Non-motor vehicle management method, electronic equipment, storage medium and product

Abstract

The invention provides a non-motor vehicle management method, electronic equipment, a storage medium and a product. The non-motor vehicle management method comprises the following steps: acquiring a detection image of a target non-motor vehicle in a camera picture, and extracting a license plate image of the target non-motor vehicle according to the detection image; acquiring license plate information based on the license plate image, wherein the license plate information comprises license plate numbers and license plate colors; and managing the target non-motor vehicle according to the license plate information and the position of the non-motor vehicle, wherein the management comprises at least one of vehicle registration, registration information query and fire detection. The non-motor vehicle can be automatically detected and the license plate of the non-motor vehicle can be identified, the non-motor vehicle can be quickly managed according to the license plate, the speed is high, the efficiency is high, the labor cost is low, errors are not prone to occurring, and the non-motor vehicle management difficulty is effectively reduced.

Description

Technical Field

[0001] This application relates to the technical field of vehicle management. Specifically, this application relates to a non-motor vehicle management method, an electronic device, a storage medium, and a product. Background Art

[0002] With the increasing congestion of urban traffic, more and more people choose to use non-motor vehicles (such as electric vehicles and motorcycles) as the main means of travel. This method is not only convenient for traveling but also relatively convenient for parking, greatly alleviating the pressure on people during daily commuting. However, the popularization of non-motor vehicles has brought new management problems and potential safety hazards.

[0003] Due to the small size of non-motor vehicles, limited parking areas, and the large number of non-motor vehicles in use, the problem of random parking of non-motor vehicles has become increasingly serious. The random parking of non-motor vehicles not only affects the overall environment of the community or park but also increases the management difficulty. In many scenarios, manual license plate recognition and non-motor vehicle management based on license plates are adopted. This management method is inefficient, time-consuming, has a high labor cost, and is prone to omissions, increasing the difficulty of non-motor vehicle management. Summary of the Invention

[0004] In view of the shortcomings of the existing methods, this application proposes a non-motor vehicle management method, an electronic device, a storage medium, and a product, which can solve the problems of low efficiency, high labor cost, and easy omission of the existing non-motor vehicle management methods, and improve the management difficulty.

[0005] According to one aspect of the embodiments of this application, the embodiments of this application provide a non-motor vehicle management method, the method including:

[0006] Obtain a detection image of a target non-motor vehicle in a camera image, and extract a license plate image of the target non-motor vehicle according to the detection image;

[0007] Obtain license plate information based on the license plate image, where the license plate information includes a license plate number and a license plate color;

[0008] Manage the target non-motor vehicle according to the license plate information and the location of the non-motor vehicle, where the management includes at least one of vehicle registration, registration information query, and fire detection.

[0009] In a possible implementation, the obtaining a detection image of a target non-motor vehicle in a camera image includes:

[0010] Obtain a camera image transmitted by the camera;

[0011] If it is determined that a non-motor vehicle is detected in the camera image, the attributes of the non-motor vehicle are identified, and a detection image of the target non-motor vehicle is obtained according to the attributes, where the attributes include vehicle type and color.

[0012] In a possible implementation, the obtaining the detection image of the target non-motor vehicle according to the attributes includes:

[0013] If it is determined according to the vehicle type that the non-motor vehicle does not belong to a bicycle, it is determined that the non-motor vehicle is the target non-motor vehicle, and the detection image of the target non-motor vehicle is obtained according to the target box of the non-motor vehicle, where the target box is determined based on the detection result of the non-motor vehicle.

[0014] In a possible implementation, the extracting the license plate image of the target non-motor vehicle from the detection image includes:

[0015] Input the detection image into a license plate key point detection model to obtain a license plate target box and license plate category information;

[0016] Obtain the license plate image according to the license plate target box and the license plate category.

[0017] In a possible implementation, the license plate category includes a single-layer license plate and a double-layer license plate, and the obtaining the license plate image according to the license plate target box and the license plate category includes:

[0018] Convert the coordinates of the preset key points of the license plate target box to preset coordinates to obtain a corrected image;

[0019] If it is determined that the license plate category is a double-layer license plate, cut the corrected image based on the cutting parameters corresponding to the license plate category to obtain a cut image;

[0020] Stitch the cut images to generate a license plate image.

