Illegal parking identification method, controller, system and shared bicycle

Abstract

The invention discloses an illegal parking identification method, a controller, a system and a shared bicycle. The method comprises the following steps: acquiring current position information of a shared bicycle; identifying a parking area according to the current position information, and determining a first identification result; if the first identification result is that the area is parked in a compliant manner, acquiring a current environment image of the shared bicycle; performing parking position identification according to the current environment image, and determining a second identification result; and if the second identification result is illegal parking, carrying out illegal parking reminding. According to the method, the current position information and the current environment image of the shared bicycle are combined for illegal parking identification, the problem that the obtained current position information is not accurate can be effectively solved, the illegal parking identification accuracy is improved, and the method does not need to input manpower and is convenient to use. The illegal parking identification cost and the operation management cost of a shared bicycle operator can be saved.

Description

Technical Field

[0001] The present invention relates to the technical field of shared vehicles, and in particular to an illegal parking identification method, a controller, a system and a shared vehicle. Background Art

[0002] Car-sharing is a new type of green travel mode that can provide users with convenient, efficient and low-carbon travel options, and is particularly suitable for solving the "last mile" problem in cities. However, with the gradual popularization of car-sharing, some management and supervision problems of car-sharing have emerged, and the most prominent management and supervision problem is the illegal parking of car-sharing.

[0003] To solve the problem of illegal parking of shared cars, it is necessary to identify and process the shared cars that are parked illegally. When solving the problem of identifying shared cars that are parked illegally, the existing technology usually adopts position positioning and video surveillance to achieve it. Position positioning in the existing technology usually obtains the parking location information through the built-in positioning module of the shared car, wherein the positioning module usually adopts a civilian positioning module, and the position positioning is limited by the accuracy of the civilian positioning module. Video surveillance in the existing technology usually sets up surveillance cameras in compliant parking areas, and determines whether the shared car is parked in the compliant parking area through image recognition. However, the installation of surveillance cameras requires poles in public areas of the city, which is costly, slow to deploy, and difficult to obtain permits. Therefore, the existing technology cannot effectively solve the problem of illegal parking identification of shared cars that is of concern to shared car operators.

[0004] To sum up, how to more accurately identify illegal parking of shared vehicles is a technical problem that needs to be urgently solved in the existing technology. Summary of the invention

[0005] The embodiments of the present invention provide a method, a controller, a system and a shared vehicle for identifying illegal parking, so as to solve the problem of how to more accurately identify illegal parking of a shared vehicle.

[0006] A first aspect of an embodiment of the present application provides a method for identifying illegal parking, including:

[0007] Get the current location information of the shared car;

[0008] Perform parking area identification according to the current location information to determine a first identification result;

[0009] If the first recognition result is regional compliance parking, obtaining a current environment image of the shared vehicle;

[0010] Perform parking posture recognition according to the current environment image to determine a second recognition result;

[0011] If the second recognition result is illegal parking, an illegal parking reminder is given.

[0012] A second aspect of an embodiment of the present application provides a controller, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned illegal parking identification method when executing the computer program.

[0013] A third aspect of the embodiments of the present application provides an illegal parking identification system, comprising a positioning module, a camera and the above-mentioned controller, wherein the controller is connected to the positioning module and the camera respectively; the positioning module is used to obtain current location information of the shared vehicle, and the controller is used to identify the parking area according to the current location information and determine a first identification result;

[0014] The camera is used to obtain the current environment image of the shared car; the controller is also used to perform parking posture recognition based on the current environment image, determine the second recognition result, and realize the recognition of illegal parking of the shared car.

[0015] A fourth aspect of an embodiment of the present application provides a shared vehicle, comprising the above-mentioned illegal parking identification system.

[0016] The above-mentioned illegal parking identification method, controller, system and shared car can identify the parking area of ​​the shared car according to the current position information and determine the first identification result. The shared car whose current position information does not meet the regional parking requirements can be excluded from the compliant parking, and the first identification result of illegal parking can be directly obtained, which can improve the recognition efficiency of illegal parking identification. When the first identification result is compliant parking, the parking posture is identified according to the current environment image to determine the second identification result, which can realize more accurate illegal parking identification of the shared car and improve the accuracy of illegal parking identification. This method combines the current position information of the shared car with the current environment image to perform illegal parking identification, which can effectively solve the problem of inaccurate current position information obtained, improve the accuracy of illegal parking identification, and this method does not require manpower investment, which can save the cost of illegal parking identification and the operating and management costs of shared car operators. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0018] Figure 1 is a flow chart of a method for identifying illegal parking in one embodiment of the present invention;

[0019] Figure 2 is another flow chart of a method for identifying illegal parking in one embodiment of the present invention;

[0020] Figure 3 is another flow chart of a method for identifying illegal parking in one embodiment of the present invention;

[0021] Figure 4 is another flow chart of a method for identifying illegal parking in one embodiment of the present invention;

[0022] Figure 5 is another flow chart of a method for identifying illegal parking in one embodiment of the present invention;

[0023] Figure 6 is another flow chart of a method for identifying illegal parking in one embodiment of the present invention;

[0024] Figure 7 is another flow chart of a method for identifying illegal parking in one embodiment of the present invention;

[0025] Figure 8 A schematic diagram of a controller in one embodiment of the present invention;

[0026] Fig. 9 It is a schematic diagram of the current parking angle and the current parking distance between the shared vehicle and the parking reference object in one embodiment of the present invention. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0028] The embodiment of the present invention provides an illegal parking identification method. Specifically, the illegal parking identification method is applied to an illegal parking identification system built into a shared car to achieve the purpose of more accurately identifying illegal parking of the shared car. The illegal parking identification system includes Figure 7 Controller shown.

