An abnormality monitoring method and device for a road traffic network and a storage medium

Abstract

The application discloses an abnormal monitoring method and device for a road traffic network and a storage medium. A graph structure representing the connectivity of roads in the road traffic network is constructed, and satellite image data used to determine whether an abnormality occurs in a traffic network node is obtained. The graph nodes and edges are matched with the traffic network nodes and roads in the satellite image data, and the attribute information of the first graph node and the associated edges in the graph structure is determined. Furthermore, based on the graph neural network, the node features of the second graph node can be determined, and the abnormality of the second graph node can be predicted based on the node features. By combining the satellite image data and the graph structure representing the topology of the road traffic network, whether an abnormality exists in the unknown traffic network node which is unknown whether an abnormality occurs is determined by the known traffic network node (and road) whether an abnormality occurs, thereby improving the timeliness and accuracy of the abnormal monitoring of the road traffic network.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical fields of satellite image, knowledge graph and road traffic analysis, in particular to a method and device for monitoring abnormality of road traffic network and a storage medium. BACKGROUND

[0002] With the continuous development of satellite technology, satellite image data collected by low-orbit satellites can be used to monitor the road surface conditions of road traffic network.

[0003] In the prior art, satellite image data can be identified by image recognition, so as to monitor the abnormality of road traffic network. For example, whether there is a traffic jam, whether the road surface is damaged, whether there is a bridge damage accident, etc.

[0004] The prior art has certain deficiencies in identifying satellite image data and monitoring whether there is abnormality in road traffic network. For example, in cloudy or foggy weather, part of the traffic network nodes may be blocked in the satellite's view, which leads to the fact that it is impossible to determine whether the traffic network nodes are abnormal under such conditions by using satellite image data. Therefore, the prior art cannot determine whether the traffic network nodes not covered by satellite image data are abnormal, thereby reducing the timeliness and accuracy of monitoring the abnormality of road traffic network.

[0005] At present, there is no effective solution to the technical problem of low timeliness and accuracy in monitoring the abnormality of road traffic network by using satellite image data in the prior art. SUMMARY

[0006] Embodiments of the present disclosure provide a method and device for monitoring abnormality of road traffic network and a storage medium, which at least solve the technical problem of low timeliness and accuracy in monitoring the abnormality of road traffic network by using satellite image data in the prior art.

[0007] According to an aspect of embodiments of the present disclosure, there is provided a method for monitoring an anomaly of a road traffic network, comprising: determining a graph structure corresponding to a road traffic network of a predetermined area, the graph structure comprising graph nodes and edges connecting between two graph nodes, wherein a graph node is used to represent a traffic network node of the road traffic network, and an edge connecting between two graph nodes is used to represent a road between the corresponding traffic network nodes; obtaining satellite image data corresponding to the predetermined area; determining attribute information corresponding to a first graph node in the graph structure and associated edges according to the satellite image data; constructing first graph data corresponding to the graph structure according to the attribute information; performing message passing on the first graph data by using a graph neural network to determine first graph features corresponding to the first graph data, and determining node features of a second graph node of the graph structure according to the first graph features, wherein the satellite image data does not contain a traffic network node corresponding to the second graph node; and determining whether the traffic network node corresponding to the second graph node has an anomaly according to the node features of the second graph node.

[0008] According to another aspect of embodiments of the present disclosure, there is also provided a storage medium comprising a stored program, wherein the program is executed by a processor when the program is run.

[0009] According to another aspect of embodiments of the present disclosure, there is also provided an apparatus for monitoring an anomaly of a road traffic network, comprising: a graph structure determination module configured to determine a graph structure corresponding to a road traffic network of a predetermined area, the graph structure comprising graph nodes and edges connecting between two graph nodes, wherein a graph node is used to represent a traffic network node of the road traffic network, and an edge connecting between two graph nodes is used to represent a road between the corresponding traffic network nodes; an image obtaining module configured to obtain satellite image data corresponding to the predetermined area; an attribute determination module configured to determine attribute information corresponding to a first graph node in the graph structure and associated edges according to the satellite image data; a graph data construction module configured to construct first graph data corresponding to the graph structure according to the attribute information; a feature determination module configured to perform message passing on the first graph data by using a graph neural network to determine first graph features corresponding to the first graph data, and determine node features of a second graph node of the graph structure according to the first graph features, wherein the satellite image data does not contain a traffic network node corresponding to the second graph node; and an anomaly determination module configured to determine whether the traffic network node corresponding to the second graph node has an anomaly according to the node features of the second graph node.

[0010] According to another aspect of the embodiments of the present disclosure, there is also provided an apparatus for monitoring anomalies of a road traffic network, comprising: a processor; and a memory connected with the processor, configured to provide the processor with instructions for processing the following processing steps: determining a graph structure corresponding to a road traffic network of a predetermined area, the graph structure comprising graph nodes and edges connecting two graph nodes, wherein the graph nodes are configured to represent traffic network nodes of the road traffic network, and the edges connecting two graph nodes are configured to represent roads between the corresponding traffic network nodes; obtaining satellite image data corresponding to the predetermined area; determining attribute information corresponding to a first graph node in the graph structure and associated edges according to the satellite image data; constructing first graph data corresponding to the graph structure according to the attribute information; performing message passing on the first graph data by using a graph neural network to determine first graph features corresponding to the first graph data, and determining node features of a second graph node of the graph structure according to the first graph features, wherein the satellite image data does not contain a traffic network node corresponding to the second graph node; and determining whether the traffic network node corresponding to the second graph node is abnormal according to the node features of the second graph node.