[0021] In a possible implementation, the obtaining the license plate information based on the license plate image includes:

[0022] Train a preset neural network to obtain a license plate recognition model, where the preset neural network includes an LPRNet network, and the LPRNet network includes a license plate color recognition branch, and the license plate color branch recognizes the license plate color based on the output of a Relu activation function at a preset position;

[0023] Input the license plate image into the license plate recognition model to obtain the license plate information.

[0024] In a possible implementation, the preset neural network is an LPRNet network, and the training of the license plate recognition model includes:

[0025] The model is trained based on a weighted cross - entropy loss function, and the weighted cross - entropy loss function is as follows:

[0026] Loss color =-(0.15y white ·log(p white ) + 0.15y yellow ·log(p yellow )

[0027] +0.35y green ·log(p green ) + 0.35y blue ·log(p blue ))

[0028] In the formula, y white is the true label of the white label, p white is the predicted value of the white label output by the license plate recognition model, y yellow is the true label of the yellow label, p yellow is the predicted value of the yellow label output by the license plate recognition model, y green is the true label of the green label, p green is the predicted value of the green label output by the license plate recognition model, y blue is the true label of the blue label, p blue is the predicted value of the blue label output by the license plate recognition model.

[0029] According to one aspect of the embodiments of the present application, an electronic device is provided, including a memory, a processor, and a computer program stored on the memory. The processor executes the computer program to implement the steps of the above - mentioned method.

[0030] According to one aspect of the embodiments of the present application, a computer - readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above - mentioned method are implemented.

[0031] According to one aspect of the embodiments of the present application, a computer program product is provided, which includes a computer program. When the computer program is executed by a processor, the steps of the above - mentioned method are implemented.

[0032] The beneficial technical effects brought by the technical solution provided by the embodiments of the present application include:

[0033] A non-motor vehicle management method provided by the present application has the beneficial effects that it acquires a detection image of a target non-motor vehicle in a camera image, extracts a license plate image of the target non-motor vehicle according to the detection image; obtains license plate information based on the vehicle image, and the license plate information includes the license plate number and the license plate color; manages the target non-motor vehicle according to the license plate information and the location where the non-motor vehicle is located, and the management includes at least one of vehicle registration, registration information query, and fire detection. The present application can automatically detect non-motor vehicles and recognize the license plates of non-motor vehicles, and quickly manage non-motor vehicles according to the license plates, with high speed, high efficiency, low labor cost and not easy to make mistakes, effectively reducing the difficulty of non-motor vehicle management.

[0034] Additional aspects and advantages of the present application will be given in part in the following description, which will become apparent from the following description, or can be understood through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The above and / or additional aspects and advantages of the present application will become apparent and easy to understand from the following description of the embodiments in conjunction with the drawings, where:

[0036] Figure 1 is a flowchart of the non-motor vehicle management method provided by the embodiment of the present application;

[0037] Figure 2 is a working flowchart of the non-motor vehicle management provided by the embodiment of the present application;

[0038] Figure 3 is a structural diagram of the electronic device provided by the embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0039] The embodiments of the present application will be described below with reference to the drawings in the present application. It should be understood that the embodiments described below in conjunction with the drawings are exemplary descriptions for explaining the technical solutions of the embodiments of the present application, and do not constitute limitations on the technical solutions of the embodiments of the present application.

[0040] Those skilled in the art can understand that, unless specifically stated, the "the" and "said" used herein may also include the plural form. It should be further understood that the term "comprising" used in the specification of this application means the presence of the described features, integers, steps, operations, elements, and / or components, but does not exclude the presence of other features, information, data, steps, operations, elements, components, and / or combinations thereof supported by the art. It should be understood that when we say that an element is "connected" or "coupled" to another element, the one element can be directly connected or coupled to the other element, or it can mean that the one element and the other element establish a connection relationship through an intermediate element. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein means at least one of the items defined by the term, for example, "A and / or B" can be implemented as "A", or implemented as "B", or implemented as "A and B".

[0041] To make the objectives, technical solutions, and advantages of this application clearer, the following will further describe the embodiments of this application in detail with reference to the accompanying drawings.

[0042] The embodiment of this application provides a non-motor vehicle management method, which can be used in mobile phones, computers, servers, clouds, and other terminals capable of non-motor vehicle management.

