A parking detection method and system combining image algorithms and physical devices

Abstract

The application discloses a parking detection method combining an image algorithm and a physical device, which acquires boundary information of a parking area through a camera bound with a vehicle body, obtains edge information of the parking area through image processing and analysis, calculates a distance between an actual parked vehicle and a specified parking line boundary and an included angle information between the vehicle body and a normal line of the parking line according to a relative relationship between the edge information of the parking area when standard parking and the edge information of the parking area when actual parking, and further guides the vehicle to be placed in a parking area in a standard manner; and the application also locks a non-motor vehicle parking position through the physical device, so that the parked vehicle is prevented from being moved by external factors. The application can guide parking and limit vehicle movement in the parking area, so that the vehicle can be placed in a neat and uniform shape, and is favorable for management of a road surface and maintenance of a city appearance.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of non-motor vehicles, in particular to a parking method and system combining image algorithm and physical device. BACKGROUND

[0002] With the popularity of shared non-motor vehicles, the problem of non-motor vehicle lane congestion and pedestrian traffic difficulty caused by random parking of vehicles is common. In order to ensure smooth traffic, save parking land, and create a neat parking appearance, the common practice is for non-motor vehicle operators to arrange vehicles in an interval mode, but this approach requires high labor costs.

[0003] Currently, non-motor vehicles generally adopt a fixed-point parking mode, which uses satellite positioning technology and electronic fence technology to simply constrain the vehicle users to park in the designated area. However, the vehicles in the designated area cannot ensure that their placement position and angle are neat and uniform, for example, the parking direction is random, the vehicle placement position is not neat, and even the vehicle part structure is located outside the designated area. SUMMARY

[0004] Therefore, the present application provides a parking detection method and system combining image algorithm and physical device, which at least partially solves the problems in the prior art.

[0005] In a first aspect, the present application provides a parking detection method combining image algorithm and physical device, comprising the following steps:

[0006] Step one: pre-process the collected original image data, and the processing process includes filtering and denoising;

[0007] Step two: take the brightness information of the boundary line of the parking area as the threshold condition for image binarization processing, and obtain a simple edge contour;

[0008] Step three: count the brightness information on the left and right sides of the original edge contour line, further confirm the physical mapping relationship between the image edge information and the actual boundary line, and determine whether the current contour line belongs to the left edge or the right edge of the reference line according to the statistical result.

[0009] Step four: connect the discontinuous line segments by means of morphological operations such as dilation and erosion to obtain a clear edge contour, and analyze the figure formed by the two contours and the upper and lower boundaries of the picture according to the edge contour line and the edge contour line of the known standard parking.

[0010] Step five: determine whether the standard parking edge contour line is parallel to the actual parking edge contour line, and based on the determination result, calculate the distance between the actual parked vehicle and the boundary line of the designated parking area and the included angle information between the vehicle body and the parking boundary normal line according to the area and contour information.

[0011] Step six: lock the non-motor vehicle foot prop with the ground buckle locking device, and limit the movement of the parked vehicle by locking.

[0012] Further, the rule in step three is to calculate the pixels of a single image;

[0013] Further, step three also includes forming a statistical conversion table by calculating the pixel difference in the image and the actual distance of the vehicle wheel to the stop line;

[0014] Further, in step four, the relationship between the actual parking edge contour line and the known standard parking edge contour line can be divided into: the actual parking edge contour line is parallel to the left side of the standard parking edge contour line, the actual parking edge contour line is parallel to the right side of the standard parking edge contour line, the vehicle body is left deflected, and the vehicle body is right deflected and inclined;

[0015] Further, in step five, the method for calculating the distance of the actual parked vehicle from the specified parking area boundary line is:

[0016] When the standard parking edge contour line is parallel to the actual parking edge contour line, the distance L is calculated as follows:

[0017]

[0018] Where S is the area formed by the actual parking edge contour line and the known standard parking edge contour line and the upper and lower boundary lines of the image; L0 is the length of the standard parking edge contour line. Then, through the left and right relative positions of the solid line A and the dashed line B, it is determined whether it is an out-of-boundary state or an insufficient state.

