Intelligent access control system for vehicle access control based on image recognition

Abstract

The invention provides a vehicle access control intelligent access control system based on image recognition, and belongs to the technical field of vehicle intelligent access control. Comprising an image acquisition assembly, a database and an automatic gate machine, the automatic gate machine is used for automatically opening and closing a crossing and facilitating entering and exiting of vehicles, the image acquisition assembly is used for acquiring image information of the vehicles, the image acquisition assembly comprises a camera, and the camera is a high-definition camera. By adopting the specific high-definition camera, the license plate can be accurately recognized under all-weather and various illumination conditions, and a vehicle image recognition engine is established, so that the recognition efficiency of the license plate of the vehicle is improved, the passing efficiency of the vehicle is improved, the entrance and exit conditions can be monitored in real time through the monitoring and alarming module, and the safety of the vehicle is improved. If abnormal conditions (such as unauthorized vehicles, license plate shielding and the like) are found, the system can automatically give an alarm to inform an administrator to intervene, so that the safety of the whole system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle intelligent access control, and in particular to an intelligent access control system for vehicle access control based on image recognition. Background Art

[0002] The vehicle access control intelligent access control system is an entrance and exit management device used to restrict the travel of motor vehicles on the road. It is now widely used in highway toll stations and parking lot systems to manage vehicle channels and to manage the entry and exit of vehicles. With the continuous development of the automobile industry and the increasing number of cars, the demand for barriers is increasing, and higher requirements are also placed on the performance of vehicle access control systems.

[0003] Some existing intelligent access control systems have low efficiency in license plate recognition and cannot respond in the first time. There are also cases where vehicles cannot be recognized, affecting the efficiency of vehicle passage. When a vehicle passes through a barrier, there is a risk of the barrier lowering, causing damage to the vehicle and the barrier. Therefore, the present application provides a vehicle entry and exit control intelligent access control system based on image recognition to meet the needs. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide an intelligent access control system for vehicle entry and exit control based on image recognition to solve the existing problems.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] A vehicle access control intelligent access control system based on image recognition, comprising: an image acquisition component, a database and an automatic gate, wherein the automatic gate is used to automatically open and close a road crossing to facilitate vehicle access, the image acquisition component is used to collect vehicle image information, the image acquisition component comprises a camera, the camera adopts a high-definition camera, and is used to capture images of vehicles entering and exiting. The camera should have a high resolution and night vision function to ensure accurate recognition under all-weather and various lighting conditions, the image acquisition component also comprises a license plate recognition module, the license plate recognition module recognizes the license plate number in real time through computer vision technology, and extracts the number characters in the license plate image to determine whether it matches the valid number stored in the database, the image acquisition component also comprises a recognition engine building module, the recognition engine building module is used to build a vehicle image recognition engine, and the database is used to store various types of vehicle data.

[0007] In some examples, the database also includes a vehicle information management module, which includes a database of authorized vehicles (such as corporate employees, long-term leased vehicles, etc.). When a vehicle passes by, the system can automatically identify the license plate number and compare it with the database to determine whether entry is allowed. For non-whitelist vehicles, temporary vehicle entry can be authorized through other means (such as text messages, APP notifications, etc.), and relevant data (such as entry time, license plate number, etc.) can be recorded.

[0008] In some examples, a background management module is also included, which is used to upload all vehicle entry and exit records to the cloud or local server for storage in real time. Administrators can view and analyze historical data, perform troubleshooting, access statistics, etc. through the background management module.

[0009] In some examples, a monitoring and alarm module is also included. The monitoring and alarm module monitors the import and export conditions in real time. If an abnormal situation is found (such as unauthorized vehicles, obscured license plates, etc.), the system can automatically issue an alarm to notify the administrator to intervene.

[0010] In some examples, the specific process of real-time recognition of license plate numbers by computer vision technology is as follows: grayscale preprocessing is performed on the image captured by the camera to improve image quality, eliminate noise, enhance contrast, etc. for subsequent processing. The grayscale preprocessing is specifically as follows:

[0011] I gray (x,y)=0.299·R(x,y)+0.587·G(x,y)+0.114·B(x,y)

[0012] Where R(x,y), G(x,y) and B(x,y) are the pixel values ​​of the red, green and blue channels at the (x,y) coordinates in the image;

[0013] After grayscale processing, the grayscale image is converted to a black and white image, which can be achieved by fixed threshold or adaptive threshold method:

[0014]

[0015] Where, T is the set threshold;

[0016] Detect the area where the license plate is located from the image through edge detection:

[0017]

[0018] in Represents the gradient strength of the image at position (x, y), and extracts the edge by setting an appropriate threshold;

[0019] Perform character segmentation to separate the characters in the license plate area from the overall image to facilitate subsequent character recognition. Characters are extracted by analyzing each connected area in the binary image.