[0029] In one embodiment, if Figure 1 As shown, a method for identifying illegal parking is provided, and the method is applied in Figure 8 The controller in the example is used to illustrate the process, including the following steps:

[0030] S101: Obtain the current location information of the shared car;

[0031] S102: Identify the parking area according to the current location information and determine a first identification result;

[0032] S103: If the first recognition result is that the shared car is parked in compliance with the area, the current environment image of the shared car is obtained;

[0033] S104: performing parking posture recognition according to the current environment image to determine a second recognition result;

[0034] S105: If the second recognition result is illegal parking, an illegal parking reminder is issued.

[0035] Among them, the current location information refers to the location information of the shared car when it is parked.

[0036] As an example, in step S101, the controller determines whether the shared car is currently in a parking state. When it is determined that the shared car is in a parking state, the current location information of the shared car is obtained. For example, the controller can obtain the current location information of the shared car through a positioning module installed on the body of the shared car. In this example, the current location information of the shared car is obtained, so that it is feasible to subsequently identify illegal parking based on the current location information of the shared car.

[0037] Among them, parking area identification refers to the process of identifying the parking area of ​​a shared car based on the current location information of the shared car to determine whether the shared car is parked illegally. The first identification result refers to the identification result of the parking area identification of the shared car based on the current location information. The first identification result includes two identification results: regional compliant parking and regional illegal parking. Among them, regional compliant parking refers to the parking state of a shared car that meets the requirements of the parking area. Regional illegal parking refers to the parking state of a shared car that does not meet the requirements of the parking area. Among them, the parking area requirements refer to the shared car being parked in a location area that meets the requirements.

[0038] As an example, in step S102, after determining to obtain the current location information of the shared car in the parking state, the controller performs parking area identification on the current location information, determines whether the shared car is illegally parked in the area, determines the first identification result, and determines that the first identification result of the shared car is illegal parking in the area or compliant parking in the area. For example, after obtaining the current location information of the shared car, the controller identifies whether the current location information is within the preset parking area range. If the current location information is within the preset parking area range, the first identification result of the shared car is determined to be compliant parking in the area; if the current location information is not within the preset parking area range, the first identification result of the shared car is determined to be illegal parking in the area. When the first identification result is determined to be illegal parking in the area, the controller sends a reminder message of illegal parking to the user of the shared car, and reminds the user of the shared car of illegal parking. In this example, the shared car is identified in the parking area according to the current location information, and the first identification result is determined. The shared car whose current location information does not meet the requirements of the parking area can be excluded from compliant parking, and the first identification result of illegal parking in the area is directly obtained, which can improve the recognition efficiency of illegal parking identification.

[0039] The current environment image refers to the surrounding environment image of the shared car at the current parking location.

[0040] As an example, in step S103, when the controller determines that the first recognition result is regional compliance parking, it obtains the current environment image around the shared car, so as to facilitate the subsequent parking posture recognition of the current environment image, and further performs parking posture recognition on the shared car whose first recognition result is regional compliance parking, so as to facilitate making the illegal parking recognition result more accurate. In this example, the controller can receive a camera set on the shared car, specifically a camera set under the front basket of the shared car, to take real-time pictures of the current environment image of the shared car, and send the current environment image to the controller, so as to facilitate the subsequent illegal parking recognition based on the current environment image again, so as to improve the recognition accuracy of illegal parking recognition. It is understandable that due to the limitation of positioning accuracy, the accuracy of the current position information obtained is not enough. Therefore, when the first recognition result is determined to be regional compliance parking based on the current position information, it is necessary to further obtain the current environment image around the shared car, so as to facilitate the subsequent determination of a more accurate second recognition result based on the current environment image, so as to achieve the purpose of improving the recognition accuracy of illegal parking recognition of shared cars. In this example, the camera of the shared car can collect the current environment image around the shared car, and the controller can obtain the current environment image collected by the camera of the shared car.

[0041] Among them, parking posture recognition refers to the process of identifying the parking environment position and parking posture of the shared car based on the current environment image of the shared car, and determining whether the shared car is parked illegally. The second recognition result is the recognition result of whether the shared car is parked illegally based on the current environment image of the shared car. The second recognition result includes two recognition results: illegal parking and compliant parking.