[0011] In the embodiments of the present disclosure, a graph structure representing the connectivity of roads in the road traffic network is constructed, and satellite image data used to determine whether the traffic network nodes are abnormal is obtained. Furthermore, by corresponding the graph nodes and edges in the graph structure to the traffic network nodes and roads that can be determined to be abnormal by the satellite image data, the attribute information of the first graph nodes and associated edges in the graph structure can be determined. Furthermore, based on the graph neural network, the node features of the second graph node, i.e., the graph node corresponding to the traffic network node whose abnormality is unknown, can be determined. Further, the abnormality of the second graph node can be predicted based on the node features. In the embodiments of the present disclosure, by combining the satellite image data and the graph structure representing the topology of the road traffic network, the abnormality of the traffic network node whose abnormality is unknown can be determined by the traffic network nodes whose abnormality is known, thereby improving the timeliness and accuracy of monitoring the anomalies of the road traffic network. BRIEF DESCRIPTION OF DRAWINGS

[0012] The accompanying drawings, which are included to provide a further understanding of the present disclosure and constitute a part of this application, illustrate certain illustrative embodiments of the present disclosure and are used to explain the present disclosure, but do not limit the present disclosure. In the drawings:

[0013] Figure 1 is a schematic diagram of a system for monitoring anomalies of a road traffic network based on satellite images according to Embodiment 1 of the present disclosure;

[0014] Figure 2A is a schematic diagram of a hardware architecture of a satellite system 10 according to Embodiment 1 of the present disclosure;

[0015] Figure 2B is a schematic diagram of a hardware architecture of the ground system 20 according to the first embodiment of the present disclosure;

[0016] Figure 3 is a flowchart of a method for monitoring anomalies of a road traffic network according to the first aspect of the first embodiment of the present disclosure;

[0017] Figure 4 is a schematic diagram of a graph structure provided by the first embodiment of the present disclosure;

[0018] Figure 5 is a schematic diagram of a display form of first graph data provided by the first embodiment of the present disclosure;

[0019] Figure 6 is a schematic diagram of a display form of second graph data provided by the first embodiment of the present disclosure;

[0020] Figure 7A is a flowchart of a first message passing provided by the first embodiment of the present disclosure;

[0021] Figure 7B is a flowchart of a second message passing provided by the first embodiment of the present disclosure;

[0022] Figure 8 is a schematic diagram of an apparatus for monitoring anomalies of a road traffic network according to the first aspect of the second embodiment of the present disclosure; and

[0023] Figure 9 is a schematic diagram of an apparatus for monitoring anomalies of a road traffic network according to the first aspect of the third embodiment of the present disclosure. DETAILED DESCRIPTION

[0024] In order to make the technical personnel of the art better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, not all the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those of ordinary skill in the art without creative labor should belong to the scope of protection of the present disclosure.

[0025] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0026] Example 1

[0027] According to this embodiment, an embodiment of a method for monitoring anomalies in a road traffic network is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0028] Figure 1 A schematic diagram of a road traffic network anomaly monitoring system based on satellite imagery according to this embodiment is shown. The system includes a ground system 20 and a satellite system 10. The satellite system 10 continuously collects satellite imagery data from the ground and sends it to the ground system 20. The ground system 20 receives the satellite imagery data sent by the satellite system 10 to monitor anomalies in the road traffic network using the satellite imagery data. How to monitor anomalies in the road traffic network using satellite imagery data will be described in detail below.

[0029] Figure 2A Further shown Figure 1 A schematic diagram of the hardware architecture of the China Satellite System 10. (Reference) Figure 2A As shown, satellite system 10 includes an integrated electronic system, which includes a processor, a memory, a bus management module, and a communication interface. The memory is connected to the processor, allowing the processor to access the memory, read program instructions stored in the memory, read data from the memory, or write data to the memory. The bus management module is connected to the processor and also to a bus such as a CAN bus. Thus, the processor can communicate with onboard peripherals connected to the bus through the bus managed by the bus management module. Furthermore, the processor also communicates with devices such as cameras, star sensors, telemetry and command transponders, and data transmission equipment via the communication interface. Those skilled in the art will understand that... Figure 2AThe structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, a satellite system may also include... Figure 2A The more or fewer components shown, or having the same Figure 2A The different configurations shown.

[0030] Figure 2B Further shown Figure 1 A schematic diagram of the hardware architecture of the ground system 20. (Reference) Figure 2B As shown, the ground system 20 may include one or more processors (processors may include, but are not limited to, microprocessors such as MCUs or programmable logic devices such as FPGAs), a memory for storing data, a transmission device for communication functions, and an input / output interface. The memory, transmission device, and input / output interface are connected to the processor via a bus. In addition, it may also include a display, keyboard, and cursor control device connected to the input / output interface. Those skilled in the art will understand that... Figure 2B The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, the ground system may also include... Figure 2B The more or fewer components shown, or having the same Figure 2B The different configurations shown.

[0031] It should be noted that, Figure 2A and Figure 2B One or more processors and / or other data processing circuits shown herein may generally be referred to as "data processing circuitry". This data processing circuitry may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuitry may be a single, independent processing module, or may be integrated, in whole or in part, into any other element in a computing device. As involved in embodiments of this disclosure, the data processing circuitry serves as processor control (e.g., selection of a variable resistor termination path connected to an interface).

[0032] Figure 2A and Figure 2B The memory shown can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the abnormal monitoring method of road traffic networks in this embodiment of the present disclosure. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby realizing the abnormal monitoring method of the road traffic network described above. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory.

[0033] It should be noted here that, in some optional embodiments, the above... Figure 2A andFigure 2B The apparatus shown can include hardware elements (including circuitry), software elements (including computer code stored on a computer readable medium), or a combination of both hardware and software elements. It should be noted that Figure 2A and Figure 2B is merely one instance of a particular, concrete example, and is intended to show the types of components that can be present in the apparatus described above.

[0034] Under the above operating environment, according to a first aspect of the embodiment, an abnormality monitoring method for a road traffic network is provided, which is implemented by the ground system 20 shown in FIG. 2. Figure 3 A flowchart of the method is shown, referring to Figure 3 As shown, the method comprises:

[0035] S302: determining a graph structure corresponding to a road traffic network of a predetermined area, the graph structure containing graph nodes and edges connecting between two graph nodes, wherein the graph nodes are used to represent traffic network nodes of the road traffic network, and the edges connecting between two graph nodes are used to represent roads between the corresponding traffic network nodes;

[0036] S304: obtaining satellite image data corresponding to the predetermined area;

[0037] S306: determining attribute information corresponding to a first graph node in the graph structure and associated edges according to the satellite image data;

[0038] S308: constructing first graph data corresponding to the graph structure according to the attribute information;

[0039] S310: performing message passing on the first graph data using a graph neural network, determining first graph features corresponding to the first graph data, and determining node features of a second graph node of the graph structure according to the first graph features, wherein the satellite image data does not contain a traffic network node corresponding to the second graph node; and

[0040] S312: determining whether the traffic network node corresponding to the second graph node exists abnormally according to the node features of the second graph node.

