Parking space corner point correction method and parking space corner point correction system, vehicle

Abstract

The present invention relates to a parking space corner point correction method and system thereof. The method comprises: an acquisition step, acquiring the parking space corner point of the parking space to be parked; and a correction step, based on the positional relationship between the parking space corner points of the parking space to be parked, performing rectangular correction on the parking space corner point to obtain the corrected parking space to be parked, wherein the acquisition step comprises: a first acquisition sub-step, acquiring the parking space corner point of the parking space to be parked; a judgment sub-step, judging whether the corner point confidence of the parking space corner point meets the first threshold and whether the corner point offset meets the second threshold, and if it is judged that the first threshold and the second threshold are met respectively, jumping to the correction step, otherwise, continuing to the second sub-acquisition step; and a second sub-acquisition step, discarding the parking space corner point detected in the first acquisition sub-step, and using the parking space corner point predicted by the odometer of the corresponding parking space corner point as the parking space corner point of the parking space to be parked. According to the present invention, the parking accuracy can be improved.

Description

Technical Field

[0001] The present invention relates to vehicle control technology, and in particular to a parking space corner point correction method, a parking space corner point correction system, and a vehicle comprising the parking space corner point correction system. Background Art

[0002] Currently, the most widely used parking space detection algorithms all detect the four corner points of a parking space and use these to construct the desired parking space. Within a single frame, as long as the detected corner points are within a certain range of the true value, the detection result is considered correct and free of jitter. However, in a video stream, if the detection results between two adjacent frames deviate significantly, the parking space will become jittery and irregular, potentially resulting in unpredictable deviations in automated parking. Summary of the Invention

[0003] In view of the above problems, the present invention aims to provide a parking space corner point correction method, a parking space corner point correction system and a vehicle including the parking space corner point correction system, which can improve parking accuracy.

[0004] A parking space corner correction method according to one aspect of the present invention is characterized by comprising:

[0005] An acquisition step of acquiring a corner point of the parking space to be parked; and

[0006] The correction step is to perform rectangular correction on the parking space corner points based on the positional relationship between the parking space corner points to obtain a corrected parking space.

[0007] Optionally, the obtaining step includes:

[0008] The first acquisition sub-step is to acquire the corner points of the parking space to be parked;

[0009] a judging sub-step of judging whether the corner point confidence of the parking space corner point satisfies a first threshold and whether the corner point offset of the parking space corner point satisfies a second threshold; if it is judged that the corner point confidence of the parking space corner point satisfies the first threshold and the corner point offset of the parking space corner point satisfies the second threshold, skipping to the correction step; otherwise, continuing to the second obtaining sub-step; and

[0010] In the second acquisition sub-step, the parking space corner point detected in the first acquisition sub-step is discarded, and the parking space corner point predicted by the odometer is used as the parking space corner point of the parking space.

[0011] Optionally, the correction step includes:

[0012] Four sides are formed according to the corner points of the parking space to be parked, and four rectangles are obtained respectively with one of the four sides as the base and the depth as the average length of the side of the side;

[0013] The rectangle with the highest confidence among the four rectangles on the heat map is taken as the corrected parking space.

[0014] Optionally, the confidence of the rectangle on the heat map is obtained by the confidence of the four corner points of the rectangle on the heat map.

[0015] Optionally, the confidence level on the heat map is obtained through a corner prediction model.

[0016] Optionally, the correction step includes:

[0017] Select two sides from the four sides formed by the corner points of the parking space to be parked, and calculate the midlines of the two sides;

[0018] Obtaining a parking space entrance line segment of the parking space according to a positional relationship between the vehicle and a corner point of the parking space to be parked;

[0019] Obtaining the intersection of the center line and the parking space entrance line segment; and

[0020] Based on the position of the intersection point and the positions of the two sides, the corner points of the parking space are corrected to be rectangular by using geometric relationships to obtain a corrected parking space.

[0021] Optionally, when the two sides are parallel, a line parallel to the two sides and equidistant from the two sides is used as the center line, a perpendicular line to the center line is drawn from the intersection point, and the intersection points of the perpendicular line with the two sides are used as the corrected parking space corner points. The corrected parking space is rectangularized based on the corrected parking space corner points to obtain the corrected parking space.

