Intelligent traffic condition information management platform and method
By combining monitoring equipment video and road sensor data, deep learning technology is used to analyze traffic conditions and generate optimization suggestions, the problem that traditional traffic management cannot cope with complex traffic conditions is solved, and all-round monitoring and optimization of road traffic conditions is achieved, improving the Traffic management efficiency and safety.
Patent Information
- Application Number
- CN202410478750.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-20
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-04-20
AI Technical Summary
Traditional road traffic management methods cannot effectively deal with road congestion in the process of urbanization, lack the ability to monitor and analyze complex traffic conditions in all aspects, and it is difficult to discover the root causes of traffic problems and propose targeted solutions.
By obtaining traffic road conditions video collected by monitoring equipment and vehicle text data collected by road pre-embedded sensor network, deep learning technology is used to perform feature extraction and correlation analysis, and traffic optimization suggestions are generated to achieve a comprehensive understanding of road traffic conditions and identify the root causes of problems. .
It improves the efficiency and safety of traffic management, alleviates traffic congestion, provides more effective traffic management strategies, improves residents' travel experience, and achieves the goal of smart city traffic management.
Smart Images

Figure CN118298637B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of traffic condition information management, and more specifically, to an intelligent traffic condition information management platform and method. Background Art
[0002] Traffic condition management aims to monitor, plan and optimize the road traffic system to improve traffic efficiency, reduce congestion and accident rates. Through real-time data collection, analysis and traffic signal control, traffic condition management can effectively regulate traffic flow, improve road capacity, improve the urban traffic environment, and promote traffic safety and convenience.
[0003] With the rapid advancement of urbanization in my country, the number of vehicles on urban roads has increased year by year, which has led to increasingly serious congestion on major roads and caused great trouble for residents. Traditional road guidance methods and simple signal light control can no longer effectively manage complex road traffic conditions.
[0004] Therefore, an intelligent traffic condition information management platform and method are desired. Summary of the invention
[0005] In order to solve the above technical problems, this application is proposed. The embodiment of this application provides a smart traffic information management platform and method, which first obtains traffic video collected by monitoring equipment and vehicle text data collected by road pre-buried sensor network, and then uses deep learning technology to extract features and perform correlation analysis on the two, and finally generates traffic optimization suggestions through a generator, thereby more comprehensively understanding road traffic conditions, thereby discovering the root causes of traffic problems and proposing targeted solutions.
[0006] According to one aspect of the present application, a smart traffic condition information management platform is provided, which includes:
[0007] Traffic condition information acquisition module, used to obtain traffic condition videos collected by monitoring equipment and vehicle text data collected by the road pre-buried sensor network;
[0008] A traffic condition information extraction module, used to extract traffic condition change-related feature vectors and vehicle information-related feature vectors from the traffic condition video collected by the monitoring device and the vehicle text data collected by the road pre-buried sensor network;
[0009] The traffic optimization suggestion generating module is used to generate traffic optimization suggestions based on the traffic condition change associated feature vector and the vehicle information associated feature vector.
[0010] According to another aspect of the present application, a method for managing intelligent traffic condition information is provided, which includes:
[0011] Obtain traffic condition videos collected by monitoring equipment and vehicle text data collected by the road pre-buried sensor network;
[0012] Extracting traffic condition change-related feature vectors and vehicle information-related feature vectors from the traffic condition video collected by the monitoring device and the vehicle text data collected by the road pre-buried sensor network;
[0013] Based on the traffic condition change-related feature vector and the vehicle information-related feature vector, a traffic optimization suggestion is generated.
[0014] Compared with the prior art, the present application provides a smart traffic condition information management platform and method, which first obtains traffic condition videos collected by monitoring equipment and vehicle text data collected by a road pre-buried sensor network, and then uses deep learning technology to perform feature extraction and correlation analysis on the two, and finally generates traffic optimization suggestions through a generator, thereby gaining a more comprehensive understanding of road traffic conditions, thereby discovering the root causes of traffic problems and proposing targeted solutions. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.
[0016] Figure 1 This is a block diagram of a smart traffic condition information management platform according to an embodiment of the present application.
