High-precision emergency stop strip intelligent perception analysis method, device and electronic equipment

Abstract

The present application provides a high-precision intelligent perception analysis method, device and electronic device for emergency parking strips, the method comprising: obtaining a parking strip image at the current moment, determining at least one target vehicle information based on a first image recognition model, and determining at least one license plate information based on a second image recognition model, determining the vehicle tracking information of the vehicle at the current moment based on the target vehicle information and the vehicle tracking information before the current moment, binding the vehicle tracking information and the license plate information, generating multiple vehicle binding information, generating a control signal based on the vehicle binding information, and sending the control signal to a control device set in the parking strip area, so that the control device outputs prompt information according to the control signal. The present application counts and manages the entire life cycle of vehicles in the emergency parking strip, thereby effectively driving the control device to make more timely and effective linkage and control.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a high-precision emergency parking strip intelligent perception analysis method, device and electronic equipment. Background Art

[0002] With the rapid development of highway construction in my country, the number of tunnels is also increasing. Tunnels are the most complex sections of highways, and are prone to traffic accidents. Accidents are difficult to handle, posing a great threat to highway driving safety. Therefore, emergency parking strips are generally set up in highway tunnels for temporary parking of vehicles when a vehicle breaks down, to ensure smooth highway traffic and reduce interference from accident vehicles and faulty vehicles.

[0003] Current parking strip systems usually use radar detection, infrared detection, geomagnetic detection and other detection methods to perform full life cycle detection on vehicles in emergency parking strips.

[0004] However, the equipment cost required for radar detection is too high. Infrared detection not only has high equipment cost, but is also easily affected by the surrounding environment. Geomagnetic detection requires pre-embedded geomagnetic detection equipment in the emergency parking lane. If there is no early planning, the road surface construction and paving will need to be re-performed, affecting traffic operations. Summary of the invention

[0005] The purpose of this application is to provide a high-precision emergency stop strip intelligent perception analysis method, device and electronic equipment to address the shortcomings of the above-mentioned prior art, so as to solve the problem of high equipment cost in the prior art.

[0006] To achieve the above objectives, the technical solutions adopted in this application are as follows:

[0007] In a first aspect, the present application provides a high-precision intelligent perception analysis method for emergency parking strips, the method comprising:

[0008] Obtaining the parking strip image at the current moment sent by the image acquisition device;

[0009] Determine at least one target vehicle information in the parking strip image based on a pre-trained first image recognition model, and determine at least one license plate information in the parking strip image based on a pre-trained second image recognition model, wherein the target vehicle information includes the number of vehicles, the vehicle image, and the vehicle position, and the license plate information includes the license plate number, the license plate position, the license plate type, and the license plate collection time;

[0010] Determine the vehicle tracking information of the vehicle at the current moment according to the target vehicle information in the parking strip image and the vehicle tracking information before the current moment, wherein the vehicle tracking information includes a vehicle tracking identifier, a vehicle type and vehicle driving information, and the vehicle driving information includes a vehicle position and a collection time corresponding to the vehicle position;

[0011] Binding the vehicle tracking information and the license plate information to generate a plurality of vehicle binding information, each of the vehicle binding information includes: the vehicle tracking information and the corresponding license plate information;

[0012] A control signal is generated according to the vehicle binding information, and the control signal is sent to a plurality of control devices arranged in the parking strip area, so that the control devices output prompt information according to the control signal.

[0013] Optionally, the binding the vehicle tracking information and the license plate information to generate vehicle binding information includes:

[0014] Traverse the collection time of each license plate in the first license plate information, and for the traversed current license plate collection time, if the collection time corresponding to the current license plate collection time and the first vehicle position in the first vehicle driving information is less than a preset time threshold, determine whether the distance between the license plate position corresponding to the current license plate collection time in the first license plate information and the first vehicle position in the first vehicle driving information is less than a preset distance. If so, bind the first license plate information and the vehicle tracking information corresponding to the first vehicle driving information to generate vehicle binding information corresponding to the first license plate information.

[0015] Optionally, the binding the vehicle tracking information and the license plate information to generate vehicle binding information includes:

[0016] Traverse the collection time of each license plate in the first license plate information, and for the current license plate collection time collected, if the collection time corresponding to the current license plate collection time and the first vehicle position in the first vehicle driving information is less than a preset time threshold, determine whether the distance between the license plate position corresponding to the current license plate collection time in the first license plate information and the first vehicle position in the first vehicle driving information is less than a preset distance; if so, determine whether the license plate type of the first license plate information and the vehicle type in the first vehicle driving information satisfy a preset matching relationship; if so, bind the first license plate information and the vehicle tracking information corresponding to the first vehicle driving information to generate vehicle binding information corresponding to the first license plate information.

[0017] Optionally, after binding the vehicle tracking information and the license plate information to generate vehicle binding information, the method further includes:

[0018] Traversing each vehicle binding information, for the traversed first vehicle binding information and second vehicle binding information, determining a collection distance between a vehicle position with the latest collection time in the first vehicle binding information and a vehicle position with the earliest collection time in the second vehicle binding information;

[0019] Determine whether the acquisition distance is less than a preset distance threshold;

[0020] If yes, determining whether the vehicle type in the first vehicle binding information and the vehicle type in the second vehicle binding information satisfy a preset matching relationship;

[0021] If so, the vehicle tracking identifier in the first vehicle binding information and the license plate number and license plate type in the license plate information are used as the vehicle tracking identifier in the second vehicle binding information and the license plate number and license plate type in the license plate information.

[0022] Optionally, after binding the vehicle tracking information and the license plate information to generate vehicle binding information, the method further includes:

[0023] Based on a multi-source shortest path algorithm, determining a first binding result according to the vehicle position with the latest collection time in each of the vehicle binding information and the vehicle position with the earliest collection time in each of the vehicle binding information, wherein the first binding result includes a plurality of vehicle binding information corresponding to each other;

[0024] Determine a target binding result according to the vehicle type in each vehicle binding information in the first binding result, wherein the target binding result includes at least one vehicle binding information before the change and vehicle binding information after the change corresponding to the vehicle binding information before the change;

[0025] The vehicle tracking identifier in each vehicle binding information before the change and the license plate number and license plate type in the license plate information are used as the vehicle tracking identifier in each vehicle binding information after the change and the license plate number and license plate type in the license plate information.

[0026] Optionally, after binding the vehicle tracking information and the license plate information to generate vehicle binding information, the method further includes:

[0027] Determining whether there is a lost vehicle based on the vehicle binding information;

[0028] If so, the binding information of the lost vehicle is determined according to the number of vehicles in the target vehicle information, the number of vehicles in the target vehicle information at the previous moment and the vehicle binding information of each vehicle, and the binding information of the lost vehicle is used as the vehicle binding information.

[0029] Optionally, generating a control signal according to the vehicle binding information includes:

[0030] Determine vehicle status information according to the vehicle position in the vehicle binding information and the collection time corresponding to the vehicle position, wherein the vehicle status information includes: vehicle entering, vehicle slowing down, vehicle stopping and vehicle exiting;

[0031] Generate an exit prompt signal according to the vehicle driving information and the vehicle state information in the vehicle binding information of each vehicle, wherein the exit prompt signal is used to control the control device to send the vehicle exit prompt information to the lane vehicle when the vehicle is in the vehicle exit state;

[0032] Generate a distance control signal according to the vehicle driving information and the vehicle status information in the vehicle binding information of each vehicle, wherein the distance control signal is used to control the control device to output distance prompt information when the distance between the vehicle and the edge area of ​​the parking strip is less than a preset distance;

[0033] The vehicle stay time is determined according to the collection time corresponding to the vehicle position in the vehicle binding information of each vehicle, and a rescue control signal is generated according to the vehicle stay time. The rescue control signal is used to control the control device to output rescue inquiry information when the vehicle stay time is greater than the preset stay time.

