Parking lot vehicle identification method and device, computer device, and storage medium

Abstract

The application relates to a parking lot vehicle identification method and device, computer equipment and a storage medium, the method comprising: monitoring the parking lot through a device group, obtaining the image of an incoming vehicle and the time when the vehicle enters when the vehicle passes through an entrance; obtaining the driving track of the incoming vehicle and detecting whether the incoming vehicle stops at an exit within a first time threshold, if yes, releasing, if not, recording the parking point of the incoming vehicle; recording the parking point of an outgoing vehicle when the vehicle moves from the parking point, and obtaining the image of the outgoing vehicle and the time when the vehicle leaves when the vehicle stops at an exit gate; matching the incoming vehicle and the outgoing vehicle by comparing the image and the parking point; calculating the presence time of the vehicle and the required payment through the time when the vehicle enters and the time when the vehicle leaves after the matching is completed, and releasing after the payment is received. The method does not need to detect the license plate and is suitable for the identification of vehicles without license plates, temporary license plates and pseudo license plates.

Description

Technical Field

[0001] This invention relates to the field of vehicle identification technology, and in particular to a method, apparatus, computer equipment, and storage medium for vehicle identification in parking lots. Background Technology

[0002] In recent years, with the continuous development of modern technology, more and more smart parking lots have appeared in cities. Smart parking lots use vehicle recognition technology to obtain the time when a vehicle is parked in the parking lot and calculate the fee to be paid. It is convenient and fast, and there is no need for manual card taking or gate opening.

[0003] Vehicle recognition technology is key to improving vehicle management efficiency in smart parking lots. Traditional vehicle recognition technology typically relies on license plate recognition. This involves analyzing dynamic video or images of entering vehicles to automatically identify the license plate number and color, extract the license plate image, and automatically segment and recognize the characters. With repeated updates and iterations, the accuracy of vehicle recognition technology has gradually increased, leading to continuous improvements in vehicle entry and exit efficiency.

[0004] However, traditional vehicle recognition technologies still have problems such as: allowing vehicles using counterfeit license plates to pass through, having no target to identify vehicles without license plates, and reporting too much or too little information after identifying vehicles using temporary license plates. Summary of the Invention

[0005] Therefore, it is necessary to provide a parking lot vehicle identification method, device, computer equipment, and storage medium that can complete the identification and control of vehicles entering and exiting the parking lot without detecting license plates, thereby solving the problem of identifying vehicles without license plates, fake license plates, and temporary license plates.

[0006] A parking lot vehicle identification method includes:

[0007] The parking lot is monitored by a group of devices, and the first image of the vehicle and the time of its entry are obtained when the vehicle enters the parking lot.

[0008] The driving trajectory of the entering vehicle is obtained, and it is detected whether the entering vehicle stops at the exit within the first time threshold. If yes, it is allowed to leave; otherwise, the parking point of the entering vehicle is recorded.

[0009] When the exiting vehicle moves from the second parking point, the parking point of the exiting vehicle is recorded, and when the exiting vehicle stops at the exit, the second image of the exiting vehicle and the exit time are obtained.

[0010] The entering vehicle is matched with the exit vehicle by comparing the first parking location with the second parking location and comparing the first image with the second image.

[0011] A parking order is generated based on the vehicle's entry and exit times after the matching is completed.

[0012] In one embodiment, it further includes:

[0013] Acquire multiple vehicle images and extract individual features from the vehicle images. The individual features include one or more of the following: color, body, logo, windows, and interior trim.

[0014] The training set is composed of individual features from multiple vehicle images. The training is performed based on the influence of these individual features on vehicle recognition, and the weight values ​​of each individual feature are obtained.

[0015] In one embodiment, the monitoring of the parking lot via a group of devices, acquiring a first image of the entering vehicle and its entry time upon entry, further includes:

[0016] Extract the first volumetric features of each item on the first image. The first volumetric features include one or more of the following: color, body, model, logo, windows, and interior trim.

[0017] The first volumetric features of each item are marked on the first image using detection boxes, and the first recognition success rate of each first volumetric feature is calculated.

[0018] The first weight of each first feature is calculated based on the first recognition success rate and the weight value.