[0022] Optionally, when the two sides are not parallel, the angle bisector of the angle formed by the extension lines of the two sides is used as the center line, a perpendicular line to the center line is drawn from the intersection point, and the intersection points of the perpendicular line with the two sides are used as the corrected parking space corner points. The corrected parking space is rectangularized based on the corrected parking space corner points to obtain the corrected parking space.

[0023] A parking space corner correction system according to one aspect of the present invention is characterized by comprising:

[0024] An acquisition module obtains the corner points of the parking space to be parked; and

[0025] The correction module performs rectangular correction on the parking space corner points based on the positional relationship between the parking space corner points to obtain a corrected parking space.

[0026] Optionally, the acquisition module includes:

[0027] The first acquisition submodule acquires the corner points of the parking space to be parked;

[0028] a judgment submodule, judging whether the corner point confidence of the parking space corner point obtained by the first obtaining submodule meets a first threshold and whether the corner point offset of the parking space corner point meets a second threshold, and outputting a judgment result; and

[0029] The second acquisition submodule is used to obtain the parking space corner point corresponding to the parking space corner point predicted by the odometer;

[0030] The output module provides the parking space corner point output by the first acquisition submodule or the parking space corner point output by the second acquisition submodule to the correction module based on the judgment result of the judgment submodule.

[0031] Optionally, in the correction module, four sides are formed according to the corner points of the parking space to be parked, and four rectangles are obtained respectively with one of the four sides as the reference and the depth as the average of the side lengths of the four sides. The rectangle with the highest confidence on the heat map among the four rectangles is taken as the corrected parking space to be parked.

[0032] Optionally, the confidence of the rectangle on the heat map is obtained by the confidence of the four corner points of the rectangle on the heat map.

[0033] Optionally, the confidence level on the heat map is obtained through a corner prediction model.

[0034] Optionally, in the correction module, two sides are selected from the four sides formed by the parking corner points of the parking space to be parked, and the center lines of the two sides are calculated. According to the positional relationship between the vehicle and the parking corner points of the parking space to be parked, the parking space entrance line segment of the parking space to be parked is obtained, and the intersection of the center line and the parking space entrance line segment is obtained. Based on the position of the intersection and the positions of the two sides, the parking space corner points are rectangularized to obtain a corrected parking space to be parked.

[0035] Optionally, when the two sides are parallel, a line parallel to the two sides and equidistant from the two sides is used as the center line, a perpendicular line to the center line is drawn from the intersection point, and the intersection points of the perpendicular line with the two sides are used as the corrected parking space corner points. The corrected parking space is rectangularized based on the corrected parking space corner points to obtain the corrected parking space.

[0036] Optionally, when the two sides are not parallel, the angle bisector of the angle formed by the extension lines of the two sides is used as the center line, a perpendicular line to the center line is drawn from the intersection point, and the intersection points of the perpendicular line and the two sides are used as the corrected parking space corner points. The corrected parking space is rectangularized based on the corrected parking space corner points to obtain the corrected parking space.

[0037] A computer-readable medium according to one aspect of the present invention stores a computer program thereon, wherein the computer program implements the parking space corner correction system when executed by a processor.

[0038] A computer device according to one aspect of the present invention includes a storage module, a processor, and a computer program stored on the storage module and executable on the processor, wherein the processor implements the parking space corner correction system when executing the computer program.

[0039] A vehicle according to one aspect of the present invention is characterized in that it includes the parking space corner point correction system according to one aspect of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 1 is a flow chart showing a method for correcting parking space corner points according to an embodiment of the present invention.

[0041] Figure 2 2 is a flow chart showing a method for correcting parking space corner points according to embodiment 1 of the present invention.

[0042] Figure 3 2 is a schematic diagram illustrating the rectangularization correction in the parking space corner point correction method according to the first embodiment of the present invention.

[0043] Figure 4 2 is a schematic diagram showing the correction result of the parking space corner point correction method according to the first embodiment of the present invention.

[0044] Figure 5 2 is a flow chart showing a method for correcting parking space corner points according to a second embodiment of the present invention.

[0045] Figure 6 1 is a schematic diagram illustrating rectangularization correction in two cases where the long sides are parallel and non-parallel in the parking space corner correction method according to embodiment 2 of the present invention.