[0034] Optionally, the training process of the first image recognition model includes:

[0035] Preprocessing the first traffic image, and using the preprocessed first traffic image and image identification information corresponding to the first traffic image as a first training sample, wherein the preprocessing includes: reducing image brightness, reducing image contrast, and increasing image blur;

[0036] Training the initial image recognition model based on the first training sample input to obtain an intermediate image recognition model;

[0037] The second traffic image and the image identification information corresponding to the second traffic image are used as second training samples, and the intermediate image recognition model is trained based on the second training samples to obtain a first image recognition model.

[0038] In a second aspect, the present application provides a high-precision intelligent perception and analysis device for emergency parking strips, the device comprising:

[0039] An acquisition module, used to acquire the parking strip image at the current moment sent by the image acquisition device;

[0040] A first determination module is used to determine at least one target vehicle information in the parking strip image based on a pre-trained first image recognition model, and to determine at least one license plate information in the parking strip image based on a pre-trained second image recognition model, wherein the target vehicle information includes the number of vehicles, the vehicle image and the vehicle position, and the license plate information includes the license plate number, the license plate position, the license plate type and the license plate collection time;

[0041] A second determination module is used to determine the vehicle tracking information of the vehicle at the current moment according to the target vehicle information in the parking strip image and the vehicle tracking information before the current moment, wherein the vehicle tracking information includes a vehicle tracking identifier, a vehicle type and vehicle driving information, and the vehicle driving information includes a vehicle position and a collection time corresponding to the vehicle position;

[0042] A binding module, used for binding the vehicle tracking information and the license plate information to generate a plurality of vehicle binding information, each of which includes: the vehicle tracking information and the corresponding license plate information;

[0043] A generating module is used to generate a control signal according to the vehicle binding information, and send the control signal to a plurality of control devices arranged in the parking strip area, so that the control device outputs prompt information according to the control signal.

[0044] Optionally, the binding module is specifically used for:

[0045] Traverse the collection time of each license plate in the first license plate information, and for the traversed current license plate collection time, if the collection time corresponding to the current license plate collection time and the first vehicle position in the first vehicle driving information is less than a preset time threshold, determine whether the distance between the license plate position corresponding to the current license plate collection time in the first license plate information and the first vehicle position in the first vehicle driving information is less than a preset distance. If so, bind the first license plate information and the vehicle tracking information corresponding to the first vehicle driving information to generate vehicle binding information corresponding to the first license plate information.

[0046] Optionally, the binding module is specifically used for:

[0047] Traverse the collection time of each license plate in the first license plate information, and for the current license plate collection time collected, if the collection time corresponding to the current license plate collection time and the first vehicle position in the first vehicle driving information is less than a preset time threshold, determine whether the distance between the license plate position corresponding to the current license plate collection time in the first license plate information and the first vehicle position in the first vehicle driving information is less than a preset distance; if so, determine whether the license plate type of the first license plate information and the vehicle type in the first vehicle driving information satisfy a preset matching relationship; if so, bind the first license plate information and the vehicle tracking information corresponding to the first vehicle driving information to generate vehicle binding information corresponding to the first license plate information.

[0048] Optionally, the binding module is further used for:

[0049] Traversing each vehicle binding information, for the traversed first vehicle binding information and second vehicle binding information, determining a collection distance between a vehicle position with the latest collection time in the first vehicle binding information and a vehicle position with the earliest collection time in the second vehicle binding information;

[0050] Determine whether the acquisition distance is less than a preset distance threshold;

[0051] If yes, determining whether the vehicle type in the first vehicle binding information and the vehicle type in the second vehicle binding information satisfy a preset matching relationship;

[0052] If so, the vehicle tracking identifier in the first vehicle binding information and the license plate number and license plate type in the license plate information are used as the vehicle tracking identifier in the second vehicle binding information and the license plate number and license plate type in the license plate information.

[0053] Optionally, after binding the vehicle tracking information and the license plate information to generate vehicle binding information, the method further includes:

[0054] Based on a multi-source shortest path algorithm, determining a first binding result according to the vehicle position with the latest collection time in each of the vehicle binding information and the vehicle position with the earliest collection time in each of the vehicle binding information, wherein the first binding result includes a plurality of vehicle binding information corresponding to each other;

[0055] Determine a target binding result according to the vehicle type in each vehicle binding information in the first binding result, wherein the target binding result includes at least one vehicle binding information before the change and vehicle binding information after the change corresponding to the vehicle binding information before the change;

[0056] The vehicle tracking identifier in each vehicle binding information before the change and the license plate number and license plate type in the license plate information are used as the vehicle tracking identifier in each vehicle binding information after the change and the license plate number and license plate type in the license plate information.

[0057] Optionally, the binding module is further used for:

[0058] Determining whether there is a lost vehicle based on the vehicle binding information;

[0059] If so, the binding information of the lost vehicle is determined according to the number of vehicles in the target vehicle information, the number of vehicles in the target vehicle information at the previous moment and the vehicle binding information of each vehicle, and the binding information of the lost vehicle is used as the vehicle binding information.

[0060] Optionally, the generating module is specifically used for:

[0061] Determine vehicle status information according to the vehicle position in the vehicle binding information and the collection time corresponding to the vehicle position, wherein the vehicle status information includes: vehicle entering, vehicle slowing down, vehicle stopping and vehicle exiting;

[0062] Generate an exit prompt signal according to the vehicle driving information and the vehicle state information in the vehicle binding information of each vehicle, wherein the exit prompt signal is used to control the control device to send the vehicle exit prompt information to the lane vehicle when the vehicle is in the vehicle exit state;

[0063] Generate a distance control signal according to the vehicle driving information and the vehicle status information in the vehicle binding information of each vehicle, wherein the distance control signal is used to control the control device to output distance prompt information when the distance between the vehicle and the edge area of ​​the parking strip is less than a preset distance;

[0064] The vehicle stay time is determined according to the collection time corresponding to the vehicle position in the vehicle binding information of each vehicle, and a rescue control signal is generated according to the vehicle stay time. The rescue control signal is used to control the control device to output rescue inquiry information when the vehicle stay time is greater than the preset stay time.

[0065] Optionally, the first determining module is specifically configured to:

[0066] Preprocessing the first traffic image, and using the preprocessed first traffic image and image identification information corresponding to the first traffic image as a first training sample, wherein the preprocessing includes: reducing image brightness, reducing image contrast, and increasing image blur;

[0067] Training the initial image recognition model based on the first training sample input to obtain an intermediate image recognition model;

[0068] The second traffic image and the image identification information corresponding to the second traffic image are used as second training samples, and the intermediate image recognition model is trained based on the second training samples to obtain a first image recognition model.

[0069] In a third aspect, the present application provides an electronic device comprising: a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the storage medium through the bus, and the processor executes the machine-readable instructions to perform the steps of the above-mentioned high-precision emergency parking strip intelligent perception analysis method.

[0070] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the high-precision emergency parking strip intelligent perception analysis method as described above are executed.