[0041] Specifically, referring to Figure 3 As shown, the ground system 20 determines a graph structure corresponding to a road traffic network of a predetermined area, the graph structure containing graph nodes and edges connecting between two graph nodes, wherein the graph nodes are used to represent traffic network nodes of the road traffic network, and the edges connecting between two graph nodes are used to represent roads between the corresponding traffic network nodes (S302).

[0042] Figure 4 is a schematic diagram of a graph structure provided by Embodiment 1 of the present disclosure.

[0043] Figure 4 The graph nodes in the graph structure shown in FIG. 1 are used to represent the traffic network nodes, Figure 4 The edges between two graph nodes in the graph structure shown in FIG. 1 are used to represent the roads between the two corresponding traffic network nodes. The graph structure can be constructed in advance. When constructing the graph structure, the correspondence between the graph nodes and the corresponding traffic network nodes, and the correspondence between the edges and the corresponding roads can be recorded, that is, the correspondence is used to indicate which traffic network node in the actual road traffic network the graph node corresponds to, and which road in the actual road traffic network the edge corresponds to. The correspondence related to the graph node can include the correspondence between the graph node and the name and geographic coordinates of the corresponding traffic network node, and the like, and similarly, the correspondence related to the edge can include the correspondence between the edge and the name and geographic coordinates of the corresponding road, and the like.

[0044] That is, in the abnormal monitoring method of the road traffic network provided in the present specification, the graph structure representing the connection relationship of the traffic network nodes and roads of the road traffic network can be combined to monitor the abnormality of the road traffic network. The traffic network nodes can be important nodes in the road traffic network, such as crossroads, T-shaped intersections, road convergence points, and bridges. The abnormality mentioned here can include the situation that the traffic is not smooth due to natural disasters, traffic accidents, and traffic congestion, etc.

[0045] Then, the ground system 20 can obtain satellite image data corresponding to a predetermined area (S304). The predetermined area mentioned here can refer to an area to be monitored by the satellite at the current time.

[0046] After obtaining the satellite image data, the ground system 20 can determine attribute information corresponding to the first graph node and the associated edges in the graph structure according to the satellite image data (S306). Specifically, the ground system 20 can first match the traffic network nodes contained in the satellite image data with the graph nodes in the graph structure, thereby determining the first graph node in the graph structure, and then determining the attribute information corresponding to the first graph node and the associated edges in the graph structure.

[0047] The first graph node mentioned here represents a traffic network node that can be determined whether an abnormality occurs through the satellite image data. The attribute information mentioned here can represent the vehicle situation in the traffic network nodes and roads corresponding to the first graph node and the associated edges. How to determine the attribute information will be described in detail below.

[0048] Then, the ground system 20 can construct first graph data corresponding to the graph structure according to the attribute information (S308). That is, on the basis of the graph structure, the first graph data also carries attribute information corresponding to the first graph nodes and the associated edges.

[0049] Figure 5 is a schematic diagram of a display form of the first graph data provided by the embodiment 1 of the present disclosure.

[0050] As can be seen, Figure 5 The structure of the first graph data shown in Figure 4 is consistent with the structure of the graph structure shown in Figure 4 , and it can be understood that the first graph data is obtained by filling the attribute information obtained from the satellite image data into the graph structure in Figure 5 The white nodes shown in

[0051] Further, the ground system 20 can use the graph neural network to perform message passing on the first graph data, determine the first graph feature corresponding to the first graph data, and determine the node feature of the second graph node of the graph structure according to the first graph feature, wherein the satellite image data does not contain the traffic network node corresponding to the second graph node (S310). Or it can be understood that the second graph node refers to a graph node that cannot determine whether the corresponding traffic network node is abnormal according to the satellite image data. The above-mentioned first graph feature can contain the node feature corresponding to each graph node, and can also contain the edge feature corresponding to each edge.

[0052] Finally, the node feature of the second graph node can be used to determine whether the traffic network node corresponding to the second graph node is abnormal (S312).

[0053] As described in the background, the existing technology for identifying satellite image data and then monitoring the road surface conditions of the road traffic network has certain deficiencies. For example, in cloudy or foggy weather, part of the traffic network nodes may be blocked in the satellite's view, which leads to the fact that it is impossible to determine whether the traffic network nodes in such a situation are abnormal through satellite image data. And, for example, by using multiple adjacent satellites to monitor the road surface conditions of the road traffic network at the same time, there may be areas where the images collected by the satellites are missing, and it is also impossible to determine whether the traffic network nodes in such areas are abnormal through satellite image data.

[0054] Therefore, in the method, a graph structure representing the road connection relationship in the road traffic network is constructed, and satellite image data used to determine whether an abnormality occurs in a traffic network node (and a road) is obtained. By corresponding the graph nodes and edges in the graph structure to the traffic network nodes and roads that can be determined to have an abnormality by the satellite image data, the attribute information of the first graph nodes and the associated edges existing in the graph structure can be determined. Then, based on the graph neural network, the node features of the second graph nodes corresponding to the graph nodes of the traffic network nodes whose abnormality is unknown can be determined. Further, based on the node features, the abnormality of the second graph nodes can be predicted.

[0055] It can be seen that, by combining the satellite image data and the graph structure representing the topology of the road traffic network, the method determines whether an abnormality exists in the traffic network nodes whose abnormality is unknown by using the traffic network nodes (and roads) whose abnormality is known, thereby improving the timeliness and accuracy of the abnormality monitoring of the road traffic network.

