A parking snapshot method based on multi-view fusion and a parking pile

Abstract

The present application relates to a kind of parking snapshot method based on multi-view fusion and a parking pile, the parking pile is equipped with a camera device and two reflector lenses;Two reflector lenses are installed in front of the lens of camera device, so that the image photographed by camera device is the splicing of the image of parking space on both sides reflected by two reflector lenses;The position and angle of reflector lens are adjusted by four distance adjusting devices connected thereto;The initial position of each distance adjusting device is recorded;After the image is gridded, the adjusting position moved to each grid is recorded;After the license plate target in the image is located in the center of the image by controlling the adjusting position of distance adjusting device, it is exported.The present application reduces the overall construction cost of parking pile, and improves the recognition and snapshot recognition ability of license plate.

Description

Technical Field

[0001] The present invention relates to the field of parking space management, and in particular to a parking capture method based on multi-view fusion and a parking pile. Background Art

[0002] With the continued growth of urban car ownership, difficulty and disorder in finding parking have become major challenges in urban traffic management. Smart parking systems leverage technologies such as the Internet of Things, cloud computing, big data, and artificial intelligence to achieve precise monitoring and intelligent management of parking areas. As key components of smart parking systems, parking piles collect real-time information about parking space status and vehicles, improving operational efficiency and service quality. Using 4G / 5G wireless communication technology, these piles transmit real-time parking status and vehicle information. Combined with intelligent analysis from a back-end management platform, these piles enable automated parking management and intelligent guidance. These piles are crucial in smart cities and are an effective solution for reducing costs and increasing efficiency.

[0003] Traditional parking piles are installed on the outside of the road, with one for each parking space, and facing the front or rear of the car. Figure 1 shown.

[0004] Traditional parking piles are built with one per parking space, resulting in high construction and maintenance costs. Furthermore, due to system construction cost constraints, traditional parking piles use fixed-focus cameras with a fixed field of view. This makes them unable to adapt to license plate recognition in different parking positions and effectively fails to recognize the license plates of vehicles that occupy spaces illegally. This reduces parking pile operating revenue. Summary of the Invention

[0005] In order to solve the above problems, the present invention proposes a parking capture method based on multi-view fusion and a parking pile.

[0006] The specific plan is as follows:

[0007] A parking capture method based on multi-view fusion includes the following steps:

[0008] S1: Every two parking spaces share a parking pile; the parking pile is equipped with a camera and two reflective lenses; the two reflective lenses are installed in front of the camera lens, so that the image captured by the camera is a mosaic of the images of the parking spaces on both sides reflected by the two reflective lenses; the position and angle of the reflective lenses are adjusted by four distance adjustment devices connected to them;

[0009] S2: Record the adjustable distance range of each distance adjustment device, and use the position when the adjustment positions of the four distance adjustment devices corresponding to each reflective lens are all in the center of the corresponding adjustable distance range as the initial position;

[0010] S3: After the areas where the parking space images on both sides are gridded, the following operations are performed on both reflective lenses: the four distance adjustment devices corresponding to the reflective lenses are adjusted so that the object initially located in the grid area where the image center is located is moved to another grid area, and the current positions of the four distance adjustment devices after the movement are recorded as the adjustment positions corresponding to the moved grid areas; this movement process is repeated until the adjustment positions corresponding to all grid areas are obtained, and the four distance adjustment devices are controlled to return to their initial positions;

[0011] S4: When a parking signal is received, the camera device corresponding to the parking space is controlled to capture an image to obtain a parking space image corresponding to the parking space;

[0012] S5: Determine whether there is a license plate target in the parking space image. If yes, proceed to S6; otherwise, proceed to S8;

[0013] S6: Determine whether the license plate target is located in the grid area where the image center is located. If so, output the parking space image; otherwise, enter S7;

[0014] S7: After searching for an adjustment position corresponding to the grid area where the license plate target is located, controlling the distance adjustment device to move to the found adjustment position, recapturing the image, and outputting the parking space image;

[0015] S8: Control the distance adjustment device to move to the adjustment position corresponding to each grid area in turn, and re-capture the image after each movement; when there is a license plate target in the parking space image, return to S6; otherwise, move to the next grid area; when there is no license plate target in the corresponding parking space image after traversing all grid areas, report an error.

[0016] Furthermore, the reflective lens is installed as follows: a fixed plate is fixedly installed in the central position in front of the lens of the camera device, the plane of the fixed plate is perpendicular to the lens surface, and two reflective lenses are respectively installed on both sides of the fixed plate, and are installed on the fixed plate through the distance adjustment device connected to them.

