Traffic sign information extraction method and system based on image recognition

Abstract

The invention discloses a traffic sign information extraction method and system based on image recognition, and relates to the technical field of image recognition and processing, and the method comprises the steps: collecting road scene data, generating a panoramic image data set and a point cloud data set, and carrying out the preprocessing of the panoramic image data set and the point cloud data set; constructing a convolutional neural network model, and identifying a panoramic image; performing point cloud data segmentation on the point cloud data set based on a panoramic image recognition result; fusing the point cloud data related to the traffic sign in the point cloud data with the panoramic image recognition result to obtain fused traffic sign data; carrying out encryption output on the fused traffic sign data; according to the method, the panoramic image is recognized by using the convolutional neural network model, point cloud data segmentation is performed on the point cloud data, and the traffic sign data is further extracted, so that the accuracy and efficiency of image recognition are improved, rapid processing of the panoramic image is realized, and the method adapts to a complex road environment and has universality and practicability.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition and processing, and in particular to a method and system for extracting traffic sign information based on image recognition. Background Art

[0002] In recent years, with the development of intelligent transportation systems, the demand for accurate perception and identification of road environments has been increasing. Image recognition technology has been widely used in the transportation field. The traditional traffic sign recognition method is image processing technology, which uses preset image features to detect traffic signs. In addition, deep learning technology, especially the application of convolutional neural networks (CNN), has greatly promoted the progress of traffic sign recognition. The CNN model can learn complex image features through large-scale data training, thereby significantly improving the recognition accuracy.

[0003] Nevertheless, the existing technology still has some significant shortcomings. First, traditional traffic sign recognition systems usually rely on single image data and ignore other forms of data, such as point cloud data, which limits the system's ability to accurately identify signs in complex environments. Point cloud data provides three-dimensional spatial information, which helps to more accurately locate and identify traffic signs, but in the existing technology, the integration and utilization of such data are not mature enough, and data security issues are gradually receiving attention. The existing technology lacks sufficient encryption measures when processing and transmitting traffic sign information, which may lead to the risk of sensitive information leakage. Summary of the invention

[0004] In view of the problems existing in the above-mentioned existing traffic sign information extraction method and system based on image recognition, the present invention is proposed.

[0005] Therefore, the problem to be solved by the present invention is that traditional traffic sign recognition relies on single image data and ignores point cloud data.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a method for extracting traffic sign information based on image recognition, which includes collecting road scene data to generate a panoramic image dataset and a point cloud dataset, and preprocessing the panoramic image dataset and the point cloud dataset; constructing a convolutional neural network model to recognize the panoramic image; performing point cloud data segmentation on the point cloud dataset based on the panoramic image recognition result; fusing the point cloud data related to the traffic sign in the point cloud data with the panoramic image recognition result to obtain fused traffic sign data; and encrypting and outputting the fused traffic sign data.

[0007] As a preferred solution of the traffic sign information extraction method based on image recognition of the present invention, wherein: the acquisition of road scene data to generate a panoramic image dataset and a point cloud dataset includes:

[0008] The VLX device is used to collect data in the road scene. The VLX device integrates a high-precision panoramic camera and a laser scanner to collect 360-degree panoramic images of the road scene and its surrounding environment and three-dimensional point cloud data of the road scene, respectively, to generate a panoramic image dataset and a point cloud dataset.

[0009] As a preferred solution of the traffic sign information extraction method based on image recognition of the present invention, the preprocessing of the panoramic image data set and the point cloud data set includes:

[0010] The preprocessing of the panoramic image data set refers to smoothing the image using a Gaussian filter and performing distortion correction using a bilinear interpolation method;

[0011] The preprocessing of the point cloud data set refers to using a statistical filtering method to remove noise points in the point cloud data, downsampling the point cloud data, and normalizing the coordinates of the point cloud data.

[0012] As a preferred solution of the traffic sign information extraction method based on image recognition of the present invention, wherein: the construction of a convolutional neural network model to recognize a panoramic image includes:

[0013] Collect historical panoramic image data and pre-process the historical panoramic image data;

[0014] Construct a convolutional neural network model, including input layer, convolution layer, pooling layer, fully connected layer and output layer, and set the input layer to panoramic image data;

[0015] Use the preprocessed historical panoramic image data as training data to input into the convolutional neural network model for model training, and use the cross entropy loss function and Adam optimization algorithm to optimize the convolutional neural network model;

[0016] The real-time panoramic image dataset is input into the input layer of the trained convolutional neural network model, and a feature vector is output, which includes the bounding box, corner points and texture features of the traffic sign.