[0042] As an example, in step S104, after acquiring the current environment image of the shared car in the parking state, the controller performs parking posture recognition on the current environment image, obtains the recognition result of whether the shared car is illegally parked, and uses the recognition result as the second recognition result. In this example, the controller has a built-in visual algorithm module, and extracts the image features of the current environment image through the deep neural network and neural network acceleration unit in the visual algorithm module. Based on the extracted image features, the shared car is recognized for the parking environment position and parking posture, and the second recognition result is determined to be illegal parking or compliant parking. For example, the controller determines whether the image features extracted based on the current environment image include fences, green belts or white lines in the parking area, determines whether the shared car meets the corresponding environment position and parking posture parking requirements based on the image features, and determines whether the second recognition result is compliant parking or illegal parking. In this example, when the first recognition result is compliant parking, the parking posture recognition is performed according to the current environment image to determine the second recognition result, which can achieve a more accurate illegal parking recognition of the shared car and improve the accuracy of illegal parking recognition.

[0043] Among them, the illegal parking reminder refers to a reminder sent to the user to remind the user to park the shared car again in compliance with the regulations when the first recognition result or the second recognition result is determined to be illegal parking.

[0044] As an example, in step S105, when the controller determines that the second recognition result is illegal parking, it sends an illegal parking reminder to the user currently using the shared car, reminding the user to park the shared car in compliance. In this example, when the controller obtains the recognition result of illegal parking, it sends a warning signal to the user as an illegal parking reminder to remind the user to park the shared car in compliance and standardize the parking order of shared cars.

[0045] The illegal parking identification method provided in this embodiment identifies the parking area of ​​the shared car according to the current position information and determines the first identification result. The shared car whose current position information does not meet the requirements of the parking area can be excluded from the compliant parking, and the first identification result of regional illegal parking can be directly obtained, which can improve the recognition efficiency of illegal parking identification. When the first identification result is regional compliant parking, the parking posture is identified according to the current environment image to determine the second identification result, which can realize more accurate illegal parking identification of the shared car and improve the accuracy of illegal parking identification. This method combines the current position information of the shared car with the current environment image to identify illegal parking, which can effectively solve the problem of inaccurate current position information obtained and improve the accuracy of illegal parking identification. This method does not require manpower investment and can save the cost of illegal parking identification and the operating and management costs of shared car operators.

[0046] In another embodiment, after step S105, that is, after the illegal parking reminder is performed if the second recognition result is illegal parking, the illegal parking recognition method further includes:

[0047] During the preset recognition time period, continue to obtain the current location information of the shared car. If the first recognition result or the second recognition result is determined to be illegal parking, obtain the current user information and the time of illegal parking, and record the current location information, current environment image, current user information and the time of illegal parking.

[0048] The preset identification time period refers to the monitoring time period for continued illegal parking identification of shared vehicles after the illegal parking reminder is sent to the user. The current user information refers to the information of the user currently using the shared vehicle, including user ID and user contact information. The illegal parking moment refers to the moment when the illegal parking identification result is finally determined to be illegal parking within the preset identification time period.

[0049] As an example, after sending an illegal parking reminder to the user, the controller continues to identify the shared car for illegal parking within a certain preset identification time period, that is, continues to execute steps S101 to S105 until it is determined that the shared car is parked in compliance with regulations, or, if no compliance parking is identified, when the preset identification time period is exceeded, the final illegal parking identification result is determined to be illegal parking, and the illegal parking identification is stopped. If the shared car is determined to be illegally parked, the current user information and illegal parking time of the shared car are obtained, and the current location information, current environment image, current user information and illegal parking time of the shared car obtained most recently are recorded in the system database to facilitate the subsequent management of the shared car and the management of the users who use the shared car. In this example, the identification cycle can be set, and steps S101 to S105 are executed on the shared car according to the identification cycle within the preset identification time period to determine whether the shared car has been re-parked in compliance with regulations by the user.

[0050] In this embodiment, when it is determined that the first recognition result or the second recognition result is illegal parking, illegal parking recognition continues in the preset recognition time period. When the first recognition result or the second recognition result is finally determined to be illegal parking, the current location information, current environment image, current user information and illegal parking time are recorded to facilitate subsequent management of shared vehicles and users.

[0051] In one embodiment, if Figure 2 As shown, step S102, i.e., identifying the parking area according to the current location information and determining a first identification result, includes:

[0052] S201: Determine whether the shared vehicle is in a legal parking area based on the current location information;

[0053] S202: If the shared vehicle is in a compliant parking area, determining the first recognition result as compliant parking;

[0054] S203: If the shared vehicle is not in the compliant parking area, the current parking distance between the shared vehicle and the compliant parking area is obtained, and a first recognition result is determined based on the current parking distance.

[0055] Among them, the compliant parking area refers to a pre-set area for preliminary identification of illegal parking of shared vehicles based on the current location information. Understandably, when identifying illegal parking of shared vehicles based on the current location information, there will usually be errors due to the limitation of positioning accuracy. For example, due to the existence of positioning errors, a compliantly parked shared vehicle may be identified as illegally parked. Therefore, it is necessary to set a larger area as the compliant parking area to reduce the identification error when performing preliminary identification of illegal parking of shared vehicles based on the current location information.

[0056] As an example, in step S201, the controller determines whether the location of the shared vehicle is in a compliant parking area based on the current location information. In this example, the controller determines whether the current location information is within the compliant parking area to facilitate subsequent determination of whether the shared vehicle is parked in compliance.