[0019] When the standard parking edge contour line is not parallel to the actual parking edge contour line, the distance L is calculated as follows:

[0020] L = (S1 + S2) / L0

[0021] Where S1 and S2 are the areas formed by the actual parking edge contour line on the upper and lower sides of the intersection point and the known standard parking edge contour line; L0 is the length of the standard parking edge contour line.

[0022] Further, in step five, the method for calculating the angle information of the actual parked vehicle body and the parking area boundary normal line is as follows:

[0023] For when the standard parking edge contour line is not parallel to the actual parking edge contour line, the angle information of the parked vehicle body is obtained by the following calculation method:

[0024] tan θ = ab / ao

[0025] Wherein, θ is the angle between the standard parking edge profile line and the actual parking edge profile line, which is equivalent to the angle information between the actual parked vehicle body and the parking area boundary normal line, ab is the distance between the upper end of the standard parking edge profile line and the actual parking edge profile line, and ao is the distance from the upper end of the standard parking edge profile line to the intersection O.

[0026] Further, the buckle locking device in step six includes a circular bottom plate, a rectangular buckle is arranged on the left and right sides of the upper surface of the circular bottom plate respectively, an arc-shaped protruding strip is arranged on the buckle, and the arc-shaped protruding strips on the left and right sides are used to fix the foot support; and a plurality of circular through holes are arranged on the circular bottom plate.

[0027] In the second aspect, the disclosure provides a parking system combining an image algorithm and a physical device, which comprises:

[0028] The image acquisition module adopts an infrared fill light and an infrared camera to acquire road surface data, and is fixed on the vehicle body directly above the wheel.

[0029] The image processing module is used for analyzing and calculating the collected images, including pre-processing the original image data;

[0030] The brightness information of the parking area boundary line obtained through large-scale statistics is used as a threshold condition for image binarization processing to obtain a simple edge profile.

[0031] The brightness information of the left and right sides of the edge of the original image is counted, and the counting rule is to calculate the pixels of a single frame of image.

[0032] The discontinuous line segments are connected by means of morphological operations such as dilation and erosion to obtain a clear edge profile.

[0033] The distance and angle calculation module calculates the pixel distance of the actual parked vehicle from the specified parking area boundary line and the angle information between the vehicle body and the parking area boundary normal line according to the edge profile line, the edge profile line of the known standard parking, and the upper and lower boundaries of the image picture, and obtains the real vehicle distance to the parking line by using the pixel distance of the picture and the conversion relationship between the pixel and the real world distance.

[0034] The locking module includes a buckle locking device for locking the foot support of a non-motor vehicle, the buckle locking device includes a circular bottom plate, a rectangular buckle is arranged on the left and right sides of the upper surface of the circular bottom plate respectively, an arc-shaped protruding strip is arranged on the buckle, and the arc-shaped protruding strips on the left and right sides are used to fix the foot support; and a plurality of circular through holes are arranged on the circular bottom plate.

[0035] Further, the image processing module further includes filtering and denoising the image.

[0036] Further, the distance and angle calculation module further comprises a judging criterion that whether the standard parking edge contour line is parallel to the actual parking edge contour line. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present disclosure, and all other drawings obtained by those of ordinary skill in the art based on these drawings without creative labor are within the protection scope of the present disclosure.

[0038] Figure 1 A schematic diagram of an image recognition algorithm provided by the embodiments of the present disclosure is shown in the following figure.

[0039] Figure 2 The standard parking edge contour line and the actual parking edge contour line in four different scenarios provided by the embodiments of the present disclosure are shown in the following figures.

[0040] Figure 3 (a) A perspective structural schematic diagram of the buckle locking device provided by the embodiments of the present disclosure is shown in the following figure.

[0041] Figure 3 (b) A top view of the buckle locking device provided by the embodiments of the present disclosure is shown in the following figure.

[0042] Figure 3 (c) An A-A cross-sectional view of the buckle locking device provided by the embodiments of the present disclosure is shown in the following figure.