[0017] Figure 2 It is a block diagram of the traffic condition information extraction module in the smart traffic condition information management platform according to the embodiment of the present application.
[0018] Figure 3 This is a block diagram of a vehicle text feature extraction unit in a smart traffic condition information management platform according to an embodiment of the present application.
[0019] Figure 4 This is a block diagram of a traffic optimization suggestion generation module in a smart traffic condition information management platform according to an embodiment of the present application.
[0020] Figure 5 This is a flowchart of a smart traffic condition information management method according to an embodiment of the present application.
[0021] Figure 6 is a block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION
[0022] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.
[0023] Exemplary Systems
[0024] Figure 1 FIG. 1 is a block diagram of a smart traffic information management platform according to an embodiment of the present application. Figure 1 As shown, the smart traffic condition information management platform 100 according to the embodiment of the present application includes: a traffic condition information acquisition module 110, which is used to obtain traffic condition videos collected by monitoring equipment and vehicle text data collected by a road pre-buried sensor network; a traffic condition information extraction module 120, which is used to extract traffic condition change-related feature vectors and vehicle information-related feature vectors from the traffic condition videos collected by the monitoring equipment and the vehicle text data collected by the road pre-buried sensor network; a traffic optimization suggestion generation module 130, which is used to generate traffic optimization suggestions based on the traffic condition change-related feature vectors and the vehicle information-related feature vectors.
[0025] In the above-mentioned intelligent traffic condition information management platform 100, the traffic condition information acquisition module 110 is used to obtain traffic condition videos collected by monitoring equipment and vehicle text data collected by the road pre-buried sensor network. It should be understood that traffic conditions refer to the comprehensive situation of the flow, speed, density and other related factors of vehicles and pedestrians on the road. It reflects the current operating status of the road system, including the degree of traffic congestion, traffic smoothness, traffic accidents, etc. By monitoring and analyzing traffic conditions, it can help traffic management departments and drivers better understand road conditions, prevent traffic congestion and accidents in advance, thereby improving road traffic efficiency and driving safety. However, with the rapid development of urbanization, the number of vehicles on urban roads has continued to increase, resulting in the worsening of congestion problems on major roads, which has brought great troubles to residents' travel. Traditional road guidance methods and simple signal light control can no longer effectively manage complex road traffic conditions. Considering that only obtaining video data may be affected by factors such as weather and light, it is easy to have recognition errors, and the lack of text information of sensor data, such as vehicle speed, vehicle type, vehicle license plate, etc., limits the ability to analyze traffic conditions in detail. Therefore, in the technical solution of the present application, traffic optimization suggestions are generated by obtaining traffic condition videos collected by monitoring equipment and vehicle text data collected by the road pre-buried sensor network, and combining deep learning technology. Among them, the traffic condition video collected by the monitoring equipment provides an intuitive display of traffic conditions, which can capture traffic congestion, accidents and other emergencies. By analyzing the video data, the system can monitor the road conditions in real time, find problems in time and take corresponding measures. The vehicle text data collected by the road pre-buried sensor network provides more detailed information about the vehicle's operating status and driving trajectory. These data may include information such as vehicle speed, density, and vehicle type. By analyzing these data, the system can better understand the dynamic changes of traffic flow, predict the possible location and time of traffic congestion, and take traffic optimization measures in advance, such as adjusting the timing of traffic lights or guiding traffic flow. By comprehensively utilizing the video collected by the monitoring equipment and the text data collected by the road sensors, the intelligent traffic management system can achieve all-round monitoring and analysis of traffic conditions. By extracting key features, the system can generate traffic optimization suggestions to help urban traffic management departments formulate more effective traffic management strategies, alleviate traffic congestion, improve traffic efficiency, and improve residents' travel experience, thereby achieving the goal of smart city traffic management.
[0026] In the above-mentioned intelligent traffic condition information management platform 100, the traffic condition information extraction module 120 is used to extract the traffic condition change associated feature vector and the vehicle information associated feature vector from the traffic condition video collected by the monitoring device and the vehicle text data collected by the road pre-buried sensor network. It should be understood that by extracting the traffic condition change associated feature vector and the vehicle information associated feature vector from the traffic condition video collected by the monitoring device and the vehicle text data collected by the road pre-buried sensor network, these feature vectors can be used to more comprehensively and accurately describe the traffic conditions, thereby realizing the optimization and decision support of the intelligent traffic management system.