[0071] The beneficial effects of the present application are: by acquiring the parking strip image at the current moment, and then determining the target vehicle information and the license plate information based on the first image recognition model and the second image recognition model respectively, and then determining the vehicle tracking information of the vehicle at the current moment according to the target vehicle information at the current moment and the vehicle tracking information before the current moment. At this time, the status of all vehicles in the emergency parking strip since entering the emergency parking strip until the current moment and the license plate information of the vehicle are obtained. Then, the vehicle tracking information and the license plate information are bound to generate vehicle binding information, and a control signal is generated according to the vehicle binding information so that multiple control devices in the parking strip area output prompt information according to the control signal, thereby effectively driving the control device to make more timely and effective linkage and control. In addition, the image acquisition device has low cost and is not easily affected by the environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0073] Figure 1 It is a scene schematic diagram of a high-precision intelligent perception analysis method for emergency parking strips proposed in an embodiment of the present application;

[0074] Figure 2 It is a flow chart of a high-precision intelligent perception analysis method for emergency parking strips provided in an embodiment of the present application;

[0075] Figure 3 is a schematic diagram of an emergency parking strip provided in an embodiment of the present application;

[0076] Figure 4 is a schematic diagram of a license plate shielding provided in an embodiment of the present application;

[0077] Figure 5 It is a schematic diagram of a process for generating vehicle binding information provided by an embodiment of the present application;

[0078] Figure 6 It is another schematic diagram of a process for generating vehicle binding information provided by an embodiment of the present application;

[0079] Figure 7 It is a schematic diagram of a flow chart of generating a control signal provided in an embodiment of the present application;

[0080] Figure 8 is a flowchart of a training process of a first image recognition model provided in an embodiment of the present application;

[0081] Fig. 9 It is a structural schematic diagram of a high-precision emergency parking strip intelligent perception and analysis device provided in an embodiment of the present application;

[0082] Fig.10 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0083] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the drawings in the present application only serve the purpose of explanation and description and are not used to limit the scope of protection of the present application. In addition, it should be understood that the schematic drawings are not drawn in real proportion. The flowchart used in this application shows the operations implemented according to some embodiments of the present application. It should be understood that the operations of the flowchart can be implemented out of sequence, and the steps without logical context can be reversed in order or implemented simultaneously. In addition, those skilled in the art can add one or more other operations to the flowchart under the guidance of the content of the present application, or remove one or more operations from the flowchart.

[0084] In addition, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0085] It should be noted that the term "comprising" will be used in the embodiments of the present application to indicate the existence of the features declared thereafter, but does not exclude the addition of other features.

[0086] In the current parking strip system, radar detection, infrared detection and geomagnetic detection are usually used to detect vehicles in the emergency parking strip throughout their life cycle. However, the equipment cost required for radar detection is too high, infrared detection is not only expensive, but also easily affected by the surrounding environment, and geomagnetic detection requires pre-embedded geomagnetic detection equipment in the emergency parking strip. If there is no planning in the early stage, the road surface construction and paving will need to be re-constructed and re-laid, affecting traffic operation.

[0087] Based on this, the present application proposes a high-precision intelligent perception and analysis method for emergency parking strips, which first obtains the parking strip image at the current moment, then determines at least one target vehicle information and at least one license plate information, and then determines the vehicle tracking information of the vehicle at the current moment based on the target vehicle information and the vehicle tracking information before the current moment, binds the vehicle tracking information and the license plate information to generate vehicle binding information, and finally generates a control signal based on the vehicle binding information, so that multiple control devices in the parking strip area output prompt information according to the control signal. The present application tracks the vehicles in the parking strip image and binds the license plates, thereby obtaining the status of the entire life cycle of each vehicle parked in the parking strip and the license plate information of the vehicle, thereby effectively driving the control device to make more timely and effective linkage and control.

[0088] Before introducing the high-precision intelligent perception analysis method for emergency stop strips, the application scenarios of this method are introduced first. Traffic accidents are very likely to occur in tunnel sections of highways, and accidents are often difficult to handle, with a large impact and a great threat to highway driving safety. Therefore, emergency stop strips are usually set up in tunnel sections of highways for use when vehicles temporarily break down, thereby ensuring smooth highway traffic and reducing interference from accident vehicles and faulty vehicles. Figure 1 As shown, Figure 1 This is a scene diagram of a high-precision intelligent perception analysis method for emergency parking strips proposed in an embodiment of the present application. In the emergency parking strip, not only an image acquisition device but also a control device can be set. Among them, the control device can be, for example, a display screen and a speaker. The present application manages vehicles parked in the emergency parking strip in real time to ensure the operational safety of the tunnel.

[0089] Figure 2 is a flow chart of a high-precision intelligent perception analysis method for emergency parking strips provided in an embodiment of the present application. Figure 2 The high-precision intelligent perception analysis method of emergency parking strip is introduced.

[0090] S201. Acquire a parking strip image at the current moment sent by an image acquisition device.

[0091] The image acquisition device may be a device with a license plate recognition function. The device with a license plate recognition function can obtain the parking strip image and the license plate information of each vehicle in the parking strip image, so that there is no need to add additional engineering or algorithms to analyze the license plate information in each image. The image acquisition device may also be a device without a license plate recognition function. This embodiment is described by taking the image acquisition device without a license plate recognition function as an example.

[0092] Optionally, the parking strip image may be acquired by an image acquisition device at a preset time interval, wherein the image acquisition device records the acquisition time of the parking strip image while acquiring the parking strip image.

[0093] Optionally, a plurality of vehicles may be included in the parking strip image.

[0094] S202. Determine at least one target vehicle information in the parking strip image based on a pre-trained first image recognition model, and determine at least one license plate information in the parking strip image based on a pre-trained second image recognition model, wherein the target vehicle information includes the number of vehicles, vehicle images, and vehicle positions, and the license plate information includes the license plate number, license plate position, license plate type, and license plate collection time.

[0095] Specifically, the parking strip image at the current moment is input into the first image recognition model, and the number of vehicles, vehicle images and vehicle positions are output. The number of vehicles is the total number of vehicles in the parking strip image at the current moment. The vehicle image is the image of each vehicle in the parking strip image, and the vehicle image is a part of the parking strip image. The vehicle position is the position of each vehicle image relative to the parking strip image.

[0096] Each parking strip image is input into the second image recognition model, and the license plate number, license plate position, license plate type and license plate acquisition time are output. The license plate number can be obtained from the front or rear license plate of the vehicle, the license plate position is the position of the license plate relative to the parking strip image, and the license plate type is the length-to-width ratio of the license plate and the license plate color. The license plate acquisition time of vehicles in the same parking strip image is the same.

[0097] As an optional implementation, if there is only a portion of the vehicle in the parking lane image, and the other portion may not have entered the parking lane or has already exited the parking lane, the vehicle is not analyzed, that is, the target vehicle information and license plate information of the vehicle are not obtained.

[0098] S203. Determine the vehicle tracking information of the vehicle at the current moment according to the target vehicle information in the parking strip image and the vehicle tracking information before the current moment, wherein the vehicle tracking information includes the vehicle tracking identification, the vehicle type and the vehicle driving information, and the vehicle driving information includes the vehicle position and the collection time corresponding to the vehicle position.

[0099] As an optional implementation, the vehicle in the parking strip image may be tracked based on an image tracking algorithm. Specifically, the image tracking algorithm may be a tracking algorithm that combines a Kalman filter with a Deep Simple Online and Realtime Tracking (DeepSort) algorithm.

[0100] Optionally, based on the image tracking algorithm, according to the target vehicle information in the parking strip image and the vehicle tracking information before the current moment, the vehicle tracking information of the vehicle at the current moment is determined. The vehicle tracking information at the current moment includes the vehicle tracking information before the current moment.

[0101] Optionally, each vehicle has a unique vehicle tracking identifier throughout its life cycle in the parking strip. Each vehicle has vehicle tracking information, which includes the vehicle tracking identifier, vehicle type, and vehicle driving information. The vehicle driving information includes all vehicle positions of the vehicle entering the parking strip throughout its life cycle and the collection time corresponding to each vehicle position.

[0102] The vehicle type may include the vehicle's length-to-width ratio, vehicle model, and vehicle color.

[0103] S204, binding the vehicle tracking information and the license plate information to generate a plurality of vehicle binding information, each of which includes: the vehicle tracking information and the corresponding license plate information.

