License plate recognition gate access management system based on multi-terminal cooperation and dynamic permission control

Abstract

The application discloses a license plate recognition gate access management system based on multi-terminal cooperation and dynamic permission control and relates to the gate management field, which comprises a collection module, a recognition module and the like.The collection module is used for collecting the license plate image of a vehicle driving into a garage, performing noise filtering and geometric calibration preprocessing operation on the license plate image, and outputting standardized license plate image data.The recognition module is used for receiving the standardized license plate image data, extracting the character contour feature and the texture feature of the license plate, completing the recognition of the license plate information based on the features, and outputting the license plate recognition result.The application improves the standardization degree through fine preprocessing of the license plate image, combines contour and texture feature weighted fusion and secondary recognition mechanism, greatly improves the recognition accuracy, reduces the recognition abnormal situation, establishes the quality check through the multi-terminal cooperation link, guarantees the stable and reliable instruction transmission, avoids the execution misplacement through the time synchronization and action delay compensation.

Description

Technical Field

[0001] This invention relates to the field of barrier gate management technology, specifically to a license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control. Background Technology

[0002] License plate recognition equipment and barrier gate control equipment are the core linkage components of the entrance and exit intelligent management system. The former obtains license plate information through image acquisition and algorithm recognition and transmits it in real time, while the latter receives the information, completes the authorization verification, and automatically drives the barrier gate to raise and lower based on the verification result, so as to quickly allow or block vehicles.

[0003] The invention patent application with application number 201010172005.4 discloses an automatic gate control system based on license plate information recognition and access management. The application aims to solve the problem that "in the prior art, the license plate recognition system is only used as an auxiliary means of the automatic gate system. It cannot achieve fully automatic control of the gate, nor can it achieve automatic management of vehicle passage. In the end, the operator still needs to manually control the gate to rise and fall after verifying the vehicle information, which greatly affects the passage efficiency of the gate".

[0004] However, in order to enhance the user's parking experience, some commercial parking garages have designated parking spaces with retractable or foldable barriers to ensure that users with designated parking spaces can park in their designated spaces after driving into the garage, and that their designated parking spaces are not occupied. Currently, most of these barrier facilities are remotely controlled by the user and do not interact with the gate.

[0005] To address this, we propose a license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control. Summary of the Invention

[0006] In view of the above-mentioned shortcomings of the existing technology, the present invention provides a license plate recognition gate access management system based on multi-terminal collaboration and dynamic permission control, which can effectively solve the problems of the existing technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions; This invention discloses a license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control, comprising: The acquisition module acquires license plate images of vehicles entering the garage, performs noise filtering and geometric calibration preprocessing on the license plate images to output standardized license plate image data. The recognition module receives the standardized license plate image data, extracts the character contour and texture features of the license plate, and identifies the license plate information based on these features, outputting the license plate recognition result. The parsing module receives the license plate recognition result, queries the preset parking space and license plate binding database to check if the license plate is bound to a specific parking space. If the query result is negative and a parking space is available, the gate opens, allowing the vehicle to enter. If the query result is positive, the gate opens, simultaneously triggering the establishment module to run. The system is used to establish a collaborative link between the gate control terminal, the dedicated parking space obstacle terminal, and the parsing module. Based on this link, the parsing module sends descent / folding control commands to the dedicated parking space obstacle terminal. The control module receives the control commands and, based on the vehicle from which the command originates, links with the parking garage's internal monitoring equipment to monitor the target's real-time location. When the target vehicle is detected to have reached the preset range of the corresponding dedicated parking space, it generates a dedicated parking space obstacle lifting / folding control signal to control the dedicated parking space obstacle terminal to complete the lifting / folding action. The management module stores license plate recognition data, parsing results, and control execution records, and provides access to and modification permissions for the parking space and license plate binding database. The acquisition module is interconnected with the identification module via a wireless network. The acquisition module and the identification module are interconnected with the parsing module via a wireless network. The parsing module is interconnected with the establishment module via a wireless network. The establishment module is interconnected with the control module via a wireless network. The control module is interconnected with the management module via a wireless network.

[0008] Furthermore, the noise filtering in the acquisition module follows the following rules: ; In the formula: This represents the pixel value at coordinates (x, y) after filtering. The radius of the filtering neighborhood; These are adaptive weighting coefficients; The original pixel value at coordinates (x+i, y+j) in the original license plate image; The geometric calibration is achieved through perspective transformation, and the calibrated coordinates (u,v) are solved using the following formula: ; In the formula: These are the parameters of the perspective transformation matrix.