[0043] Figure 3 (d) A B-B cross-sectional view of the buckle locking device provided by the embodiments of the present disclosure is shown in the following figure. DETAILED DESCRIPTION

[0044] The embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0045] The embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0046] It is important to note that the various aspects described throughout this disclosure can be implemented in any number of forms, and that the specifics set forth in any particular implementation are merely exemplary. Based on the disclosure herein, one skilled in the art should appreciate that an aspect described herein can be implemented independently of any other aspects and that an aspect described herein can be implemented both as any stand-alone aspect or in combination with other aspects in a multiple aspect implementation.

[0047] It is also important to note that the present disclosure can be carried out in ways not specifically set forth herein without departing from essential characteristics of the present disclosure. Thus, the present disclosure should not be limited to the implementations set forth herein, but should be given the widest scope consistent with the principles and features disclosed herein.

[0048] Furthermore, in the following description, numerous specific details are set forth in order to provide a thorough understanding of the examples. However, it will be apparent to one skilled in the art that the aspects can be practiced without these specific details.

[0049] At present, the constraint condition of non-motor vehicle parking is to park the vehicle in the designated area, but the position and angle information of the vehicle in the parking area cannot be regulated. The main reason is that the vehicle cannot perceive the spatial position and body angle information of the region where it is located. In order to regulate the vehicle placement, the embodiment introduces an image algorithm, compares and analyzes the image of the specified parking area and the road surface information of the known area, and obtains more detailed data information of the vehicle in the parking area, that is, the spatial position coordinates and the body angle data.

[0050] A parking detection method combining image algorithm and physical device, as shown in Figure 1 includes the following steps:

[0051] S1: Preprocessing the original image data, including filtering and denoising, to reduce the influence of the shooting area being disturbed by sand, fallen leaves and other objects;

[0052] Considering that the vehicle parking area is mostly an open area, and the time when the vehicle is used is not limited to daytime, the present application uses an infrared fill light and an infrared camera to obtain road surface data, reduces the influence of the sharp fluctuation of the picture brightness caused by light, and further reduces the technical complexity of image preprocessing. Therefore, after obtaining the infrared image data, the data needs to be preprocessed first.

[0053] S2: Take the brightness information of the parking area boundary line after large-scale statistics as the threshold condition of image binary processing, and obtain a simple edge contour;

[0054] S3: Statistically analyze the brightness information of the left and right sides of the edge on the original image, and the statistical rule is to calculate the pixels of a single frame image;

[0055] Due to the limitation of the field of view angle of the infrared camera and the installation position of the camera, it is necessary to statistically analyze the brightness information of the left and right sides of the edge on the original image;

[0056] Further confirm the physical mapping relationship between the image edge information and the actual boundary line, form a statistical conversion table by statistically analyzing the pixel difference in the picture and the distance from the actual wheel to the parking line, and confirm whether the current contour line belongs to the left edge or the right edge of the reference line according to the statistical result.

[0057] S4: Connect the discontinuous line segments by means of morphological operations including dilation and erosion to obtain clear edge contours, and analyze the figure formed by the two contours and the upper and lower boundaries of the picture according to the edge contour line and the edge contour line of the known standard parking.

[0058] The figure formed by the possible existing contour and the upper and lower boundaries of the picture is shown in Figure 2 , wherein 1, 2, 3 and 4 respectively represent different states of actual parking, the solid line A in the figure represents the standard parking edge contour line, the dashed line B represents the actual parking edge contour line, S, S1 and S2 represent the envelope area of the standard parking and actual parking contour line. 1 indicates that the current state is insufficient compared with the standard parking state, and needs to continue to move forward; 2 indicates that the current state is out of range compared with the standard parking state, and needs to move backward; 3 indicates that the current parking state has left inclination compared with the standard parking state, and needs to be inclined to the right; 4 indicates that the current parking state has right inclination compared with the standard parking state, and needs to be inclined to the left.

[0059] S5: Determine whether the standard parking edge contour line and the actual parking edge contour line are parallel, and based on the determination result, calculate the pixel distance of the actual parked vehicle from the specified parking area boundary line and the angle information of the vehicle body and the parking boundary normal line according to the area and contour length information, and obtain the real vehicle distance to the parking line distance by using the pixel distance and the pixel to real world distance conversion relationship.

[0060] Specifically, it includes: calculating the pixel-level distance of the actual parked vehicle from the specified parking area boundary line according to the area and contour length information, and obtaining the actual distance of the current vehicle from the standard parking according to the conversion relationship between the pixel distance and the actual physical distance.