[0043] As Figure 1 、 Figure 2 shown, the non-motor vehicle management method of this application includes:

[0044] S101: Obtain the detection image of the target non-motor vehicle in the camera image, and extract the license plate image of the target non-motor vehicle according to the detection image.

[0045] Optionally, the camera can be installed in various areas where non-motor vehicles need to be managed, such as communities and industrial parks, or at the vehicle entrances and exits of these areas. The camera captures the vehicles or objects that enter the shooting range to generate a camera image.

[0046] Optionally, the camera can collect images regularly, or can collect images after detecting that the current image acquisition condition is met to generate a camera image, where the image acquisition condition can be at least one of being in the image acquisition time period, receiving an image acquisition instruction, the area currently captured by the camera belonging to the image acquisition area, and there being a vehicle in the predetermined area.

[0047] Optionally, the image captured by the camera can be directly transmitted to the terminal executing the method of this application, or the camera can detect the non-motor vehicles in the camera image, and when the detection image of the target non-motor vehicle is obtained, transmit the detection image to the terminal.

[0048] Optionally, the camera image can be transmitted to the terminal by wired or wireless transmission. The camera image can be transmitted to the terminal in real time, at regular intervals, or after receiving a screen transmission instruction sent by the terminal, and then the camera image is transmitted.

[0049] Optionally, obtaining a detection image of a target non-motor vehicle in the camera image includes: obtaining the camera image transmitted by the camera; if it is determined that a non-motor vehicle is detected in the camera image, identifying the attributes of the non-motor vehicle, and obtaining the detection image of the target non-motor vehicle according to the attributes, where the attributes include vehicle type and color.

[0050] Optionally, a non-motor vehicle detection model can be used for the detection of non-motor vehicles. The neural network used by the non-motor vehicle detection model can be YoloV5, YoloV8, YoloV11, DETR, and other models that can be used for neural network detection. When performing model training, images related to non-motor vehicles can be used as samples to train the non-motor vehicle detection model.

[0051] In one embodiment, a non-motor vehicle detection model is used to detect non-motor vehicles in the camera image. After detecting a non-motor vehicle, the non-motor vehicle detection model can use a target box to identify the non-motor vehicle in the camera image, and the identified non-motor vehicle is located within the target box. Obtain the camera image with the target box, and crop out the image of the non-motor vehicle through the target box in the camera image.

[0052] Optionally, obtaining the detection image of the target non-motor vehicle according to the attributes includes: if it is determined that the non-motor vehicle does not belong to a bicycle according to the vehicle type, determining that the non-motor vehicle is the target non-motor vehicle, and obtaining the detection image of the target non-motor vehicle according to the target box of the non-motor vehicle, where the target box is determined based on the detection result of the non-motor vehicle.

[0053] Optionally, the types include bicycles, animal-drawn vehicles, motorcycles, wheelchairs, tricycles, electric bicycles, and other types belonging to the category of non-motor vehicles.

[0054] Optionally, a pre-trained non-motor vehicle attribute recognition model can be used to obtain the type and color of non-motor vehicles. When performing recognition, the image of the non-motor vehicle can be input into the non-motor vehicle attribute recognition model, and the vehicle type and color of the non-motor vehicle can be determined through the recognition information output by the non-motor vehicle attribute recognition model.

[0055] Optionally, the license plate vertical recognition model can be VGG, ResNet18, ResNet50, and other models that can recognize non-motor vehicle attributes.

[0056] In one embodiment, non-motor vehicle types are classified. If the category of the non-motor vehicle is a bicycle, license plate recognition and subsequent operations are not performed. There are two reasons for this processing method: 1. Bicycles do not have license plates; 2. Bicycles do not spontaneously combust and have almost no safety hazards. When the vehicle type is recognized as a motorcycle, an electric vehicle, or other categories with license plates or capable of spontaneous combustion, subsequent license plate recognition operations are performed, and vehicle management is carried out based on the license plate.

[0057] Optionally, when it is determined that the non-motor vehicle in the image is a bicycle based on the vehicle type, subsequent operations may not be performed. It is also possible to detect whether the color of the non-motor vehicle is a preset color (such as a color that cannot be parked) when it is determined that the non-motor vehicle in the image is a bicycle based on the vehicle type. If so, a preset manager is reminded to clean up or move the non-motor vehicle.