[0009] Furthermore, in the character contour feature extraction stage of the recognition module, edge detection is used to obtain the character edge pixels, and contour feature parameters are calculated: ; In the formula: These are contour feature parameters; This represents the total number of pixels at the character edge. , The difference between the x-coordinate and y-coordinate of the m-th edge pixel and its adjacent pixels; Let be the angle between the gradient direction of the m-th edge pixel and the horizontal centerline of the character; Texture feature extraction: ; The recognition module performs a weighted fusion of contour feature parameter L and texture feature value T to obtain the recognition confidence level. ; In the formula: , where is the texture feature value; D and A are the number of distance values ​​and the number of angle values ​​in the gray-level co-occurrence matrix, respectively. Let the distance be d and the angle be... The gray-level co-occurrence matrix element values ​​at that time; The mean of all elements in the gray-level co-occurrence matrix; The standard deviation of all elements in the gray-level co-occurrence matrix; These are the feature weight coefficients; These are the preset maximum values ​​for contour feature parameters and preset maximum values ​​for texture feature parameters; when When the result exceeds the preset confidence threshold, the corresponding license plate recognition result is output.

[0010] Furthermore, when the recognition module recognizes license plate information, if the recognition results are inconsistent after a preset number of consecutive recognition attempts or the recognition confidence level is lower than a preset confidence threshold, a secondary recognition mechanism is triggered. The feature extraction range for the secondary recognition is adjusted using the following formula: ; In the formula: This refers to the feature extraction range for secondary recognition. This refers to the feature extraction range for the initial identification. This is a range adjustment factor; This represents the minimum permissible confidence level for identification. This represents the current confidence level.

[0011] Furthermore, the determination of whether there are vacant parking spaces in the garage in the analysis module is based on the interaction between the analysis module and the garage management terminal. The garage management terminal counts the vehicles entering and leaving in real time based on the gate to measure the number of remaining vacant parking spaces. Among them, the number of reserved parking spaces in the garage is not included in the range of the number of remaining vacant parking spaces in the garage management terminal.

[0012] Furthermore, during the collaborative link establishment phase of the establishment module, the legitimacy of the barrier gate control terminal and the dedicated parking space obstacle terminal is verified simultaneously. After successful authentication, a stable collaborative link is established based on link quality parameters, the link quality of which is characterized as follows: ; In the formula: These are link quality parameters; , , This refers to the link quality weighting coefficient. , , For link transmission bit error rate, link transmission bandwidth, and link transmission delay; The preset maximum allowable bit error rate, maximum transmission bandwidth, and maximum allowable transmission delay; when When the quality exceeds the preset link quality threshold, the collaborative link is confirmed to be successfully established. The control command issued by the parsing module to the dedicated parking space obstacle terminal includes a command identifier, a parking space identifier, and action parameters. The control command is synchronously verified for integrity using the following formula: ; In the formula: This is a verification code; To control the number of fields in the command; The value of the t-th field; Let t be the validation weight of the t-th field; The preset verification base; After receiving the control command, the dedicated parking space obstacle terminal calculates the check code. If the calculated check code matches the check code carried in the command, the corresponding action is executed. Otherwise, the request module executes the control command issuance operation again.

[0013] Furthermore, the preset range of dedicated parking spaces in the control module is a dynamically adjustable range, conforming to: ; In the formula: The dynamic range radius; The real-time speed of the target vehicle; This is the preset system response time; The base preset radius; When a vehicle is parked in a designated parking space, the obstacle terminal for the designated parking space will reset synchronously after the vehicle leaves, based on the built-in sensors.

[0014] Furthermore, the collaborative link between the barrier gate control terminal, the dedicated parking space obstacle terminal, and the analysis module is configured with a time synchronization mechanism: Local time correction for each terminal: ; In the formula: This is the corrected terminal time; The local time of the terminal; This is the difference between the terminal's local time and the system's base time. This is the synchronization correction coefficient; When the control module generates a dedicated parking space obstacle lifting / folding control signal, it incorporates action execution delay compensation, resulting in a compensated action execution time. ; In the formula: The estimated time for the target vehicle to arrive at the designated parking space within a preset range; To control signal transmission delay; The standard time taken for a parking space obstacle to complete the lifting / folding action.

[0015] Compared with the known prior art, the technical solution provided by this invention has the following beneficial effects: This invention improves the standardization of license plate images through fine preprocessing, and significantly improves recognition accuracy and reduces recognition anomalies by combining contour and texture feature weighted fusion and secondary recognition mechanism. Through the establishment and quality verification of multi-terminal collaborative links, it ensures stable and reliable instruction transmission. Time synchronization and action delay compensation avoid execution misalignment. The smoothly dynamically adjusted parking space preset range adapts to the vehicle driving status, making parking space obstacle control more timely and accurate. Furthermore, the classification and management of dedicated parking spaces and vacant parking spaces improves garage utilization efficiency and optimizes the overall vehicle entry and exit process. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0017] Figure 1 This is a schematic diagram of the license plate recognition gate access management system based on multi-terminal collaboration and dynamic access control. Detailed Implementation

[0018] 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, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0019] The present invention will be further described below with reference to embodiments. Example