[0046] Figure 7 2 is a schematic diagram showing an example of a correction result using the parking space corner point correction method according to the second embodiment of the present invention.

[0047] Figure 8 FIG. 1 is a schematic diagram showing another example of correction results using the parking space corner point correction method according to the second embodiment of the present invention.

[0048] Figure 9 2 is a block diagram showing the structure of the parking space corner correction system of the present invention. DETAILED DESCRIPTION

[0049] The following describes some of the various embodiments of the present invention, which are intended to provide a basic understanding of the present invention, but are not intended to identify the key or decisive elements of the present invention or to limit the scope of protection.

[0050] For the purposes of brevity and illustration, the principles of the present invention are described herein primarily with reference to exemplary embodiments thereof. However, those skilled in the art will readily recognize that the same principles are equally applicable to and can be implemented in all types of parking space corner correction methods, parking space corner correction systems, and vehicles incorporating such parking space corner correction systems, and that any such variations do not depart from the true spirit and scope of this patent application.

[0051] The inventors of the present invention have discovered that in the case of vertical, horizontal and small-angle inclined parking spaces, the following problems still exist:

[0052] The parking space detection results in the video are jittery;

[0053] The model detection result is a non-standard rectangle, which increases the difficulty of parking path planning; and

[0054] Fixed an issue where parking in a slightly tilted parking space was misjudged as a vertical / horizontal space, causing the vehicle to park crookedly.

[0055] In order to solve the above problems, the present invention proposes a parking space corner point correction method. The parking space corner point correction method of the present invention comprises the following steps:

[0056] An acquisition step of acquiring a corner point of the parking space to be parked; and

[0057] The correction step is to perform rectangular correction on the parking space corner points based on the positional relationship between the parking space corner points to obtain a corrected parking space.

[0058] Here, the parking space corner points are obtained by detecting the coordinates of the parking space corner points in the image captured by the camera, for example, by using a corner point detection model (based on a deep learning network). Here, preferably, the rationality of the parking space corner point status can be judged.

[0059] A parking space corner point correction method according to one embodiment of the present invention is described below.

[0060] Figure 1 1 is a flow chart showing a method for correcting parking space corner points according to an embodiment of the present invention.

[0061] like Figure 1 As shown, a parking space corner correction method according to one embodiment of the present invention includes:

[0062] Step S100: Detecting the corner points of the parking space to be parked;

[0063] Step S200: Determine the status of the parking space corner point. Specifically, for example, determine whether the corner point confidence of the parking space corner point meets a preset first threshold and whether the corner point offset meets a preset second threshold. If it is determined that the corner point confidence meets the first threshold and the corner point offset meets the second threshold, proceed to step S300; otherwise, proceed to step S400.

[0064] Step S300: retaining the parking space corner points detected in step S100;

[0065] Step S400: discarding the parking space corner point detected in step S100 and using the parking space corner point predicted by the odometer as the parking space corner point of the parking space to be parked;

[0066] Step S500: determining four parking space corner points based on step S300 or step S400; and

[0067] Step S600: Based on the positional relationship between the parking space corner points of the parking space to be parked, the parking space corner points are corrected to be rectangular to obtain a corrected parking space to be parked.

[0068] Here, judging the status of parking space corner points primarily refers to determining whether the corner points detected by the model are reasonable. For example, if the odometer tracking indicates that a corner point should be at the (x, y) position, but the model detection result deviates significantly (by a pre-set threshold), the model-detected corner point is considered inaccurate and the detection result for that corner point is discarded. The judgment criteria can be set as follows: corner point confidence > first threshold and corner point offset < second threshold. The confidence level is output by the detection model, which indicates the possible (x, y) coordinates of a point belonging to a parking space corner point and the likelihood (confidence) of those coordinates belonging to the corner point. The corner point offset refers to the distance between the detected corner point and the corner point predicted based on the previous result.