[0058] Optionally, when non-motor vehicles are stored in different colors, if it is recognized that the non-motor vehicle is not a bicycle, the color of the non-motor vehicle can also be obtained, and its storage location is determined based on the color. If the storage location is the same as the current location of the non-motor vehicle, the next step of processing is continued. If not, a preset manager or the person using the non-motor vehicle can be reminded.

[0059] Optionally, when it is recognized that the non-motor vehicle is not a bicycle, it is also possible to obtain whether the color of the position where the license plate is placed on the non-motor vehicle (such as the front or rear of the vehicle) includes a specified color (such as the color of the license plate). If so, subsequent processing is performed. If not, a preset manager can be provided or an image of the non-motor vehicle can be stored for the manager's manual management.

[0060] Optionally, the license plate image of the target non-motor vehicle is extracted from the detected image, including: inputting the detected image into a license plate key point detection model to obtain a license plate target box and license plate category information; obtaining the license plate image according to the license plate target box and license plate category. The boundary of the license plate corresponds to the license plate target box.

[0061] Optionally, the license plate key point detection model can be a model such as YoloV5, YoloV8, PlateDetect, YoloV11, etc. that can be used to recognize the license plates of non-motor vehicles.

[0062] Optionally, the license plate category includes a single-layer license plate and a double-layer license plate. Obtaining the license plate image according to the license plate target box and license plate category includes: converting the coordinates of the preset key points of the license plate target box into preset coordinates to obtain a corrected image; if it is determined that the license plate category is a double-layer license plate, the corrected image is cut based on the cutting parameters corresponding to the license plate category to obtain a cut image; the cut images are spliced to generate a license plate image.

[0063] Optionally, the preset key points can be the coordinates of the four corners of the license plate target box, or the coordinates of the center points of each font within the target box.

[0064] In one embodiment, the image of the target non-motor vehicle is input into the license plate key point detection model, which identifies the license plate of the target non-motor vehicle and outputs the license plate target box [x p1 ,y p1 ,x p2 ,y p2 ; license plate category L p ; confidence Conf p ; the coordinates of the key point at the upper left corner of the license plate (x ul ,y ul ), the coordinates of the key point at the upper right corner (x ur ,y ur ), the coordinates of the key point at the lower right corner (x lr ,y lr ), and the coordinates of the key point at the lower left corner (x ll ,y ll ). x p1 can be the coordinates of the first vertex of the license plate target box (which can be the vertex at the upper left corner), y p1 can be the coordinates of the second vertex of the license plate target box (which can be the vertex at the lower left corner), x p2 can be the coordinates of the third vertex of the license plate target box (which can be the vertex at the lower right corner), y p2 can be the coordinates of the fourth vertex of the license plate target box (which can be the vertex at the upper right corner). The license plate category includes two categories, namely single-layer license plates and double-layer license plates.

[0065] Optionally, the coordinates of the preset key points can be the same as the vertex coordinates of the license plate target box.

[0066] Optionally, from the camera's perspective, the license plate in the license plate image is tilted, and ordinary object detection can only crop out the tilted license plate, which is not conducive to subsequent recognition. Therefore, the four preset key points of the license plate are used for correction. Specifically, the perspective transformation of opencv can be used for correction.

[0067] In one embodiment, the size of the corrected image corresponds to the position of the preset coordinates. Specifically, the size of the corrected image can be (280, 100), and the coordinate transformation method is as follows:

[0068] (x ul ,y ul ) → (0, 0)

[0069] (x ur ,y ur ) → (280, 0)

[0070] (x lr ,y lr )→(280,100)

[0071] (x ll ,y ll )→(0,100)

[0072] Optionally, when the license plate is a double-layer license plate, the characters in the first row of the license plate can be centered, and the characters in the second row can be composed of numbers and letters. When cutting the corrected image, these two rows can be separated, and the separated images can be spliced to obtain the license plate image.

[0073] Optionally, the cutting parameters of the corrected image can be determined according to the height of the two rows of characters in the corrected image. It can also be determined according to the double-layer license plate size information of the area where the current target non-motor vehicle is located.