[0104] As an optional implementation, when the license plate collection time and the collection time corresponding to the vehicle position are the same, the distance between the vehicle and its license plate should be less than the preset distance, and the vehicle type and license plate type meet the preset matching relationship. If the vehicle tracking information and license plate information meet the above conditions, they can be bound, that is, the vehicle binding information is generated.

[0105] S205: Generate a control signal according to the vehicle binding information, and send the control signal to a plurality of control devices arranged in the parking strip area, so that the control devices output prompt information according to the control signal.

[0106] Optionally, the control device may include a display screen and a speaker. As an optional implementation, a control signal may be generated according to the vehicle binding information, the preset distance and the preset stay time, so that the control device outputs prompt information according to the control signal. The prompt information may include an exit prompt signal, a distance control signal and a rescue control signal. The exit prompt signal is used to send an exit prompt signal to the lane vehicle when the vehicle exits the parking strip, the distance control signal is used to prompt the vehicle that is too close to the edge of the parking strip, and the rescue control signal is used to send a rescue inquiry to the vehicle that stays for a long time and provide rescue contact information.

[0107] Optionally, each vehicle binding information is uploaded to the business processing platform as a record. In addition, the vehicle driving status can be determined based on the vehicle position in the vehicle driving information and the collection time corresponding to the vehicle position, and the vehicle driving status can be uploaded to the business processing platform. The business processing platform stores the vehicle binding information and the corresponding vehicle driving status. The business processing platform can also generate a control signal based on the vehicle binding information and the vehicle driving status. Among them, the vehicle driving status includes vehicle entering, vehicle slowing down, vehicle stopping, vehicle reversing and vehicle exiting.

[0108] As an optional implementation, the control signal may be sent to each control device by way of wireless communication.

[0109] In this embodiment, by acquiring the parking strip image at the current moment, the target vehicle information and the license plate information are determined based on the first image recognition model and the second image recognition model, respectively, and then the vehicle tracking information of the vehicle at the current moment is determined based on the target vehicle information at the current moment and the vehicle tracking information before the current moment. At this time, the status of all vehicles in the emergency parking strip since entering the emergency parking strip until the current moment and the license plate information of the vehicle are obtained. Then, the vehicle tracking information and the license plate information are bound to generate vehicle binding information, and a control signal is generated according to the vehicle binding information so that multiple control devices in the parking strip area output prompt information according to the control signal, thereby effectively driving the control device to make more timely and effective linkage and control. In addition, the image acquisition device has low cost and is not easily affected by the environment.

[0110] There are two implementation methods for binding vehicle tracking information and license plate information in this application. As a first optional implementation method, the process of binding vehicle tracking information and license plate information to generate vehicle binding information is as follows:

[0111] Traverse the collection time of each license plate in the first license plate information, and for the current license plate collection time traversed, if the collection time corresponding to the current license plate collection time and the first vehicle position in the first vehicle driving information is less than a preset time threshold, determine whether the distance between the license plate position corresponding to the current license plate collection time in the first license plate information and the first vehicle position in the first vehicle driving information is less than a preset distance. If so, bind the first license plate information and the vehicle tracking information corresponding to the first vehicle driving information to generate vehicle binding information corresponding to the first license plate information.

[0112] Optionally, if the image acquisition device has a license plate recognition function, it is necessary to determine the binding of the vehicle tracking information and the license plate information based on a preset time threshold of the image acquisition time, and it is necessary to determine whether the vehicle position and license plate position distance collected in the same time period are less than a preset distance.

[0113] Specifically, the license plate collection time in the first license plate information is traversed, and for the current license plate collection time traversed, if the current license plate collection time and the collection time corresponding to the first vehicle position in the first vehicle driving information are less than the preset time threshold, the license plate information corresponding to the current license plate collection time and the target vehicle information corresponding to the first vehicle position can be regarded as collected in the same time period. On this basis, if the distance between the license plate position corresponding to the current license plate collection time in the first license plate information and the first vehicle position in the first vehicle driving information is less than the preset distance, the first license plate information and the vehicle tracking information corresponding to the first vehicle driving information are bound to generate vehicle binding information corresponding to the first license plate information.

[0114] The first license plate information is any one of a plurality of license plate information, the first vehicle driving information is any one of a plurality of vehicle driving information, and the first vehicle position is any one of the first vehicle driving information.

[0115] Optionally, if the image acquisition device does not have a license plate recognition function, the preset time threshold can be set to 0, and the license plate information and the vehicle position are derived from the same parking strip image. It should be understood that in the same parking strip image, the license plate position and the license plate position of the same vehicle should overlap or be at a small distance. Therefore, a preset distance is set, and when it is determined that the distance between the license plate position in the first license plate information and the vehicle position in the first vehicle driving information is less than the preset distance, it means that the first license plate information matches the first vehicle driving information.

[0116] It is worth mentioning that the preset time threshold can be determined based on a specific usage scenario. When the image acquisition device has a license plate recognition function, the preset time threshold set is smaller than the preset time threshold set when the image acquisition device does not have a license plate recognition function.

[0117] It is worth noting that in the emergency parking strip, due to the position of the image acquisition device and the position of each vehicle, some vehicles may overlap or the license plate may be blocked. Therefore, when the distance between the license plate position in more than one license plate information and the vehicle position in the vehicle information is less than the preset distance, the following second implementation method needs to be executed for binding. For example, Figure 3 Schematic diagram of an emergency parking zone provided in an embodiment of the present application. Figure 3 As shown, if the license plate position of vehicle A and the license plate position of vehicle B are both less than the preset distance from the vehicle position of vehicle A, the following second implementation method needs to be executed for binding. When calculating the distance between the license plate position and the vehicle position, the distance can be calculated with the center point of the vehicle as the reference point.

[0118] As a second optional implementation, the process of binding vehicle tracking information and license plate information to generate vehicle binding information is as follows:

[0119] Traverse the collection time of each license plate in the first license plate information, and for the current license plate collection time collected, if the collection time corresponding to the current license plate collection time and the first vehicle position in the first vehicle driving information is less than a preset time threshold, then determine whether the distance between the license plate position corresponding to the current license plate collection time in the first license plate information and the first vehicle position in the first vehicle driving information is less than a preset distance. If so, determine whether the license plate type of the first license plate information and the vehicle type in the first vehicle driving information satisfy a preset matching relationship. If so, bind the first license plate information and the vehicle tracking information corresponding to the first vehicle driving information to generate vehicle binding information corresponding to the first license plate information.

[0120] Optionally, the preset matching relationship may be that the length-to-width ratio of the license plate matches the vehicle type.

[0121] As an optional implementation, when the current license plate collection time and the collection time corresponding to the first vehicle position are less than a preset time threshold, in addition to determining the relationship between the license plate position corresponding to the current license plate collection time in the first license plate information and the first vehicle position in the first vehicle driving information, and the relationship between the license plate type and the vehicle type, it is also possible to determine whether the vehicle color matches the license plate color, thereby binding the vehicle tracking information corresponding to the first license plate information and the first vehicle driving information according to the matching result.

[0122] It should be understood that when capturing images of the same vehicle in the same environment, the vehicle color and the license plate color have a certain relationship. For example, the contrast of the vehicle in the vehicle color is the same as the contrast of the license plate in the license plate color.

[0123] In this embodiment, the vehicle binding information is generated by binding the license plate information, vehicle driving information and the vehicle tracking identifier corresponding to the vehicle driving information according to the relationship between the license plate position and the vehicle position, and the relationship between the license plate type and the vehicle type, so as to timely obtain the vehicle binding information of each vehicle in the emergency parking lane, so as to control the control device to output prompt information for the vehicle binding information of each vehicle.