[0017] Furthermore, four distance adjustment devices are connected to four vertices of the reflective lens.

[0018] Furthermore, the grid is divided into 9 grids of 3*3.

[0019] Furthermore, before step S3, it also includes measuring the actual adjustable distance range of each distance adjustment device, and judging whether the maximum value in the actual adjustable distance range is less than the maximum value of the theoretical adjustable distance range. If so, if the adjustment position corresponding to a recorded grid area contains the maximum value of the theoretical adjustable distance range, the maximum value in the measured actual adjustable distance range is used to replace the maximum value of the theoretical adjustable distance range; at the same time, it is judged whether the minimum value in the actual adjustable distance range is greater than the minimum value of the theoretical adjustable distance range. If so, if the adjustment position corresponding to a recorded grid area contains the minimum value of the theoretical adjustable distance range, the minimum value in the measured actual adjustable distance range is used to replace the minimum value of the theoretical adjustable distance range.

[0020] Furthermore, step S7 adopts a step-by-step adjustment method when controlling the distance adjustment device to move to the found adjustment position, specifically: after taking the found adjustment position of each distance adjustment device as the target position, the difference between the target position of each distance adjustment device and the current position is taken as the adjustment distance, and according to the set number of steps, the quotient of the adjustment distance and the number of steps is used as the adjustment step length of each step; when controlling the distance adjustment to move to the found adjustment position, the movement is performed in steps, one adjustment step length each time, and after each movement, an image is captured, and it is determined whether the license plate target in the parking space image is located in the grid area where the image center is located. If so, the movement is ended and the parking space image is output; otherwise, the next movement is performed until the number of movements reaches the number of steps.

[0021] A parking pile is provided, on which a camera device and two reflective lenses are installed; the two reflective lenses are installed in front of the lens of the camera device, so that the image captured by the camera device is a splicing of parking space images on both sides reflected by the two reflective lenses; the reflective lenses adjust their positions and angles through four distance adjustment devices connected to them; the parking pile is also provided with a control device, which includes a processor, a memory, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the above-mentioned method of the embodiment of the present invention when executing the computer program.

[0022] The present invention adopts the above technical solution and has the following beneficial effects:

[0023] (1) Reduce the overall construction cost of parking piles. Compared with the traditional parking pile construction method, the parking piles through multi-perspective fusion can reduce the number of parking piles by nearly half, reduce the construction cost, and have obvious cost advantages.

[0024] (2) Improve the parking pile's ability to capture license plates of various types of vehicles. By adjusting the reflective lens, it can capture license plate images in a wider range. It can also capture images of special vehicles with different license plate hanging positions. It also improves the ability to capture license plates of vehicles that deliberately do not park according to regulations, reduces the probability of evasion, and increases revenue. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 Shown is a schematic diagram of the installation location of a traditional parking pile.

[0026] Figure 2 Shown is a schematic diagram of the installation position of the parking pile in embodiment 1 of the present invention.

[0027] Figure 3 Shown is a flow chart of the method in embodiment 1 of the present invention.

[0028] Figure 4 FIG. 1 is a schematic diagram showing the installation positions of the camera device and the reflective lens in the first embodiment of the present invention.

[0029] Figure 5 FIG. 1 is a schematic diagram of a grid structured into 9 grids in the first embodiment of the present invention.

[0030] Figure 6 The figure shows a schematic diagram of the license plate target detected in the first embodiment of the present invention. In the figure, [1] represents the pixel coordinates [Xc, Yc] of the image center, [2] represents the pixel coordinates [Xid, Yid] of the center of the license plate area, [3] represents the pixel length pixL of the image screen, and [4] represents the pixel height pixH of the image screen. DETAILED DESCRIPTION

[0031] To further illustrate various embodiments, the present invention provides accompanying drawings. These drawings form part of the present disclosure and are primarily used to illustrate the embodiments and, in conjunction with the relevant description in the specification, to explain the operating principles of the embodiments. By referring to these drawings, those skilled in the art will be able to understand other possible implementations and the advantages of the present invention.

[0032] The present invention will now be further described with reference to the accompanying drawings and specific embodiments.