[0017] As a preferred solution of the traffic sign information extraction method based on image recognition of the present invention, the point cloud data segmentation of the point cloud data set based on the panoramic image recognition result includes:

[0018] Based on the panoramic image recognition results, the traffic sign feature points are mapped to three-dimensional point cloud data through spatial projection, and the three-dimensional coordinates corresponding to each pixel in the panoramic image are calculated using the camera's content parameters;

[0019] Use the mapped recognition results to locate the corresponding area in the point cloud data;

[0020] Based on the point cloud area located by the panoramic image recognition information, the RANSAC algorithm is used to segment the point cloud data, including:

[0021] Construct a plane model, randomly select three non-collinear points from the point cloud data set as the point set S, set S = {p1, p2, p3}, p1 = (x1, y1, z1), p2 = (x2, y2, z2), p3 = (x3, y3, z3);

[0022] Perform plane fitting on the selected point set S to obtain the plane equation;

[0023] Substitute all point cloud data into the fitted plane equation and calculate the distance d from each point to the fitted plane;

[0024] Set a distance threshold and compare it with the distance d from each point to the fitting plane. If d≤, the point is judged as an inner point; if d>, the point is judged as an outer point.

[0025] Perform multiple iterations, each time selecting different points as the point set for model fitting, compare the fitting results, and select the plane model with the largest number of internal points for point cloud data segmentation, and segment out the point cloud data related to the traffic sign, including the geometric shape and coordinate data of the traffic sign.

[0026] As a preferred solution of the method for extracting traffic sign information based on image recognition according to the present invention, the method of extracting traffic signs from point cloud data based on segmented point cloud data and fusing the traffic sign data obtained by fusing the traffic sign data with the panoramic image recognition result comprises:

[0027] The coordinate data of the traffic sign is obtained by fusing the bounding box data in the panoramic image recognition and the three-dimensional coordinates in the point cloud segmentation;

[0028] The weighted average method is used to fuse the panoramic image recognition results and the data extracted from the point cloud data;

[0029] The fused traffic sign data are integrated to generate a traffic sign data package, including the three-dimensional coordinates, shape and texture features of the traffic sign.

[0030] As a preferred solution of the method for extracting traffic sign information based on image recognition according to the present invention, the method of encrypting and outputting the fused traffic sign data packet comprises:

[0031] The fused traffic sign data is converted into XML format for output, the output XML data is encrypted using asymmetric encryption technology, a pair of public and private keys are generated, the public key is used to encrypt the XML data, the private key is used to decrypt the data when receiving the XML data, the original XML data is restored, and the encrypted XML data is encrypted and transmitted via the transport layer security protocol TLS.

[0032] Another object of the present invention is to provide a traffic sign information extraction system based on image recognition, which comprises:

[0033] A data acquisition module is used to collect road scene data, generate panoramic image data sets and point cloud data sets, and perform preprocessing;

[0034] Image recognition module, used to build a convolutional neural network model to recognize panoramic images;

[0035] Point cloud segmentation module, used to segment traffic sign related point cloud data;

[0036] A data fusion module is used to fuse the traffic sign point cloud data with the panoramic image recognition result to generate a traffic sign data packet;

[0037] The data encryption module is used to format the traffic sign data packet and perform data encryption.

[0038] A computer device comprises: a memory and a processor; the memory stores a computer program, and the processor implements the steps of a method for extracting traffic sign information based on image recognition when executing the computer program.

[0039] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a method for extracting traffic sign information based on image recognition.

[0040] The beneficial effects of the present invention are as follows: the present invention generates a panoramic image dataset and a point cloud dataset by collecting road scene data, and adopts a convolutional neural network model to recognize the panoramic image, performs point cloud data segmentation on the point cloud data, and further extracts traffic sign data, thereby improving the accuracy and efficiency of traffic sign image recognition. At the same time, combined with advanced point cloud segmentation, the rapid processing of panoramic photos is realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0042] Figure 1 Schematic diagram of the flow of the traffic sign information extraction method based on image recognition.

[0043] Figure 2 Schematic diagram of the process of traffic sign data acquisition and fusion.

[0044] Figure 3 This is the structural diagram of traffic sign information extraction based on image recognition. DETAILED DESCRIPTION

[0045] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0046] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0047] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0048] Example 1

[0049] Reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and which provides a method for extracting traffic sign information based on image recognition. The method for extracting traffic sign information based on image recognition includes:

[0050] S1, collecting road scene data to generate a panoramic image dataset and a point cloud dataset, and preprocessing the panoramic image dataset and the point cloud dataset;

[0051] Furthermore, collecting road scene data to generate panoramic image datasets and point cloud datasets includes:

[0052] The VLX device is used to collect data in the road scene. The VLX device integrates a high-precision panoramic camera and a laser scanner to collect 360-degree panoramic images of the road scene and its surrounding environment and three-dimensional point cloud data of the road scene, respectively, to generate a panoramic image dataset and a point cloud dataset.