[0019] In one embodiment, the process of recording the second parking position of the departing vehicle as it moves from the second parking position, and acquiring a second image of the departing vehicle and its departure time when it stops at the exit, further includes:

[0020] Extract the second volume features from the second image. The second volume features include one or more of the following: color, body, model, logo, windows, and interior trim.

[0021] The second features are marked on the second image using detection boxes, and the second recognition success rate of each second volume feature is calculated.

[0022] The second weight of each second feature is calculated based on the second recognition success rate and the weight value.

[0023] In one embodiment, the step of matching the entering vehicle with the exiting vehicle by comparing the first image and the second image based on the first parking location and the second parking location further includes:

[0024] Obtain the first weight of the first volumetric feature of each item in the first image;

[0025] Obtain the second weight of each of the second volumetric features in the second image;

[0026] The vehicles entering the venue are matched with the vehicles leaving the venue after calculation based on the first and second proportions.

[0027] In one embodiment, after calculating the second weight of each second volume feature based on the second recognition success rate and the weight value, the method further includes:

[0028] Calculate the average value of the first proportion of the first volumetric feature of each item in the first image of the entering vehicle and record it as Pin; and calculate the average value of the second proportion of the second volumetric feature of each item in the second image of the exiting vehicle and record it as Pout.

[0029] The second image and second parking location of the exiting vehicle are compared with the first image and first parking location of the entering vehicle in the system to confirm the entering vehicle corresponding to the exiting vehicle.

[0030] The Pin value of the entering vehicle and the Pout value of the exiting vehicle are compared successively by feature error. If the error value is within a preset range, the matching is completed; if the error value is not within the preset range, the vehicle is not allowed to pass.

[0031] In one embodiment, the process of calculating the vehicle's on-site time and required fee based on the vehicle's entry and exit times after matching, and releasing the vehicle upon receiving payment, further includes:

[0032] Clear the first image, entry time, and first parking point of the entering vehicle, and the second image, exit time, and second parking point of the exiting vehicle.

[0033] A parking lot vehicle identification device includes:

[0034] The acquisition module is used to monitor the parking lot through a group of devices, and to acquire the first image of the vehicle and the time of entry when the vehicle enters the parking lot.

[0035] The data acquisition module is also used to acquire the driving trajectory of the entering vehicle and detect whether the entering vehicle stops at the exit within a first time threshold. If yes, it is allowed to leave; otherwise, the first stopping point of the entering vehicle is recorded.

[0036] The acquisition module is also used to record the second parking point of the exiting vehicle when the exiting vehicle moves from the second parking point, and to acquire the second image of the exiting vehicle and the exit time when the exiting vehicle stops at the exit.

[0037] The comparison module is used to match the entering vehicle with the exit vehicle by comparing the first image and the second image based on the first parking location and the second parking location;

[0038] The calculation module is used to generate parking orders by matching the entry and exit times of the vehicles.

[0039] In one embodiment, a computer device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to perform the following steps:

[0040] The parking lot is monitored by a group of devices, and the first image of the vehicle and the time of its entry are obtained when the vehicle enters the parking lot.

[0041] The driving trajectory of the entering vehicle is obtained, and it is detected whether the entering vehicle stops at the exit within the first time threshold. If yes, it is allowed to leave; otherwise, the first stopping point of the entering vehicle is recorded.

[0042] When the exiting vehicle moves from the second parking point, the second parking point of the exiting vehicle is recorded, and when the exiting vehicle stops at the exit, the second image of the exiting vehicle and the exit time are obtained.

[0043] The entering vehicle is matched with the exit vehicle by comparing the first parking location with the second parking location and comparing the first image with the second image.

[0044] A parking order is generated based on the vehicle's entry and exit times after the matching is completed.

[0045] In one embodiment, a computer-readable storage medium stores a computer program, characterized in that, when executed by a processor, the computer program performs the following steps:

[0046] The parking lot is monitored by a group of devices, and the first image of the vehicle and the time of its entry are obtained when the vehicle enters the parking lot.

[0047] The driving trajectory of the entering vehicle is obtained, and it is detected whether the entering vehicle stops at the exit within the first time threshold. If yes, it is allowed to leave; otherwise, the first stopping point of the entering vehicle is recorded.