[0033] Example 1:

[0034] The embodiment of the present invention provides a parking capture method based on multi-view fusion, such as Figure 2 As shown, the method includes the following steps:

[0035] S1: Every two parking spaces share one parking pile; a camera device and two reflective lenses are installed on the parking pile; the two reflective lenses are installed in front of the lens of the camera device, so that the image captured by the camera device is a splicing of the images of the parking spaces on both sides reflected by the two reflective lenses; the position and angle of the reflective lenses are adjusted by the four distance adjustment devices connected to them.

[0036] like Figure 3 The figure shows a schematic diagram of roadside parking spaces in this embodiment, in which a parking pile is set between two parking spaces. The camera device installed on the parking pile can simultaneously capture the vehicle registration information of the two parking spaces on the left and right. With the same number of monitored parking spaces, the number of parking piles can be reduced exponentially (i.e., halved), thereby reducing construction and maintenance costs.

[0037] Figure 4 The figure shows a schematic diagram of the installation positions of the camera device and reflective lens in this embodiment. As can be seen from the figure, a fixed plate 2 is fixedly mounted in the center in front of the camera lens 4. The plane of the fixed plate 2 is perpendicular to the surface of the lens 4. The reflective lens 1 is mounted on the fixed plate 2 via a distance adjustment device 3 connected by its four vertices. Two reflective lenses 1 are mounted on either side of the fixed plate 2. In this embodiment, the distance adjustment device 3 uses a telescopic rod. By controlling the telescopic distance of the telescopic rod, the position and angle of the reflective lens 1 relative to the fixed plate 2 can be adjusted. When installing the camera device, it is necessary to ensure that the surface of the lens 4 is perpendicular to the horizontal plane and that the plane of the fixed plate 2 is parallel to the dividing line between the two parking spaces.

[0038] In this embodiment, the four distance adjustment devices are preferably connected to the four vertices of the reflective lens.

[0039] S2: Record the adjustable distance range of each distance adjustment device ([mind, maxd], where mind represents the minimum value of the adjustable distance range and maxd represents the maximum value of the adjustable distance range), and use the position when the adjustment positions of the four distance adjustment devices corresponding to each reflective lens are all in the center of the corresponding adjustable distance range as the initial position.

[0040] The theoretical adjustable range of each distance adjustment device within a product batch is fixed and determined during product design. In this embodiment, the initial position of each distance adjustment device is determined when it is at the center of its adjustable range (assuming maxd - mind = datD, then the center of the adjustable range is represented by mind + datD / 2).

[0041] S3: After the areas where the parking space images on both sides are located are gridded, the following operations are performed on the two reflective lenses: the four distance adjustment devices corresponding to the reflective lenses are adjusted so that the object that is initially in the grid area where the image center is located is moved to other grid areas, and the current positions of the four distance adjustment devices after the movement are recorded as the adjustment positions corresponding to the moved grid areas; the movement process is repeated until the adjustment positions corresponding to all grid areas are obtained, and the four distance adjustment devices are controlled to return to the initial positions.

[0042] Gridding is to split the image into grids of the same size. In this embodiment, it is preferably split into 9 grids of 3*3, such as Figure 5 As shown, in other embodiments, it can also be divided into other numbers of grids, which is not limited here.

[0043] The adjustment positions corresponding to the 9 grids obtained in this embodiment are shown in Table 1. The four columns A, B, C, and D correspond to the adjustment positions of the four distance adjustment devices, respectively. For example, the content of the row where grid number 1 is located represents the current positions of the four distance adjustment devices when moving to the grid area with grid number 1.

[0044] Table 1

[0045]

[0046] As can be seen from Table 1, when moving to the grid areas containing grids numbered 1 and 3, the corresponding adjustment device positions are at the boundaries of the adjustable distance range. Considering that the maximum or minimum values ​​of the adjustable distance ranges of each distance adjustment device may have errors during actual production, when the maximum or minimum values ​​are less than the maximum value or greater than the minimum value, the distance adjustment device may not be able to move to the position recorded in Table 1. To address this technical problem, this embodiment also includes measuring the actual maximum and minimum values ​​of the adjustable distance range of each distance adjustment device before each product is put into use. When the adjustment position corresponding to a certain grid area is the maximum or minimum value of the adjustable distance range, the actual maximum or minimum value is used.

[0047] S4: After receiving a signal that the vehicle is parked, the camera device corresponding to the parking space of the signal is controlled to capture an image to obtain a parking space image corresponding to the parking space.

[0048] The vehicle parking signal is sent when a vehicle target is detected in an image captured by the camera device.