[0056] Optionally, the operation of determining the first graph feature corresponding to the first graph data by message passing of the graph neural network comprises: performing message passing of the first graph data by using a plurality of pre-set graph neural networks respectively to obtain a plurality of second graph features corresponding to the respective graph neural networks; determining weight values corresponding to the plurality of graph neural networks according to the second graph features; and performing weighted summation on the plurality of second graph features according to the weight values to determine the first graph feature.

[0057] The plurality of pre-set graph neural networks mentioned above are different graph neural networks. For example, the plurality of graph neural networks are of the same type but have different network parameters; for another example, the plurality of graph neural networks are of different types. The graph neural network can be a neural network for a knowledge graph, for example, a graph convolution network, and the specific type of graph neural network is not limited. How to determine the weight values corresponding to the plurality of graph neural networks according to the second graph features will be described below.

[0058] After the plurality of second graph features are determined, the weight values corresponding to the plurality of graph neural networks can be determined. Each graph neural network corresponds to a second graph feature and a weight value. By performing weighted summation on the second graph features corresponding to the plurality of graph neural networks through the weight values of the plurality of image neural networks, the first graph feature can be obtained.

[0059] Optionally, according to the second feature, the operation of determining the weight values corresponding to the plurality of graph neural networks respectively includes: extracting at least one third graph node from the first graph node, and masking the attribute information corresponding to the third graph node in the first graph data to construct second graph data; performing message passing on the second graph data by using the plurality of graph neural networks respectively to obtain a plurality of third first graph features corresponding to each graph neural network respectively; determining a plurality of second node features corresponding to the plurality of second features and the third graph node respectively, and determining a plurality of third node features corresponding to the plurality of third graph features and the third graph node respectively; and determining the weight values corresponding to the plurality of graph neural networks respectively according to the similarity between the second node features and the third node features corresponding to each graph neural network.

[0060] First, the second graph data is graphically illustrated. Figure 6 is a schematic diagram of the display form of the second graph data provided by the embodiment 1 of the present disclosure.

[0061] Referring to Figure 6 , Figure 6 and Figure 5 The difference between and is that Figure 5 part of the white nodes in is changed to gray nodes, that is, Figure 6 contains three kinds of nodes, white nodes, black nodes and gray nodes, and the gray nodes are the third graph nodes mentioned in the above content. From Figure 5 and Figure 6 It can be seen that the difference between the first graph data and the second graph data is that the second graph data is obtained by covering the attribute information originally possessed by some first graph nodes (i.e., third graph nodes) in the first graph data.

[0062] Among them, the third graph node can be randomly extracted from the first graph node.

[0063] Figure 7A is a flowchart of the first message passing provided by the embodiment 1 of the present disclosure.

[0064] Figure 7B is a flowchart of the second message passing provided by the embodiment 1 of the present disclosure.

[0065] Referring to Figure 7A and 7B , when determining the weight values corresponding to the plurality of graph neural networks, the ground system 20 can perform message passing on the graph data twice by using the plurality of graph neural networks. The first message passing is message passing on the first graph data to obtain a plurality of graph features (second graph features) corresponding to the first graph data. The second message passing is message passing on the second graph data to obtain a plurality of graph features (third graph features) corresponding to the second graph data.

[0066] For a third graph node, the node feature of the first graph node corresponding to the third graph node is included in the second graph feature as a second node feature corresponding to the third graph node. The node feature of the third graph node is included in the third graph feature as a third node feature corresponding to the third graph node. Thus, the second node features corresponding to each graph neural network include the second node features corresponding to each third graph node, and the third node features corresponding to each graph neural network include the third node features corresponding to each third graph node.

[0067] Then, the similarity between the second node feature and the third node feature of the same third graph node output by the same graph neural network (i.e., the similarity between the second node feature and the third node feature corresponding to each graph neural network in the above description) can be determined. Then, according to the aforementioned similarity, the weight value corresponding to each graph neural network can be determined. The similarity can be determined by, for example, cosine distance.

[0068] That is, for a graph neural network, the number of third graph nodes is equal to the number of similarities that can be calculated. For example, if there is only one third graph node, there is one similarity (the similarity between the second node feature and the third node feature of the third graph node output by the graph neural network) corresponding to each graph neural network. The similarities can be normalized as the weight values corresponding to each graph neural network.

[0069] For example, if there are multiple third graph nodes, multiple similarities corresponding to a graph neural network can be determined. For example, the multiple similarities can be averaged (or maximized) to obtain a similarity average value. Then, according to the similarity average values corresponding to each graph neural network, the weight values corresponding to each graph neural network.

[0070] After determining the weight values of each graph neural network, the ground system 20 can weight the corresponding second graph features according to the weight values of each graph neural network, and then sum the weighted second graph features to obtain the first graph feature. Thus, by this method, the graph neural network with better feature modeling effect through message passing can play a greater role in determining the first graph feature, thereby improving the accuracy of the final abnormality judgment through the node features of the second graph nodes included in the first graph feature.

[0071] Since the abnormality monitoring in the present specification mainly needs to be performed on the traffic network nodes of unknown abnormal situations, the ground system 20 can determine the node features corresponding to the second graph nodes from the first graph features, so as to determine whether the traffic network nodes corresponding to the corresponding second graph nodes are abnormal according to the node features corresponding to the second graph nodes. Specifically, the ground system 20 can input the node features corresponding to the second graph nodes into the pre-trained classifier for abnormality judgment, so as to determine whether the traffic network nodes corresponding to the corresponding second graph nodes are abnormal through the classifier.

[0072] It should be noted that the classifier in the present method can be pre-trained through supervised training, and the graph neural network in the present method can be pre-trained together with the classifier. When training the graph neural network and the classifier, the loss of the classifier can be shared with the graph neural network, so that the two are jointly trained.

[0073] Of course, the first graph features determined through the above-mentioned manner can also contain edge features (i.e., when the message passing is performed through the graph neural network, not only the node features are updated, but also the edge features are updated), and the edge features of the corresponding edges can be determined according to the first graph features, and whether the roads of the corresponding unknown abnormal situations are abnormal can be determined according to the edge features.