[0074] In one embodiment, the img p image is cut, and two images will be cut out. Image 1 is intercepted according to the parameters of the upper left corner coordinates (56, 0) of the license plate target box, width 168, and height 50. The intercepted image is:

[0075] img p1 = img p [0:50, 56:224]

[0076] Image 2 is intercepted according to the parameters of the upper left corner coordinates (0, 30) of the license plate target box, width 280, and height 70. The intercepted image is:

[0077] img p2 = img p [30:100, 0:280]

[0078] Scale img p1 to (63, 48), and scale img p2 to (105, 48). Finally, place img p1 on the left and img p2 on the right, and splice them into an image img pf of (168, 48).

[0079] S102: Obtain license plate information based on the license plate image.

[0080] Optionally, the license plate information includes the license plate number and the license plate color. Obtaining license plate information based on the vehicle image includes: training a preset neural network to obtain a license plate recognition model, where the preset neural network includes the LPRNet network, and the LPRNet network includes a license plate color recognition branch, and the license plate color branch recognizes the license plate color based on the output of the Relu activation function at a preset position; inputting the license plate image into the license plate recognition model to obtain the license plate information.

[0081] Optionally, the license plate recognition model can also be PaddleOCR with a license plate color recognition branch added, and the license plate color recognition branch can be provided with at least one of a fully connected layer, a convolutional layer, and a transformer.

[0082] In one embodiment, the preset position is the penultimate Relu activation function, and the output of this Relu activation function is provided to the license plate color recognition branch. Among them, the license plate recognition branch can include a convolutional layer and a fully connected layer. During training, multiple images including license plates can be used for model training. And during the training process, all the weight parameters of the LPRNet model can be frozen, and only the weight parameters of the license plate recognition branch are trained.

[0083] Optionally, the preset neural network is the LPRNet network, and the training of the license plate recognition model includes:

[0084] Model training is performed based on the weighted cross-entropy loss function, and the weighted cross-entropy loss function is:

[0085] Loss color = -(0.15y white · log(p white ) + 0.15y yellow · log(p yellow ))

[0086] + 0.35y green · log(p green ) + 0.35y blue · log(p blue ))

[0087] In the formula, y white is the true label of the white label, p white is the predicted value of the white label output by the license plate recognition model, y yellow is the true label of the yellow label, p yellow is the predicted value of the yellow label output by the license plate recognition model, y green is the true label of the green label, p green is the predicted value of the green label output by the license plate recognition model, y blue is the true label of the blue label, pblue It is the predicted value of the blue label output by the license plate recognition model.

[0088] Optionally, for the weighted cross-entropy loss function, the weights of the white label, yellow label, green label, and blue label (i.e., 0.15 and 0.35 in the formula) can be determined according to the number of white license plates, yellow license plates, green license plates, and blue license plates in the dataset used to train the license plate recognition model during the training of the license plate recognition model. Among them, for license plates with a smaller number, their weights are larger than those of license plates with a larger number. By adjusting the weights (which can be adjusted to other values than 0.15 and 0.35), the degree to which the model focuses on each category is balanced, thereby improving the overall color recognition accuracy.

[0089] S103: Manage the target non-motor vehicle according to the license plate information and the location of the non-motor vehicle.

[0090] Optionally, the management includes at least one of vehicle registration, registration information query, and fire detection.

[0091] Optionally, after obtaining the license plate information, the license plate number can be input into the database to detect whether the license plate number in the license plate information is stored in the database. If so, it is determined that the target non-motor vehicle is a registered vehicle; if not, it is determined that the target non-motor vehicle is an illegal vehicle or a dangerous vehicle. Among them, the database can be a non-motor vehicle information database, and the information in this database can be manually registered or transmitted by external devices.

[0092] Optionally, when it is determined that the license plate number of the target non-motor vehicle fails to match and it is determined to be an illegal vehicle or a dangerous vehicle, it can be intercepted or the person using the target non-motor vehicle can be reminded not to park in the current area.

[0093] Optionally, after determining vehicle registration based on the license plate number, it can also be detected whether the vehicle type, license plate color, and vehicle color of the target non-motor vehicle are consistent with the information registered in the database. If so, it is determined that the registration information detection passes and the identity verification is successful; if one or more of them are different, the contact object corresponding to the license plate number in the database can be used or it can be marked as an illegal vehicle or a dangerous vehicle, thereby preventing the occurrence of license plate cloning.