[0020] For character recognition, deep learning (such as convolutional neural network CNN) is used for character recognition. Assume that the input image is the character area I char (x,y), it can be classified by the convolutional neural network model:

[0021]

[0022] in is the predicted character, f is the trained model, θ is the model parameter, the recognized license plate characters are spliced, the license plate sequence is corrected, and the license plate number is checked for legality;

[0023] In some examples, the steps for building the vehicle image recognition engine are:

[0024] Collect a large dataset of vehicle images, typically containing various types of vehicles, different angles, lighting, and environmental conditions, e.g., using publicly available vehicle image datasets;

[0025] Preprocess the image, including but not limited to image scaling: scaling the image to a fixed size, image normalization: usually scaling the pixel value of the image from [0,255] to [0,1] or [-1,1], data enhancement: to improve the robustness of the model, random rotation, flipping, cropping and other methods can be used to increase data diversity;

[0026] A deep convolutional neural network (CNN) is used for feature extraction. The convolution operation formula in CNN is:

[0027]

[0028] Where X is the input image, W is the convolution kernel (filter), and Y is the output feature map; for each layer of convolution, an activation function (such as ReLU) is performed:

[0029] f(x)=max(0,x)

[0030] During training, the cross-entropy loss function is usually used as the loss function, and optimizers such as Adam or SGD are used to update the weights:

[0031] Cross entropy loss function:

[0032]

[0033] Where C is the number of categories, yi is the true label, is the probability predicted by the model;

[0034] Optimizer (Adam optimizer)

[0035]

[0036] Where θ is the model parameter, α is the learning rate, β1 and β2 are hyperparameters, usually set to 0.9 and 0.999;

[0037] After training the model, it needs to be evaluated. Common evaluation indicators include: Accuracy: the proportion of correctly classified samples to the total samples, Recall: the proportion of correctly identified positive samples to all positive samples, F1 value: the weighted average of accuracy and recall;

[0038]

[0039] in

[0040] Recall = the number of correctly predicted positive classes / the total number of actual positive classes;

[0041] Precision = the number of correctly predicted positive classes / the total number of predicted positive classes;

[0042] Save the trained model as a .h5 or .pb file and deploy it to a server or mobile terminal for real-time inference.

[0043] A working method of a vehicle access control intelligent access control system based on image recognition, comprising the following steps:

[0044] S1, the vehicle approaches the entrance, and the camera automatically captures the license plate image;

[0045] S2, the system identifies the license plate number through image recognition technology and compares it with the license plate information in the database;

[0046] S3. If it is a whitelist vehicle, the system automatically confirms and opens the gate; if it is a temporary vehicle or an unfamiliar vehicle, the system sends a notification to the management backend, and the administrator can authorize or reject it as needed;

[0047] S4. After the vehicle successfully passes, the system records the time and updates the data;

[0048] S5. When the vehicle arrives at the exit, the camera recognizes the license plate system again and checks the vehicle entry record to confirm whether the vehicle has been registered in the system;

[0049] S6. If everything is in compliance, the system controls the gate or access control to automatically unlock and allow the vehicle to exit. If there is no entry record, the system contacts the backend administrator through the monitoring and alarm module for processing.

[0050] Compared with the prior art, the present invention has at least the following beneficial effects:

[0051] In the above scheme, by adopting specific high-definition cameras, accurate license plate recognition is guaranteed under all weather and various lighting conditions, and a vehicle image recognition engine is established to improve the recognition efficiency of vehicle license plates and the efficiency of vehicle passage. The import and export conditions can be monitored in real time through the monitoring and alarm module. If any abnormal situation is found (such as unauthorized vehicles, obstructed license plates, etc.), the system can automatically issue an alarm to notify the administrator to intervene, thereby improving the security of the entire system. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The accompanying drawings, which are incorporated herein and constitute a part of the specification, illustrate embodiments of the present disclosure and, together with the description, further serve to explain the principles of the present disclosure and to enable those skilled in the relevant art to make and use the present disclosure.