[0057] As an example, in step S202, when the controller determines that the shared vehicle is in the compliant parking area, it directly determines the first recognition result as compliant parking. In this example, if the shared vehicle is in the compliant parking area, the first recognition result is determined to be compliant parking, which can improve the efficiency of illegal parking recognition and facilitate further and more accurate recognition of whether illegal parking is occurring.

[0058] Among them, the current parking distance refers to the distance between the current parking position of the shared car and the compliant parking area.

[0059] As an example, in step S203, when the controller determines that the shared car is not in the compliant parking area, it obtains the current parking distance between the shared car and the compliant parking area, and determines the first recognition result when the shared car is not in the compliant parking area based on the current parking distance. Understandably, due to the inaccuracy of positioning, misjudgment may occur. Therefore, the shared car whose current location information is outside the compliant parking area is further taken into account, and the first recognition result is further determined based on the current parking distance between the shared car and the compliant parking area, so as to reduce the recognition error rate and improve the recognition accuracy.

[0060] The illegal parking identification method provided in this embodiment determines whether the shared car is in the compliant parking area based on the current location information. If the shared car is in the compliant parking area, the first identification result is determined to be compliant parking, which can improve the efficiency of illegal parking identification and facilitate further and more accurate identification of illegal parking. If the shared car is not in the compliant parking area, the current parking distance between the shared car and the compliant parking area is obtained, and the first identification result is determined based on the current parking distance between the shared car and the compliant parking area, so as to reduce the identification error rate and improve the identification accuracy.

[0061] In one embodiment, if Figure 3 As shown, step S203, i.e. determining a first recognition result based on the current parking distance, includes:

[0062] S301: If the current parking distance does not exceed the preset distance threshold, determining the first recognition result as compliant parking;

[0063] S302: If the current parking distance exceeds a preset distance threshold, determining that the first recognition result is illegal parking.

[0064] Among them, the preset distance threshold refers to the preset distance value used to determine whether the shared car is parked in compliance with regulations.

[0065] As an example, in step S301, after obtaining the current parking distance of the shared car outside the compliant parking area, the controller compares the size relationship between the current parking distance and the preset distance threshold, and directly determines the first recognition result as compliant parking when it is determined that the current parking distance does not exceed the preset distance threshold. Understandably, if the current parking distance does not exceed the preset distance threshold, it indicates that the current location information of the shared car is close to the location of the compliant parking area. At this time, the first recognition result is determined to be compliant parking, which facilitates the subsequent further identification of illegal parking based on the current environment image, thereby improving the accuracy of identifying illegal parking of shared cars.

[0066] As an example, in step S302, after obtaining the current parking distance of the shared car outside the compliant parking area, the controller compares the size relationship between the current parking distance and the preset distance threshold, and directly determines the first recognition result as illegal parking when it is determined that the current parking distance exceeds the preset distance threshold. Understandably, if the current parking distance exceeds the preset distance threshold, it indicates that the current location information of the shared car is far from the location of the compliant parking area. At this time, determining the first recognition result as illegal parking can improve the efficiency of identifying illegal parking of shared cars and save the recognition cost of illegal parking of shared cars.

[0067] The illegal parking identification method provided in this embodiment determines that the first identification result is compliant parking when the shared car is not in the compliant parking area if the current parking distance does not exceed the preset distance threshold, thereby improving the accuracy of the identification of illegal parking of the shared car. If the current parking distance exceeds the preset distance threshold, the first identification result is determined to be illegal parking, which can improve the efficiency of the identification of illegal parking of the shared car and save the identification cost of the identification of illegal parking of the shared car.

[0068] In one embodiment, if Figure 4 As shown, step S104, i.e., performing parking posture recognition according to the current environment image to determine a second recognition result, includes:

[0069] S401: performing image segmentation on the current environment image to obtain at least one segmented image and a segmentation recognition feature corresponding to each segmented image;

[0070] S402: When the segmentation recognition feature includes the target compliance feature, determining the segmentation image including the target compliance feature as the target segmentation image, and performing parking posture recognition according to the target segmentation image to determine a second recognition result;

[0071] S403: When none of the segmented recognition features include the target compliance feature, determining that the second recognition result is illegal parking.

[0072] The segmented image refers to the image generated after image segmentation and recognition of the current environment image. The segmented recognition features refer to the image features corresponding to the segmented image. The segmented recognition features obtained by segmentation include image features such as white lines in the parking area, curbs, and wheels.

[0073] As an example, in step S401, the controller performs image segmentation on the acquired current environment image of the shared car to obtain at least one segmented image, and identifies each segmented image to obtain image features corresponding to each segmented image, and uses the image features as segmentation recognition features. For example, the controller calls a pre-trained deep neural network-based image segmentation algorithm and a neural network acceleration unit to perform image segmentation and recognition on the current environment image, obtain a segmented image corresponding to the current environment image, and identify the segmentation recognition features corresponding to the segmented image, for example, to obtain segmentation recognition features including but not limited to image features such as white lines, curbs, and wheels in the parking area. In this example, image segmentation and recognition are performed on the current environment image to obtain a segmented image for illegal parking recognition and a segmentation recognition feature corresponding to the segmented image, so that it is feasible to obtain a second recognition result based on the segmented image in the subsequent process.