[0027] Figure 2 FIG. 1 is a block diagram of a traffic information extraction module in a smart traffic information management platform according to an embodiment of the present application. Figure 2 As shown, in a specific embodiment of the present application, the traffic information extraction module 120 includes: a traffic video feature extraction unit 121, which is used to extract features from the traffic video collected by the monitoring device to obtain multiple traffic feature matrices; a traffic feature encoding unit 122, which is used to feature encode the multiple traffic feature matrices to obtain the traffic change associated feature vector; a vehicle text feature extraction unit 123, which is used to extract features from the vehicle text data collected by the road pre-buried sensor network to obtain the vehicle information associated feature vector. It should be understood that the traffic feature matrix is a matrix obtained by extracting features from the traffic video, in which each element represents a specific traffic feature. These features may include information such as vehicle density, speed distribution, road occupancy, traffic flow, etc. By extracting these features, the changes and characteristics of the road conditions can be described more comprehensively, providing more reference basis for traffic management decisions.
[0028] Furthermore, feature encoding is the process of converting raw feature data into a more representative and interpretable feature representation. In the field of transportation, feature encoding of multiple traffic condition feature matrices can fuse and extract various road condition feature information, thereby generating a more compact and informative traffic condition change associated feature vector.
[0029] Furthermore, the road pre-buried sensor network is a sensor network used to monitor vehicles passing through the road, which can collect text data of vehicles, such as vehicle type, speed, vehicle location and other information. These data contain rich information about vehicle behavior and road use, which can be converted into more meaningful and operational feature vectors through feature extraction. In the vehicle information associated feature vector, each feature may represent different vehicle attributes or behaviors, such as speed distribution, vehicle density, vehicle type distribution, etc. Through feature extraction, these key features can be extracted from the original text data to form a comprehensive feature vector, which reflects the behavior and characteristics of the vehicle on the road.
[0030] In a specific embodiment of the present application, the traffic condition video feature extraction unit 121 includes: extracting video key frames from the traffic condition video collected by the monitoring device to obtain multiple traffic condition key frames; passing the multiple traffic condition key frames through a traffic condition convolutional neural network including a deep and shallow fusion module to obtain the multiple traffic condition feature matrices. It should be understood that the traffic condition video data is usually huge, and extracting key frames can compress the video data into representative frames of key information, thereby reducing storage space and speeding up data retrieval. In addition, the video key frame is the frame that best represents the entire video content in the video sequence. Extracting these key frames can help quickly browse and understand the video content without watching the entire video frame by frame, so as to help simplify data processing, improve information extraction efficiency, strengthen monitoring and analysis of traffic conditions, and provide more effective decision support and the ability to respond to emergencies for traffic management departments. Specifically, multiple traffic condition key frames are extracted from the traffic condition video collected by the monitoring device at a predetermined sampling frequency.
[0031] Furthermore, in order to use deep learning technology to extract key features in traffic conditions, multiple traffic condition key frames are passed through a traffic condition convolutional neural network containing a deep and shallow fusion module to obtain multiple traffic condition feature matrices. It should be understood that the deep and shallow fusion module is a module that combines deep and shallow features, aiming to make full use of feature information at different levels and improve the representation and generalization capabilities of the model. Through this module, multiple levels of feature representation can be effectively fused to more comprehensively capture the complex features of traffic conditions. Among them, the traffic condition convolutional neural network is a neural network model specially designed to process traffic scene data, with strong feature extraction and representation capabilities. By inputting multiple traffic condition key frames into such a neural network, the network can learn key features such as spatiotemporal information and vehicle operation status in traffic conditions. Multiple traffic condition key frames are respectively input into the traffic condition convolutional neural network containing a deep and shallow fusion module. The deep and shallow fusion module can combine feature information at different levels to improve the representation ability of features, so that the model can better understand the complexity of traffic conditions. Specifically, a shallow feature map is extracted from the Mth layer of the traffic road condition convolutional neural network including the deep and shallow fusion module, wherein M is greater than or equal to 1 and less than or equal to 6; a deep feature map is extracted from the Nth layer of the traffic road condition convolutional neural network including the deep and shallow fusion module, wherein N / M is greater than or equal to 5 and less than or equal to 10; the shallow feature map and the deep feature map are fused using the deep and shallow feature fusion module of the traffic road condition convolutional neural network including the deep and shallow fusion module to obtain a fused feature map; and, the fused feature map is globally pooled along the channel dimension to obtain the multiple traffic road condition feature matrices.