[0053] Figure 1 It is a structural diagram of the intelligent access control system for vehicle access control based on image recognition;

[0054] Figure 2 This is the workflow diagram of the intelligent access control system for vehicle access control based on image recognition. DETAILED DESCRIPTION

[0055] The following is a detailed description of a vehicle access control intelligent access control system based on image recognition provided by the present invention in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following implementations are listed as the best and preferred embodiments, and technicians in some known technical fields may also adopt other alternative methods to implement; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0056] It should be noted that the references to "one embodiment", "embodiment", "exemplary embodiment", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0057] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0058] It will be understood that the meaning of “on,” “over,” and “above” in this disclosure should be interpreted in the broadest manner, so that “on” not only means “directly on” something, but also includes the meaning of being “on” something with intervening features or layers therebetween, and “on” or “over” not only means “on” or “above” something, but also includes the meaning of being “on” or “above” something with no intervening features or layers therebetween.

[0059] Additionally, spatially relative terms such as "under," "beneath," "lower," "above," "upper," and the like may be used herein for descriptive convenience to describe the relationship of one element or feature to another element or features, as shown in the accompanying drawings. Spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the accompanying drawings. The device may be oriented in other ways, and the spatially relative descriptors used herein may be similarly interpreted accordingly.

[0060] Example 1

[0061] like Figure 1 As shown, an embodiment of the present invention provides an intelligent access control system for vehicle access control based on image recognition, including an image acquisition component, a database and an automatic gate. The automatic gate is used to automatically open and close the gate to facilitate vehicle access. The image acquisition component is used to collect vehicle image information. The image acquisition component includes a camera. The camera adopts a high-definition camera to capture images of vehicles entering and exiting. The camera should have a high resolution and night vision function to ensure accurate recognition under all weather and various lighting conditions. The image acquisition component also includes a license plate recognition module. The license plate recognition module uses computer vision technology to perform real-time recognition of license plate numbers, and extracts number characters in the license plate image to determine whether it matches the valid numbers stored in the database. The image acquisition component also includes a recognition engine building module. The recognition engine building module is used to build a vehicle image recognition engine. The database is used to store various types of vehicle data.

[0062] It should be further explained in this embodiment that the database also includes a vehicle information management module, which includes a database of authorized vehicles (such as corporate employees, long-term leased vehicles, etc.). When a vehicle passes by, the system can automatically identify the license plate number and compare it with the database to determine whether entry is allowed. For non-whitelist vehicles, temporary vehicle entry can be authorized through other means (such as text messages, APP notifications, etc.), and relevant data (such as entry time, license plate number, etc.) can be recorded.

[0063] It should be further explained that this embodiment also includes a background management module, which is used to upload all vehicle entry and exit records to the cloud or local server for storage in real time. The administrator can view and analyze historical data, perform troubleshooting, access statistics, etc. through the background management module.

[0064] It should be further explained that this embodiment also includes a monitoring and alarm module, which monitors the import and export conditions in real time. If an abnormal situation is found (such as unauthorized vehicles, blocked license plates, etc.), the system can automatically issue an alarm to notify the administrator to intervene.

[0065] It should be further explained in this embodiment that the specific process of real-time recognition of license plate numbers using computer vision technology is as follows: grayscale preprocessing is performed on the image captured by the camera to improve image quality, eliminate noise, enhance contrast, etc., for subsequent processing. The grayscale preprocessing is specifically as follows:

[0066] I gray (x,y)=0.299·R(x,y)+0.587·G(x,y)+0.114·B(x,y)

[0067] Where R(x,y), G(x,y) and B(x,y) are the pixel values ​​of the red, green and blue channels at the (x,y) coordinates in the image;

[0068] After grayscale processing, the grayscale image is converted to a black and white image, which can be achieved by fixed threshold or adaptive threshold method:

[0069]

[0070] Where, T is the set threshold;

[0071] Detect the area where the license plate is located from the image through edge detection:

[0072]

[0073] in Represents the gradient strength of the image at position (x, y), and extracts the edge by setting an appropriate threshold;

[0074] Perform character segmentation to separate the characters in the license plate area from the overall image to facilitate subsequent character recognition. Characters are extracted by analyzing each connected area in the binary image.

[0075] For character recognition, deep learning (such as convolutional neural network CNN) is used for character recognition. Assume that the input image is the character area I char (x,y), it can be classified by the convolutional neural network model:

[0076]

[0077] in is the predicted character, f is the trained model, θ is the model parameter, the recognized license plate characters are spliced, the license plate sequence is corrected, and the license plate number is checked for legality;

[0078] It should be further explained in this embodiment that the steps of building the vehicle image recognition engine are:

[0079] Collect a large dataset of vehicle images, typically containing various types of vehicles, different angles, lighting, and environmental conditions, e.g., using publicly available vehicle image datasets;