[0074] Among them, the target compliance feature refers to the segmentation recognition feature that meets the parking environment requirements of the shared car. For example, the target compliance features include curbs, fences, white lines in the parking area, and green belts. Understandably, if the segmentation recognition features corresponding to the segmented image include target compliance features such as curbs, fences, white lines in the parking area, or green belts, it indicates that the parking of the shared car is relatively standardized and meets the surrounding environment requirements for the parking of the shared car. The target segmentation image refers to a segmentation image containing the target compliance feature.

[0075] As an example, in step S402, after acquiring each segmented image and the segmented identification features corresponding to each segmented image, the controller first performs image feature recognition on the segmented identification features corresponding to each segmented image, determines whether the segmented image contains the target compliance feature, and when it is determined that there is a segmented image containing the target compliance feature, screens the segmented image containing the target compliance feature, determines the segmented image containing the target compliance feature as the target segmented image, and performs parking posture recognition based on the target segmented image to determine the second recognition result. For example, after performing image recognition on each segmented image, the controller acquires the segmented identification features corresponding to each segmented image, performs image feature recognition on the segmented identification features, determines whether there are image features such as curbs, fences, white lines in parking areas, and green belts in the segmented identification features, and when it is determined that there are any target compliance features among the image features such as curbs, fences, white lines in parking areas, and green belts in the segmented identification features, determines the segmented image containing the target compliance feature as the target segmented image, and further performs illegal parking recognition based on the target segmented image to obtain the second recognition result. In this example, it is determined whether the segmentation recognition feature corresponding to the segmented image includes the target compliance feature. If the segmentation recognition feature includes the target compliance feature, the segmented image including the target compliance feature is determined to be the target segmented image, and the second recognition result is determined based on the target segmented image, which can make the obtained second recognition result more accurate.

[0076] As an example, in step S403, after the controller performs image feature recognition on the segmentation recognition feature corresponding to the segmented image, if it is determined that the segmentation recognition feature does not include the target compliance feature, the controller directly determines the second recognition result as illegal parking. In this example, if the segmentation recognition feature does not include the target compliance feature, the second recognition result is directly determined to be illegal parking, which can not only achieve more accurate illegal parking recognition, but also improve recognition efficiency and save the recognition cost of illegal parking recognition.

[0077] The illegal parking identification method provided in this embodiment determines whether the segmentation identification feature corresponding to the segmentation image contains the target compliance feature. If the segmentation identification feature contains the target compliance feature, the segmentation image containing the target compliance feature is determined as the target segmentation image, and the parking posture recognition is performed according to the target segmentation image to determine the second recognition result, which can make the obtained second recognition result more accurate. If the segmentation identification feature does not contain the target compliance feature, the second recognition result is directly determined to be illegal parking, which can not only achieve more accurate illegal parking identification, but also improve the recognition efficiency and save the recognition cost of illegal parking identification.

[0078] In one embodiment, if Figure 5 As shown, step S402, i.e., performing parking posture recognition according to the target segmented image to determine a second recognition result, includes:

[0079] S501: Determine the wheel road surface type according to the target segmentation image;

[0080] S502: If the wheel road surface type is the target road surface type, the current parking posture of the shared vehicle is obtained, and a second recognition result is determined according to the current parking posture;

[0081] S503: If the wheel road surface type is not the target road surface type, determine that the second recognition result is illegal parking.

[0082] The wheel surface type refers to the type of the road surface where the wheels of the shared vehicle are currently parked. Understandably, due to the different parking positions of the shared vehicles, the wheel surface type corresponding to the road surface where the wheels are currently parked includes asphalt roads, cement roads, and road curbs paved with floor tiles.

[0083] As an example, in step S501, when the controller determines that there is a segmentation recognition feature that includes a target compliance feature, the controller determines the wheel road surface type under the parking position of the wheel according to the target segmentation image that includes the target compliance feature, so as to facilitate the subsequent determination of whether the wheel road surface type is the target road surface type. For example, the controller selects the segmentation image that includes the road surface under the wheel from the segmentation image that includes the target compliance feature, identifies the segmentation image of the road surface under the wheel, and determines the wheel road surface type.

[0084] Among them, the target road surface type refers to the road surface type used to determine whether the wheel road surface type meets the parking standards. In this example, the target road surface type is a road curb paved with floor tiles. It can be understood that for shared cars, their parking position when parked in compliance should be on the road curb paved with floor tiles on the side of the road, rather than on the road surface including asphalt roads and cement roads. Therefore, the target road surface type in this example is a road curb paved with floor tiles.

[0085] The current parking posture refers to the parking angle of the shared car and the parking distance between the shared car and the parking reference object in front of the wheel. The parking reference object refers to the reference object used to help users park in compliance with regulations, including reference objects such as white lines in the parking area.