[0032] In a specific embodiment of the present application, the traffic condition feature encoding unit 122 includes: aggregating the multiple traffic condition feature matrices along the channel dimension to obtain a three-dimensional traffic condition input tensor; passing the three-dimensional traffic condition input tensor through a three-dimensional traffic condition convolution feature encoder to obtain the traffic condition change associated feature vector. It should be understood that in deep learning, neural networks usually require data in the form of tensors as input. By integrating multiple feature matrices into a tensor as input, richer information can be provided to the neural network, which helps to improve the performance and generalization ability of the model.
[0033] Furthermore, the traffic condition three-dimensional convolution feature encoder is a convolutional neural network model specially designed for processing traffic condition data, and its structure and parameter settings can effectively extract key feature information from the three-dimensional input tensor of traffic conditions. By inputting the three-dimensional input tensor of traffic conditions into the feature encoder, the feature encoder can capture the changing trend and related features of traffic conditions, help identify the differences in traffic conditions at different time points or regions, and thus effectively extract and encode the key features of traffic conditions and capture the change-related information of traffic conditions. Specifically, the traffic condition three-dimensional convolution feature encoder is used to perform the following operations on the input data in the forward pass of the layer: perform three-dimensional convolution processing on the input data based on the three-dimensional convolution kernel to obtain a convolution feature map; perform mean pooling processing on each feature matrix along the channel dimension on the convolution feature map to obtain a pooled feature map; and perform nonlinear activation on the pooled feature map to obtain an activated feature map; wherein the output of the last layer of the traffic condition three-dimensional convolution feature encoder is the traffic condition change-related feature vector, and the input of the first layer of the traffic condition three-dimensional convolution feature encoder is the traffic condition three-dimensional input tensor.
[0034] Figure 3 FIG. 1 is a block diagram of a vehicle text feature extraction unit in a smart traffic information management platform according to an embodiment of the present application. Figure 3As shown, in a specific embodiment of the present application, the vehicle text feature extraction unit 123 includes: a vehicle text information semantic encoding subunit 1231, which is used to semantically encode the vehicle text data collected by the road pre-buried sensor network to obtain multiple vehicle information feature vectors; a vehicle information association convolution encoding subunit 1232, which is used to pass the multiple vehicle information feature vectors through the vehicle information association convolution encoding model to obtain the vehicle information association feature vector. It should be understood that the vehicle text data collected by the road pre-buried sensor network may contain various information about the vehicle, such as vehicle model, color, speed, position, etc. Through semantic encoding, these text data can be converted into continuous vector representations. In the process of semantic encoding, semantic information needs to be extracted from the vehicle text data first. This may involve the use of natural language processing techniques, such as word embedding models or pre-trained language models, to understand the meaning and key information in the text. Then, the extracted semantic information will be converted into a vector representation. This vector representation is usually a high-dimensional numerical vector, in which each dimension corresponds to a specific feature or concept in the semantic space. In this way, the semantic information in the text data is mapped to the vector space. After semantic encoding, the text data of each vehicle will be converted into a feature vector, which contains the semantic representation of vehicle information. Among them, the text data of each vehicle will be encoded into a feature vector, so semantic encoding of the vehicle text data in the entire data set will obtain multiple vehicle information feature vectors, each of which represents the semantic information of a vehicle. By semantically encoding the vehicle text data collected by the road pre-buried sensor network to obtain multiple vehicle information feature vectors, semantic understanding and representation of vehicle information can be achieved.