[0080] Preprocess the image, including but not limited to image scaling: scaling the image to a fixed size, image normalization: usually scaling the pixel value of the image from [0,255] to [0,1] or [-1,1], data enhancement: to improve the robustness of the model, random rotation, flipping, cropping and other methods can be used to increase data diversity;

[0081] A deep convolutional neural network (CNN) is used for feature extraction. The convolution operation formula in CNN is:

[0082]

[0083] Where X is the input image, W is the convolution kernel (filter), and Y is the output feature map; for each layer of convolution, an activation function (such as ReLU) is performed:

[0084] f(x)=max(0,x)

[0085] During training, the cross-entropy loss function is usually used as the loss function, and optimizers such as Adam or SGD are used to update the weights:

[0086] Cross entropy loss function:

[0087]

[0088] Where C is the number of categories, y i is the true label, is the probability predicted by the model;

[0089] Optimizer (Adam optimizer)

[0090]

[0091] Where θ is the model parameter, α is the learning rate, β1 and β2 are hyperparameters, usually set to 0.9 and 0.999;

[0092] After training the model, it needs to be evaluated. Common evaluation indicators include: Accuracy: the proportion of correctly classified samples to the total samples, Recall: the proportion of correctly identified positive samples to all positive samples, F1 value: the weighted average of accuracy and recall;

[0093]

[0094] in

[0095] Recall = the number of correctly predicted positive classes / the total number of actual positive classes;

[0096] Precision = the number of correctly predicted positive classes / the total number of predicted positive classes;

[0097] Save the trained model as a .h5 or .pb file and deploy it to a server or mobile terminal for real-time inference.

[0098] Example 2

[0099] like Figure 2 As shown, a working method of a vehicle access control intelligent access control system based on image recognition includes the following steps:

[0100] S1, the vehicle approaches the entrance, and the camera automatically captures the license plate image;

[0101] S2, the system identifies the license plate number through image recognition technology and compares it with the license plate information in the database;

[0102] S3. If it is a whitelist vehicle, the system automatically confirms and opens the gate; if it is a temporary vehicle or an unfamiliar vehicle, the system sends a notification to the management backend, and the administrator can authorize or reject it as needed;

[0103] S4. After the vehicle successfully passes, the system records the time and updates the data;

[0104] S5. When the vehicle arrives at the exit, the camera recognizes the license plate system again and checks the vehicle entry record to confirm whether the vehicle has been registered in the system;

[0105] S6. If everything is in compliance, the system controls the gate or access control to automatically unlock and allow the vehicle to exit. If there is no entry record, the system contacts the backend administrator through the monitoring and alarm module for processing.

[0106] The technical solution provided by the present invention adopts a specific high-definition camera to ensure accurate license plate recognition under all weather and various lighting conditions, and establishes a vehicle image recognition engine, thereby improving the recognition efficiency of vehicle license plates and the efficiency of vehicle passage. The import and export conditions can be monitored in real time through the monitoring and alarm module. If an abnormal situation is found (such as unauthorized vehicle, blocked license plate, etc.), the system can automatically issue an alarm to notify the administrator to intervene, thereby improving the security of the entire system.

[0107] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0108] A person skilled in the art will appreciate that all or part of the steps in the above-mentioned embodiment method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc.

[0109] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. An intelligent access control system for vehicle access control based on image recognition, characterized in that: include: An image acquisition component, a database and an automatic gate, wherein the automatic gate is used to automatically open and close the road crossing to facilitate vehicle entry and exit, the image acquisition component is used to collect vehicle image information, the image acquisition component includes a camera, and the camera adopts a high-definition camera to capture images of vehicles entering and exiting. The camera should have a high resolution and night vision function to ensure accurate recognition under all-weather and various lighting conditions. The image acquisition component also includes a license plate recognition module, which uses computer vision technology to perform real-time recognition of license plate numbers, and extracts number characters in the license plate image to determine whether it matches the valid number stored in the database. The image acquisition component also includes a recognition engine building module, which is used to build a vehicle image recognition engine, and the database is used to store various types of vehicle data.

2. According to claim 1, the vehicle access control intelligent access control system based on image recognition is characterized in that: The database also includes a vehicle information management module, which contains a database of authorized vehicles (such as corporate employees, long-term leased vehicles, etc.). When a vehicle passes by, the system can automatically identify the license plate number and compare it with the database to determine whether entry is allowed. For non-whitelist vehicles, temporary vehicle entry can be authorized through other means (such as text messages, APP notifications, etc.), and relevant data (such as entry time, license plate number, etc.) can be recorded.