[0020] This embodiment describes a license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control, such as... Figure 1 As shown, it includes: The acquisition module is used to acquire license plate images of vehicles entering the garage, and to perform preprocessing operations such as noise filtering and geometric calibration on the license plate images to output standardized license plate image data. The noise filtering in the acquisition module follows the following rules: ; In the formula: This represents the pixel value at coordinates (x, y) after filtering. The radius of the filtering neighborhood; These are adaptive weighting coefficients; The original pixel value at coordinates (x+i, y+j) in the original license plate image; The above formula is used to accurately filter out noise in license plate images while preserving key character information of the license plate to the greatest extent. By introducing an adaptive weight coefficient, the coefficient is dynamically adjusted according to the mean and standard deviation of pixels in the neighborhood. The smaller the difference between the pixel value and the mean, the greater the weight, and the greater the difference, the smaller the weight. This effectively suppresses noise interference and avoids the loss of details caused by fixed weight filtering, making the filtered image more suitable for subsequent recognition needs. in, ,and ,in These represent the mean and standard deviation of pixels within the neighborhood, respectively. Geometric calibration is achieved through perspective transformation, and the calibrated coordinates (u,v) are solved using the following formula: ; In the formula: These are the parameters of the perspective transformation matrix; The above formula addresses the tilt and distortion issues that may occur when license plates are photographed from different angles. By fitting the coordinates of four preset corner points in the actual license plate image with the corresponding corner points of the standard license plate, the perspective transformation matrix parameters are obtained. These parameters are then used to perform coordinate transformation on the distorted license plate, correcting it into a standardized and regular image, thus laying an accurate image foundation for subsequent character feature extraction and recognition. in, It is obtained by fitting the coordinates of four pre-defined corner points in the license plate image with the corner point coordinates of a standard license plate; The recognition module is used to receive standardized license plate image data, extract the character contour features and texture features of the license plate, complete the recognition of license plate information based on the features, and output the license plate recognition result; In the character contour feature extraction stage of the recognition module, edge detection is used to obtain the character edge pixels, and contour feature parameters are calculated. ; In the formula: These are contour feature parameters; This represents the total number of pixels at the character edge. , The difference between the x-coordinate and y-coordinate of the m-th edge pixel and its adjacent pixels; Let be the angle between the gradient direction of the m-th edge pixel and the horizontal centerline of the character; The above formula comprehensively and accurately captures the shape features of license plate characters. After obtaining the edge pixels of the characters through edge detection, it comprehensively considers the total number of edge pixels, the difference between the horizontal and vertical coordinates of each edge pixel and its adjacent pixels, and the angle between the gradient direction of the pixel and the horizontal central axis of the character. This information is integrated and calculated to form quantitative parameters that can completely represent the character outline, providing support for character recognition. Specifically, The cosine distance represents the Euclidean distance between the m-th edge pixel and its adjacent pixels, reflecting the spatial continuity of the character outline; The cosine of the angle between the gradient direction of the m-th edge pixel and the horizontal midline of the character is given when the gradient direction is parallel to the horizontal midline. When the value is close to 1, the pixel contributes significantly to the contour features. When the gradient direction is perpendicular to the horizontal midline, cos... A value close to 0 indicates a small contribution weight for that pixel. This design is based on the typical features of license plate characters, namely that the strokes of the characters mostly extend horizontally. Horizontal contour information has a higher discriminative power for character recognition. This is achieved through cosine... Weighting can highlight the horizontal contour features, suppress the interference of vertical edge noise, and improve the ability of the contour feature parameter L to represent the character shape; the weighted distance of all edge pixels is averaged to form a quantitative parameter that can fully reflect the spatial distribution and directional characteristics of the character contour. Texture feature extraction: ; The above formula, based on the core characteristics of the gray-level co-occurrence matrix and combined with different values ​​of distance and angle, quantifies the texture distribution characteristics of characters by calculating the correlation between the matrix element values ​​and the mean and standard deviation of all elements under each value, forming unique texture feature values ​​that complement contour features to improve the comprehensiveness of recognition; specifically, the gray-level co-occurrence matrix... The co-occurrence frequency of gray-level value pairs in the image along a direction with a distance of d and an angle of α was statistically analyzed, which can reflect the spatial distribution pattern of character texture; where The matrix element values ​​at different distances and angles are normalized to ensure the weight balance of texture features in each direction; the exponential term constitutes a Gaussian weight function, where the matrix element values... close to the mean When the exponent value is close to 1, the element contributes significantly to the texture features. When the value of the exponent term is far from the mean, the value of the exponent term approaches 0, and the contribution weight of the element is suppressed. This weight design can enhance the representative texture distribution pattern in the gray-level co-occurrence matrix, while suppressing abnormal texture values ​​and noise interference, forming a robust texture feature value T. By summing the weighted matrix element values ​​under multiple distances and angles, multi-directional texture information is integrated, so that the texture feature value T can fully reflect the internal texture distribution characteristics of the character, complementing the contour feature parameter L, and jointly improving the accuracy and stability of license plate character recognition. The recognition module performs a weighted fusion of contour feature parameter L and texture feature value T to obtain the recognition confidence level. ; The above formula takes into account both the shape and internal texture information of the character. It weights and fuses the contour feature parameters and texture feature values, and sets the sum of the weight coefficients to 1 to ensure a reasonable ratio. At the same time, it introduces the preset maximum value of the two for normalization processing. The recognition result is output only when the confidence of the fused result exceeds the threshold to ensure that the recognition judgment result is reliable. In the formula: , where is the texture feature value; D and A are the number of distance values ​​and the number of angle values ​​in the gray-level co-occurrence matrix, respectively. Let the distance be d and the angle be... The gray-level co-occurrence matrix element values ​​at that time; The mean of all elements in the gray-level co-occurrence matrix; The standard deviation of all elements in the gray-level co-occurrence matrix; These are the feature weight coefficients; These are the preset maximum values ​​for contour feature parameters and preset maximum values ​​for texture feature parameters; when When the result exceeds the preset confidence threshold, the corresponding license plate recognition result will be output. in, All are positive numbers and their sum is 