[0035] Furthermore, multiple vehicle information feature vectors are passed through a vehicle information association convolutional coding model, aiming to encode and associate vehicle information through a convolutional neural network, thereby capturing the association features between different vehicles. It should be understood that the vehicle information association convolutional coding model is a neural network model specially designed for processing vehicle information, and its structure and parameter settings can effectively extract associated feature information from multiple vehicle information feature vectors, helping the model to better understand the relationship and interaction between vehicles. Through convolution operations and feature mapping, the vehicle information association convolutional coding model can fuse multiple vehicle information feature vectors together to form a more comprehensive and representative vehicle information association feature vector to effectively extract and model the association features between vehicles. Specifically, each layer of the vehicle information association convolutional coding model is used to perform convolution processing, pooling processing and nonlinear activation processing on the input data in the forward pass of the layer to output the vehicle information association feature vector from the last layer of the vehicle information association convolutional coding model, wherein the input of the vehicle information association convolutional coding model is the multiple vehicle information feature vectors.
[0036] In a specific embodiment of the present application, the vehicle text information semantic encoding subunit 1231 includes: performing word embedding encoding on the vehicle text data collected by the road pre-buried sensor network to obtain multiple vehicle information data items; passing the multiple vehicle information data items through the vehicle information semantic encoding model to obtain the multiple vehicle information feature vectors. It should be understood that word embedding is a technology that maps words to a continuous vector space, and converts text information into a dense vector representation by learning the semantic relationship between words. In vehicle text data, word embedding technology can help convert key words and concepts in vehicle information into vector form. Through word embedding encoding, each word or phrase will be mapped to a vector, which contains the semantic information of the word. For each word or phrase in the vehicle text data, the corresponding vector representation can be obtained through word embedding encoding, thereby realizing the semantic representation of vehicle information. By performing word embedding encoding on the vehicle text data collected by the road pre-buried sensor network to obtain multiple vehicle information data items, the semantic representation of vehicle information can be realized.
[0037] Furthermore, the vehicle information semantic encoding model is a deep learning model that aims to convert vehicle information data items into feature vectors with semantic meaning by learning the semantic associations between these information items. In this way, the model can better understand the meaning and context of vehicle information to extract rich semantic features from multiple vehicle information data items. These feature vectors contain important semantic information of vehicle information, which helps the system understand vehicle data more comprehensively. Specifically, the embedding layer of the vehicle information semantic encoding model is used to map each of the multiple vehicle information data items to a word embedding vector to obtain a sequence of word embedding vectors; the converter-based Bert model of the vehicle information semantic encoding model is used to perform global contextual semantic encoding on the sequence of word embedding vectors to obtain multiple vehicle information feature vectors.
[0038] In the above-mentioned intelligent traffic condition information management platform 100, the traffic optimization suggestion generation module 130 is used to generate traffic optimization suggestions based on the traffic condition change associated feature vector and the vehicle information associated feature vector. It should be understood that by performing correlation analysis on the traffic condition change associated feature vector and the vehicle information associated feature vector, exploring the relationship and mutual influence between them, it is possible to find the impact of traffic condition changes on vehicle information and the feedback relationship of vehicle information on traffic conditions. Then, based on the analysis and correlation results of the feature vectors, the system can generate traffic optimization suggestions. These suggestions can cover aspects such as traffic signal optimization, route planning, and traffic control, aiming to improve traffic efficiency, reduce congestion, and enhance traffic safety. For example, the system can adjust the timing of traffic lights, optimize vehicle routes, etc. according to vehicle information and road condition changes.
[0039] Figure 4 FIG. 1 is a block diagram of a traffic optimization suggestion generation module in a smart traffic condition information management platform according to an embodiment of the present application. Figure 4 As shown, in a specific embodiment of the present application, the traffic optimization suggestion generation module 130 includes a traffic information change feature fusion unit 131, which is used to fuse the traffic condition change associated feature vector and the vehicle information associated feature vector to obtain a traffic condition judgment feature vector; an optimization unit 132, which is used to perform implicit group optimization with spatial sparsity restrictions on the traffic condition judgment feature vector to obtain an optimized traffic condition judgment feature vector; and an optimization suggestion judgment generation unit 133, which is used to pass the optimized traffic condition judgment feature vector through a generator to obtain a traffic optimization suggestion. It should be understood that the traffic condition change associated feature vector and the vehicle information associated feature vector respectively contain important features about traffic conditions and vehicle information. Fusion of these two types of feature vectors can integrate information from different aspects, so that the system can comprehensively consider traffic conditions and vehicle information. During the fusion process, the system will merge or combine the traffic condition change associated feature vector and the vehicle information associated feature vector to generate a new traffic condition judgment feature vector. This new feature vector will comprehensively reflect the relationship between traffic conditions and vehicle information, providing a more comprehensive information basis. The resulting traffic condition feature vector can be used to support traffic decision-making and optimization strategy formulation. The system can provide real-time traffic optimization suggestions based on this feature vector to help improve traffic flow, reduce congestion and improve traffic efficiency.