[0061] Referring to Figure 2 , the distance calculation method of the actual parked vehicle from the specified parking area boundary line is as follows:

[0062] When the standard parking edge profile line is parallel to the actual parking edge profile line, the distance L is calculated as follows:

[0063]

[0064] Where S is the area formed by the actual parking edge profile line and the known standard parking edge profile line; L0 is the length of the standard parking edge profile line. Then, by judging the relative position of the solid line A and the dashed line B, it is determined whether it is the out-of-bound state or the insufficient state.

[0065] When the standard parking edge profile line is not parallel to the actual parking edge profile line, the distance L is calculated as follows:

[0066] L=(S1+S2) / L0

[0067] Where S1 and S2 are the areas formed by the actual parking edge profile line on the upper and lower sides of the intersection point and the known standard parking edge profile line; L0 is the length of the standard parking edge profile line.

[0068] Since the infrared camera is fixed to the vehicle body, the profile line angle can be mapped and converted with the angle formed by the vehicle body and the normal of the specified parking area boundary. With a large amount of statistical information, the angle information of the electric vehicle body parking can be obtained by the profile line angle.

[0069] For the angle information of the vehicle body parking, when the standard parking edge profile line is parallel to the actual parking edge profile line, the angle of the vehicle body parking can be determined;

[0070] For when the standard parking edge profile line is not parallel to the actual parking edge profile line, the angle information of the vehicle body parking is obtained by the following calculation method:

[0071] tanθ=ab / ao

[0072] Where θ is the angle between the standard parking edge profile line and the actual parking edge profile line, which is equivalent to the angle information between the actual parked vehicle body and the normal of the parking area boundary, ab is the distance between the upper end of the standard parking edge profile line and the actual parking edge profile line, and ao is the distance from the upper end of the standard parking edge profile line to the intersection point O.

[0073] According to the prior parking area boundary information and the profile line distance and angle information obtained by the image algorithm, the parking coordinates and the placement orientation data of the vehicle can be accurately obtained, and it is informed to the parking personnel whether the vehicle is correctly placed. For the vehicle that does not meet the specifications, the parking personnel can be guided to accurately park.

[0074] S6: The non-motor vehicle foot support is locked by using the ground buckle locking device, and the method of locking is used to limit the front, rear, left and right movements of the parked vehicle. After the vehicle is accurately parked, the user may interfere with the already parked vehicle when taking other vehicles, for example, moving other vehicles due to the narrow space during the vehicle taking process, causing the front and rear movements and left and right shaking of the already parked vehicle, causing the neatly arranged vehicle to become disordered. In order to prevent the vehicle from being moved again, the conventional vehicle locking method can be to build a parking station to fix the vehicle, but for the shared electric vehicle with a larger size, the site demand of the parking station is larger, which will have a higher cost pressure on the operation company. In order to solve the problem, the buckle locking device of the foot support and the ground is used to limit the movement of the parked vehicle by locking.

[0075] The foot support device is a unified structure of the non-motor vehicle, and the press button buckle device is introduced in the designated parking area, and the structure is shown in Figure 3 The hardware cost, construction and maintenance cost of the scheme are low, and the scheme can be quickly and efficiently popularized. At the same time, the invention considers the maintenance of the device, and the buckle device is made of a material that is not easy to corrode and wear; the structure design also considers the invasion of sand and other foreign matters, so the external structure is convenient to disassemble and clean.

[0076] The buckle locking device comprises a circular bottom plate, a long rectangular buckle is arranged on each of the left and right sides of the upper plane of the circular bottom plate, an arc-shaped protruding strip is arranged on the buckle, and the foot support is fixed by the arc-shaped protruding strips on the left and right sides; and a plurality of circular through holes are arranged on the circular bottom plate, and the circular through holes are used to facilitate washing out sand and other foreign matters.