[0048] When the exiting vehicle moves from the second parking point, the second parking point of the exiting vehicle is recorded, and when the exiting vehicle stops at the exit, the second image of the exiting vehicle and the exit time are obtained.

[0049] The entering vehicle is matched with the exit vehicle by comparing the first parking location with the second parking location and comparing the first image with the second image.

[0050] A parking order is generated based on the vehicle's entry and exit times after the matching is completed.

[0051] In one embodiment, a computer program product includes a computer program that, when executed by a processor, performs the following steps:

[0052] The parking lot is monitored by a group of devices, and the first image of the vehicle and the time of its entry are obtained when the vehicle enters the parking lot.

[0053] The driving trajectory of the entering vehicle is obtained, and it is detected whether the entering vehicle stops at the exit within the first time threshold. If yes, it is allowed to leave; otherwise, the first stopping point of the entering vehicle is recorded.

[0054] When the exiting vehicle moves from the second parking point, the second parking point of the exiting vehicle is recorded, and when the exiting vehicle stops at the exit, the second image of the exiting vehicle and the exit time are obtained.

[0055] The entering vehicle is matched with the exit vehicle by comparing the first parking location with the second parking location and comparing the first image with the second image.

[0056] A parking order is generated based on the vehicle's entry and exit times after the matching is completed.

[0057] The aforementioned parking lot vehicle identification method, device, computer equipment, and storage medium acquire a first image, entry time, and second parking position of an entering vehicle. Vehicles leaving the parking lot within a first time threshold are directly allowed entry. For vehicles that have not left the parking lot within the first time threshold, the second parking position of the exiting vehicle, its second image at the exit, and its exit time are acquired to match the corresponding entering vehicle. The parking time is calculated using the exit and entry times for charging. This application can complete vehicle entry and exit identification and control without license plate detection, and can identify vehicles without license plates or using temporary license plates. For vehicles using counterfeit license plates, because the second image and parking position of the counterfeit license plate vehicle are inconsistent with the first image and parking position of the vehicle with the counterfeit license plate, it is not allowed to enter. Therefore, this application can solve the problem of parking lot entry and exit identification of vehicles without license plates, with counterfeit license plates, and with temporary license plates. Attached Figure Description

[0058] Figure 1 This is one of the flowcharts for the parking lot vehicle identification method provided in this application;

[0059] Figure 2The second flowchart of the parking lot vehicle identification method provided in this application;

[0060] Figure 3 The third flowchart of the parking lot vehicle identification method provided for this application;

[0061] Figure 4 The fourth flowchart of the parking lot vehicle identification method provided for this application;

[0062] Figure 5 The fifth flowchart of the parking lot vehicle identification method provided in this application;

[0063] Figure 6 The sixth flowchart of the parking lot vehicle identification method provided in this application;

[0064] Figure 7 A schematic diagram of the parking lot vehicle identification device module provided in this application;

[0065] Figure 8 An internal structural diagram of the computer device provided in this application. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0067] like Figure 1 As shown, in one embodiment, a parking lot vehicle identification method includes the following steps:

[0068] Step S110: Monitor the parking lot through the equipment group, and acquire the first image of the entering vehicle and the entry time when the vehicle enters the parking lot.

[0069] Specifically, the parking lot is monitored by cameras or other devices. When a vehicle stops at the entrance, it is marked as an entering vehicle. The camera with built-in target detection algorithm or other algorithms captures the vehicle and integrates the multiple frames to form the first image. The time when the vehicle enters the parking lot is recorded as the entry time.

[0070] Step S120: Obtain the driving trajectory of the entering vehicle and detect whether the entering vehicle stops at the exit within the first time threshold. If yes, allow it to leave; otherwise, record the first parking point of the entering vehicle.

[0071] Specifically, after a vehicle enters the parking lot, a camera with a built-in target tracking algorithm or other algorithms continuously tracks the vehicle to obtain its driving trajectory. If the vehicle leaves the parking lot within the first time threshold, it is allowed to leave directly, and the first image and entry time data are cleared. If the vehicle stays in the parking lot for more than the first time threshold, the parking location of the vehicle is recorded.