[0040] In particular, in the technical solution of the present application, it is considered that in the feature extraction process of traffic condition video and vehicle text data, all key features may not be fully captured. If the important features of traffic condition and vehicle information are not accurately extracted or represented, the information aggregation degree will be low. In the fusion process of traffic condition feature matrix and vehicle information feature vector, there may be unreasonable fusion mode or insufficient balance. If the feature fusion is inappropriate, it will affect the full integration of information, resulting in low information aggregation degree. If the feature representation method between traffic condition change associated feature vector and vehicle information associated feature vector is not unified or the scale is inconsistent, the information aggregation degree will also be reduced. Due to the low information aggregation degree, the generated traffic condition judgment feature vector may not be able to fully reflect the correlation between traffic condition and vehicle information, resulting in the lack of accuracy and practicality of the generated traffic optimization suggestions. If the correlation features between traffic condition and vehicle information cannot be accurately captured, it is impossible to effectively plan traffic routes and traffic flow control, which may lead to problems such as traffic jams and accidents. The low information aggregation degree will lead to the decline of the quality of the traffic condition judgment feature vector, so that the generated traffic optimization suggestions cannot effectively improve traffic efficiency, which may lead to traffic congestion and reduced efficiency. In order to solve the problem of low information aggregation of the traffic condition judgment feature vector, in the technical solution of the present application, implicit group optimization with spatial sparsity restriction is performed on the traffic condition judgment feature vector.
[0041] Specifically, the traffic condition judgment feature vector is subjected to implicit group optimization with spatial sparsity restriction to obtain an optimized traffic condition judgment feature vector, including: calculating an implicit group optimization factor with spatial sparsity restriction of the traffic condition judgment feature vector; weighting the traffic condition judgment feature vector using the implicit group optimization factor with spatial sparsity restriction as a weight to obtain the optimized traffic condition judgment feature vector.
[0042] Furthermore, the implicit group optimization factor of the spatial sparsity restriction of the traffic condition judgment feature vector is calculated, including: summing the absolute values of the eigenvalues of all positions of the traffic condition judgment feature vector to obtain the total value of the traffic condition judgment feature vector; calculating the norm of the traffic condition judgment feature vector; calculating the variance of the eigenvalue set of the traffic condition judgment feature vector; calculating the logarithmic function value with base 2 of the total value of the traffic condition judgment feature vector, multiplying it by the norm of the traffic condition judgment feature vector, and then dividing it by the variance of the eigenvalue set of the traffic condition judgment feature vector to obtain the implicit group optimization factor of the spatial sparsity restriction of the traffic condition judgment feature vector.
[0043] Here, the traffic condition judgment feature vector is subjected to implicit group optimization with spatial sparsity constraints, which evaluates the fitness of each particle in the traffic condition judgment feature vector in a high-dimensional feature space, and updates the traffic condition judgment feature vector according to the optimal solution of each particle in the traffic condition judgment feature vector and the group optimal solution so that the updated traffic condition judgment feature vector satisfies the spatial sparsity constraint. In this way, the aggregation degree of the feature manifold of the traffic condition judgment feature vector is enhanced by restoring the basic information in the full-precision information representation space, thereby improving the expression effect of the traffic condition judgment feature vector.
[0044] Furthermore, by optimizing the traffic condition judgment feature vector through the generator, the system can generate targeted traffic optimization suggestions based on the optimized traffic condition judgment feature vector. These suggestions may involve traffic signal adjustment, route planning optimization, traffic control measures, etc., aiming to improve traffic efficiency, reduce congestion, and enhance traffic safety, thereby providing more intelligent and convenient support for urban traffic management and individual travel.