[0069] The purpose of judging the status of the parking space corner points is to ensure that the four corner points are as accurate as possible. For the parking space corner points at time t, there are three sets of data that can be used, namely (1) the parking space corner points predicted by the odometer from the beginning (the lowest accuracy, but not too far from the target parking space); (2) the parking space corner points detected by the corner point detection model (the highest accuracy can be achieved, but there may be a few corner points that deviate); (3) the parking space corner points predicted by the odometer based on the detection results at time t-1 (i.e. the previous moment) (higher accuracy and relatively stable). The "corner points at the corresponding position tracked by the odometer" mentioned above refers to the parking space corner points predicted at the previous moment in case (3).

[0070] Next, two embodiments of the parking space corner point correction method of the present invention are described.

[0071] Figure 2 2 is a flow chart showing a method for correcting parking space corner points according to embodiment 1 of the present invention.

[0072] The parking space corner point correction method of embodiment 1 of the present invention includes the following steps:

[0073] Step S1: Detecting the corner points of the parking space to be parked;

[0074] Step S2: Determine the status of the parking space corner point. Specifically, for example, determine whether the corner point confidence of the parking space corner point meets a preset first threshold and whether the corner point offset meets a preset second threshold. If it is determined that the corner point confidence meets the preset first threshold and the corner point offset meets the preset second threshold, proceed to step S3; otherwise, proceed to step S4.

[0075] Step S3: retaining the parking space corner points detected in step S1;

[0076] Step S4: discarding the parking space corner point detected in step S1 and using the parking space corner point predicted by the odometer as the parking space corner point of the parking space to be parked;

[0077] Step S5: determining four parking space corner points; and

[0078] Step S6: constructing four sides according to the corner points of the parking space to be parked, taking one of the four sides as a reference and the depth as the average of the side lengths of the side, to obtain four rectangles respectively; and

[0079] Step S7: Take the rectangle with the highest confidence on the heat map among the four rectangles as the corrected parking space (i.e., the predicted parking space), wherein the confidence of the rectangle on the heat map can be obtained through the confidence of the four corner points of the rectangle on the heat map (the confidence on the heat map is obtained through the corner point prediction model).

[0080] Figure 3 2 is a schematic diagram illustrating the rectangularization correction in the parking space corner point correction method according to the first embodiment of the present invention.

[0081] exist Figure 3 In the heat map, from left to right, the top, right, bottom, and left edges are used as the references. A side of the parking space is selected and, with this side as the reference, the depth is the average length of the side edges, forming four rectangles. For each rectangle, the confidence level of its four corner points on the heat map is used to determine the rectangle's confidence level. By comparing the confidence levels of the four rectangles, the largest rectangle is selected as the detection result, achieving rectangularization.

[0082] Figure 4 2 is a schematic diagram showing the correction result of the parking space corner point correction method according to the first embodiment of the present invention.

[0083] The parking space corner point correction method of embodiment 1 of the present invention is mainly aimed at the situation where the points obtained from the heat map lack the shape information of the parking space when there are interference points, and there may be a parking space corner point detection deviation. Figure 4 In the figure, there is a deviation in the upper left corner of the parking space. By using the correction method of Example 1, it can be corrected to the upper left corner of the parking space in the right figure.

[0084] Therefore, as described above, the parking space corner point correction method according to the first embodiment of the present invention can reduce the influence of the detection result deviation and improve the parking accuracy.

[0085] Figure 5 2 is a flow chart showing a method for correcting parking space corner points according to a second embodiment of the present invention.

[0086] The parking space corner point correction method according to embodiment 2 of the present invention includes the following steps:

[0087] Step S11: Detecting the corner points of the parking space to be parked;

[0088] Step S12: Determine the status of the parking space corner point. Specifically, for example, determine whether the corner point confidence of the parking space corner point meets a preset first threshold and whether the corner point offset meets a preset second threshold. If it is determined that the corner point confidence meets the first threshold and the corner point offset meets the second threshold, proceed to step S13; otherwise, proceed to step S14.

[0089] Step S13: retaining the parking space corner points detected in step S11;

[0090] Step S14: discarding the parking space corner point detected in step S11 and using the parking space corner point predicted by the odometer as the parking space corner point of the parking space to be parked;

[0091] Step S15: determining four parking space corner points according to step S13 or step S14; and

[0092] Step S16: Determine whether the long sides of the parking spaces are parallel. If so, proceed to step S17. If not, proceed to step S18.