[0094] Optionally, when it is determined that the license plate number does not exist, the license plate number, vehicle type, license plate color, and vehicle color of the target non-motor vehicle can also be stored in the database, or the information of the target non-motor vehicle can be sent to a preset management personnel to remind them to handle it.

[0095] Optionally, when determining that the target non-motor vehicle has a risk of spontaneous combustion according to the vehicle type, after the license plate number is recognized, the camera image with the license plate number can be transmitted to the fire detection model, and the fire detection model can be used for fire detection. In this way, the situation of spontaneous combustion of non-motor vehicles can be detected in the first time, and a warning can be sent to remind relevant personnel to rescue; also, according to the location of the fire, the fire can be traced back to quickly investigate which non-motor vehicle has spontaneous combustion.

[0096] Optionally, the fire detection model can be YoloV5, YoloV8, YoloV11, DETR, and other types of models that can be used for fire detection.

[0097] The non-motor vehicle management method provided by the embodiments of the present application obtains the detection image of the target non-motor vehicle in the camera image, and extracts the license plate image of the target non-motor vehicle according to the detection image; obtains the license plate information based on the vehicle image, and the license plate information includes the license plate number and the license plate color; manages the target non-motor vehicle according to the license plate information and the location of the non-motor vehicle, and the management includes at least one of vehicle registration, registration information query, and fire detection. The present application can automatically detect non-motor vehicles and identify the license plates of non-motor vehicles, and quickly manage non-motor vehicles according to the license plates, with high speed, high efficiency, low labor cost and not easy to make mistakes, effectively reducing the difficulty of non-motor vehicle management.

[0098] In an alternative embodiment, an electronic device is provided, as Figure 3 shown. Figure 3 The electronic device 4000 shown in the figure includes: a processor 4001 and a memory 4003. Among them, the processor 4001 and the memory 4003 are connected, such as connected through a bus 4002. Optionally, the electronic device 4000 may further include a transceiver 4004, and the transceiver 4004 can be used for data interaction between the electronic device and other electronic devices, such as sending and / or receiving data, etc. It should be noted that in practical applications, the transceiver 4004 is not limited to one, and the structure of the electronic device 4000 does not constitute a limitation to the embodiments of the present application.

[0099] The processor 4001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logical blocks, modules, and circuits described in connection with the disclosure of this application. The processor 4001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0100] The bus 4002 may include a path for transmitting information between the above components. The bus 4002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus 4002 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only a thick line is shown in the figure, but it does not mean that there is only one bus or one type of bus.

[0101] The memory 4003 may be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or it may also be an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory), or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), magnetic disk storage media, other magnetic storage devices, or any other medium that can be used to carry or store computer programs and can be read by a computer, which is not limited herein.

[0102] The memory 4003 is used to store the computer program for implementing the embodiments of this application and is controlled by the processor 4001 for execution. The processor 4001 is used to execute the computer program stored in the memory 4003 to implement the steps shown in the foregoing method embodiments.

[0103] Among them, the electronic device can be any kind of electronic product that can perform human-computer interaction with an object. For example, a personal computer, a tablet computer, a smart phone, a personal digital assistant (PDA), a game console, an Internet Protocol Television (IPTV), a smart wearable device, etc.

[0104] The electronic device may further include a network device and / or an object device. Among them, the network device includes, but is not limited to, a single network server, a server group composed of multiple network servers, or a cloud composed of a large number of hosts or network servers based on cloud computing.

[0105] The network where the electronic device is located includes, but is not limited to, the Internet, a wide area network, a metropolitan area network, a local area network, a virtual private network (VPN), etc.

[0106] The embodiment of the present application provides a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the steps and corresponding contents of the foregoing method embodiment can be implemented.

[0107] The embodiment of the present application also provides a computer program product. The computer program product includes a computer program. When the computer program is executed by a processor, the steps and corresponding contents of the foregoing method embodiment can be implemented.

[0108] Those skilled in the art of the present technology can understand that the steps, measures, and solutions in the various operations, methods, and processes discussed in the present application can be alternated, changed, combined, or deleted. Further, the other steps, measures, and solutions in the various operations, methods, and processes discussed in the present application can also be alternated, changed, rearranged, decomposed, combined, or deleted. Further, the steps, measures, and solutions in the related technologies that are the same as those disclosed in the present application can also be alternated, changed, rearranged, decomposed, combined, or deleted.