[0049] The parking space image is half of the image taken by the camera device. For example, the parking space image corresponding to the left parking space is the left half of the image taken by the camera device, and the parking space image corresponding to the right parking space is the right half of the image taken by the camera device. The specific corresponding relationship needs to be set.

[0050] S5: Determine whether there is a license plate target in the parking space image. If yes, proceed to S6; otherwise, proceed to S8.

[0051] To determine whether a license plate target exists in the parking space image, an existing license plate target recognition model may be used, and no limitation is imposed here.

[0052] S6: Determine whether the license plate target is located in the grid area where the image center is located. If so, output the parking space image; otherwise, enter S7.

[0053] like Figure 5 As shown, in this embodiment, the grid area where the image center is located is the area where the grid numbered 2 is located.

[0054] S7: After searching for an adjustment position corresponding to the grid area where the license plate target is located, the distance adjustment device is controlled to move to the found adjustment position, and the image is captured again to output the parking space image.

[0055] like Figure 6 As shown, the grid area where the license plate target is located is the grid area with sequence number 4. Find the row with grid sequence number 4 from Table 1. Its contents mind+datD*3 / 4, mind+datD / 4, mind+datD, mind+datD / 2 are the adjustment positions of the four distance adjustment devices numbered A, B, C, and D. Then control the four distance adjustment devices so that their adjustment positions are at the found positions.

[0056] Furthermore, this embodiment also provides a step-by-step adjustment method, the specific process is as follows:

[0057] After the adjustment position of each distance adjustment device is found as the target position destD, the difference between the target position destD and the current position d of each distance adjustment device is used as the adjustment distance. According to the set number of steps n (such as 5 steps), the quotient of the adjustment distance and the number of steps is used as the adjustment step length stepD of each step (stepD = (destD-d) / n).

[0058] While controlling the distance adjustment to move it to the found adjustment position, it moves in steps, one adjustment step at a time. After each movement, an image is captured, and it is determined whether the license plate target in the parking space image is located in the grid area where the image center is located. If so, the movement ends and the parking space image is output; otherwise, the next movement is performed until the number of moves reaches the step number.

[0059] By using the above-mentioned step-by-step adjustment method, the probability that the license plate target in the output parking space image is located at the center of the image can be increased.

[0060] S8: Control the distance adjustment device to move to the adjustment position corresponding to each grid area in turn, and re-capture the image after each movement; when there is a license plate target in the parking space image, return to S6; otherwise, move to the next grid area; when there is no license plate target in the corresponding parking space image after traversing all grid areas, report an error.

[0061] In this embodiment, the grids are moved in order of sequence 1-9 to reduce the overall moving distance.

[0062] The method of this embodiment adjusts the position of the reflective lens by identifying the area where the license plate target is located in the image, so that the license plate target in the output image is always in a relatively centered position, which helps to improve the accuracy of vehicle license plate recognition.

[0063] The embodiments of the present invention can improve the field of view of the camera device's shooting area (including the horizontal and vertical ranges), thereby effectively solving the problem of identifying license plates that adapt to different hanging positions, and also solving the problem of identifying license plates of vehicles that do not park according to regulations and occupy parking spaces.

[0064] Example 2:

[0065] The present invention also provides a parking pile, which adopts the structural features described in Example 1 and is also equipped with a control device, which includes a memory, a processor, and a computer program stored in the memory and run on the processor. When the processor executes the computer program, steps S2-S8 in the above method embodiment in Example 1 of the present invention are implemented.

[0066] Furthermore, as an executable solution, the processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices. The general-purpose processor may be a microprocessor or any conventional processor. The processor serves as the control center of the multi-view fusion-based parking capture terminal device, connecting various components of the multi-view fusion-based parking capture terminal device using various interfaces and circuits.

[0067] The memory can be used to store the computer programs and / or modules. The processor implements the various functions of the multi-view fusion-based parking capture terminal device by running or executing the computer programs and / or modules stored in the memory and accessing the data stored in the memory. The memory may primarily include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function; the data storage area may store data created based on the use of the processor. Furthermore, the memory may include high-speed random access memory and non-volatile memory, such as a hard disk, internal memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0068] Although the present invention has been particularly shown and described in conjunction with preferred embodiments, it will be understood by those skilled in the art that various changes in form and details may be made to the present invention without departing from the spirit and scope of the invention as defined in the appended claims, and all such changes are within the scope of protection of the present invention.