[0086] As an example, in step S502, after the controller obtains the wheel pavement type according to the segmented image, it determines whether the wheel pavement type is the target pavement type. When it is determined that the wheel pavement type is the target pavement type, it continues to obtain the current parking posture of the shared car, and identifies the current parking posture of the shared car to determine whether the current parking posture of the shared car meets the corresponding requirements. According to whether the current parking posture of the shared car meets the corresponding requirements, the second recognition result is determined. For example, after the controller obtains the wheel pavement type, it determines whether the wheel pavement type is a curb paved with floor tiles. If it is determined that the wheel pavement type is a curb paved with floor tiles, the current parking posture including the parking angle of the shared car and the parking distance between the parking reference objects in front of the wheel is obtained, and the second recognition result is determined according to whether the parking angle and the parking distance between the parking reference objects in front of the wheel meet the corresponding parking requirements. In this example, when the wheel pavement type is the target pavement type, the current parking posture of the shared car is obtained again, and the second recognition result is determined according to the current parking posture, which can effectively improve the accuracy of illegal parking recognition.

[0087] As an example, in step S503, after the controller obtains the wheel pavement type according to the segmented image, it determines whether the wheel pavement type is the target pavement type. When it is determined that the wheel pavement type is not the target pavement type, the second recognition result is directly determined to be illegal parking. For example, after the controller obtains the wheel pavement type, it determines whether the wheel pavement type is a road curb paved with floor tiles. If it is determined that the wheel pavement type is a road curb such as an asphalt road or a cement road, but not a road curb paved with bricks, the second recognition result is directly determined to be illegal parking. Understandably, in order to prevent affecting normal road traffic, shared cars cannot be parked on the pavement of the road, and generally need to be parked on the road curb paved with bricks. Therefore, if it is identified that the wheel pavement type is not the target pavement type, the second recognition result can be directly determined to be illegal parking. In this example, when the wheel pavement type is not the target pavement type, the second recognition result is determined to be illegal parking, which can improve the recognition efficiency of illegal parking recognition and save the recognition cost of illegal parking recognition.

[0088] The illegal parking identification method provided in this embodiment determines the wheel-pavement type based on the target segmentation image containing the target compliance features, and when the wheel-pavement type is the target pavement type, obtains the current parking posture of the shared car, and determines the second identification result based on the current parking posture, which can effectively improve the accuracy of illegal parking identification. When the wheel-pavement type is not the target pavement type, the second identification result is determined to be illegal parking, which can improve the identification efficiency of illegal parking identification and save the identification cost of illegal parking identification.

[0089] In one embodiment, in step S502, the current parking posture includes a current parking angle between the shared vehicle and the parking reference object, and a current parking distance between the shared vehicle and the parking reference object.

[0090] In this embodiment, the current parking angle refers to the Y-direction angle between the shared vehicle and the parking reference object. The current parking distance refers to the distance between the front end of the shared vehicle's wheels and the parking reference object. Understandably, the Y-direction angle refers to the angle between the shared vehicle along the radial (Y-direction) direction of the vehicle body and the parking reference object. For example, Fig. 9 It is a schematic diagram of the current parking angle and the current parking distance between the shared car and the parking reference object. Among them, the current parking angle is determined by the angle between the parking reference object and the body of the shared car in the Y direction. The current parking distance is determined by the vertical distance between the front end of the shared car's wheel and the parking reference object. In this example, the angle between the radial direction (Y direction) of the shared car and the parking reference object is obtained, and the angle is determined as the current parking angle; the minimum bounding box including the shared car is determined, and the minimum bounding box is used as the wheel target box, and the vertical distance between the center of the top edge of the front of the wheel corresponding to the wheel target box and the parking reference object is used as the current parking distance. In this example, the parking reference objects include the white line of the parking area, the fence and the bottom edge of the curb.

[0091] In one embodiment, if Figure 6 As shown, step S502, i.e. determining a second recognition result according to the current parking posture, includes:

[0092] S601: Compare the current parking angle with the preset angle range, and compare the current parking distance with the body length of the shared car;

[0093] S602: If the current parking angle is within the preset angle range and the current parking distance is less than or equal to the vehicle body length, determining the second recognition result as compliant parking;

[0094] S603: If the current parking angle exceeds the preset angle range, and / or the current parking distance exceeds the vehicle body length, the second recognition result is determined to be illegal parking.

[0095] Among them, the preset angle range refers to the angle range used to determine whether the current parking angle meets the parking requirements for compliant parking. The body length refers to the radial (Y-direction) length of the shared car. Understandably, the body length is the maximum length of the shared car.

[0096] As an example, in step S601, after obtaining the current parking angle and the current parking distance between the shared car and the parking reference object, the controller compares the current parking angle with the preset angle range and the current parking distance with the body length of the shared car. In this example, the current parking angle is compared with the preset angle range and the current parking distance is compared with the body length, so that the subsequent determination of the second recognition result is feasible.

[0097] As an example, in step S602, when the controller determines that the current parking angle is within the preset angle range and the current parking distance is less than or equal to the body length of the shared vehicle, the controller determines that the second recognition result is a compliant parking. ° Up to 120 ° It is understandable that if the current parking angle is less than 60 ° , or the current parking angle is greater than 120 ° , it means that the current parking angle exceeds the preset angle range. ° Up to 120 ° In this example, the controller determines that the current parking angle does not exceed 60 degrees. ° Up to 120 °, and when the current parking distance is less than or equal to the vehicle body length, the second recognition result is determined to be compliant parking. In this example, determining the second recognition result based on the current parking angle and the current parking distance can improve the accuracy of illegal parking recognition, make the second recognition result more accurate, and make the second recognition result more consistent with the actual parking situation of the shared car.