[0109] In the description of the present application, the directions or positional relationships indicated by the words "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. are the exemplary directions or positional relationships based on the drawings, which are for the convenience of describing or simplifying the embodiments of the present application, rather than indicating or implying that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation to the present application.

[0110] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of this application, unless otherwise specified, the meaning of "a plurality" is two or more.

[0111] In the description of this application, it should be noted that, unless otherwise clearly specified and defined, the terms "mounted", "connected" and "coupled" should be construed broadly. For example, it may be a fixed connection, a detachable connection or an integral connection; it may be a direct connection or an indirect connection through an intermediate medium, and it may be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0112] In the description of this specification, specific features, structures, materials or characteristics may be combined in any one or more embodiments or examples in a suitable manner.

[0113] The above are only some embodiments of this application. It should be pointed out that for those of ordinary skill in the art, without departing from the technical concept of the solution of this application, other similar implementation means based on the technical idea of this application also belong to the protection scope of the embodiments of this application.

Claims

1. A non-motor vehicle management method, characterized in that: The method comprises: Acquire a detection image of a target non-motor vehicle in a camera image, and extract a license plate image of the target non-motor vehicle based on the detection image; Acquire license plate information based on the license plate image, where the license plate information includes a license plate number and a license plate color; The target non-motor vehicle is managed according to the license plate information and the location of the non-motor vehicle, and the management includes at least one of vehicle registration, registration information query, and fire detection.

2. The non-motor vehicle management method according to claim 1, characterized in that: The step of obtaining a detection image of a target non-motor vehicle in a camera image includes: Get the camera image transmitted by the camera; If it is determined that a non-motor vehicle is detected from the camera image, the attributes of the non-motor vehicle are identified, and a detection image of the target non-motor vehicle is acquired according to the attributes, wherein the attributes include vehicle type and color.

3. The non-motor vehicle management method according to claim 2, characterized in that: The step of acquiring a detection image of the target non-motor vehicle according to the attribute comprises: If it is determined according to the vehicle type that the non-motor vehicle is not a bicycle, the non-motor vehicle is determined to be a target non-motor vehicle, and a detection image of the target non-motor vehicle is obtained according to a target frame of the non-motor vehicle, wherein the target frame is determined based on the detection result of the non-motor vehicle.

4. The non-motor vehicle management method according to claim 1, characterized in that: The step of extracting the license plate image of the target non-motor vehicle according to the detection image comprises: Input the detection image into the license plate key point detection model to obtain the license plate target frame and license plate category information; A license plate image is acquired according to the license plate target frame and the license plate category.

5. The non-motor vehicle management method according to claim 4, characterized in that: The license plate categories include single-layer license plates and double-layer license plates, and acquiring a license plate image according to the license plate target frame and the license plate category includes: Converting the coordinates of the preset key points of the license plate target frame into preset coordinates to obtain a corrected image; If it is determined that the license plate type is a double-layer license plate, cutting the corrected image based on the cutting parameters corresponding to the license plate type to obtain a cut image; The cut images are spliced ​​to generate a license plate image.

6. The non-motor vehicle management method according to claim 1, characterized in that: The acquiring the license plate information based on the license plate image comprises: Training a preset neural network to obtain a license plate recognition model, wherein the preset neural network includes an LPRNet network, and the LPRNet network includes a license plate color recognition branch, and the license plate color branch recognizes the license plate color based on the output of a Relu activation function at a preset position; The license plate image is input into the license plate recognition model to obtain the license plate information.

7. The non-motor vehicle management method according to claim 6, characterized in that: The preset neural network is an LPRNet network, and the training of the license plate recognition model includes: The model training is performed based on the weighted cross entropy loss function, which is: Loss color =-(0.15y white ·log(p white )+0.15y yellow ·log(p yellow )+0.35y green ·log(p green )+0.35y blue ·log(p blue )) In the formula, y white Is the real label of white label, p white is the white label prediction value output by the license plate recognition model, y yellow Is the real label of the yellow label, p yellow is the yellow label prediction value output by the license plate recognition model, y green is the true label of the green label, p green is the green label prediction value output by the license plate recognition model, y blue is the true label of the blue label, p blue is the blue label prediction output by the license plate recognition model.

8. An electronic device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images