[0045] In summary, the embodiment of the present application first obtains the traffic condition video collected by the monitoring equipment and the vehicle text data collected by the road pre-buried sensor network, and then uses deep learning technology to perform feature extraction and correlation analysis on the two, and finally generates traffic optimization suggestions through a generator, thereby gaining a more comprehensive understanding of the road traffic conditions, thereby discovering the root causes of traffic problems and proposing targeted solutions.
[0046] As described above, the smart traffic information management platform 100 according to the embodiment of the present application can be implemented in various terminal devices, such as a server deployed with a smart traffic information management control algorithm. In one example, the smart traffic information management platform 100 can be integrated into the terminal device as a software module and / or a hardware module. For example, the smart traffic information management platform 100 can be a software module in the operating system of the terminal device, or it can be an application developed for the terminal device; of course, the smart traffic information management platform 100 can also be one of the many hardware modules of the terminal device.
[0047] Alternatively, in another example, the smart traffic condition information management platform 100 and the terminal device may also be separate devices, and the smart traffic condition information management platform 100 may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.
[0048] Exemplary Methods
[0049] Figure 5Flow chart of the intelligent traffic condition information management method according to the embodiment of the present application. Figure 5 As shown, according to the intelligent traffic condition information management method of the embodiment of the present application, it includes: S110, acquiring the traffic condition video collected by the monitoring equipment and the vehicle text data collected by the road pre-buried sensor network; S120, extracting the traffic condition change associated feature vector and the vehicle information associated feature vector from the traffic condition video collected by the monitoring equipment and the vehicle text data collected by the road pre-buried sensor network; S130, generating traffic optimization suggestions based on the traffic condition change associated feature vector and the vehicle information associated feature vector.
[0050] Here, those skilled in the art can understand that the specific operations of each step in the above-mentioned intelligent traffic condition information management method have been referred to above. Figures 1 to 4 It has been introduced in detail in the description of the intelligent traffic condition information management platform, and therefore, its repeated description will be omitted.
[0051] Exemplary Electronic Devices
[0052] Below, reference Figure 6 To describe an electronic device according to an embodiment of the present application.
[0053] like Figure 6 As shown, the electronic device 10 includes an input device 11, an input interface 12, a central processing unit 13, a memory 14, an output interface 15, an output device 16, and a bus 17. The input interface 12, the central processing unit 13, the memory 14, and the output interface 15 are connected to each other through the bus 17, and the input device 11 and the output device 16 are connected to the bus 17 through the input interface 12 and the output interface 15, respectively, and then connected to other components of the electronic device 10.
[0054] Specifically, the input device 11 receives input information from the outside and transmits the input information to the central processing unit 13 through the input interface 12; the central processing unit 13 processes the input information based on the computer executable instructions stored in the memory 14 to generate output information, temporarily or permanently stores the output information in the memory 14, and then transmits the output information to the output device 16 through the output interface 15; the output device 16 outputs the output information to the outside of the electronic device 10 for user use.
[0055] In one embodiment, Figure 6 The electronic device 10 shown can be implemented as a network device, which may include: a memory configured to store programs; a processor configured to run the programs stored in the memory to execute any one of the smart traffic road condition information management methods described in the above embodiments.
[0056] According to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program tangibly contained on a machine-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network, and / or installed from a removable storage medium.
[0057] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In hardware implementations, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.
[0058] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present application, but the present application is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and substance of the present application, and these modifications and improvements are also considered to be within the scope of protection of the present application.