1; When the recognition module recognizes license plate information, if the recognition results are inconsistent after a preset number of recognition attempts or the recognition confidence level is lower than a preset confidence threshold, a secondary recognition mechanism is triggered. The feature extraction range for the secondary recognition is adjusted using the following formula: ; In the formula: This refers to the feature extraction range for secondary recognition. This refers to the feature extraction range for the initial identification. This is a range adjustment factor; This represents the minimum permissible confidence level for identification. The current identification confidence level; The above formula addresses situations where the confidence level of the initial recognition is insufficient or the results are inconsistent. To improve the effectiveness of the secondary recognition, the feature extraction range is dynamically adjusted based on the character region of the initial recognition, according to the difference between the current recognition confidence level and the minimum allowable confidence level. The smaller the difference, the larger the adjustment coefficient and the wider the extraction range. The new range expands outward from the center of the initial recognition region to ensure coverage of key features that may be missed, such as gradient pixels at the character edges and local deformed textures. At the same time, the feature calculation algorithm remains consistent, and only the extraction range is optimized to improve recognition accuracy. in, Based on the license plate character area, a rectangular area defined by pixel coordinates is defined, including two dimensions: horizontal coverage width and vertical coverage height. ∈ (0,1], current recognition confidence With minimum allowable identification confidence The smaller the difference, the larger its value; the larger the difference, the smaller its value. It should be noted that the reason for triggering secondary recognition is that the recognition confidence level is lower than the minimum allowable recognition confidence level. When the above dynamic adjustment formula is used to calculate the feature extraction range for secondary recognition, the expansion range is dynamically determined by the confidence difference. When secondary recognition is triggered because the recognition results for a preset number of consecutive recognition attempts are inconsistent but the confidence levels are all higher than a certain threshold... When the system determines that character feature extraction is insufficient, it will uniformly expand the feature extraction range according to a preset ratio β (1.2-1.5). =β× Features are re-extracted and secondary recognition is performed; Feature extraction range for secondary recognition The specific application process is as follows: Calculated based on the above formula Then, using the center of the license plate character area located during the initial recognition as a reference, according to... The corresponding horizontal expansion width and vertical expansion height redefine the effective area for feature extraction in the secondary recognition. This area must completely cover the feature extraction area of ​​the initial recognition and expand outwards by a preset proportion of pixels to ensure that key features that may have been missed in the initial recognition, such as character edge gradient pixels and local deformation textures, are included. Based on the feature extraction logic, in Within the defined effective area, character contour features and texture features are re-extracted, and contour feature parameters and texture feature values ​​for secondary recognition are calculated respectively. During the extraction process, the feature calculation algorithm is kept consistent with the initial recognition, and only the range of feature extraction area is changed. Based on the contour feature parameters and texture feature values ​​obtained from the secondary recognition, the recognition confidence is recalculated. When the confidence is greater than the preset confidence threshold, the corresponding license plate recognition result is output. When the confidence is still lower than the preset confidence threshold, the recognition module sends a recognition anomaly signal to the management module. After receiving the recognition anomaly signal, the management module triggers the manual verification process and records the time of the anomaly, the original license plate image data, the feature extraction region parameters, feature values ​​and confidence data of the first and second recognitions. The manual verification result is synchronously updated to the license plate recognition result and the storage data of the management module, and the locally cached license plate-recognition result association record is also updated. The parsing module is used to receive the license plate recognition result and query the preset parking space and license plate binding database to see if the license plate is bound to a dedicated parking space. If the query result is no and there is a vacant parking space in the garage, the gate will be opened to allow the vehicle to enter. If the query result is yes, the gate will be opened and the creation module will be triggered to run simultaneously. The determination of whether there are vacant parking spaces in the garage in the parsing module is based on the interaction between the parsing module and the garage management terminal. The garage management terminal counts the number of vehicles entering and leaving the gate in real time to measure the number of remaining vacant parking spaces. Among them, the dedicated parking spaces in the garage are not included in the range of the number of remaining vacant parking spaces in the garage management terminal; The module is used to establish a collaborative link between the barrier gate control terminal, the dedicated parking space obstacle terminal, and the analysis module. Based on the link, the analysis module sends descent / folding control commands to the dedicated parking space obstacle terminal. During the module's collaborative link establishment phase, the legitimacy of the barrier gate control terminal and the dedicated parking space obstacle terminal is verified simultaneously. After successful authentication, a stable collaborative link is established based on link quality parameters. The link quality is characterized as follows: ; In the formula: These are link quality parameters; , , This refers to the link quality weighting coefficient. , , For link transmission bit error rate, link transmission bandwidth, and link transmission delay; The preset maximum allowable bit error rate, maximum transmission bandwidth, and maximum allowable transmission delay; The above formula comprehensively considers three core indicators of link transmission error rate, bandwidth and delay. Dynamic weighting coefficients are set according to different communication environments and system requirements. When the transmission error rate has a significant impact on stability, the corresponding weight is increased. When the transmission rate requirement is high, the bandwidth weight is emphasized. When the timing coordination requirement is high, the delay weight is increased. The link quality parameter is formed by weighted calculation. The link is confirmed to be successfully established only when the parameter exceeds the threshold, thus providing a reliable connection guarantee for multi-terminal collaboration. when When the quality exceeds the preset link quality threshold, the collaborative link is confirmed to be successfully established. The control commands sent by the parsing module to the dedicated parking space obstacle terminal include the command identifier, parking space identifier, and action parameters. The control commands are synchronously verified for integrity using the following formula: ; In the formula: This is a verification code; To control the number of fields in the command; The value of the t-th field; Let t be the validation weight of the t-th field; The preset verification base; The above formula sets differentiated verification weights based on the importance of different fields in the instruction. It assigns higher weights to key fields such as instruction identifier and parking space identifier, and appropriately reduces the weights of non-key fields. At the same time, it integrates and calculates the verification code by combining the total number of fields, the range