[0053] By using the VLX device that integrates a high-precision panoramic camera and laser scanner, all possible traffic signs and related facilities in the road scene are captured to ensure that important information is not missed during the image recognition process. At the same time, panoramic image data and point cloud data are collected. The point cloud data provides three-dimensional spatial information of the road scene, accurately reflects the shape, coordinates and distance of the object, and provides data support for subsequent recognition and processing steps, enhancing the system's adaptability to complex road scenes and improving recognition accuracy and stability.

[0054] Furthermore, the preprocessing of the panoramic image dataset and the point cloud dataset includes,

[0055] The preprocessing of the panoramic image data set refers to smoothing the image using a Gaussian filter and performing distortion correction using a bilinear interpolation method;

[0056] The preprocessing of the point cloud data set refers to using a statistical filtering method to remove noise points in the point cloud data, downsampling the point cloud data, and normalizing the coordinates of the point cloud data.

[0057] By preprocessing the panoramic image dataset and point cloud dataset, noise and redundant information can be effectively removed, including image blur, lighting changes, isolated points and abnormal points in the point cloud data that may be introduced during the acquisition process. This preprocessing step ensures the purity and consistency of the data, thus laying a solid foundation for subsequent image recognition and point cloud processing. Preprocessing can also optimize the quality of image and point cloud data and improve the recognition accuracy of the entire system.

[0058] S2, build a convolutional neural network model to recognize panoramic images;

[0059] Furthermore, a convolutional neural network model is constructed to recognize panoramic images including:

[0060] Collect historical panoramic image data and pre-process the historical panoramic image data;

[0061] Construct a convolutional neural network model, including input layer, convolution layer, pooling layer, fully connected layer and output layer, and set the input layer to panoramic image data;

[0062] Use the preprocessed historical panoramic image data as training data to input into the convolutional neural network model for model training, and use the cross entropy loss function and Adam optimization algorithm to optimize the convolutional neural network model;

[0063] The real-time panoramic image dataset is input into the input layer of the trained convolutional neural network model, and a feature vector is output, which includes the bounding box, corner points and texture features of the traffic sign.

[0064] By constructing a convolutional neural network model to recognize panoramic images, the function of automatically extracting traffic signs and their classification information from road scenes is realized. The convolutional neural network model can effectively extract multi-level features from images and is suitable for processing complex image data. It ensures that the system can accurately identify traffic signs under a variety of environmental conditions, provides high-confidence recognition results for subsequent point cloud data segmentation, and provides key reference information for further processing of point cloud data. Through this step, the precise positioning and classification of traffic signs in panoramic images are achieved, thereby improving the efficiency and accuracy of traffic sign recognition.

[0065] S3, performing point cloud data segmentation on the point cloud data set based on the panoramic image recognition result;

[0066] Furthermore, the point cloud data segmentation of the point cloud data set based on the panoramic image recognition result includes:

[0067] Based on the panoramic image recognition results, the traffic sign feature points are mapped to three-dimensional point cloud data through spatial projection, and the three-dimensional coordinates corresponding to each pixel in the panoramic image are calculated using the camera's content parameters;

[0068] Use the mapped recognition results to locate the corresponding area in the point cloud data;

[0069] Based on the point cloud area located by the panoramic image recognition information, the RANSAC algorithm is used to segment the point cloud data, including:

[0070] Construct a plane model, randomly select three non-collinear points from the point cloud data set as the point set S, set S = {p1, p2, p3}, p1 = (x1, y1, z1), p2 = (x2, y2, z2), p3 = (x3, y3, z3);

[0071] Perform plane fitting on the selected point set S to obtain the plane equation;

[0072] Substitute all point cloud data into the fitted plane equation and calculate the distance d from each point to the fitted plane;

[0073] Set a distance threshold and compare it with the distance d from each point to the fitting plane. If d≤, the point is determined to be an interior point; if d>, the point is determined to be an exterior point. The determination is set by experts based on the relevant technical field content and customer requirements;

[0074] Perform multiple iterations, each time selecting different points as the point set for model fitting, compare the fitting results, and select the plane model with the largest number of internal points for point cloud data segmentation, and segment out the point cloud data related to the traffic sign, including the geometric shape and coordinate data of the traffic sign.