[0072] It should be noted that the camera that acquires the first image or dynamic image can be replaced by one or more interconnected drones or a group of cruise robots. The drones or cruise robots need to be able to withstand sufficient traffic statistics analysis, such as JS Super Eye, JS High-Speed ​​Tracking Camera, JS Adaptive Telescopic Camera, and Xiaojie Cruise Robot.

[0073] It should be noted that the definition of entering vehicles mentioned above has a time domain, from the moment the vehicle enters the parking lot to the moment the vehicle is detected stopping at the exit within the first time threshold. The time domain of exiting vehicles mentioned below is from the moment a vehicle that has exceeded the first time threshold moves from its parking position to the moment it stops at the exit.

[0074] Step S130: When the exiting vehicle moves from the second parking point, record the second parking point of the exiting vehicle, and when the exiting vehicle stops at the exit, obtain the second image of the exiting vehicle and the exit time.

[0075] Specifically, during the monitoring of the parking lot, when a vehicle is detected moving from its parking spot, it is marked as an exiting vehicle. The second parking spot of the exiting vehicle is recorded. When the exiting vehicle stops at the exit, the camera with a built-in target detection algorithm or other algorithms captures the exiting vehicle, and the multiple frames are integrated into an image. The time when the exiting vehicle stops at the exit is recorded as the exit time.

[0076] Step S140: Match the entering vehicle with the exit vehicle by comparing the first parking location with the second parking location and the first image with the second image.

[0077] Specifically, based on the second image and second parking point of the exiting vehicle, the first image and first parking point of the corresponding entering vehicle are found in the system to complete the matching of exiting and entering vehicles.

[0078] Step S150: Generate a parking order by matching the vehicle's entry and exit times.

[0079] Specifically, the exiting and entering vehicles after matching are the same vehicle. The difference between the exit time and the entry time of the vehicle is the vehicle's time on site. The fee required for the vehicle is calculated according to the parking lot's charging standard and displayed to the vehicle owner. Once the fee is received from the vehicle owner, the vehicle can be released.

[0080] The aforementioned parking lot vehicle identification method acquires the first image, entry time, and first parking position of an entering vehicle. Vehicles leaving the parking lot within a first time threshold are directly allowed to enter. For vehicles that have not left the parking lot within the first time threshold, the method acquires the second parking position of the exiting vehicle, as well as the second image and exit time of the exiting vehicle, to match the corresponding entering vehicle and generate a parking order based on the exit and entry times. This method can complete the identification and control of vehicle entry and exit without license plate detection and can identify vehicles without license plates or using temporary license plates. However, for vehicles using counterfeit license plates, since the second image and parking position of the counterfeit license plate vehicle are inconsistent with the first image and parking position of the vehicle with the counterfeit license plate, it is not allowed to enter. Therefore, this method can solve the problem of parking lot entry and exit identification of vehicles without license plates, counterfeit license plates, and temporary license plates.

[0081] It should be noted that current technologies typically match entering and exiting vehicles by recognizing license plate numbers, but this cannot control vehicles without license plates. Using image recognition to match entering and exiting vehicles allows for the management of vehicles without license plates entering and exiting the venue.

[0082] It should be noted that counterfeiting license plates refers to forging the license plates of vehicles with registration information within the venue by means of handwriting, taking photos with a mobile phone, etc., and then showing the counterfeit license plate to the exit gate when the vehicle leaves, thereby evading payment.

[0083] Furthermore, it is determined whether the first parking point and the second parking point are the same. If they are the same, the parking points of the entering vehicle and the exiting vehicle are matched. If they are different, it is checked whether the vehicle that moved from the first parking point to the second parking point is the same vehicle. If they are the same, the parking points of the entering vehicle and the exiting vehicle are matched.

[0084] It should be noted that if a vehicle enters and parks at the first parking spot, and then moves to the second parking spot after a period of time, and then drives out from the second parking spot and parks at the exit after a period of time, the system may detect that the first and second parking spots are different and may report an error. Therefore, the system checks whether the vehicle moving from the first parking spot to the second parking spot is the same vehicle and matches the vehicle at the first parking spot with the vehicle leaving from the second parking spot.