[0124] Furthermore, in the process of managing the complete life cycle of a vehicle entering the emergency parking strip, there may be a situation where the license plate is blocked or the entire vehicle is blocked, such as Figure 4 As shown. Among them, Figure 4 This is a schematic diagram of a license plate shielding provided by an embodiment of the present application. This situation may result in two vehicle tracking marks on the same vehicle, or the vehicle disappearing.

[0125] Specifically, the reason for the same vehicle to have two vehicle tracking marks may be that the vehicle is blocked while driving or stopping in the parking strip, and a new vehicle tracking mark is set for it when it appears again, thereby forming new vehicle tracking information. In this case, the following two implementation methods can be performed after the vehicle binding information is generated.

[0126] As an optional implementation, Figure 5 Schematic diagram of a process for generating vehicle binding information provided by an embodiment of the present application. Figure 5 As shown:

[0127] S501, traversing each vehicle binding information, and for the traversed first vehicle binding information and second vehicle binding information, determining a collection distance between a vehicle position with the latest collection time in the first vehicle binding information and a vehicle position with the earliest collection time in the second vehicle binding information.

[0128] Optionally, if there is no new vehicle entering or exiting the current emergency stop zone, steps S501 to S504 may be performed to make the matching result accurate.

[0129] It should be understood that when traversing the binding information of each vehicle in the current emergency parking strip, if there is only one vehicle in the current emergency parking strip, the vehicle tracking identifier of the vehicle will not change, and there is no need to make the next judgment. If two or more vehicle binding information can be traversed, then for the traversed first vehicle binding information and the second vehicle binding information, determine the collection distance between the vehicle position with the latest collection time in the first vehicle binding information and the vehicle position with the earliest collection time in the second vehicle binding information. Among them, the first vehicle binding information and the second vehicle binding information can be any two of the vehicle binding information corresponding to the vehicles existing in the current emergency parking strip.

[0130] S502: Determine whether the collection distance is less than a preset distance threshold.

[0131] It should be understood that if two vehicle tracking identifiers are determined for the same vehicle in the parking strip due to reasons such as being blocked, the collection distance corresponding to the adjacent collection times between the two vehicle binding information corresponding to the vehicle should be less than the preset distance threshold. Exemplarily, if the first vehicle binding information and the second vehicle binding information are traversed to include vehicle binding identifier a and vehicle binding identifier b respectively. Among them, the first vehicle binding information includes a collection time of 1 minute and 1 second, and the corresponding collection location is x, a collection time of 1 minute and 2 seconds, and the corresponding collection location is y, a collection time of 1 minute and 3 seconds, and the corresponding collection location is z, and the second vehicle binding information includes a collection time of 1 minute and 5 seconds, and the corresponding collection location u, a collection time of 1 minute and 6 seconds, and the corresponding collection location v, a collection time of 1 minute and 7 seconds, and the corresponding collection time w, and it is determined whether the collection location z corresponding to 1 minute and 3 seconds in the first vehicle binding information and the collection location u corresponding to 1 minute and 5 seconds in the second vehicle binding information are less than the preset distance threshold. If so, it means that the first vehicle binding information and the second vehicle binding information may be information of the same vehicle.

[0132] S503: If yes, determine whether the vehicle type in the first vehicle binding information and the vehicle type in the second vehicle binding information satisfy a preset matching relationship.

[0133] Optionally, on the basis that the collection distance corresponding to the first vehicle binding information and the second vehicle binding information is less than a preset distance threshold, if the vehicle type in the first vehicle binding information and the vehicle type in the second vehicle binding information satisfy a preset matching relationship, then it means that the vehicle corresponding to the first vehicle binding information and the vehicle corresponding to the second vehicle binding information are the same vehicle.

[0134] S504: If yes, use the vehicle tracking identifier in the first vehicle binding information and the license plate number and license plate type in the license plate information as the vehicle tracking identifier in the second vehicle binding information and the license plate number and license plate type in the license plate information.

[0135] As an optional implementation, if the same vehicle has multiple identification changes, the vehicle tracking identification corresponding to the earliest vehicle binding information and the license plate number and license plate type in the license plate information will be used as the vehicle tracking identification corresponding to the vehicle binding information corresponding to all subsequent vehicles and the license plate number and license plate type in the license plate information.

[0136] In this embodiment, each vehicle binding information is traversed, and the collection distance between the vehicle position with the latest collection time in the first vehicle binding information and the vehicle position with the earliest collection time in the second vehicle binding information is determined for the traversed first vehicle binding information and the second vehicle binding information, and it is determined whether the collection distance is less than a preset distance threshold. If so, it is determined whether the vehicle type satisfies a preset matching relationship. If so, the vehicle tracking identifier in the first vehicle binding information and the license plate number and license plate type in the license plate information are used as the vehicle tracking identifier in the second vehicle binding information and the license plate number and license plate type in the license plate information, thereby avoiding multiple records of the entire life cycle of the same vehicle entering the emergency parking strip, which affects the management effect.

[0137] As an optional implementation, Figure 6 This is another flow chart of generating vehicle binding information provided by an embodiment of the present application. Figure 6 As shown:

[0138] S601. Based on a multi-source shortest path algorithm, determine a first binding result according to the vehicle position with the latest collection time in each vehicle binding information and the vehicle position with the earliest collection time in each vehicle binding information, wherein the first binding result includes a plurality of vehicle binding information corresponding to each other.

[0139] In the first binding result, the vehicle binding information with an earlier collection time in each group of vehicle binding information may be the vehicle binding information before the selected change, and the vehicle binding information with a later collection time may be the vehicle binding information after the selected change. The sum of the distances between the vehicle position corresponding to the latest collection time and the vehicle position corresponding to the latest collection time of each group of vehicle binding information before the selected change is the smallest.

[0140] Among them, the multi-source shortest path algorithm is used to find the shortest distance between the vehicle position with the latest collection time among multiple vehicle binding information and the shortest distance between the earliest collection time among multiple vehicle binding information, and the two vehicle binding information corresponding to the shortest distance are used as the vehicle binding information before and after the change to be selected.

[0141] S602: Determine a target binding result according to the vehicle type in each vehicle binding information in the first binding result, wherein the target binding result includes at least one vehicle binding information before the change and vehicle binding information after the change corresponding to the vehicle binding information before the change.

[0142] Specifically, it is determined whether the vehicle types in the two corresponding vehicle binding information in the first binding result satisfy a preset matching relationship. If so, the two corresponding vehicle binding information are used as a pre-change vehicle binding information and a post-change vehicle binding information corresponding to the pre-change vehicle binding information.

[0143] S603: Use the vehicle tracking identifier in each vehicle binding information before the change and the license plate number and license plate type in the license plate information as the vehicle tracking identifier in each vehicle binding information after the change and the license plate number and license plate type in the license plate information.

[0144] Exemplarily, if there is vehicle binding information A1 before the change, vehicle binding information A2 after the change corresponding to the vehicle binding information A1 before the change, vehicle binding information B1 before the change, vehicle binding information B2 after the change corresponding to the vehicle binding information B1 before the change, vehicle binding information C1 before the change, and vehicle binding information C2 after the change corresponding to the vehicle binding information C1 before the change, then the vehicle tracking identifier and the license plate number and license plate type in the license plate information of A1 are used as the vehicle tracking identifier and the license plate number and license plate type in the license plate information of A2, the vehicle tracking identifier and the license plate number and license plate type in the license plate information of B1 are used as the vehicle tracking identifier and the license plate number and license plate type in the license plate information of B2, and the vehicle tracking identifier and the license plate number and license plate type in the license plate information of C1 are used as the vehicle tracking identifier and the license plate number and license plate type in the license plate information of C2.