[0093] Step S17: Obtain the midlines of the two parallel long sides;

[0094] Step S18: Obtain the angle bisector of the angle formed by the two long sides;

[0095] Step S19: Correct the positions of the two corner points of the short side; and

[0096] Step S20: rectangularizing the parking space according to the two corner points. Figure 61 is a schematic diagram illustrating rectangularization correction in two cases where the long sides are parallel and non-parallel in the parking space corner correction method according to embodiment 2 of the present invention.

[0097] In the case where the long sides are not parallel, such as Figure 6 As shown in the left figure, the four corner points are C, B, 1, and 2 in clockwise order, with arrows indicating the vehicle's direction of travel. B1 is the parking space entrance segment calculated based on the relative position of the vehicle and the parking space. Select parking space lines BC and I2 on either side of parking space entrance segment B1, calculate the intersection point A of their extensions, and obtain the angle bisector AD of angle BA1. Point D is the intersection of the angle bisector and parking space entrance segment B1. Draw a perpendicular line 0'1' through point D to AD, where 0' is the intersection of perpendicular line 0'1' and AC, and 1' is the intersection of perpendicular line 0'1' and A2.

[0098] Here, let the coordinates of point A be A(x A ,y A ), the coordinates of point B are B(x B ,y B ), the coordinates of point C are C(x C ,y C ), the coordinates of point D are D(x D ,y D ), the coordinates of the 0' point are 0'(x 0' ,y 0' ), the coordinates of point 1' are 1'(x 1' ,y 1' ), the coordinates of point 0' and point 1' can be calculated based on the coordinates of point A, point B, point C, and point D.

[0099] The following formula represents the coordinates of the 0' point 0' (x 0' ,y 0' ) 0' and y 0' The calculation process is to find the coordinates of point 1'(x 1' ,y 1' ) 1' and y 1' The calculation process is similar to finding the coordinates of point 0'(x 0' ,y 0' ) process.

[0100]

[0101]

[0102] (x D -x A )·x D +(y D -yA )·y D =A1

[0103] (y C -y B )·x B -(x C -x B )·y B =A2

[0104]

[0105] When the long sides are parallel, Figure 6 As shown in the figure on the right, the four corner points are C, B, 1, and 2 in clockwise order, and the arrows indicate the direction of vehicle travel. Among them, B1 is the parking space entrance line segment calculated based on the relative position of the vehicle and the parking space. Select the parking space line BC and parking space line 12 on both sides of the parking space entrance line segment B1. Since the parking space line BC and the parking space line 12 are parallel, draw the midline AD of these two parallel lines. The intersection of the midline AD and the parking space entrance line segment B1 is point D. Draw the perpendicular line 0'1' of AD through point D, where 0' is the intersection of the perpendicular line 0'1' and the parking space line BC, and 1' is the intersection of the perpendicular line 0'1' and the parking space line 12. Similarly, the coordinates of points 0' and 1' can also be calculated based on points A, B, C, and D.

[0106] In this way, after correcting the positions of the two corner points of the short side according to the parallel and non-parallel situations, the parking space can be rectangularized according to the corrected positions of the two corner points to obtain a corrected parking space.

[0107] Figure 7 2 is a schematic diagram showing an example of a correction result using the parking space corner point correction method according to the second embodiment of the present invention.

[0108] Figure 7 The figure shows the situation of a small angle oblique parking space. Here, the so-called "small angle oblique parking space" means that the angle between the vehicle and the parking space entrance is not an obtuse angle. As an example, when the inclination angle of the oblique parking space is less than 75 degrees, the oblique parking space at this time is as follows: Figure 7 As shown in the left figure, the small angle parking space category is misjudged as a vertical or parallel parking space. Directly relying on the effect of rectangularization of the first two corner points will result in errors. After correction by the correction method of Example 2, Figure 7 The correction effect is shown in the figure on the right.

[0109] Figure 8 FIG. 1 is a schematic diagram showing another example of correction results using the parking space corner point correction method according to the second embodiment of the present invention. Figure 8 Indicates the case of corner point misdetection of parking space, such as Figure 8The upper left corner point of the left side of the figure is detected incorrectly, resulting in a parking space error. After correction using the correction method of Example 2, the result is Figure 8 The correction effect is shown in the figure on the right.

[0110] Next, the parking space corner point correction system of the present invention will be described.