[0085] like Figure 2As shown, in this embodiment, the parking lot vehicle identification method further includes the following steps:

[0086] Step S101: Acquire multiple vehicle images and extract individual features from the vehicle images. The individual features include one or more of the following: color, body, logo, windows, and interior trim.

[0087] Specifically, vehicle images contain multiple individual features, including headlights, license plates, grille information, body color, vehicle logo, front and rear of the vehicle, windows, interior decorations, and interior furnishings. Under normal circumstances, if vehicles of the same specifications and model exist, matching can be achieved by comparing license plates.

[0088] Step S102: The individual features in multiple vehicle images are combined into a training set and trained according to the influence of the individual features on vehicle recognition to obtain the weight values ​​of each individual feature.

[0089] Specifically, the impact of each individual feature on vehicle recognition varies. Vehicle body color, grille information, and license plate have a significant impact. Grille information includes the grille shape and logo; license plate information includes the license plate color. Features are extracted from each vehicle image, preprocessed, and then divided into training and test sets. A convolutional neural network model is used to train the model on the training set and validated on the test set. The success rate of the model predicting vehicle matching based on each feature is calculated, resulting in the weight values ​​of each individual feature. For example, the weight of vehicle body color is 30%, and the sum of the weight values ​​of all individual features is 100%. This application is not limited to the figures mentioned and may be adjusted according to specific circumstances.

[0090] Specifically, when a vehicle enters or exits, its individual features are marked on the vehicle image using detection boxes. The system can quickly select each individual feature and compare it, improving the speed of recognition and release.

[0091] like Figure 3 As shown, in this embodiment, the monitoring of the parking lot via a group of devices, acquiring a first image of the entering vehicle and its entry time upon entry, is followed by the following steps:

[0092] Step S111: Extract the first volumetric features of each item on the first image. The first volumetric features include one or more of the following: color, body, model, logo, window, and interior trim.

[0093] Specifically, when a vehicle enters and stops at the entrance, an image of the vehicle is extracted to obtain the first image, and various individual features on the first image are extracted to obtain the first volume feature;

[0094] Step S112: Mark the first volumetric features of each item on the first image using the detection box, and calculate the first recognition success rate of each first volumetric feature;

[0095] Specifically, the recognition of each first feature is performed. Due to the different vehicle models, deviations in the fixed position when parked at the entrance, and the possible differences in the position and proportion of each first feature in the image, the recognition success rate of each first feature is calculated to obtain the first recognition success rate.

[0096] Step S113: Calculate the first weight of each first volume feature based on the first recognition success rate and the weight value.

[0097] Specifically, the first recognition success rate of each first volume feature is calculated along with its corresponding weight value to obtain the first proportion of each first volume feature. For example, the weight value of the vehicle body color among the features is 30%. When recognizing the first image, the first recognition success rate of the vehicle body color among the first volume features is 96%. The product of the weight value of the vehicle body color and the first recognition success rate of the vehicle body color is calculated to obtain the first proportion of the vehicle body color in the first image. Similarly, the first proportions corresponding to the first features such as vehicle body, vehicle model, vehicle logo, and vehicle windows in the first image are calculated separately.

[0098] like Figure 4 As shown, in this embodiment, when the exiting vehicle moves from the second parking point, the second parking point of the exiting vehicle is recorded, and when the exiting vehicle stops at the exit, a second image of the exiting vehicle and the exit time are acquired. The method further includes:

[0099] Step 131: Extract the second volume features of each item on the second image. The second volume features include one or more of the following: color, body, model, logo, windows, and interior trim.

[0100] Specifically, when a vehicle exits and stops at the exit, an image of the exiting vehicle is extracted to obtain a second image, and various individual features on the second image are extracted to obtain a second volume feature;

[0101] Step 132: Mark each second feature on the second image using the detection box, and calculate the second recognition success rate of each second volume feature;

[0102] Specifically, the recognition success rate of each second feature is calculated based on the different vehicle models, deviations in the fixed position when parked at the entrance, the different positions and proportions of each second feature in the image, and the influence of ambient light.

[0103] Step 133: Calculate the second weight of each second volume feature based on the second recognition success rate and the weight value.