3. The vehicle access control intelligent access control system based on image recognition according to claim 1 is characterized in that: It also includes a background management module, which is used to upload all vehicle entry and exit records to the cloud or local server for storage in real time. Administrators can view and analyze historical data, perform troubleshooting, access statistics, etc. through the background management module.

4. The vehicle access control intelligent access control system based on image recognition according to claim 1 is characterized in that: It also includes a monitoring and alarm module, which monitors the import and export conditions in real time. If an abnormal situation is found (such as unauthorized vehicles, blocked license plates, etc.), the system can automatically issue an alarm to notify the administrator to intervene.

5. The vehicle access control intelligent access control system based on image recognition according to claim 1 is characterized in that: The specific process of real-time recognition of license plate numbers using computer vision technology is as follows: grayscale preprocessing is performed on the image captured by the camera to improve image quality, eliminate noise, enhance contrast, etc., for subsequent processing. The grayscale preprocessing is specifically as follows: I gray (x,y)=0.299·R(x,y)+0.587·G(x,y)+0.114·B(x,y) Where R(x,y), G(x,y) and B(x,y) are the pixel values ​​of the red, green and blue channels at the (x,y) coordinates in the image; After grayscale processing, the grayscale image is converted to a black and white image, which can be achieved by fixed threshold or adaptive threshold method: Where, T is the set threshold; Detect the area where the license plate is located from the image through edge detection: in Represents the gradient strength of the image at position (x, y), and extracts the edge by setting an appropriate threshold; Perform character segmentation to separate the characters in the license plate area from the overall image to facilitate subsequent character recognition. Characters are extracted by analyzing each connected area in the binary image. For character recognition, deep learning (such as convolutional neural network CNN) is used for character recognition. Assume that the input image is the character area I char (x,y), it can be classified by the convolutional neural network model: in is the predicted character, f is the trained model, θ is the model parameter, and the recognized license plate characters are spliced, the license plate sequence is corrected, and the license plate number is checked for legality.

6. The vehicle access control intelligent access control system based on image recognition according to claim 1 is characterized in that: The steps for building the vehicle image recognition engine are: Collect a large dataset of vehicle images, typically containing various types of vehicles, different angles, lighting, and environmental conditions, e.g., using publicly available vehicle image datasets; Preprocess the image, including but not limited to image scaling: scaling the image to a fixed size, image normalization: usually scaling the pixel value of the image from [0,255] to [0,1] or [-1,1], data enhancement: to improve the robustness of the model, random rotation, flipping, cropping and other methods can be used to increase data diversity; A deep convolutional neural network (CNN) is used for feature extraction. The convolution operation formula in CNN is: Where X is the input image, W is the convolution kernel (filter), and Y is the output feature map; for each layer of convolution, an activation function (such as ReLU) is performed: f(x)=max(0,x) During training, the cross-entropy loss function is usually used as the loss function, and optimizers such as Adam or SGD are used to update the weights: Cross entropy loss function: Where C is the number of categories, y i is the true label, is the probability predicted by the model; Optimizer (Adam optimizer) Where θ is the model parameter, α is the learning rate, β1 and β2 are hyperparameters, usually set to 0.9 and 0.999; After training the model, it needs to be evaluated. Common evaluation indicators include: Accuracy: the proportion of correctly classified samples to the total samples, Recall: the proportion of correctly identified positive samples to all positive samples, F1 value: the weighted average of accuracy and recall; in Recall = the number of correctly predicted positive classes / the total number of actual positive classes; Precision = the number of correctly predicted positive classes / the total number of predicted positive classes; Save the trained model as a .h5 or .pb file and deploy it to a server or mobile terminal for real-time inference.

7. A working method of a vehicle access control intelligent access control system based on image recognition, applied to a vehicle access control intelligent access control system based on image recognition as described in any one of claims 1 to 6, characterized in that: The following steps are involved: S1, the vehicle approaches the entrance, and the camera automatically captures the license plate image; S2, the system identifies the license plate number through image recognition technology and compares it with the license plate information in the database; S3. If it is a whitelist vehicle, the system automatically confirms and opens the gate; if it is a temporary vehicle or an unfamiliar vehicle, the system sends a notification to the management backend, and the administrator can authorize or reject it as needed; S4. After the vehicle successfully passes, the system records the time and updates the data; S5. When the vehicle arrives at the exit, the camera recognizes the license plate system again and checks the vehicle entry record to confirm whether the vehicle has been registered in the system; S6. If everything is in compliance, the system controls the gate or access control to automatically unlock and allow the vehicle to exit. If there is no entry record, the system contacts the backend administrator through the monitoring and alarm module for processing.

Images