[0098] As an example, in step S603, when the controller determines that the current parking angle exceeds the preset angle range and / or the current parking distance exceeds the length of the shared vehicle, the controller determines that the second recognition result is illegal parking. ° Up to 120 ° Understandably, if the current parking angle is less than 60 ° , or the current parking angle is greater than 120 ° , it indicates that the current parking angle exceeds the preset angle range, that is, only when the current parking angle is within 60 ° Up to 120 ° In this example, the controller determines that the current parking angle exceeds 60 degrees. ° Up to 120 ° , and / or, when the current parking distance does not exceed the vehicle body length, the second recognition result is determined to be illegal parking. In this example, determining the second recognition result based on the current parking angle and the current parking distance can improve the accuracy of illegal parking recognition, make the second recognition result more accurate, make the second recognition result more consistent with the actual parking situation of the shared car, and reduce the probability of misjudgment.

[0099] The illegal parking identification method provided in this embodiment compares the current parking angle with the preset angle range, compares the current parking distance with the body length of the shared car, and determines the second identification result based on the current parking angle and the current parking distance. This can improve the accuracy of illegal parking identification, make the second identification result more accurate, make the second identification result more in line with the actual parking situation of the shared car, and reduce the probability of misjudgment.

[0100] In another embodiment, if Figure 7 As shown, before step S103, that is, before obtaining the current environment image of the shared car, the illegal parking identification method also includes:

[0101] S701: Obtain the current driving status of the shared car;

[0102] S702: If the current driving state is the parking state, turn on the camera to collect the current environment image of the shared car through the camera;

[0103] The current driving state refers to whether the shared car is in a driving state. The current driving state includes the driving state and the parking state.

[0104] As an example, in step S701, the controller obtains the current driving state of the shared car, so as to determine the on and off state of the camera based on the current driving state of the shared car.

[0105] Among them, the camera refers to a camera device installed on a shared car to obtain images of the current environment.

[0106] As an example, in step S702, when the controller determines that the current driving state is the parking state, the controller controls the camera installed on the shared car to turn on, so as to collect the current environment image of the shared car through the camera. In this example, the camera is turned on after determining that the current driving state of the shared car is the parking state, and the current environment image of the shared car is collected through the camera, which can save power resources, avoid resource waste, and reduce management costs.

[0107] The illegal parking identification method provided in this embodiment turns on the camera only when the current driving state is the parking state, and collects the current environment image of the shared car through the camera. This method can save power resources, avoid resource waste, and reduce management costs.

[0108] In another embodiment, after step S104, that is, after determining the second recognition result, the illegal parking recognition method further includes:

[0109] If the second recognition result is compliant parking, and / or the camera is turned on for longer than a preset time threshold, the camera is turned off.

[0110] The preset time threshold refers to a preset threshold for the start time of the monitoring camera.

[0111] As an example, when the controller determines that the second recognition result is compliant parking and the camera on time does not exceed the preset time threshold, the controller controls the camera to turn off. It is understandable that if the shared car is parked in compliance, it means that there is no need to identify illegal parking. At this time, the camera can be turned off to save resources. Alternatively, the controller does not detect that the second recognition result is compliant parking, and the controller monitors that the camera on time has reached the preset time threshold. At this time, the controller controls the camera to turn off to save resources. It is understandable that when the camera on time reaches the preset time threshold, it indicates that the maximum single working time of the camera has been exceeded, and the camera can be turned off. Alternatively, the controller detects that the second recognition result is compliant parking, and the controller monitors that the camera on time has reached the preset time threshold. At this time, the controller controls the camera to turn off to save resources. In this example, when the second recognition result is compliant parking, and / or the camera on time exceeds the preset time threshold, turning off the camera can save resources and reduce management costs.

[0112] The illegal parking identification method provided in this embodiment can save resources, avoid resource waste, and reduce management costs by turning off the camera when the second identification result is compliant parking and / or the camera is turned on for more than a preset time threshold.

[0113] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0114] In one embodiment, a controller is provided. The controller may be a server, and its internal structure diagram may be as follows: Figure 8 As shown. The controller includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the controller is used to provide computing and control capabilities. The memory of the controller includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the controller is used to store data used or generated during the execution of the illegal parking identification method. The network interface of the controller is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for identifying illegal parking is implemented.

[0115] In one embodiment, a controller is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the illegal parking identification method in the above embodiment is implemented, for example Figure 1 S101-S105 as shown, or Figures 2 to 7To avoid repetition, it will not be described here.

[0116] In one embodiment, a parking violation identification system is provided, comprising a positioning module, a camera and the above-mentioned controller, wherein the controller is connected to the positioning module and the camera respectively; the positioning module is used to obtain the current location information of the shared car; the controller is used to identify the parking area according to the current location information and determine a first identification result;

[0117] The camera is used to obtain the current environment image of the shared car; the controller is also used to identify the parking posture according to the current environment image, determine the second recognition result, and realize the recognition of illegal parking of the shared car.