of values, and the fixed base preset according to the verification fault tolerance rate. The receiving end confirms that the instruction is correct by comparing the verification code before executing the action, so as to avoid execution abnormalities caused by errors or loss of key information. After receiving the control command, the dedicated parking space obstacle terminal calculates a checksum. If the calculated checksum matches the checksum carried in the command, the corresponding action is executed; otherwise, it requests the establishment module to re-execute the control command issuance operation. When the checksum calculated by the dedicated parking space obstacle terminal after receiving the control command does not match the checksum carried in the command, the terminal sends a verification failure signal to the establishment module, requesting the establishment module to re-execute the control command issuance operation. After receiving the verification failure signal, the establishment module automatically regenerates a control command containing the same command identifier, parking space identifier, and action parameters, recalculates the checksum, and sends it to the dedicated parking space obstacle terminal. This retry process is performed a maximum of 3 times. If the dedicated parking space obstacle terminal still reports a failure after 3 retries... If the verification fails, the module determines that the abnormality is due to a transmission anomaly in the collaborative link or a terminal reception anomaly. It then automatically triggers a degradation scheme: the module sends a transmission anomaly alarm to the management module and simultaneously sends a control command failure status back to the parsing module. Upon receiving the failure status, the parsing module keeps the gate open, allowing vehicles to enter the garage normally, but no longer triggers automatic control of the dedicated parking space obstacle. Vehicle owners must manually operate the obstacle via remote control to raise or lower it. The management module records detailed data such as the time of the anomaly, the license plate information involved, the dedicated parking space number, and the number of retries. This data is used by system administrators to troubleshoot link or terminal failures, ensuring that the system can still guarantee normal vehicle entry and exit from the garage even in the event of a partial communication failure. in, , , All are positive numbers, and , , The sum is, and when the garage communication environment has strong interference and transmission errors have a more significant impact on the stability of the cooperative link, The larger the value, the smaller the value; when the system has high requirements for the transmission rate of control commands, insufficient bandwidth can easily lead to command transmission stuttering or loss. The larger the value, the smaller the value; when multi-terminal collaborative actions require precise timing coordination, and transmission delays can easily cause misalignment in action execution. The larger the value, the smaller the value; The default value range is [0.5, 2]. When the field in the control command that plays a key role in the effectiveness of the action execution (such as the command identifier, parking space identifier) ​​is... The larger the value, the more likely it is to be a non-critical field that serves only as supplementary information. The smaller the value; Based on the total number of fields in the control command, the maximum value range of each field, and the system's allowed error tolerance rate, the values ​​are preset to fixed positive integers through statistical analysis. The control module is used to receive control commands, and based on the vehicle from which the command originates, it links the monitoring equipment inside the garage to monitor the real-time location of the target. When the target vehicle is detected to have arrived within the preset range of the corresponding dedicated parking space, it generates a dedicated parking space obstacle lifting / folding control signal to control the dedicated parking space obstacle terminal to complete the lifting / folding action. The preset range of dedicated parking spaces in the control module is a dynamically adjustable range, conforming to: ; In the formula: The dynamic range radius; The real-time speed of the target vehicle; This is the preset system response time; The base preset radius; The above formula is designed to adapt to the needs of vehicles parking in dedicated parking spaces at different driving speeds. It dynamically adjusts the preset monitoring range of the dedicated parking space by multiplying the real-time driving speed of the vehicle by the system's reaction time and then adding the basic preset radius to form a dynamic range radius. This ensures that no matter how fast or slow the vehicle is, the system can reserve enough time in advance to trigger the obstacle action, so that the vehicle can park smoothly when it arrives, thus improving the convenience of using dedicated parking spaces. Among them, when a vehicle is parked in a designated parking space, the designated parking space obstacle terminal resets synchronously after sensing that the vehicle has left, based on the built-in sensor. The management module is used to store license plate recognition data, parsing results, and control execution records, and provides access to and modification permissions for the database that binds parking spaces and license plates. The collaborative link configuration time synchronization mechanism between the barrier gate control terminal, the dedicated parking space obstacle terminal, and the analysis module is as follows: Local time correction for each terminal: ; In the formula: This is the corrected terminal time; The local time of the terminal; This is the difference between the terminal's local time and the system's base time. For synchronization correction coefficients, ∈ (0,1], which is dynamically adjusted according to the link quality parameter. The larger the link quality parameter, the better. The closer the value is to 1; The above formula calculates the difference between the terminal's local time and the system's reference time, and combines it with a synchronization correction coefficient that is dynamically adjusted based on the link quality parameters. The better the link quality, the closer the correction coefficient is to 1. This corrects the terminal's local time, ensuring that the time of each terminal is as consistent as possible with the system's reference time, thus avoiding misalignment of actions due to time deviation. When the control module generates a dedicated parking space obstacle lifting / folding control signal, it incorporates action execution delay compensation, resulting in a more efficient action execution time. ; In the formula: The estimated time for the target vehicle to arrive at the designated parking space within a preset range; To control signal transmission delay; The standard time taken for a parking space obstacle to complete a lifting / folding action; The above formula dynamically calculates the vehicle's estimated arrival time based on the vehicle's current position, real-time speed, and preset range boundary coordinates. Then, it deducts the control signal transmission delay and the standard time for the obstacle to complete its action to obtain the final action execution time. This compensates for the time loss during transmission and action, ensuring that the obstacle completes its corresponding action just as the vehicle arrives, thereby improving the accuracy of coordinated action. in, The garage's internal monitoring equipment, linked by the control module, collects the target vehicle's current position, real-time speed, and the boundary coordinates of the preset range of the exclusive parking space in real time. It dynamically calculates the estimated time for the vehicle to travel from its current position to the boundary of the preset range, while also adapting to the speed fluctuations during the vehicle's journey. The acquisition module is interconnected with the identification module via a wireless network. The acquisition module and the identification module are interconnected with the parsing module via a wireless network. The parsing module is interconnected with the establishment module via a wireless network. The establishment module is interconnected with the control module via a wireless network. The control module is interconnected with the management module via a wireless network.