[0077] In the embodiment, the distance difference of the vehicle from the standard parking position and the body placement angle difference data are obtained by image analysis algorithm, combined with prior parking area boundary line data, calculated and mapped by image analysis algorithm, to assist the parking personnel to standardize the placement of the vehicle; finally, the vehicle parking is confirmed by the ground supporting foot support locking device, the device can ensure that the vehicle parking position is accurate, and the placed vehicle is locked to prevent the vehicle from being moved. Under the double protection of the parking point locking and the body placement angle, the vehicle can be accurately placed to the neat and uniform shape, which is conducive to the management of the low-cost road surface road and the maintenance of the city appearance.

[0078] The embodiment of the application also provides a parking system combining an image algorithm and a physical device, which comprises:

[0079] An image acquisition module is used to acquire road surface data by using an infrared fill light and an infrared camera, and is fixed on the vehicle body directly above the wheel. The infrared fill light and the infrared camera are used to solve the problem that the system has difficulty in acquiring the road surface image of the vehicle parking area in the case that there is no obvious light source at night.

[0080] The image processing module is used for analyzing and calculating the collected image, including pre-processing of the original image data, including filtering and denoising, reducing the influence of the shooting area being disturbed by sand, fallen leaves and other objects;

[0081] The brightness information of the parking area boundary line after large-scale statistics is also included as the threshold condition of image binarization processing to obtain a simple edge contour.

[0082] The brightness information of the left and right sides of the original image is counted, and the counting rule is to calculate the pixels of a single frame of image.

[0083] The discontinuous line segments are connected by means of morphological operations such as dilation and erosion to obtain a clear edge contour, and the edge contour line is compared with the known standard parking edge contour line.

[0084] The distance and angle calculation module is used for analyzing the figure formed by the edge contour line and the known standard parking edge contour line based on the edge contour line and the known standard parking edge contour line, judging whether the standard parking edge contour line and the actual parking edge contour line are parallel, and calculating the angle information between the actual parked vehicle body and the normal line of the parking area boundary based on the area and contour length information according to the judgment result.

[0085] The locking module includes a buckle locking device for locking the foot support of the non-motor vehicle, the buckle locking device includes a circular bottom plate, a rectangular buckle is arranged on the left and right sides of the circular bottom plate respectively, an arc-shaped protruding strip is arranged on the buckle, and the foot support is fixed by the left and right arc-shaped protruding strips; a plurality of circular through holes are arranged on the circular bottom plate.

[0086] The above is only a specific embodiment of the present disclosure, but the protection scope of the present disclosure is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present disclosure, which should be covered within the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be subject to the protection scope of the claims.