[0074] Optionally, before determining the attribute information corresponding to the first graph nodes and the associated edges in the graph structure according to the satellite image data, the method further comprises: performing abnormality identification on the satellite image data for the traffic network nodes and the roads between the traffic network nodes to obtain an abnormality identification result; determining the first traffic network nodes according to the abnormality identification result, the first traffic network nodes being the traffic network nodes for which whether there is an abnormality can be identified through the satellite image data; and determining the first graph nodes corresponding to the first traffic network nodes in the graph structure.

[0075] That is, before determining the attribute information corresponding to the first graph nodes and the associated edges, it is necessary to determine which are the first graph nodes. The ground system 20 can first perform abnormality identification on the traffic network nodes existing in the satellite image data through image recognition, and take the graph nodes corresponding to the traffic network nodes for which the abnormality can be determined through the satellite image data as the first graph nodes.

[0076] Optionally, the operation of determining the attribute information corresponding to the first graph node and the associated edge in the graph structure according to the satellite image data comprises: determining whether a traffic network node corresponding to the first graph node or a road corresponding to the associated edge is damaged according to the satellite image data; in the case that the corresponding traffic network node or the corresponding road is damaged, determining the attribute information of the first graph node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road according to a preset minimum value of a vehicle minimum distance and / or a vehicle minimum speed; in the case that the corresponding traffic network node or the corresponding road is not damaged and there is a vehicle on the corresponding traffic network node or the corresponding road, determining an actual vehicle minimum distance and / or an actual vehicle minimum speed in the corresponding traffic network node and the corresponding road according to the satellite image data, and determining the attribute information of the first graph node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road according to the actual vehicle minimum distance and / or the actual vehicle minimum speed; and in the case that the corresponding traffic network node or the corresponding road is not damaged and there is no vehicle on the corresponding traffic network node or the corresponding road, determining the attribute information of the first graph node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road according to a preset maximum value of the vehicle minimum distance and / or a vehicle maximum speed.

[0077] In the determination of the attribute information corresponding to the first graph node and the corresponding edge, the attribute information can be determined according to different cases. As described above, the cases can be divided into three kinds. The first case is that the traffic network node or the road is damaged; the second case is that the traffic network node or the road is not damaged and there is a vehicle; and the third case is that the traffic network node or the road is not damaged and there is no vehicle.

[0078] In the first case, the traffic network node or the road is damaged, so the attribute information can be a preset minimum value of a vehicle minimum distance and / or a vehicle minimum speed. The vehicle minimum distance and / or the vehicle minimum speed can be determined as a minimum value of a statistical vehicle minimum distance and / or a statistical vehicle minimum speed in all historical times. Of course, the minimum value can also be preset artificially.

[0079] In the second case, the traffic network node or the road is not damaged and there is a vehicle, so the actual vehicle minimum distance and / or the actual vehicle minimum speed can be directly used as the attribute information.

[0080] Optionally, the operation of determining the actual vehicle minimum speed in the corresponding traffic network node and the corresponding road according to the satellite image data comprises: determining the actual vehicle minimum speed in the corresponding traffic network node and the corresponding road according to the satellite image data and satellite image data collected within a preset time interval between the satellite image data and the historical satellite image data.

[0081] That is, the actual minimum speed of a vehicle can be determined by using satellite imagery data acquired at similar times in the past.

[0082] In the third scenario, where the traffic network nodes or roads are undamaged and there are no vehicles (which can be considered an extreme case), the maximum value of the pre-set minimum vehicle spacing and / or minimum vehicle speed can be determined as attribute information. This maximum value can be determined using a method similar to the minimum value mentioned above. For example, it can be determined as the maximum value of the minimum vehicle spacing and / or minimum vehicle speed statistically analyzed over all historical periods. Alternatively, the maximum values ​​of the minimum vehicle spacing and / or minimum vehicle speed can be manually preset.

[0083] It should be noted that for the second graph nodes and the edges corresponding to roads where it is unknown whether anomalies have occurred, the attribute information of the corresponding graph nodes and edges (i.e., the attribute information that needs to be determined through satellite imagery data) can be set to default attribute information. For example, the specific content of the attribute information can be filled with 0; another example is to set the specific content of the attribute information to a manually defined element (such as a letter).

[0084] By using the attribute information of each graph node and edge, the initial features of the corresponding graph node or edge can be obtained. After message passing through the graph neural network, the initial features of each graph node and edge are updated (this can be done in one round or multiple rounds of message passing), thus obtaining graph features containing the node features of each graph node and the edge features of each edge (regardless of whether it is the first graph feature, the second graph feature, or the third graph feature, their format is consistent, so they are all referred to as graph features here).

[0085] In addition, refer to Figure 1 As shown, according to a third aspect of this embodiment, a storage medium is provided. The storage medium includes a stored program, wherein, when the program is executed, a processor performs any of the methods described above.

[0086] Therefore, according to this embodiment, by combining graph structure and satellite imagery data, it is possible to predict whether anomalies have occurred in traffic network nodes whose anomalies are unknown, by referring to traffic network nodes whose anomalies are known to exist. Furthermore, based on this, the weights corresponding to each graph neural network are determined, and the second graph features derived from each graph neural network are weighted to obtain the first graph features. This allows graph neural networks, which are better at feature modeling through message passing, to play a greater role in determining the first graph features, thereby improving the accuracy of anomaly detection based on the node features of the second graph nodes contained in the first graph features.