[0021] In this embodiment, the acquisition module acquires license plate images of vehicles entering the garage, performs noise filtering and geometric calibration preprocessing on the license plate images to output standardized license plate image data. The recognition module then receives the standardized license plate image data, extracts the character contour features and texture features of the license plate, and performs license plate information recognition based on these features, outputting the license plate recognition result. The parsing module then receives the license plate recognition result and queries a preset parking space and license plate binding database to check if the license plate is bound to a specific parking space. If the query result is negative and a parking space in the garage is available, the gate opens, allowing the vehicle to enter. If the query result is positive, the gate opens, simultaneously triggering the establishment module to run. The module further establishes a collaborative link between the barrier gate control terminal, the dedicated parking space obstacle terminal, and the analysis module. Based on this link, the analysis module sends a lowering / folding control command to the dedicated parking space obstacle terminal. The control module synchronously receives the control command and, based on the vehicle from which the command originates, links the internal monitoring equipment of the parking garage to monitor the real-time location of the target. When the target vehicle is detected to have arrived within the preset range of the corresponding dedicated parking space, a lifting / folding control signal for the dedicated parking space obstacle is generated, controlling the dedicated parking space obstacle terminal to complete the lifting / folding action. Finally, the management module stores the license plate recognition data, analysis results, and control execution records, providing access to and modification permissions for the parking space and license plate binding database.