[0111] Figure 9 2 is a block diagram showing the structure of the parking space corner correction system of the present invention.

[0112] like Figure 9 As shown, the parking space corner correction system of the present invention is characterized by comprising:

[0113] An acquisition module 100 acquires a corner point of a parking space to be parked; and

[0114] The correction module 200 corrects the parking space corner points into a rectangular shape based on the positional relationship between the parking space corner points to obtain a corrected parking space.

[0115] The acquisition module 100 includes:

[0116] The first acquisition submodule 110 acquires the parking space corner point of the parking space to be parked;

[0117] a judgment submodule 120, which judges whether the corner point confidence of the parking space corner point obtained by the first obtaining submodule meets a first threshold and whether the corner point offset of the parking space corner point meets a second threshold, and outputs a judgment result; and

[0118] The second acquisition submodule 130 is used to acquire the parking space corner point corresponding to the parking space corner point predicted by the odometer;

[0119] The output module 140 provides the parking space corner point output by the first acquisition submodule or the parking space corner point output by the second acquisition submodule to the correction module based on the judgment result of the judgment submodule.

[0120] Optionally, in the correction module 200, four edges are formed based on the corner points of the parking space to be parked. Four rectangles are obtained, each with one of the four edges as a reference and a depth equal to the average length of the side of the edge. The rectangle with the highest confidence on the heat map is selected as the corrected parking space to be parked. The confidence of the rectangle on the heat map is obtained by calculating the confidence of the four corner points of the rectangle on the heat map. The confidence on the heat map is obtained using a corner prediction model.

[0121] Optionally, in the correction module 200, two sides are selected from the four sides formed by the corner points of the parking space to be parked, the midlines of the two sides are calculated, the parking space entrance line segment of the parking space to be parked is obtained based on the positional relationship between the vehicle and the corner points of the parking space to be parked, the intersection point of the midline and the parking space entrance line segment is obtained, and based on the position of the intersection point and the positions of the two sides, the correction module 200 is used to calculate the midlines of the two sides. Figure 6 The corner points of the parking space are rectangularized based on the geometric relationship to obtain the corrected parking space.

[0122] Among them, when the two sides are parallel, a line parallel to the two sides and equidistant from the two sides is used as the center line, a perpendicular line to the center line is drawn from the intersection point, and the intersection points of the perpendicular line with the two sides are used as the corrected parking space corner points. The corrected parking space corner points are used to perform rectangularization to obtain the corrected parking space.

[0123] Among them, when the two sides are not parallel, the angle bisector of the angle formed by the extension lines of the two sides is used as the median line, a perpendicular line to the median line is drawn from the intersection point, and the intersection points of the perpendicular line and the two sides are used as the corrected parking space corner points. The corrected parking space is rectangularized based on the corrected parking space corner points to obtain the corrected parking space.

[0124] The present invention also provides a computer-readable medium having a computer program stored thereon, wherein the computer program implements the above-mentioned parking space corner correction system when executed by a processor.

[0125] The present invention also provides a computer device, comprising a storage module, a processor, and a computer program stored on the storage module and executable on the processor, wherein the processor implements the above-mentioned parking space corner correction system when executing the computer program.

[0126] The present invention also provides a vehicle, characterized in that it comprises the above-mentioned parking space corner point correction system.

[0127] The above examples primarily illustrate the parking space corner correction method, parking space corner correction system, and vehicle incorporating the parking space corner correction system of the present invention. While only certain specific embodiments of the present invention have been described, those skilled in the art will appreciate that the present invention may be embodied in numerous other forms without departing from its spirit and scope. Therefore, the presented examples and embodiments are to be considered illustrative rather than restrictive, and the present invention may encompass various modifications and substitutions without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A parking space corner correction method, characterized in that: include: An acquisition step is to acquire the corner point of the parking space to be parked; as well as The correction step is to correct the parking space corner points into a rectangular shape based on the positional relationship between the parking space corner points to obtain a corrected parking space. Wherein, the correction step comprises: Select two sides from the four sides formed by the corner points of the parking space to be parked, and calculate the median lines of the two sides; According to the positional relationship between the vehicle and the parking space corner point of the parking space to be parked, an entrance line segment of the parking space to be parked is obtained; Obtaining the intersection of the center line and the parking space entrance line segment; and Based on the position of the intersection point and the positions of the two sides, the corner points of the parking space are corrected to be rectangular by using geometric relationships to obtain a corrected parking space. When the two sides are parallel, a line parallel to the two sides and equidistant from the two sides is used as the center line, a perpendicular line to the center line is drawn from the intersection point, and the intersection points of the perpendicular line with the two sides are used as corrected parking space corner points. The corrected parking space is rectangularized based on the corrected parking space corner points to obtain a corrected parking space.