[0118] As an example, the illegal parking identification system includes a positioning module, a camera and a controller. In this example, the positioning module adopts a civilian positioning module, which is connected to the controller of the shared car. The positioning module is used to obtain the current location information of the shared car; the controller is used to identify the parking area according to the obtained current location information and obtain a first identification result. The camera adopts a high-definition USB camera and is installed under the front basket of the shared car. It is connected to the controller of the shared car and is used to collect the current environment image and transmit the current environment image to the controller of the shared car; the controller is used to analyze and process the current environment image, obtain the second identification result, and realize the identification of illegal parking of the shared car.

[0119] The illegal parking identification system provided in this embodiment uses a relatively simple hardware device, combined with current location information and current environment images, to achieve relatively accurate illegal parking identification of shared vehicles.

[0120] In one embodiment, a shared vehicle is provided, comprising the above-mentioned illegal parking identification system.

[0121] The shared car provided in this embodiment adopts the above-mentioned illegal parking identification system to implement the above-mentioned illegal parking identification method, and adopts a relatively simple hardware system to combine the current location information and the current environment image to identify the illegal parking of the shared car, which can improve the accuracy of illegal parking identification. In addition, there is no need to improve the illegal parking identification system, and no human participation is required, which can save hardware costs and management costs and avoid cost waste.

[0122] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0123] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for identifying illegal parking, characterized in that: include: Get the current location information of the shared car; Perform parking area identification according to the current location information to determine a first identification result; If the first recognition result is regional compliance parking, obtaining a current environment image of the shared vehicle; Perform parking posture recognition according to the current environment image to determine a second recognition result; If the second recognition result is illegal parking, an illegal parking reminder is given.

2. The method for identifying illegal parking according to claim 1, characterized in that: The step of identifying the parking area according to the current location information and determining a first identification result includes: Determine whether the shared vehicle is in a legal parking area according to the current location information; If the shared vehicle is in the compliant parking area, determining the first recognition result as compliant parking; If the shared vehicle is not in the compliant parking area, a current parking distance between the shared vehicle and the compliant parking area is obtained, and the first recognition result is determined based on the current parking distance.

3. The method for identifying illegal parking according to claim 2, characterized in that: The determining the first recognition result based on the current parking distance includes: If the current parking distance does not exceed the preset distance threshold, determining the first recognition result as compliant parking; If the current parking distance exceeds a preset distance threshold, the first recognition result is determined to be illegal parking.

4. The method for identifying illegal parking according to any one of claims 1 to 3, characterized in that: The performing parking posture recognition according to the current environment image to determine the second recognition result includes: Performing image segmentation on the current environment image to obtain at least one segmented image and a segmentation recognition feature corresponding to each segmented image; In the case where the segmentation recognition feature includes a target compliance feature, determining a segmented image including the target compliance feature as a target segmented image, and performing parking posture recognition according to the target segmented image to determine a second recognition result; In a case where the segmented identification feature does not include a target compliance feature, the second identification result is determined to be illegal parking.

5. The method for identifying illegal parking according to claim 4, characterized in that: The step of performing parking posture recognition according to the target segmented image to determine a second recognition result includes: Determining the wheel-pavement type according to the target segmentation image; If the wheel road surface type is the target road surface type, obtaining the current parking posture of the shared vehicle, and determining a second recognition result according to the current parking posture; If the wheel road surface type is not the target road surface type, the second recognition result is determined to be illegal parking.

6. The method for identifying illegal parking according to claim 5, characterized in that: The current parking posture includes a current parking angle between the shared vehicle and the parking reference object, and a current parking distance between the shared vehicle and the parking reference object; The determining a second recognition result according to the current parking posture includes: Compare the current parking angle with a preset angle range, and compare the current parking distance with the body length of the shared vehicle; If the current parking angle is within the preset angle range and the current parking distance is less than or equal to the vehicle body length, determining the second recognition result as compliant parking; If the current parking angle exceeds the preset angle range, and / or the current parking distance exceeds the vehicle body length, the second recognition result is determined to be illegal parking.

7. The illegal parking identification method according to any one of claims 1 to 6, characterized in that: Before obtaining the current environment image of the shared vehicle, the illegal parking identification method further includes: Obtaining the current driving status of the shared vehicle; If the current driving state is a parking state, the camera is turned on to collect a current environment image of the shared car through the camera; After determining the second recognition result, the illegal parking recognition method further includes: If the second recognition result is compliant parking, and / or the camera on time exceeds a preset time threshold, the camera is turned off.

8. A controller comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the illegal parking identification method as described in any one of claims 1 to 7 is implemented.

9. A parking violation identification system, characterized in that: It comprises a positioning module, a camera and the controller as claimed in claim 8, wherein the controller is connected to the positioning module and the camera respectively; the positioning module is used to obtain the current position information of the shared car; The controller is used to identify the parking area according to the current position information and determine a first identification result; The camera is used to obtain the current environment image of the shared car; The controller is also used to perform parking posture recognition based on the current environment image, determine a second recognition result, and realize the recognition of illegal parking of shared vehicles.

10. A shared car, characterized in that: Including the illegal parking identification system as described in claim 9.

Images