[0022] The system in the above embodiments is accurate and efficient in identifying vehicles entering and exiting the garage, reasonably allocates vacant parking spaces, ensures the use of dedicated parking spaces, and ensures smooth equipment coordination and accurate and timely action execution, reducing waiting and misoperation. It not only greatly improves the efficiency of garage management, but also makes it more convenient and worry-free for car owners to enter and exit, while ensuring the parking experience and the safety of parking space use.

[0023] Application example: The underground parking garage of the XX commercial complex has introduced this license plate recognition barrier gate access management system to improve vehicle entry and exit efficiency and the security of dedicated parking space management.

[0024] When a car registered to parking space number 15 in Zone B enters the garage entrance, the acquisition module at the entrance immediately captures the vehicle's license plate image. The system first performs noise filtering on the image, ultimately obtaining an image with uniform pixel distribution and effectively suppressed interference signals. Subsequently, geometric calibration is performed through perspective transformation. Based on the pre-set correspondence between the four corner points in the license plate image and the corner points of a standard license plate, fitting parameters correct the tilted license plate image into a regular standard license plate image, outputting standardized license plate image data.

[0025] After receiving the standardized image, the recognition module first obtains the character edge pixels through edge detection and calculates the contour feature parameters; then, it obtains the texture feature value through gray-level co-occurrence matrix correlation calculation. After fusing the two features according to the preset weight, the recognition confidence is 0.93, which is higher than the system's preset confidence threshold of 0.85. Therefore, the license plate recognition result "XA12345" is directly output.

[0026] After receiving the license plate recognition result, the parsing module queries the preset parking space and license plate binding database to confirm that the license plate is bound to a dedicated parking space. It then controls the gate to open, allowing the vehicle to enter, and simultaneously triggers the establishment module. The establishment module first verifies the legitimacy of the gate control terminal and the obstacle terminal of parking space 15 in Zone B. After successful verification, it calculates the link quality parameters. The results show that the link transmission error rate, bandwidth, and latency all meet the system requirements, indicating that the link quality is up to standard. A collaborative link between the three is successfully established, and a control command containing the instruction identifier, parking space identifier, and action parameters is sent to the obstacle terminal of the dedicated parking space. After receiving the command, the terminal verifies its integrity and waits to execute subsequent actions.

[0027] The control module, in conjunction with the garage's internal monitoring equipment, identifies the vehicle as a monitoring target and tracks its location and driving status in real time. Calculations show the vehicle's real-time speed is 4 km / h. Combining this with the preset system reaction time and basic preset radius, the preset dynamic range radius for the designated parking space is determined to be 3 meters. Simultaneously, the system dynamically calculates the estimated time for the vehicle to reach the preset range based on its current position, real-time speed, and the boundary coordinates of the designated parking space. This timeframe is then subtracted for control signal transmission delays and the standard time required for obstacle lifting / lowering actions to determine the execution time. When the vehicle is detected entering the 3-meter range of designated parking space number 15 in Zone B, the control module immediately generates a folding control signal. Upon receiving the signal, the obstacle terminal quickly completes the folding action, allowing the vehicle to smoothly enter and park. After the vehicle leaves, the obstacle terminal's built-in sensor detects the departure and simultaneously resets.

[0028] During this process, the barrier gate control terminal, the dedicated parking space obstacle terminal, and the parsing module synchronize their local time using a time synchronization mechanism. After correction, the time of each terminal deviates very little from the system's reference time, ensuring precise timing of multi-terminal coordinated actions. The management module stores the license plate recognition data, parsing results, control command content, and execution records in real time. It also allows administrators, after authorization, to access the parking space and license plate binding database to modify and maintain parking space binding information.

[0029] When another car without a designated parking space enters, the data collection and recognition module completes the standardized processing and recognition as well. After querying, the analysis module confirms that there is no designated parking space and learns from the interaction with the garage management terminal that there are still vacant parking spaces in the public area of ​​the garage. Then, it controls the gate to open, allowing the vehicle to enter the garage to find a public vacant parking space. The management module records the relevant data of the vehicle's entry and exit at the same time.

[0030] In summary, the system in the above embodiments improves the standardization of license plate images through fine preprocessing, significantly improves recognition accuracy and reduces recognition anomalies by combining contour and texture feature weighted fusion and secondary recognition mechanisms, ensures stable and reliable instruction transmission through the establishment and quality verification of multi-terminal collaborative links, avoids execution misalignment through time synchronization and action delay compensation, adapts the preset parking space range to the vehicle driving status, makes parking space obstacle control more timely and accurate, and improves garage utilization efficiency through the classification and management of dedicated parking spaces and vacant parking spaces, thus optimizing the overall vehicle entry and exit process.

[0031] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control, characterized in that, include: The acquisition module is used to acquire license plate images of vehicles entering the garage, and to perform preprocessing operations such as noise filtering and geometric calibration on the license plate images to output standardized license plate image data. The recognition module is used to receive standardized license plate image data, extract the character contour features and texture features of the license plate, complete the recognition of license plate information based on the features, and output the license plate recognition result; The parsing module is used to receive the license plate recognition result and query the preset parking space and license plate binding database to see if the license plate is bound to a dedicated parking space. If the query result is no and there is a vacant parking space in the garage, the gate will be opened to allow the vehicle to enter. If the query result is yes, the gate will be opened and the creation module will be triggered to run simultaneously. The module is used to establish a collaborative link between the barrier gate control terminal, the dedicated parking space obstacle terminal, and the analysis module. Based on the link, the analysis module sends descent / folding control commands to the dedicated parking space obstacle terminal. The control module is used to receive control commands, and based on the vehicle from which the command originates, it links the monitoring equipment inside the garage to monitor the real-time location of the target. When the target vehicle is detected to have arrived within the preset range of the corresponding dedicated parking space, it generates a dedicated parking space obstacle lifting / folding control signal to control the dedicated parking space obstacle terminal to complete the lifting / folding action. The management module stores license plate recognition data, parsing results, and control execution records, and provides access to and modification permissions for the database that binds parking spaces and license plates.