[0087] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0088] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0089] Example 2

[0090] Figure 8 An anomaly monitoring device 800 for a road traffic network according to a first aspect of this embodiment is shown, which corresponds to the method described according to the first aspect of Embodiment 1. Reference Figure 8 As shown, the device 800 includes: a graph structure determination module 810, used to determine a graph structure corresponding to a road traffic network in a predetermined area, the graph structure including graph nodes and edges connecting two graph nodes, wherein the graph nodes represent traffic network nodes of the road traffic network, and the edges connecting two graph nodes represent roads between the corresponding traffic network nodes; an image acquisition module 820, used to acquire satellite image data corresponding to the predetermined area; an attribute determination module 830, used to determine attribute information corresponding to a first graph node and associated edges in the graph structure based on the satellite image data; a graph data construction module 840, used to construct first graph data corresponding to the graph structure based on the attribute information; a feature determination module 850, used to use a graph neural network to perform message passing on the first graph data, determine the first graph features corresponding to the first graph data, and determine the node features of a second graph node in the graph structure based on the first graph features, wherein the satellite image data does not contain traffic network nodes corresponding to the second graph node; and an anomaly determination module 860, used to determine whether there is an anomaly in the traffic network node corresponding to the second graph node based on the node features of the second graph node.

[0091] Optionally, the feature determination module 850 is configured to perform message passing on the first graph data using a plurality of preset graph neural networks respectively to obtain a plurality of second graph features respectively corresponding to the graph neural networks; determine weight values respectively corresponding to the graph neural networks according to the second graph features; and perform weighted summation on the plurality of second graph features according to the weight values to determine the first graph feature.

[0092] Optionally, the feature determination module 850 is configured to extract at least one third graph node from the first graph node, mask attribute information in the first graph data corresponding to the third graph node, construct second graph data, perform message passing on the second graph data using the plurality of graph neural networks respectively to obtain a plurality of third first graph features respectively corresponding to the graph neural networks; determine a plurality of second node features respectively corresponding to the plurality of second graph features and the third graph node, and determine a plurality of third node features respectively corresponding to the plurality of third graph features and the third graph node; and determine weight values respectively corresponding to the plurality of graph neural networks according to similarities between the second node features and the third node features corresponding to each graph neural network.

[0093] Optionally, before determining attribute information corresponding to the first graph node and the associated edge in the graph structure according to the satellite image data, the attribute determination module 830 is further configured to perform anomaly identification on the satellite image data for traffic network nodes and roads between traffic network nodes to obtain an anomaly identification result; determine the first traffic network node according to the anomaly identification result, the first traffic network node being a traffic network node for which it can be identified from the satellite image data whether an anomaly exists; and determine the first graph node corresponding to the first traffic network node in the graph structure.

[0094] Optionally, the attribute determining module 830 is configured to determine, according to the satellite image data, whether a traffic network node corresponding to the first graph node or a road corresponding to the associated edge has been damaged; in a case where the corresponding traffic network node or the corresponding road has been damaged, determine attribute information of the first graph node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road according to a preset minimum value of the minimum vehicle distance and / or the minimum vehicle speed; in a case where the corresponding traffic network node or the corresponding road has not been damaged and there is a vehicle on the corresponding traffic network node or the corresponding road, determine an actual minimum vehicle distance and / or an actual minimum vehicle speed in the corresponding traffic network node and the corresponding road according to the satellite image data, and determine the attribute information of the first graph node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road according to the actual minimum vehicle distance and / or the actual minimum vehicle speed; and in a case where the corresponding traffic network node or the corresponding road has not been damaged and there is no vehicle on the corresponding traffic network node or the corresponding road, determine the attribute information of the first graph node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road according to a preset maximum value of the minimum vehicle distance and / or the maximum vehicle speed.

[0095] Optionally, the attribute determining module 830 is configured to determine the actual minimum vehicle speed in the corresponding traffic network node and the corresponding road according to the satellite image data and satellite image data collected within a preset time interval from the satellite image data.

[0096] According to the present embodiment, the graph structure and the satellite image data are combined, so that the traffic network nodes with known abnormal situations can be referred to to predict whether the traffic network nodes with unknown abnormal situations have been abnormal. Furthermore, the weights corresponding to the graph neural networks are determined to weight the second graph features obtained by the graph neural networks to obtain the first graph features. Thus, the graph neural networks with better feature modeling effect through message passing can play a greater role in determining the first graph features, thereby improving the accuracy of the abnormality judgment through the node features of the second graph nodes included in the first graph features.

[0097] Embodiment 3

[0098] Figure 9 An abnormality monitoring device 900 for a road traffic network according to a first aspect of the present embodiment is shown, which corresponds to the method according to the first aspect of the present embodiment. Reference is made to the description of the method according to the first aspect of the present embodiment. Figure 9As shown, the apparatus 900 comprises: a processor 910; and a memory 920 connected with the processor 910, configured to provide the processor 910 with instructions to process the following processing steps: determining a graph structure corresponding to a road traffic network of a predetermined area, the graph structure containing graph nodes and edges connecting between two graph nodes, wherein the graph nodes are used to represent traffic network nodes of the road traffic network, and the edges connecting between two graph nodes are used to represent roads between the corresponding traffic network nodes; obtaining satellite image data corresponding to the predetermined area; determining attribute information corresponding to a first graph node and associated edges in the graph structure according to the satellite image data; constructing first graph data corresponding to the graph structure according to the attribute information; performing message passing on the first graph data using a graph neural network to determine first graph features corresponding to the first graph data, and determining node features of a second graph node of the graph structure according to the first graph features, wherein the satellite image data does not contain a traffic network node corresponding to the second graph node; and determining whether the traffic network node corresponding to the second graph node exists abnormally according to the node features of the second graph node.

[0099] Optionally, the operation of performing message passing on the first graph data using a graph neural network to determine first graph features corresponding to the first graph data comprises: performing message passing on the first graph data using a plurality of pre-set graph neural networks respectively to obtain a plurality of second graph features respectively corresponding to each graph neural network; determining weight values respectively corresponding to the plurality of graph neural networks according to the second graph features; and performing weighted summation on the plurality of second graph features according to the weight values to determine the first graph features.

[0100] Optionally, the operation of determining weight values respectively corresponding to the plurality of graph neural networks according to the second graph features comprises: extracting at least one third graph node from the first graph node, and masking attribute information corresponding to the third graph node in the first graph data to construct second graph data, performing message passing on the second graph data using the plurality of graph neural networks respectively to obtain a plurality of third first graph features respectively corresponding to each graph neural network; determining a plurality of second node features respectively corresponding to the plurality of second graph features and the third graph node, and determining a plurality of third node features respectively corresponding to the plurality of third graph features and the third graph node; and determining the weight values respectively corresponding to the plurality of graph neural networks according to similarities between the second node features and the third node features corresponding to each graph neural network.