[0075] By performing point cloud data segmentation on the point cloud data set according to the panoramic image recognition results, the segmentation process is made more accurate and efficient. The panoramic image recognition results provide preliminary coordinate data of traffic signs to guide the segmentation process of point cloud data, so that the segmentation range is more focused on the area containing traffic signs, reducing interference and calculation in irrelevant areas. By combining the two-dimensional data of image recognition with the three-dimensional point cloud data, the complexity of point cloud data cutting is greatly reduced, which can reduce the cases of mis-segmentation and missed segmentation in point cloud data and ensure the accuracy and reliability of the segmentation results.

[0076] S4, fusing the point cloud data related to the traffic sign in the point cloud data with the panoramic image recognition result to obtain fused traffic sign data;

[0077] Furthermore, the point cloud data related to the traffic sign in the point cloud data is fused with the panoramic image recognition result to obtain fused traffic sign data including:

[0078] The coordinate data of the traffic sign is obtained by fusing the bounding box data in the panoramic image recognition and the three-dimensional coordinates in the point cloud segmentation;

[0079] The weighted average method is used to fuse the panoramic image recognition results and the data extracted from the point cloud data;

[0080] The fused traffic sign data are integrated to generate a traffic sign data package, including the three-dimensional coordinates, shape and texture features of the traffic sign.

[0081] By fusing the results of point cloud data extraction with the results of panoramic image recognition, the advantages of two-dimensional and three-dimensional data are complemented, and the advantages of image data and point cloud data are integrated to generate complete and accurate traffic sign data. In practical applications, road scenes are often complex. For example, traffic signs may be partially obscured, tilted, and have environmental noise. By using the fused traffic sign data, the system's adaptability in complex road scenes can be improved, ensuring the accurate extraction and recognition of traffic signs, and enhancing the robustness and practicality of the system.

[0082] S5, encrypting and outputting the fused traffic sign data packet;

[0083] Furthermore, encrypting and outputting the fused traffic sign data packet includes:

[0084] The fused traffic sign data is converted into XML format for output, the output XML data is encrypted using asymmetric encryption technology, a pair of public and private keys are generated, the public key is used to encrypt the XML data, the private key is used to decrypt the data when receiving the XML data, the original XML data is restored, and the encrypted XML data is encrypted and transmitted via the transport layer security protocol TLS.

[0085] By encrypting the fused traffic sign data, data protection is achieved. Especially during data transmission, it can effectively prevent data from being intercepted, tampered with, and accessed without authorization, ensuring the confidentiality and integrity of traffic sign information throughout the system operation, avoiding potential security threats and information leakage, and thus significantly improving the overall security of the system. Traffic sign data can be converted into multiple formats and can support output in multiple vector file formats. Users can select and edit according to their needs, thereby improving customer satisfaction.

[0086] Example 2

[0087] Reference Figure 3 , which is a second embodiment of the present invention, and which is different from the previous embodiment, provides a traffic sign information extraction system based on image recognition, comprising:

[0088] A data acquisition module is used to collect road scene data, generate panoramic image data sets and point cloud data sets, and perform preprocessing;

[0089] Image recognition module, used to build a convolutional neural network model to recognize panoramic images;

[0090] Point cloud segmentation module, used to segment traffic sign related point cloud data;

[0091] A data fusion module is used to fuse the traffic sign related point cloud data with the panoramic image recognition result to generate a traffic sign data packet;

[0092] The data encryption module is used to format the traffic sign data packet and perform data encryption.

[0093] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0094] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0095] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0096] It should be understood that the various parts of the present invention can be implemented with hardware, software, firmware or a combination thereof. In the above embodiments, a plurality of steps or methods can be implemented with software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented with hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0097] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for extracting traffic sign information based on image recognition, characterized in that: include, Collect road scene data to generate a panoramic image dataset and a point cloud dataset, and pre-process the panoramic image dataset and the point cloud dataset; Build a convolutional neural network model to recognize panoramic images; Perform point cloud data segmentation on the point cloud data set based on the panoramic image recognition results; The point cloud data related to the traffic sign in the point cloud data is fused with the panoramic image recognition result to obtain fused traffic sign data; The fused traffic sign data is encrypted and output.

2. The method for extracting traffic sign information based on image recognition as claimed in claim 1, characterized in that: The collecting of road scene data to generate a panoramic image dataset and a point cloud dataset includes: The VLX device is used to collect data in the road scene. The VLX device integrates a high-precision panoramic camera and a laser scanner to collect 360-degree panoramic images of the road scene and its surrounding environment and three-dimensional point cloud data of the road scene, respectively, to generate a panoramic image dataset and a point cloud dataset.