2. The license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control as described in claim 1, characterized in that, The noise filtering in the acquisition module follows the following rules: ； In the formula: This represents the pixel value at coordinates (x, y) after filtering. The radius of the filtering neighborhood; These are adaptive weighting coefficients; The original pixel value at coordinates (x+i, y+j) in the original license plate image; The geometric calibration is achieved through perspective transformation, and the calibrated coordinates (u,v) are solved using the following formula: ； In the formula: These are the parameters of the perspective transformation matrix.

3. The license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control as described in claim 1, characterized in that, In the character contour feature extraction stage of the recognition module, edge detection is used to obtain character edge pixels, and contour feature parameters are calculated. ； In the formula: These are contour feature parameters; This represents the total number of pixels at the character edge. , The difference between the x-coordinate and y-coordinate of the m-th edge pixel and its adjacent pixels; Let be the angle between the gradient direction of the m-th edge pixel and the horizontal centerline of the character; Texture feature extraction: ; The recognition module performs a weighted fusion of contour feature parameter L and texture feature value T to obtain the recognition confidence level. ; In the formula: , where is the texture feature value; D and A are the number of distance values ​​and the number of angle values ​​in the gray-level co-occurrence matrix, respectively. Let the distance be d and the angle be... The gray-level co-occurrence matrix element values ​​at that time; The mean of all elements in the gray-level co-occurrence matrix; The standard deviation of all elements in the gray-level co-occurrence matrix; These are the feature weight coefficients; These are the preset maximum values ​​for contour feature parameters and preset maximum values ​​for texture feature parameters; when When the result exceeds the preset confidence threshold, the corresponding license plate recognition result is output.

4. The license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control as described in claim 1, characterized in that, When the recognition module recognizes license plate information, if the recognition results are inconsistent after a preset number of consecutive recognition attempts or the recognition confidence level is lower than a preset confidence threshold, a secondary recognition mechanism is triggered. The feature extraction range for the secondary recognition is adjusted using the following formula: ； In the formula: This refers to the feature extraction range for secondary recognition. This refers to the feature extraction range for the initial identification. This is a range adjustment factor; The minimum allowable confidence level for identification; This represents the current confidence level.

5. The license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control as described in claim 1, characterized in that, The determination of whether there are vacant parking spaces in the garage in the analysis module is based on the interaction between the analysis module and the garage management terminal. The garage management terminal counts the number of vehicles entering and leaving the gate in real time to measure the number of remaining vacant parking spaces. Among them, the number of reserved parking spaces in the garage is not included in the range of the number of remaining vacant parking spaces in the garage management terminal.

6. The license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control as described in claim 1, characterized in that, During the collaborative link establishment phase of the module, the legitimacy of the barrier gate control terminal and the dedicated parking space obstacle terminal is verified simultaneously. After successful authentication, a stable collaborative link is established based on link quality parameters. The link quality is characterized as follows: ； In the formula: These are link quality parameters; , , This refers to the link quality weighting coefficient. , , For link transmission bit error rate, link transmission bandwidth, and link transmission delay; The preset maximum allowable bit error rate, maximum transmission bandwidth, and maximum allowable transmission delay; when When the quality exceeds the preset link quality threshold, the collaborative link is confirmed to be successfully established. The control command issued by the parsing module to the dedicated parking space obstacle terminal includes a command identifier, a parking space identifier, and action parameters. The control command is synchronously verified for integrity using the following formula: ； In the formula: This is a verification code; To control the number of fields in the command; The value of the t-th field; Let t be the validation weight of the t-th field; The preset verification base; After receiving the control command, the dedicated parking space obstacle terminal calculates the check code. If the calculated check code matches the check code carried in the command, the corresponding action is executed. Otherwise, the request module executes the control command issuance operation again.

7. The license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control as described in claim 1, characterized in that, The preset range of dedicated parking spaces in the control module is a dynamically adjustable range, conforming to: ； In the formula: The dynamic range radius; The real-time speed of the target vehicle; This is the preset system response time; The base preset radius; When a vehicle is parked in a designated parking space, the obstacle terminal for the designated parking space will reset synchronously after the vehicle leaves, based on the built-in sensors.

8. The license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control as described in claim 1, characterized in that, The collaborative link configuration time synchronization mechanism between the barrier gate control terminal, the dedicated parking space obstacle terminal, and the analysis module is as follows: Local time correction for each terminal: ; In the formula: This is the corrected terminal time; The local time of the terminal; This is the difference between the terminal's local time and the system's base time. This is the synchronization correction coefficient; When the control module generates a dedicated parking space obstacle lifting / folding control signal, it incorporates action execution delay compensation, resulting in a compensated action execution time. ; In the formula: The estimated time for the target vehicle to arrive at the designated parking space within a preset range; To control signal transmission delay; The standard time taken for a parking space obstacle to complete the lifting / folding action.

9. The license plate recognition barrier gate access management system based on multi-terminal collaboration and dynamic access control as described in claim 1, characterized in that, The acquisition module is interconnected with the identification module via a wireless network. The acquisition module and the identification module are interconnected with the parsing module via a wireless network. The parsing module is interconnected with the establishment module via a wireless network. The establishment module is interconnected with the control module via a wireless network. The control module is interconnected with the management module via a wireless network.

Images