[0145] In this embodiment, the first binding result is determined based on a multi-source shortest path algorithm, so that multiple vehicles whose vehicle tracking identifications have changed are associated respectively, thereby avoiding multiple records of multiple vehicles located in the emergency parking zone during their entire life cycle due to uncontrollable reasons such as sudden changes in light.

[0146] Furthermore, the reason for the vehicle disappearing may be that the vehicle is always blocked during the stopping process, resulting in the inability to obtain the vehicle information at the current time and a period of time before the current time, that is, the vehicle disappears. In this case, the following steps can be performed after generating the vehicle binding information:

[0147] Optionally, determine whether there is a lost vehicle based on the vehicle binding information.

[0148] Specifically, it is determined whether there is a lost vehicle based on the number of vehicles in the target vehicle information corresponding to the vehicle binding information at the current moment. It is worth noting that when the vehicle position in the vehicle driving information in the vehicle binding information ends at the exit position of the parking strip, it means that the vehicle has exited the parking strip normally, that is, the vehicle is not a lost vehicle. If the vehicle does not exit the parking strip normally, and the number of vehicles at the current moment is less than the number of vehicles in the target vehicle information at the previous moment, it means that there is a lost vehicle.

[0149] Optionally, if so, the binding information of the lost vehicle is determined based on the number of vehicles in the target vehicle information, the number of vehicles in the target vehicle information at the previous moment, and the vehicle binding information of each vehicle, and the binding information of the lost vehicle is used as one vehicle binding information.

[0150] Optionally, if there is a lost vehicle, the number of lost vehicles is first determined based on the number of vehicles at the current moment and the number of vehicles in the target vehicle information at the previous moment. Then, the binding information of the lost vehicle is determined based on the number of lost vehicles and the vehicle binding information of all vehicles.

[0151] Specifically, the current vehicle position and the previous vehicle position in the vehicle binding information of all vehicles are compared. If the vehicle disappears, there is a vehicle position of the lost vehicle at the previous moment, and the vehicle position data of the lost vehicle at the current moment is empty, and the number of empty data is equal to the number of lost vehicles.

[0152] Optionally, the vehicle position of the lost vehicle at the previous moment is used as the vehicle position at the current moment, the current time is used as the collection time of the vehicle position, and the vehicle position at the current moment, the collection time, the vehicle tracking identification and license plate information at the previous moment are used as a new vehicle binding information.

[0153] Optionally, since the vehicle position at the current moment is the same as the vehicle position at the previous moment, it indicates that the vehicle is currently in a stopped state. Until the lost vehicle reappears, the vehicle binding information to be merged is determined according to the license plate number of the vehicle that appears, and the vehicle tracking identifier is updated.

[0154] In this embodiment, by associating vehicle tracking identifiers and compensating for missing vehicles, the integrity of the entire life cycle of the vehicle from entering the emergency parking zone to exiting the emergency parking zone can be ensured, so as to achieve vehicle statistics and management.

[0155] As an optional implementation, refer to Figure 7 The specific steps of generating a control signal according to the vehicle binding information in the above step S205 are introduced. Figure 7 It is a flow chart of generating a control signal provided in an embodiment of the present application.

[0156] S701. Determine vehicle status information according to the vehicle position in the vehicle binding information and the collection time corresponding to the vehicle position. The vehicle status information includes: vehicle entering, vehicle slowing down, vehicle stopping, and vehicle exiting.

[0157] Specifically, the direction and speed of the vehicle can be calculated based on the distance between the vehicle positions at adjacent collection times, and the vehicle status information can be determined based on the direction and speed of the vehicle. Specifically, when the vehicle moves from the entrance side of the parking strip to the parking side of the parking strip, and the front of the vehicle faces the parking side of the parking strip, the vehicle status information is that the vehicle enters, and when the vehicle position corresponding to the adjacent collection times does not change, the vehicle status information is that the vehicle stops. When the vehicle moves from the parking side of the parking strip to the exit side of the parking strip, and the front of the vehicle faces the exit side of the parking strip, the vehicle status information is that the vehicle exits.

[0158] As an optional implementation, the vehicle status information may also include vehicle reverse movement, which may be determined based on the direction and speed of the vehicle. Specifically, when the vehicle moves from the entrance side of the parking strip to the parking side of the parking strip, and the front of the vehicle faces the entrance side of the parking strip, or when the vehicle moves from the parking side of the parking strip to the exit side of the parking strip, and the front of the vehicle faces the parking side of the parking strip, the vehicle status information is vehicle reverse movement.

[0159] S702. Generate an exit prompt signal according to the vehicle driving information and vehicle status information in the vehicle binding information of each vehicle. The exit prompt signal is used to control the control device to send an exit vehicle prompt information to the lane vehicle when the vehicle is in the vehicle exit state.

[0160] Optionally, the distance between the vehicle position in the vehicle driving information in the vehicle binding information of each vehicle and the parking lane exit is determined in real time. If the distance is less than the preset exit distance and the vehicle is in an exit state, the control device is controlled to send an exit vehicle prompt information to the lane vehicle.

[0161] Among them, the vehicle prompt information is used to remind the vehicles in the lane that the emergency parking lane is about to leave the vehicle, so please pay attention to avoid it.

[0162] S703. Generate a distance control signal according to the vehicle driving information and vehicle status information in the vehicle binding information of each vehicle, wherein the distance control signal is used to control the control device to output distance prompt information when the distance between the vehicle and the edge area of ​​the parking strip is less than a preset distance.

[0163] Optionally, it is determined in real time whether the distance between the vehicle position in the vehicle driving information in the vehicle binding information and the edge area of ​​the parking strip is less than a preset distance. If so, and the vehicle status information indicates that the vehicle is not in a stopped state, a distance control signal is generated so that the control device outputs distance prompt information according to the distance control signal, thereby prompting occupants of the vehicle that the vehicle is too close to the edge area of ​​the parking strip, which may cause danger.

[0164] As an example, the control device for outputting the distance prompt information may be a speaker.

[0165] S704. Determine the vehicle stay time according to the vehicle position corresponding to the collection time in the vehicle binding information of each vehicle, and generate a rescue control signal according to the vehicle stay time. The rescue control signal is used to control the control device to output rescue inquiry information when the vehicle stay time is greater than the preset stay time.

[0166] Optionally, the difference between the earliest collection time and the latest collection time among the collection times corresponding to the vehicle positions of each vehicle is used as the vehicle residence time. If the vehicle residence time of a vehicle exceeds the preset residence time, the vehicle may need rescue. At this time, a rescue control signal is generated and sent to the control device, so that the control device outputs rescue inquiry information.

[0167] Optionally, the rescue inquiry information may be, for example, outputting a rescue inquiry voice to inquire whether the person in the vehicle needs rescue.

[0168] As another optional implementation, based on a pre-trained third image recognition model, it is determined whether there is a person outside the vehicle standing in a preset danger zone in the parking image. If so, a reminder signal is output. The reminder signal is used to remind the person outside the vehicle to leave the danger zone as soon as possible to ensure the safety of the people.

[0169] In this embodiment, by generating a distance control signal and a rescue control signal, the control device outputs distance prompt information and rescue inquiry information, thereby ensuring the safety of personnel and vehicles and participating in the rescue operation in a timely manner.

[0170] Figure 8 is a flowchart of a training process of a first image recognition model provided in an embodiment of the present application. Figure 8 The training process of the first image recognition model in step S202 is introduced as follows:

[0171] S801 , preprocessing a first traffic image, and using the preprocessed first traffic image and image identification information corresponding to the first traffic image as a first training sample, wherein the preprocessing includes: reducing image brightness, reducing image contrast, and increasing image blur.

[0172] Due to insufficient optical fiber in the tunnel, the parking strip images captured by the image acquisition device are often too dark and of poor quality. Therefore, in the process of training the first image recognition model, the first traffic image needs to be preprocessed to better recognize the image.