[0104] Specifically, the second recognition success rate of each second volumetric feature is calculated along with its corresponding weight value to obtain the second proportion of each second volumetric feature. For example, the weight value of the vehicle body color among the features is 30%. When recognizing the second image, the second recognition success rate of the vehicle body color among the second volumetric features is 96%. The product of the weight value of the vehicle body color and the second recognition success rate of the vehicle body color is calculated to obtain the second proportion of the vehicle body color in the second image. Similarly, the second proportions corresponding to the second features such as vehicle body, vehicle model, vehicle logo, and vehicle windows in the second image are calculated separately.

[0105] like Figure 5 As shown, in this embodiment, the process of matching the entering vehicle with the exiting vehicle by comparing the first image and the second image based on the first parking location and the second parking location further includes:

[0106] Step S141: Obtain the first weight of the first volumetric feature of each item in the first image;

[0107] Step S142: Obtain the second weight of each of the second volumetric features in the second image;

[0108] Step S143: Match the entering vehicles with the exit vehicles after calculating based on the first and second proportions.

[0109] Specifically, the first weight and the second weight of each individual feature are calculated to obtain a matching value. If the matching value is within a preset range, the entering vehicle corresponding to the first image and the exit vehicle corresponding to the second image are successfully matched.

[0110] like Figure 6 As shown, in this embodiment,

[0111] Specifically, each individual feature has a different impact on vehicle recognition. For example, the body color has a greater impact on the recognition result than the window trim, meaning the body color has a larger weight. Based on the system's preset parameters, the weight of each individual feature in the acquired image is denoted as p.

[0112] Step S410: Calculate the average value of the first proportion of the first volumetric feature of each item in the first image of the entering vehicle and record it as Pin; and calculate the average value of the second proportion of the second volumetric feature of each item in the second image of the exiting vehicle and record it as Pout.

[0113] Specifically, the average value of the first proportion of each first volumetric feature in the first image of the entering vehicle is calculated and denoted as Pin, and the average value of the second proportion of each second volumetric feature in the second image of the exiting vehicle is calculated and denoted as Pout.

[0114] Step S420: Compare the second image and second parking point of the exiting vehicle with the first image and first parking point of the entering vehicle in the system to confirm the entering vehicle corresponding to the exiting vehicle.

[0115] Specifically, each exiting vehicle corresponds to an entering vehicle, and the images and parking locations of the two are the same. By comparing the images and parking locations, the system identifies the entering vehicle that corresponds to the exiting vehicle at the parking exit and matches the two.

[0116] It should be noted that if a vehicle enters and parks at the first parking spot, and then moves to the second parking spot after a period of time, and then drives out from the second parking spot and parks at the exit after a period of time, the system may detect that the first and second parking spots are different and may report an error. Therefore, the system checks whether the vehicle moving from the first parking spot to the second parking spot is the same vehicle and matches the vehicle at the first parking spot with the vehicle leaving from the second parking spot.

[0117] Furthermore, it is determined whether the first parking point and the second parking point are the same. If they are the same, the parking points of the entering vehicle and the exiting vehicle are matched. If they are different, it is checked whether the vehicle that moved from the first parking point to the second parking point is the same vehicle. If they are the same, the parking points of the entering vehicle and the exiting vehicle are matched.

[0118] Step S430: Perform successive feature error comparison between the Pin value of the entering vehicle and the Pout value of the exiting vehicle. If the error value is within the preset range, the matching is completed; if the error value is not within the preset range, the vehicle is not allowed to proceed.

[0119] Specifically, the successive feature errors are calculated using the root mean square error (RMSE) formula, as shown below:

[0120]

[0121] Where X is obs.i Observation, X model.i If the absolute difference between RMESPin and RMESPout is between 0.1 and 0.25, a match is completed; otherwise, the match is not allowed.

[0122] In this embodiment, a parking order is generated by matching the vehicle's entry and exit times, and the vehicle's on-site time and required fees are calculated. The vehicle is released after payment is received. The process also includes:

[0123] Clear the first image, entry time, and first parking point of the entering vehicle, and the second image, exit time, and second parking point of the exiting vehicle.

[0124] Specifically, after a vehicle pays its fee and leaves the parking lot, its image, entry time, exit time, and parking location are cleared to reduce system cache and prevent incorrect recognition when the vehicle re-enters the parking lot.