Claims

1. A parking capture method based on multi-view fusion, characterized in that: The following steps are involved: S1: Every two parking spaces share a parking pile; the parking pile is equipped with a camera and two reflective lenses; the two reflective lenses are installed in front of the camera lens, so that the image captured by the camera is a mosaic of the images of the parking spaces on both sides reflected by the two reflective lenses; the position and angle of the reflective lenses are adjusted by four distance adjustment devices connected to them; S2: Record the adjustable distance range of each distance adjustment device, and use the position when the adjustment positions of the four distance adjustment devices corresponding to each reflective lens are all in the center of the corresponding adjustable distance range as the initial position; S3: After the areas where the parking space images on both sides are gridded, the following operations are performed on both reflective lenses: four distance adjustment devices corresponding to the reflective lenses are adjusted so that the object initially located within the grid area where the image center is located is moved to another grid area, and the current positions of the four distance adjustment devices after the movement are recorded as the adjustment positions corresponding to the moved grid area; Repeat this movement process until the adjustment positions corresponding to all grid areas are obtained, and then control the four distance adjustment devices to return to their initial positions; S4: When a parking signal is received, the camera device corresponding to the parking space is controlled to capture an image to obtain a parking space image corresponding to the parking space; S5: Determine whether there is a license plate target in the parking space image. If yes, proceed to S6; otherwise, proceed to S8; S6: Determine whether the license plate target is located in the grid area where the image center is located. If so, output the parking space image; Otherwise, go to S7; S7: After searching for an adjustment position corresponding to the grid area where the license plate target is located, controlling the distance adjustment device to move to the found adjustment position, recapturing the image, and outputting the parking space image; S8: Control the distance adjustment device to move to the adjustment position corresponding to each grid area in sequence, and re-capture the image after each movement; if a license plate target exists in the parking space image, return to S6; otherwise, move to the next grid area; When no license plate target exists in the corresponding parking space image after all grid areas are traversed, an error is reported.

2. The parking capture method based on multi-view fusion according to claim 1, characterized in that: The reflective lens is installed as follows: a fixing plate is fixedly installed in the center position in front of the lens of the camera device, and the plane of the fixing plate is perpendicular to the lens surface. Two reflective lenses are respectively installed on both sides of the fixing plate and are installed on the fixing plate through the distance adjustment device connected to them.

3. The parking capture method based on multi-view fusion according to claim 1, characterized in that: Four distance adjustment devices are connected to the four vertices of the reflective lens.

4. The parking capture method based on multi-view fusion according to claim 1, characterized in that: The grid is divided into 9 grids of 3*3.

5. The parking capture method based on multi-view fusion according to claim 1, characterized in that: Before step S3, it also includes measuring the actual adjustable distance range of each distance adjustment device, and judging whether the maximum value in the actual adjustable distance range is less than the maximum value of the theoretical adjustable distance range. If so, if the adjustment position corresponding to a recorded grid area contains the maximum value of the theoretical adjustable distance range, the maximum value in the measured actual adjustable distance range is used to replace the maximum value of the theoretical adjustable distance range; at the same time, it is judged whether the minimum value in the actual adjustable distance range is greater than the minimum value of the theoretical adjustable distance range. If so, if the adjustment position corresponding to a recorded grid area contains the minimum value of the theoretical adjustable distance range, the minimum value in the measured actual adjustable distance range is used to replace the minimum value of the theoretical adjustable distance range.

6. The parking capture method based on multi-view fusion according to claim 1, characterized in that: Step S7 adopts a step-by-step adjustment method when controlling the distance adjustment device to move to the found adjustment position, specifically: after taking the found adjustment position of each distance adjustment device as the target position, the difference between the target position of each distance adjustment device and the current position is taken as the adjustment distance, and according to the set number of steps, the quotient of the adjustment distance and the number of steps is taken as the adjustment step length of each step; when controlling the distance adjustment to move to the found adjustment position, the movement is performed in steps, one adjustment step length each time, and after each movement, an image is captured, and it is determined whether the license plate target in the parking space image is located in the grid area where the image center is located. If so, the movement is ended and the parking space image is output; otherwise, the next movement is performed until the number of movements reaches the number of steps.

7. A parking pile, characterized by: A camera device and two reflective lenses are installed on the parking pile; the two reflective lenses are installed in front of the lens of the camera device, so that the image captured by the camera device is a splicing of the parking space images on both sides reflected by the two reflective lenses; the reflective lenses adjust their positions and angles through four distance adjustment devices connected to them; a control device is also installed on the parking pile, which includes a processor, a memory, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, steps S2-S8 of the method as described in any one of claims 1 to 6 are implemented.

Images