3. The method for extracting traffic sign information based on image recognition as claimed in claim 2, characterized in that: The preprocessing of the panoramic image dataset and the point cloud dataset includes: The preprocessing of the panoramic image data set refers to smoothing the image using a Gaussian filter and performing distortion correction using a bilinear interpolation method; The preprocessing of the point cloud data set refers to using a statistical filtering method to remove noise points in the point cloud data, downsampling the point cloud data, and normalizing the coordinates of the point cloud data.

4. The method for extracting traffic sign information based on image recognition as claimed in claim 3, characterized in that: The construction of the convolutional neural network model to recognize the panoramic image includes: Collect historical panoramic image data and pre-process the historical panoramic image data; Construct a convolutional neural network model, including input layer, convolution layer, pooling layer, fully connected layer and output layer, and set the input layer to panoramic image data; Use the preprocessed historical panoramic image data as training data to input into the convolutional neural network model for model training, and use the cross entropy loss function and Adam optimization algorithm to optimize the convolutional neural network model; The real-time panoramic image dataset is input into the input layer of the trained convolutional neural network model, and a feature vector is output, which includes the bounding box, corner points and texture features of the traffic sign.

5. The method for extracting traffic sign information based on image recognition as claimed in claim 4, characterized in that: The point cloud data segmentation of the point cloud data set based on the panoramic image recognition result includes: Based on the panoramic image recognition results, the traffic sign feature points are mapped to three-dimensional point cloud data through spatial projection, and the three-dimensional coordinates corresponding to each pixel in the panoramic image are calculated using the camera's content parameters; Use the mapped recognition results to locate the corresponding area in the point cloud data; Based on the point cloud area located by the panoramic image recognition information, the RANSAC algorithm is used to segment the point cloud data, including: Construct a plane model, randomly select three non-collinear points from the point cloud data set as the point set S, set S = {p1, p2, p3}, p1 = (x1, y1, z1), p2 = (x2, y2, z2), p3 = (x3, y3, z3); Perform plane fitting on the selected point set S to obtain the plane equation; Substitute all point cloud data into the fitted plane equation and calculate the distance d from each point to the fitted plane; Set a distance threshold and compare it with the distance d from each point to the fitting plane. If d≤, the point is judged as an inner point; if d>, the point is judged as an outer point. Perform multiple iterations, each time selecting different points as the point set for model fitting, compare the fitting results, and select the plane model with the largest number of internal points for point cloud data segmentation, and segment out the point cloud data related to the traffic sign, including the geometric shape and coordinate data of the traffic sign.

6. The method for extracting traffic sign information based on image recognition as claimed in claim 5, characterized in that: The step of fusing the point cloud data related to the traffic sign in the point cloud data with the panoramic image recognition result to obtain the fused traffic sign data includes: The coordinate data of the traffic sign is obtained by fusing the bounding box data in the panoramic image recognition and the three-dimensional coordinates in the point cloud segmentation; The weighted average method is used to fuse the panoramic image recognition results and the data extracted from the point cloud data; The fused traffic sign data are integrated to generate a traffic sign data package, including the three-dimensional coordinates, shape and texture features of the traffic sign.

7. The method for extracting traffic sign information based on image recognition as claimed in claim 6, characterized in that: The encrypted output of the fused traffic sign data packet includes: The fused traffic sign data is converted into XML format for output, the output XML data is encrypted using asymmetric encryption technology, a pair of public and private keys are generated, the public key is used to encrypt the XML data, the private key is used to decrypt the data when receiving the XML data, the original XML data is restored, and the encrypted XML data is encrypted and transmitted via the transport layer security protocol TLS.

8. A traffic sign information extraction system based on image recognition according to the traffic sign information extraction method based on image recognition according to any one of claims 1 to 7, characterized in that: include, A data acquisition module is used to collect road scene data, generate panoramic image data sets and point cloud data sets, and perform preprocessing; Image recognition module, used to build a convolutional neural network model to recognize panoramic images; Point cloud segmentation module, used to segment traffic sign related point cloud data; A data fusion module is used to fuse the traffic sign related point cloud data with the panoramic image recognition result to generate a traffic sign data packet; The data encryption module is used to format the traffic sign data packet and perform data encryption.

9. A computer device comprising: A memory and a processor; the memory stores a computer program, characterized in that: when the processor executes the computer program, the steps of the traffic sign information extraction method based on image recognition described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for extracting traffic sign information based on image recognition according to any one of claims 1 to 7 are implemented.

Images