Claims
1. A smart traffic information management platform, characterized in that: include: Traffic condition information acquisition module, used to obtain traffic condition videos collected by monitoring equipment and vehicle text data collected by the road pre-buried sensor network; A traffic condition information extraction module, used to extract traffic condition change-related feature vectors and vehicle information-related feature vectors from the traffic condition video collected by the monitoring device and the vehicle text data collected by the road pre-buried sensor network; A traffic optimization suggestion generating module, used for generating a traffic optimization suggestion based on the traffic condition change associated feature vector and the vehicle information associated feature vector; Wherein, the traffic optimization suggestion generation module includes: A traffic information change feature fusion unit, used for fusing the traffic condition change associated feature vector and the vehicle information associated feature vector to obtain a traffic condition judgment feature vector; An optimization unit, used for performing implicit group optimization with spatial sparsity restriction on the traffic condition judgment feature vector to obtain an optimized traffic condition judgment feature vector; An optimization suggestion judgment generating unit, used for passing the optimized traffic condition judgment feature vector through a generator to obtain a traffic optimization suggestion; Wherein, the optimization unit comprises: Calculating an implicit group optimization factor constrained by spatial sparsity of the traffic condition judgment feature vector; The traffic condition judgment feature vector is weighted by using the implicit group optimization factor of the spatial sparsity restriction as a weight to obtain the optimized traffic condition judgment feature vector; The calculation of the implicit group optimization factor of the spatial sparsity restriction of the traffic condition judgment feature vector includes: Summing the absolute values of the eigenvalues at all positions of the traffic condition judgment eigenvector to obtain a total value of the traffic condition judgment eigenvector; Calculating a norm of the traffic condition judgment feature vector; Calculating the variance of the eigenvalue set of the traffic condition judgment eigenvector; The logarithmic function value with base 2 of the sum value of the traffic condition judgment feature vector is calculated, multiplied by a norm of the traffic condition judgment feature vector, and then divided by the variance of the eigenvalue set of the traffic condition judgment feature vector to obtain an implicit group optimization factor restricted by the spatial sparsity of the traffic condition judgment feature vector.
2. The intelligent traffic condition information management platform according to claim 1 is characterized in that: The traffic condition information extraction module comprises: A traffic condition video feature extraction unit, used for extracting features from the traffic condition video collected by the monitoring device to obtain a plurality of traffic condition feature matrices; A traffic condition feature encoding unit, used for performing feature encoding on the plurality of traffic condition feature matrices to obtain the traffic condition change associated feature vector; The vehicle text feature extraction unit is used to extract features from the vehicle text data collected by the road pre-buried sensor network to obtain the vehicle information associated feature vector.
3. The intelligent traffic condition information management platform according to claim 2 is characterized in that: The traffic condition video feature extraction unit comprises: Extracting video key frames from the traffic condition video collected by the monitoring device to obtain a plurality of traffic condition key frames; The multiple traffic condition key frames are respectively passed through a traffic condition convolutional neural network including a deep and shallow fusion module to obtain the multiple traffic condition feature matrices.
4. The intelligent traffic condition information management platform according to claim 3 is characterized in that: The traffic condition feature encoding unit comprises: Aggregating the multiple traffic condition feature matrices along the channel dimension to obtain a three-dimensional traffic condition input tensor; The three-dimensional traffic condition input tensor is passed through a three-dimensional traffic condition convolution feature encoder to obtain the traffic condition change-related feature vector.
5. The intelligent traffic condition information management platform according to claim 4 is characterized in that: The vehicle text feature extraction unit comprises: A vehicle text information semantic encoding subunit, used for semantically encoding the vehicle text data collected by the road pre-buried sensor network to obtain a plurality of vehicle information feature vectors; The vehicle information associated convolutional coding subunit is used to pass the multiple vehicle information feature vectors through a vehicle information associated convolutional coding model to obtain the vehicle information associated feature vector.
6. The intelligent traffic condition information management platform according to claim 5 is characterized in that: The vehicle text information semantic encoding subunit includes: Performing word embedding encoding on the vehicle text data collected by the road pre-buried sensor network to obtain a plurality of vehicle information data items; The multiple vehicle information data items are passed through a vehicle information semantic encoding model to obtain the multiple vehicle information feature vectors.
7. A smart traffic information management method for the smart traffic information management platform according to claim 1, characterized in that: include: Obtain traffic condition videos collected by monitoring equipment and vehicle text data collected by the road pre-buried sensor network; Extracting traffic condition change-related feature vectors and vehicle information-related feature vectors from the traffic condition video collected by the monitoring device and the vehicle text data collected by the road pre-buried sensor network; Based on the traffic condition change-related feature vector and the vehicle information-related feature vector, a traffic optimization suggestion is generated.