[0173] The first traffic image may include an image of a non-tunnel section.

[0174] Optionally, the image identification information is pre-identified information.

[0175] S802: Train the initial image recognition model based on the first training sample input to obtain an intermediate image recognition model.

[0176] S803: Use the second traffic image and the image identification information corresponding to the second traffic image as a second training sample, and train the intermediate image recognition model based on the second training sample to obtain a first image recognition model.

[0177] Optionally, the second traffic image is an image of a parking strip in a tunnel section. Using the second traffic image as a second training sample to train the intermediate image recognition model again can ensure that the first image recognition model outputted in the end has a better target detection effect, and the model has good adaptability and robustness.

[0178] In this embodiment, the first traffic image is preprocessed and then input into the initial image recognition model for training to obtain an intermediate image recognition model, and then the second traffic image is input into the intermediate image recognition model for training to obtain the first image recognition model, so that the first image recognition model has better target detection effect, and the model adaptability and robustness are good.

[0179] Based on the same inventive concept, the embodiments of the present application also provide a high-precision emergency parking strip intelligent perception analysis device corresponding to the high-precision emergency parking strip intelligent perception analysis method. Since the principle of solving the problem by the device in the embodiments of the present application is similar to the above-mentioned high-precision emergency parking strip intelligent perception analysis method in the embodiments of the present application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be repeated.

[0180] Reference Fig. 9 As shown, it is a structural schematic diagram of a high-precision emergency parking strip intelligent perception analysis device provided in an embodiment of the present application, and the device includes: an acquisition module 901, a first determination module 902, a second determination module 903, a binding module 904 and a generation module 905; wherein:

[0181] An acquisition module 901 is used to acquire the parking strip image at the current moment sent by an image acquisition device;

[0182] A first determination module 902 is used to determine at least one target vehicle information in the parking strip image based on a pre-trained first image recognition model, and to determine at least one license plate information in the parking strip image based on a pre-trained second image recognition model, wherein the target vehicle information includes the number of vehicles, the vehicle image and the vehicle position, and the license plate information includes the license plate number, the license plate position, the license plate type and the license plate collection time;

[0183] The second determination module 903 is used to determine the vehicle tracking information of the vehicle at the current moment according to the target vehicle information in the parking strip image and the vehicle tracking information before the current moment, the vehicle tracking information includes the vehicle tracking identification, the vehicle type and the vehicle driving information, and the vehicle driving information includes the vehicle position and the collection time corresponding to the vehicle position;

[0184] The binding module 904 is used to bind the vehicle tracking information and the license plate information to generate a plurality of vehicle binding information, each of which includes: the vehicle tracking information and the corresponding license plate information;

[0185] The generating module 905 is used to generate a control signal according to the vehicle binding information, and send the control signal to a plurality of control devices arranged in the parking strip area, so that the control devices output prompt information according to the control signal.

[0186] Optionally, the binding module 904 is specifically used for:

[0187] Traverse the collection time of each license plate in the first license plate information, and for the current license plate collection time traversed, if the collection time corresponding to the current license plate collection time and the first vehicle position in the first vehicle driving information is less than a preset time threshold, determine whether the distance between the license plate position corresponding to the current license plate collection time in the first license plate information and the first vehicle position in the first vehicle driving information is less than a preset distance. If so, bind the first license plate information and the vehicle tracking information corresponding to the first vehicle driving information to generate vehicle binding information corresponding to the first license plate information.

[0188] Optionally, the binding module 904 is specifically used for:

[0189] Traverse the collection time of each license plate in the first license plate information, and for the current license plate collection time collected, if the collection time corresponding to the current license plate collection time and the first vehicle position in the first vehicle driving information is less than a preset time threshold, then determine whether the distance between the license plate position corresponding to the current license plate collection time in the first license plate information and the first vehicle position in the first vehicle driving information is less than a preset distance. If so, determine whether the license plate type of the first license plate information and the vehicle type in the first vehicle driving information satisfy a preset matching relationship. If so, bind the first license plate information and the vehicle tracking information corresponding to the first vehicle driving information to generate vehicle binding information corresponding to the first license plate information.

[0190] Optionally, the binding module 904 is further configured to:

[0191] Traversing each vehicle binding information, for the traversed first vehicle binding information and second vehicle binding information, determining a collection distance between a vehicle position with the latest collection time in the first vehicle binding information and a vehicle position with the earliest collection time in the second vehicle binding information;

[0192] Determine whether the collection distance is less than a preset distance threshold;

[0193] If so, determining whether the vehicle type in the first vehicle binding information and the vehicle type in the second vehicle binding information satisfy a preset matching relationship;

[0194] If so, the vehicle tracking identifier in the first vehicle binding information and the license plate number and license plate type in the license plate information are used as the vehicle tracking identifier in the second vehicle binding information and the license plate number and license plate type in the license plate information.

[0195] Optionally, the binding module 904 is further configured to:

[0196] Based on a multi-source shortest path algorithm, determining a first binding result according to the vehicle position with the latest collection time in each of the vehicle binding information and the vehicle position with the earliest collection time in each of the vehicle binding information, wherein the first binding result includes a plurality of vehicle binding information corresponding to each other;

[0197] Determine a target binding result according to the vehicle type in each vehicle binding information in the first binding result, wherein the target binding result includes at least one vehicle binding information before the change and vehicle binding information after the change corresponding to the vehicle binding information before the change;

[0198] The vehicle tracking identifier in each vehicle binding information before the change and the license plate number and license plate type in the license plate information are used as the vehicle tracking identifier in each vehicle binding information after the change and the license plate number and license plate type in the license plate information.

[0199] Optionally, the binding module 904 is further configured to:

[0200] Determine whether there is a lost vehicle based on the vehicle binding information;

[0201] If so, the binding information of the lost vehicle is determined according to the number of vehicles in the target vehicle information, the number of vehicles in the target vehicle information at the previous moment and the vehicle binding information of each vehicle, and the binding information of the lost vehicle is used as one vehicle binding information.

[0202] Optionally, the generating module 905 is specifically used for:

[0203] Determine the vehicle status information according to the vehicle position in the vehicle binding information and the collection time corresponding to the vehicle position, the vehicle status information includes: vehicle entering, vehicle slowing down, vehicle stopping and vehicle exiting;

[0204] generating an exit prompt signal according to the vehicle driving information and vehicle state information in the vehicle binding information of each vehicle, wherein the exit prompt signal is used to control the control device to send the vehicle exit prompt information to the lane vehicle when the vehicle is in the vehicle exit state;

[0205] Generate a distance control signal according to the vehicle driving information and vehicle status information in the vehicle binding information of each vehicle, the distance control signal is used to control the control device to output distance prompt information when the distance between the vehicle and the edge area of ​​the parking strip is less than a preset distance;

[0206] The vehicle stay time is determined according to the collection time corresponding to the vehicle position in the vehicle binding information of each vehicle, and a rescue control signal is generated according to the vehicle stay time. The rescue control signal is used to control the control device to output rescue inquiry information when the vehicle stay time is greater than the preset stay time.

[0207] Optionally, the first determining module 902 is specifically configured to:

[0208] Preprocessing the first traffic image, and using the preprocessed first traffic image and image identification information corresponding to the first traffic image as a first training sample, wherein the preprocessing includes: reducing image brightness, reducing image contrast, and increasing image blur;

[0209] Training the initial image recognition model based on the first training sample input to obtain an intermediate image recognition model;

[0210] The second traffic image and the image identification information corresponding to the second traffic image are used as second training samples, and the intermediate image recognition model is trained based on the second training samples to obtain a first image recognition model.

[0211] For descriptions of the processing flow of each module in the device and the interaction flow between each module, reference may be made to the relevant descriptions in the above method embodiment, which will not be described in detail here.