[0101] Optionally, before the attribute information corresponding to the first graph node and the associated edge in the graph structure is determined according to the satellite image data, the memory 920 is further configured to provide the processor 910 with instructions to process the following processing steps: performing anomaly identification on the satellite image data for the traffic network nodes and the roads between the traffic network nodes to obtain an anomaly identification result; determining the first traffic network node according to the anomaly identification result, the first traffic network node being a traffic network node for which whether an anomaly exists can be identified according to the satellite image data; and determining the first graph node corresponding to the first traffic network node in the graph structure.

[0102] Optionally, the operation of determining the attribute information corresponding to the first graph node and the associated edge in the graph structure according to the satellite image data includes: determining, according to the satellite image data, whether damage occurs to the traffic network node corresponding to the first graph node or the road corresponding to the associated edge; in the case where damage occurs to the corresponding traffic network node or the corresponding road, determining the attribute information of the first graph node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road according to a preset minimum value of a minimum distance between vehicles and / or a minimum speed of vehicles; in the case where no damage occurs to the corresponding traffic network node or the corresponding road and there are vehicles on the corresponding traffic network node or the corresponding road, determining, according to the satellite image data, an actual minimum distance between vehicles and / or an actual minimum speed of vehicles in the corresponding traffic network node and the corresponding road, and determining the attribute information of the first graph node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road according to the actual minimum distance between vehicles and / or the actual minimum speed of vehicles; and in the case where no damage occurs to the corresponding traffic network node or the corresponding road and there are no vehicles on the corresponding traffic network node or the corresponding road, determining the attribute information of the first graph node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road according to a preset maximum value of the minimum distance between vehicles and / or a maximum speed of vehicles.

[0103] Optionally, the operation of determining the actual minimum speed of vehicles in the corresponding traffic network node and the corresponding road according to the satellite image data includes: determining the actual minimum speed of vehicles in the corresponding traffic network node and the corresponding road according to the satellite image data and satellite image data collected within a preset time interval between the satellite image data and historical satellite image data.

[0104] Thus, according to the present embodiment, by combining the graph structure and the satellite image data, it can be determined whether an abnormal situation occurs at a traffic network node whose abnormal situation is unknown by referring to a traffic network node whose abnormal situation is known. Moreover, on this basis, the weight corresponding to each graph neural network is also determined to weight the second graph features obtained by each graph neural network to obtain the first graph features. Thus, the graph neural network with better feature modeling effect through message passing can play a greater role in determining the first graph features, thereby improving the accuracy of the abnormality judgment through the node features of the second graph nodes contained in the first graph features.

[0105] The above-mentioned embodiment numbers of the present application are only for description, not representing the advantages or disadvantages of the embodiments.

[0106] In the above-mentioned embodiments of the present application, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0107] In several embodiments provided in the present application, it should be understood that the disclosed technical contents can be implemented by other ways. Among them, the above-mentioned device embodiments are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division way, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed units can be indirect coupling or communication connection through some interfaces, units or modules, which can be electrical or other forms.

[0108] The units described as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units, that is, they can be located in one place, or they can be distributed on multiple network units. According to actual needs, part or all of the units can be selected to achieve the purpose of the present embodiment.

[0109] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be realized in the form of hardware or in the form of software functional unit.

[0110] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application, essentially or in other words, the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to make a computer device (which can be a personal computer, a server or a network device, etc.) execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: a U disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0111] The above is only the preferred embodiment of the present application, it should be pointed out that, for those skilled in the art, without departing from the principles of the present application, can make a number of improvements and refinements, these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. A method for monitoring anomalies in a road traffic network, characterized in that, include: Determine the graph structure corresponding to the road traffic network of the predetermined area. The graph structure includes graph nodes and edges connecting two graph nodes. The graph nodes are used to represent traffic network nodes of the road traffic network, and the edges connecting two graph nodes are used to represent roads between the corresponding traffic network nodes. Acquire satellite imagery data corresponding to the predetermined area; Based on the satellite image data, determine the attribute information corresponding to the first graph node and its associated edges in the graph structure; Construct first graph data corresponding to the graph structure based on the attribute information; A graph neural network is used to perform message passing on the first graph data, determine the first graph features corresponding to the first graph data, and determine the node features of the second graph nodes of the graph structure based on the first graph features, wherein the satellite image data does not contain the traffic network nodes corresponding to the second graph nodes; and Determining whether there are anomalies in the traffic network nodes corresponding to the second graph nodes based on the node characteristics of the second graph nodes, and wherein the operation of determining the attribute information corresponding to the first graph nodes and associated edges in the graph structure based on the satellite image data includes: Based on the satellite imagery data, determine whether the traffic network node corresponding to the first map node or the road corresponding to the associated edge is damaged; if the corresponding traffic network node or road is damaged, determine the attribute information of the first map node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road based on the minimum value of the preset minimum vehicle spacing and / or minimum vehicle speed; if the corresponding traffic network node or road is not damaged and there are vehicles in the corresponding traffic network node or road, determine the actual minimum vehicle spacing and / or actual minimum vehicle speed in the corresponding traffic network node and road based on the satellite imagery data, and determine the attribute information of the first map node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road based on the actual minimum vehicle spacing and / or actual minimum vehicle speed; and if the corresponding traffic network node or road is not damaged and there are no vehicles in the corresponding traffic network node or road, determine the attribute information of the first map node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road based on the maximum value of the preset minimum vehicle spacing and / or minimum vehicle speed.

2. The method according to claim 1, characterized in that, The operation of using a graph neural network to perform message passing on the first graph data and determine the first graph features corresponding to the first graph data includes: Multiple pre-set graph neural networks are used to perform message passing on the first graph data to obtain multiple second graph features corresponding to each graph neural network. Based on the features of the second graph, determine the weight values ​​corresponding to the multiple graph neural networks respectively; and The first graph feature is determined by weighting and summing the multiple second graph features according to the weight values.