[0125] like Figure 7 As shown, in one embodiment, a parking lot vehicle identification device includes:

[0126] The acquisition module 510 is used to monitor the parking lot through the device group and acquire the first image of the vehicle and the entry time when the vehicle enters the parking lot.

[0127] Specifically, the parking lot is monitored by cameras or other devices. When a vehicle stops at the entrance, it is marked as an entering vehicle. The camera with built-in target detection algorithm or other algorithms captures the vehicle and integrates the multiple frames to form an image. The time when the vehicle enters the parking lot is recorded as the entry time.

[0128] The acquisition module 510 is also used to acquire the driving trajectory of the entering vehicle and detect whether the entering vehicle stops at the exit within a first time threshold. If yes, it is allowed to leave; otherwise, the first parking point of the entering vehicle is recorded.

[0129] Specifically, after a vehicle enters the parking lot, a camera with a built-in target tracking algorithm or other algorithms continuously tracks the vehicle to obtain its driving trajectory. If the vehicle leaves the parking lot within the first time threshold, it is allowed to leave directly, and the first image and entry time data are cleared. If the vehicle stays in the parking lot for more than the first time threshold, the first parking position of the vehicle is recorded.

[0130] The acquisition module 510 is also used to record the second parking position of the exiting vehicle when the exiting vehicle moves from the second parking position, and to acquire the second image of the exiting vehicle and the exit time when the exiting vehicle stops at the exit.

[0131] Specifically, during the monitoring of the parking lot, when a vehicle is detected moving from the second parking point, the vehicle is marked as an exit vehicle, the second parking point of the exit vehicle is recorded, and when the exit vehicle stops at the exit, the exit vehicle is captured by a camera with a built-in target detection algorithm or other algorithms, and multiple frames are integrated to form a second image. The time when the exit vehicle stops at the exit gate is recorded as the exit time.

[0132] The comparison module 520 is used to match the entering vehicle with the exit vehicle by comparing the first image and the second image based on the first parking location and the second parking location.

[0133] Specifically, based on the second image and second parking point of the exiting vehicle, the first image and first parking point of the corresponding entering vehicle are found in the system to complete the matching of exiting and entering vehicles.

[0134] The calculation module 530 is used to generate parking orders by matching the entry and exit times of the vehicles.

[0135] Specifically, the exiting and entering vehicles after matching are the same vehicle. The difference between the exit time and the entry time of the vehicle is the vehicle's time on site. The fee required for the vehicle is calculated according to the parking lot's charging standard and displayed to the vehicle owner. Once the fee is received from the vehicle owner, the vehicle can be released.

[0136] In one embodiment, a computer device is provided, which may be a smart terminal, and its internal structure diagram may be as follows: Figure 8 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a parking lot vehicle identification method.

[0137] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0138] In one embodiment, a computer device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps in the above method embodiments.

[0139] In one embodiment, a computer storage medium stores a computer program that, when executed by a processor, implements the steps described in the above method embodiments.

[0140] In one embodiment, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and executes the computer instructions, causing the computer device to perform the steps in the above method embodiments.

[0141] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0142] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0143] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for vehicle identification in a parking lot, characterized in that, The method includes: The parking lot is monitored by a group of devices, and the first image of the vehicle and the time of its entry are obtained when the vehicle enters the parking lot. Multiple vehicle images are acquired, and individual features are extracted from the vehicle images. The individual features from the multiple vehicle images are combined into a training set and trained according to the influence of the individual features on vehicle recognition to obtain the weight values ​​of each individual feature. The driving trajectory of the entering vehicle is obtained, and it is detected whether the entering vehicle stops at the exit within the first time threshold. If yes, it is allowed to leave; otherwise, the first stopping point of the entering vehicle is recorded. When the exiting vehicle moves from the second parking point, the second parking point of the exiting vehicle is recorded, and when the exiting vehicle stops at the exit, the second image of the exiting vehicle and the exit time are obtained. Extract each second volumetric feature from the second image, mark each second feature on the second image using a detection box, and calculate the second recognition success rate of each second volumetric feature; calculate the second weight of each second volumetric feature based on the second recognition success rate and the weight value. Matching the entering vehicle with the exiting vehicle by comparing the first image and the second image based on the first parking location and the second parking location; including: calculating the average value of the first proportion of each first volumetric feature in the first image of the entering vehicle and recording it as Pin, and calculating the average value of the second proportion of each second volumetric feature in the second image of the exiting vehicle and recording it as Pout; comparing the second image of the exiting vehicle and the second parking location with the first image and the first parking location of the entering vehicle in the system to confirm the entering vehicle corresponding to the exiting vehicle; performing successive feature error comparisons on the Pin value of the entering vehicle and the Pout value of the exiting vehicle, and if the error value is within a preset range, the matching is completed; if the error value is not within the preset range, the vehicle is not allowed to proceed. A parking order is generated based on the vehicle's entry and exit times after the matching is completed.