Claims

1. A parking detection method combining an image algorithm and a physical device, characterized in that, It comprises the following steps: Step one: pre-process the original image data collected, the processing process including filtering and denoising; Step two: take the brightness information of the parking area boundary line after large-scale statistics as the threshold condition for image binarization processing, and obtain the edge contour line; Step three: count the brightness information of the left and right sides of the edge contour line on the original image, further confirm the physical mapping relationship between the image edge information and the actual boundary line, and confirm whether the edge contour line belongs to the left edge or the right edge of the reference line according to the statistical results; Step four: connect the discontinuous line segments by means of morphological operations including dilation and erosion to obtain clear edge contours, analyze the figure formed by the two contours and the upper and lower boundaries of the picture according to the edge contour line and the known standard parking edge contour line, and analyze the figure formed by the two contours and the upper and lower boundaries of the picture; Step five: judge whether the standard parking edge contour line and the actual parking edge contour line are parallel, and based on the judgment result, calculate the pixel distance of the actual parked vehicle from the specified parking area boundary line and the angle information of the vehicle body and the parking area boundary normal line according to the area and contour length information, and obtain the real vehicle-to-parking line distance by using the pixel distance and the pixel-to-real world distance conversion relationship; The area is the area formed by the actual parking edge contour line, the known standard parking edge contour line and the upper and lower boundary lines of the image picture; Step six: lock the non-motor vehicle foot support by using the ground buckle locking device, and limit the forward, backward, left and right movement of the parked vehicle by locking. 2.The parking detection method combining an image algorithm and a physical device according to claim 1, wherein, The rule for counting in step three is to calculate the pixels of a single image. 3.The parking detection method combining an image algorithm and a physical device according to claim 2, wherein, In step three, a statistical conversion table is formed by counting the pixel difference in the picture and the actual distance of the vehicle wheel to the parking line. 4.The parking detection method combining an image algorithm and a physical device according to claim 1, wherein, In step four, the relationship between the actual parking edge contour line and the known standard parking edge contour line can be divided into: the actual parking edge contour line is parallel to the left side of the standard parking edge contour line, the actual parking edge contour line is parallel to the right side of the standard parking edge contour line, the vehicle body is left deflected, and the vehicle body is right deflected. 5.The parking detection method combining an image algorithm and a physical device according to claim 1, wherein, The method for calculating the distance and angle of the actual parked vehicle from the specified parking area boundary line in step five is divided into when the standard parking edge contour line is parallel to the actual parking edge contour line and when it is not. 6.The parking detection method combining an image algorithm and a physical device according to claim 5, wherein, The distance of the actual parked vehicle from the specified parking area boundary line in step five is calculated as follows: When the standard parking edge contour line is parallel to the actual parking edge contour line, the distance L is calculated as follows: ； Wherein, S is the area formed by the actual parking edge profile line, the known standard parking edge profile line and the upper and lower boundary lines of the image frame; is the length of the standard parking edge profile line, and the left and right relative positions of the standard parking edge profile line and the actual parking edge profile line are used to determine whether the actual parking edge profile line is out of the boundary or insufficient. When the standard parking edge contour line is not parallel to the actual parking edge contour line, the distance L is calculated as follows: ； Wherein, S1, S2 is the area formed by the actual parking edge profile line on the upper and lower sides of the intersection and the known standard parking edge profile line; is the length of the standard parking edge profile line. 7.The parking detection method combining an image algorithm and a physical device according to claim 5, wherein, The angle information of the actual parked vehicle body and the parking area boundary normal line in step five is calculated as follows: For when the standard parking edge contour line is not parallel to the actual parking edge contour line, the angle information of the parked vehicle body is obtained by the following calculation method: ； wherein, is the angle between the standard parking edge profile line and the actual parking edge profile line, which is equivalent to the angle information between the actual parked vehicle body and the normal of the parking area boundary, ab is the distance between the upper end of the standard parking edge profile line and the actual parking edge profile line, and ao is the distance from the upper end of the standard parking edge profile line to the intersection O. 8.The parking detection method combining an image algorithm and a physical device according to claim 1, wherein, The buckle locking device in step six includes a circular bottom plate, a rectangular buckle is arranged on the left and right sides of the circular bottom plate, and a circular arc protruding strip is arranged on the buckle, which fixes the foot support through the left and right circular arc protruding strips.

9. A parking system combining image algorithm and physical device, comprising: The image acquisition module adopts an infrared light supplement lamp and an infrared camera to acquire road surface data, and is fixed on the vehicle body directly above the wheel; The image processing module is used for analyzing and calculating the collected images, including pre-processing the original image data; The brightness information of the boundary line of the parking area obtained through large-scale statistics is taken as a threshold condition for image binarization processing to obtain the edge contour line; The brightness information of the left and right sides of the edge on the original graph is counted, and the counting rule is to calculate the pixels of a single frame of image; The morphological operation including dilation and corrosion is used to connect the discontinuous line segments to obtain the clear edge contour; The distance and angle calculation module, The standard parking edge contour line and the actual parking edge contour line are judged whether they are parallel, and based on the judgment result, the pixel distance of the actual parked vehicle from the specified parking area boundary line and the included angle information between the vehicle body and the parking area boundary normal line are calculated according to the area and contour length information formed by the actual parking edge contour line, the known standard parking edge contour line and the upper and lower boundary lines of the image picture, the real vehicle-to-parking line distance is obtained by using the pixel distance of the picture and the conversion relationship between the pixel and the real world distance; The locking module includes a buckle locking device for locking the foot support of the non-motor vehicle, the buckle locking device includes a circular bottom plate, a rectangular buckle is arranged on the left and right sides of the circular bottom plate respectively, and an arc-shaped protruding strip is arranged on the buckle, and the foot support is fixed through the left and right arc-shaped protruding strips.

10. The parking system incorporating both image algorithms and physical devices of claim 9, wherein, A plurality of circular through holes are arranged on the circular bottom plate.

Images