3. The method according to claim 2, characterized in that, The operation of determining the weight values ​​corresponding to multiple graph neural networks based on the features of the second graph includes: Extract at least one third graph node from the first graph node, and mask the attribute information corresponding to the third graph node in the first graph data to construct the second graph data; The second graph data is processed by the multiple graph neural networks to obtain multiple third graph features corresponding to each graph neural network. Each of the plurality of second graph features is determined to correspond to a plurality of second node features corresponding to a node in the third graph, and each of the plurality of third graph features is determined to correspond to a plurality of third node features corresponding to a node in the third graph; and Based on the similarity between the second node features and the third node features corresponding to each graph neural network, the weight values ​​corresponding to the plurality of graph neural networks are determined respectively.

4. The method according to claim 1, characterized in that, Before determining the attribute information corresponding to the first graph node and associated edges in the graph structure based on the satellite imagery data, the method further includes: Anomaly identification is performed on satellite imagery data targeting traffic network nodes and roads between traffic network nodes, yielding anomaly identification results; Based on the anomaly identification results, a first traffic network node is determined. The first traffic network node refers to a traffic network node whose anomalies can be identified through the satellite image data. Determine the first graph node in the graph structure that corresponds to the first traffic network node.

5. The method according to claim 1, characterized in that, The operation of determining the minimum actual vehicle speed at corresponding traffic network nodes and on corresponding roads based on the satellite imagery data includes: Based on the satellite imagery data and historical satellite imagery data whose acquisition time intervals with the satellite imagery data do not exceed a preset duration, the minimum actual vehicle speeds at corresponding traffic network nodes and on corresponding roads are determined.

6. A storage medium, characterized in that, The storage medium includes a stored program, wherein, when the program is executed, the method described in any one of claims 1 to 5 is performed by a processor.

7. A device for monitoring anomalies in a road traffic network, characterized in that, include: The graph structure determination module is used to determine the graph structure corresponding to the road traffic network of a predetermined area. The graph structure includes graph nodes and edges connecting two graph nodes. The graph nodes are used to represent traffic network nodes of the road traffic network, and the edges connecting two graph nodes are used to represent roads between the corresponding traffic network nodes. The image acquisition module is used to acquire satellite image data corresponding to the predetermined area; An attribute determination module is used to determine attribute information corresponding to a first graph node and its associated edges in the graph structure based on the satellite image data. Specifically, the attribute determination module is used to: determine whether a traffic network node corresponding to a first graph node or a road corresponding to an associated edge is damaged, based on the satellite image data; if the corresponding traffic network node or road is damaged, determine the attribute information of the first graph node or the edge corresponding to the corresponding road based on a pre-set minimum vehicle spacing and / or minimum vehicle speed; if the corresponding traffic network node or road is not damaged, and the corresponding traffic network node or the associated edge is damaged... When vehicles are present on the road, the minimum distance between actual vehicles and / or the minimum speed of actual vehicles in the corresponding traffic network node and the corresponding road are determined based on satellite imagery data. Based on the minimum distance between actual vehicles and / or the minimum speed of actual vehicles, the attribute information of the first graph node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road is determined. When the corresponding traffic network node or the corresponding road is not damaged and there are no vehicles in the corresponding traffic network node or the corresponding road, the attribute information of the first graph node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road is determined based on the maximum value of the preset minimum distance between vehicles and / or the maximum speed of vehicles. The graph data construction module is used to construct first graph data corresponding to the graph structure based on the attribute information; A feature determination module is used to perform message passing on the first graph data using a graph neural network, determine the first graph features corresponding to the first graph data, and determine the node features of the second graph nodes of the graph structure based on the first graph features, wherein the satellite image data does not contain traffic network nodes corresponding to the second graph nodes; and The anomaly detection module is used to determine whether there are any anomalies in the traffic network nodes corresponding to the nodes in the second graph based on the node characteristics of the nodes in the second graph.

8. A device for monitoring anomalies in a road traffic network, characterized in that, include: processor; as well as A memory, connected to the processor, for providing the processor with instructions to perform the following processing steps: Determine the graph structure corresponding to the road traffic network of the predetermined area. The graph structure includes graph nodes and edges connecting two graph nodes. The graph nodes are used to represent traffic network nodes of the road traffic network, and the edges connecting two graph nodes are used to represent roads between the corresponding traffic network nodes. Acquire satellite imagery data corresponding to the predetermined area; Based on the satellite image data, determine the attribute information corresponding to the first graph node and its associated edges in the graph structure; Construct first graph data corresponding to the graph structure based on the attribute information; A graph neural network is used to perform message passing on the first graph data, determine the first graph features corresponding to the first graph data, and determine the node features of the second graph nodes of the graph structure based on the first graph features, wherein the satellite image data does not contain the traffic network nodes corresponding to the second graph nodes; and Determining whether there are anomalies in the traffic network nodes corresponding to the second graph nodes based on the node characteristics of the second graph nodes, and wherein the operation of determining the attribute information corresponding to the first graph nodes and associated edges in the graph structure based on the satellite image data includes: Based on the satellite imagery data, determine whether the traffic network node corresponding to the first map node or the road corresponding to the associated edge is damaged; if the corresponding traffic network node or road is damaged, determine the attribute information of the first map node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road based on the minimum value of the preset minimum vehicle spacing and / or minimum vehicle speed; if the corresponding traffic network node or road is not damaged and there are vehicles in the corresponding traffic network node or road, determine the actual minimum vehicle spacing and / or actual minimum vehicle speed in the corresponding traffic network node and road based on the satellite imagery data, and determine the attribute information of the first map node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road based on the actual minimum vehicle spacing and / or actual minimum vehicle speed; and if the corresponding traffic network node or road is not damaged and there are no vehicles in the corresponding traffic network node or road, determine the attribute information of the first map node corresponding to the corresponding traffic network node or the edge corresponding to the corresponding road based on the maximum value of the preset minimum vehicle spacing and / or minimum vehicle speed.

Images