2. The parking lot vehicle identification method according to claim 1, characterized in that, The individual characteristics include one or more of the following: color, body, logo, windows, and interior trim.

3. The parking lot vehicle identification method according to claim 2, characterized in that, The monitoring of the parking lot via a group of devices, acquiring a first image of the entering vehicle and its entry time upon entry, further includes: Extract the first volumetric features of each item on the first image. The first volumetric features include one or more of the following: color, body, model, logo, windows, and interior trim. The first volumetric features of each item are marked on the first image using detection boxes, and the first recognition success rate of each first volumetric feature is calculated. The first weight of each first feature is calculated based on the first recognition success rate and the weight value.

4. The parking lot vehicle identification method according to claim 3, characterized in that, The second individual characteristic includes one or more of the following: color, body, model, logo, windows, and interior trim.

5. The parking lot vehicle identification method according to claim 4, characterized in that, The process of matching the entering vehicle with the exit vehicle by comparing the first image and the second image based on the first parking location and the second parking location further includes: Obtain the first weight of the first volumetric feature of each item in the first image; Obtain the second weight of each of the second volumetric features in the second image; The vehicles entering the venue are matched with the vehicles leaving the venue after calculation based on the first and second proportions.

6. The parking lot vehicle identification method according to claim 5, characterized in that, The process involves calculating the vehicle's on-site time and required fees based on the vehicle's entry and exit times after matching, and releasing the vehicle upon receipt of payment. This process also includes: Clear the first image, entry time, and first parking point of the entering vehicle, and the second image, exit time, and second parking point of the exiting vehicle.

7. A parking lot vehicle identification device, characterized in that, include: The acquisition module is used to monitor the parking lot through a group of devices, and to acquire the first image of the vehicle and the time of entry when the vehicle enters the parking lot. Multiple vehicle images are acquired, and individual features are extracted from the vehicle images. The individual features from the multiple vehicle images are combined into a training set and trained according to the influence of the individual features on vehicle recognition to obtain the weight values ​​of each individual feature. The data acquisition module is also used to acquire the driving trajectory of the entering vehicle and detect whether the entering vehicle stops at the exit within a first time threshold. If yes, it is allowed to leave; otherwise, the first stopping point of the entering vehicle is recorded. The acquisition module is also used to record the second parking point of the exiting vehicle when it moves from the second parking point, and to acquire the second image of the exiting vehicle and the exit time when it stops at the exit; to extract the second volumetric features of each item on the second image, to mark each second feature on the second image using a detection box, and to calculate the second recognition success rate of each second volumetric feature; and to calculate the second weight of each second volumetric feature based on the second recognition success rate and the weight value. The comparison module is used to match the entering vehicle with the exiting vehicle by comparing the first image and the second image based on the first parking location and the second parking location; including: calculating the average value of the first proportion of each first volumetric feature in the first image of the entering vehicle and recording it as Pin, and calculating the average value of the second proportion of each second volumetric feature in the second image of the exiting vehicle and recording it as Pout; comparing the second image of the exiting vehicle and the second parking location with the first image and the first parking location of the entering vehicle in the system to confirm the entering vehicle corresponding to the exiting vehicle; performing successive feature error comparison between the Pin value of the entering vehicle and the Pout value of the exiting vehicle, and if the error value is within a preset range, the matching is completed; if the error value is not within the preset range, the vehicle is not allowed to proceed. The calculation module is used to generate parking orders by matching the entry and exit times of the vehicles.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images