2. The parking space corner correction method according to claim 1, characterized in that: The obtaining step comprises: The first acquisition sub-step is to acquire the parking space corner point of the parking space to be parked; a judging sub-step, judging whether the corner point confidence of the parking space corner point satisfies a first threshold and whether the corner point offset of the parking space corner point satisfies a second threshold, and if it is judged that the corner point confidence of the parking space corner point satisfies the first threshold and the corner point offset of the parking space corner point satisfies the second threshold, jumping to the correction step, otherwise, continuing to the second sub-acquisition step; and In the second sub-acquisition step, the parking space corner point detected in the first acquisition sub-step is discarded, and the parking space corner point predicted by the odometer is used as the parking space corner point of the parking space.

3. The parking space corner correction method according to claim 1, characterized in that: When the two sides are not parallel, the angle bisector of the angle formed by the extension lines of the two sides is used as the median line, a perpendicular line to the median line is drawn from the intersection point, and the intersection points of the perpendicular line with the two sides are used as corrected parking space corner points. The corrected parking space is rectangularized based on the corrected parking space corner points to obtain a corrected parking space.

4. A parking space corner correction system, characterized in that: include: An acquisition module is used to acquire the corner points of the parking space to be parked; as well as The correction module corrects the parking space corner points into a rectangular shape based on the positional relationship between the parking space corner points to obtain the corrected parking space. In the correction module, two sides are selected from the four sides formed by the parking space corner points of the parking space to be parked, and the midlines of the two sides are calculated. According to the positional relationship between the vehicle and the parking space corner points of the parking space to be parked, the parking space entrance line segment of the parking space to be parked is obtained, and the intersection point of the midline and the parking space entrance line segment is obtained. Based on the position of the intersection point and the positions of the two sides, the parking space corner points are corrected to be rectangularized to obtain a corrected parking space to be parked. When the two sides are parallel, a line parallel to the two sides and equidistant from the two sides is used as the center line, a perpendicular line to the center line is drawn from the intersection point, and the intersection points of the perpendicular line with the two sides are used as the corrected parking space corner points. The corrected parking space is rectangularized based on the corrected parking space corner points to obtain the corrected parking space.

5. The parking space corner correction system according to claim 4, characterized in that: The acquisition module comprises: The first acquisition submodule is used to acquire the parking space corner point of the parking space to be parked; a judgment submodule, configured to judge whether the corner point confidence of the parking space corner point obtained by the first obtaining submodule satisfies a first threshold and whether the corner point offset of the parking space corner point satisfies a second threshold and output a judgment result; and The second sub-acquisition module is used to acquire the parking space corner point corresponding to the parking space corner point predicted by the odometer; An output module is used to provide the parking space corner point output by the first acquisition submodule or the parking space corner point output by the second acquisition submodule to the correction module based on the judgment result from the judgment submodule.

6. The parking space corner correction system according to claim 4, characterized in that: When the two sides are not parallel, the angle bisector of the angle formed by the extension lines of the two sides is used as the median line, a perpendicular line to the median line is drawn from the intersection point, and the intersection points of the perpendicular line with the two sides are used as corrected parking space corner points. The corrected parking space is rectangularized based on the corrected parking space corner points to obtain a corrected parking space.

7. A computer readable medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the parking space corner point correction method according to any one of claims 1 to 3 is implemented.

8. A computer device comprising a storage module, a processor, and a computer program stored in the storage module and executable on the processor, characterized in that: When the processor executes the computer program, the parking space corner point correction method according to any one of claims 1 to 3 is implemented.

9. A vehicle, characterized in that: It comprises the parking space corner point correction system as described in any one of claims 4 to 6.

Images