground is (X o , Y o ), the scaling ratio of the image is K, and the image is a bird's eye view, then the calculation formula of the distance △y is .

7. The method of claim 1, wherein In the first step, the first height is 35 cm.

Images