[0212] The present application also provides an electronic device, such as Fig.10 1 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application, including: a processor 1001, a memory 1002 and a bus. The memory 1002 stores machine-readable instructions executable by the processor 1001 (for example, Fig. 9 In the device, the acquisition module 901, the first determination module 902, the second determination module 903, the binding module 904 and the generation module 905 corresponding to the execution instructions, etc.), when the computer device is running, the processor 1001 and the memory 1002 communicate through the bus, and when the machine-readable instructions are executed by the processor 1001, the above-mentioned high-precision emergency parking strip intelligent perception analysis method is executed.

[0213] The embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the above-mentioned high-precision emergency parking strip intelligent perception analysis method are executed.

[0214] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, the specific working process of the system and device described above can refer to the corresponding process in the method embodiment, and will not be repeated in this application. In the several embodiments provided in this application, it should be understood that the disclosed system, device and method can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0215] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or part of the technical solution that contributes to the prior art or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccess Memory), disk or optical disk and other media that can store program code.

[0216] The above are only specific implementation methods of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be covered by the protection scope of the present application.

Claims

1. A high-precision intelligent perception analysis method for emergency parking strips, characterized in that: The method comprises: Obtaining the parking strip image at the current moment sent by the image acquisition device; Determine at least one target vehicle information in the parking strip image based on a pre-trained first image recognition model, and determine at least one license plate information in the parking strip image based on a pre-trained second image recognition model, wherein the target vehicle information includes the number of vehicles, the vehicle image, and the vehicle position, and the license plate information includes the license plate number, the license plate position, the license plate type, and the license plate collection time; Determine the vehicle tracking information of the vehicle at the current moment according to the target vehicle information in the parking strip image and the vehicle tracking information before the current moment, wherein the vehicle tracking information includes a vehicle tracking identifier, a vehicle type and vehicle driving information, and the vehicle driving information includes a vehicle position and a collection time corresponding to the vehicle position; Binding the vehicle tracking information and the license plate information to generate a plurality of vehicle binding information, each of the vehicle binding information includes: the vehicle tracking information and the corresponding license plate information; generating a control signal according to the vehicle binding information, and sending the control signal to a plurality of control devices arranged in the parking strip area, so that the control devices output prompt information according to the control signal; The generating a control signal according to the vehicle binding information comprises: Determine vehicle status information according to the vehicle position in the vehicle binding information and the collection time corresponding to the vehicle position, wherein the vehicle status information includes: vehicle entering, vehicle slowing down, vehicle stopping and vehicle exiting; Generate an exit prompt signal according to the vehicle driving information and the vehicle state information in the vehicle binding information of each vehicle, wherein the exit prompt signal is used to control the control device to send the vehicle exit prompt information to the lane vehicle when the vehicle is in the vehicle exit state; Generate a distance control signal according to the vehicle driving information and the vehicle status information in the vehicle binding information of each vehicle, wherein the distance control signal is used to control the control device to output distance prompt information when the distance between the vehicle and the edge area of ​​the parking strip is less than a preset distance; The vehicle stay time is determined according to the collection time corresponding to the vehicle position in the vehicle binding information of each vehicle, and a rescue control signal is generated according to the vehicle stay time. The rescue control signal is used to control the control device to output rescue inquiry information when the vehicle stay time is greater than the preset stay time.

2. The high-precision emergency parking strip intelligent perception analysis method according to claim 1 is characterized in that: The step of binding the vehicle tracking information and the license plate information to generate vehicle binding information includes: Traverse the collection time of each license plate in the first license plate information, and for the traversed current license plate collection time, if the collection time corresponding to the current license plate collection time and the first vehicle position in the first vehicle driving information is less than a preset time threshold, determine whether the distance between the license plate position corresponding to the current license plate collection time in the first license plate information and the first vehicle position in the first vehicle driving information is less than a preset distance. If so, bind the first license plate information and the vehicle tracking information corresponding to the first vehicle driving information to generate vehicle binding information corresponding to the first license plate information.

3. The high-precision emergency parking zone intelligent perception analysis method according to claim 1 is characterized in that: The step of binding the vehicle tracking information and the license plate information to generate vehicle binding information includes: Traverse the collection time of each license plate in the first license plate information, and for the current license plate collection time collected, if the collection time corresponding to the current license plate collection time and the first vehicle position in the first vehicle driving information is less than a preset time threshold, determine whether the distance between the license plate position corresponding to the current license plate collection time in the first license plate information and the first vehicle position in the first vehicle driving information is less than a preset distance; if so, determine whether the license plate type of the first license plate information and the vehicle type in the first vehicle driving information satisfy a preset matching relationship; if so, bind the first license plate information and the vehicle tracking information corresponding to the first vehicle driving information to generate vehicle binding information corresponding to the first license plate information.

4. The high-precision intelligent perception analysis method for emergency parking strips according to claim 1 is characterized in that: After binding the vehicle tracking information and the license plate information to generate vehicle binding information, the method further includes: Traversing each vehicle binding information, and for the traversed first vehicle binding information and second vehicle binding information, determining a collection distance between a vehicle position with the latest collection time in the first vehicle binding information and a vehicle position with the earliest collection time in the second vehicle binding information; Determine whether the acquisition distance is less than a preset distance threshold; If yes, determining whether the vehicle type in the first vehicle binding information and the vehicle type in the second vehicle binding information satisfy a preset matching relationship; If so, the vehicle tracking identifier in the first vehicle binding information and the license plate number and license plate type in the license plate information are used as the vehicle tracking identifier in the second vehicle binding information and the license plate number and license plate type in the license plate information.

5. The high-precision intelligent perception analysis method for emergency parking strips according to claim 1 is characterized in that: After binding the vehicle tracking information and the license plate information to generate vehicle binding information, the method further includes: Based on a multi-source shortest path algorithm, determining a first binding result according to the vehicle position with the latest collection time in each of the vehicle binding information and the vehicle position with the earliest collection time in each of the vehicle binding information, wherein the first binding result includes a plurality of vehicle binding information corresponding to each other; Determine a target binding result according to the vehicle type in each vehicle binding information in the first binding result, wherein the target binding result includes at least one vehicle binding information before the change and vehicle binding information after the change corresponding to the vehicle binding information before the change; The vehicle tracking identifier in each vehicle binding information before the change and the license plate number and license plate type in the license plate information are used as the vehicle tracking identifier in each vehicle binding information after the change and the license plate number and license plate type in the license plate information.

6. The high-precision emergency parking zone intelligent perception analysis method according to claim 1 is characterized in that: After binding the vehicle tracking information and the license plate information to generate vehicle binding information, the method further includes: Determining whether there is a lost vehicle based on the vehicle binding information; If so, the binding information of the lost vehicle is determined according to the number of vehicles in the target vehicle information, the number of vehicles in the target vehicle information at the previous moment and the vehicle binding information of each vehicle, and the binding information of the lost vehicle is used as the vehicle binding information.

7. The high-precision intelligent perception analysis method for emergency parking strips according to claim 1 is characterized in that: The training process of the first image recognition model includes: Preprocessing the first traffic image, and using the preprocessed first traffic image and image identification information corresponding to the first traffic image as a first training sample, wherein the preprocessing includes: reducing image brightness, reducing image contrast, and increasing image blur; Training the initial image recognition model based on the first training sample input to obtain an intermediate image recognition model; The second traffic image and the image identification information corresponding to the second traffic image are used as second training samples, and the intermediate image recognition model is trained based on the second training samples to obtain a first image recognition model.

8. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor executes the machine-readable instructions to perform the steps of the high-precision emergency stop strip intelligent perception analysis method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, executes the steps of the high-precision emergency stop strip intelligent perception analysis method as claimed in any one of claims 1 to 6.

Images