Traffic jam identification method under low-altitude view angle and related equipment thereof

Abstract

The invention discloses a traffic jam identification method under a low-altitude view angle and related equipment thereof, and the method comprises the steps: obtaining a road image of a to-be-identified road section; performing target detection on the vehicles in the road image by using a preset deep learning model to obtain frames of the vehicles in the road image; according to the frame of each vehicle in the road image, determining the dense degree of the vehicles in the road image and the number of dense vehicles; and identifying and obtaining a traffic jam result of the road section to be identified according to the density degree between the vehicles and the number of the dense vehicles. The method can be widely applied to traffic road congestion monitoring, is especially suitable for road sections where vehicles run slowly, is accurate in judgment, and greatly improves the accuracy of traffic road congestion monitoring.

Description

Technical Field

[0001] The present application relates to the technical field of traffic congestion monitoring, and in particular to a method for identifying traffic congestion from a low-altitude perspective and related equipment. Background Art

[0002] At present, the main methods for monitoring whether there is congestion on traffic roads include monitoring methods based on vehicle speed and based on GPS positioning.

[0003] Among them, the monitoring method based on vehicle speed generally uses the vehicle speed to judge the traffic congestion situation. For example, when the driving speed of most vehicles on the road is lower than the preset driving threshold, it is considered that the road is congested. This method can reflect the traffic smoothness to a certain extent, but it mainly relies on the accuracy and real-time nature of speed data. When the traffic flow is large and the speed is slow, this method cannot fully reflect the severity of the congestion, especially when some vehicles suddenly slow down due to complex traffic conditions (such as traffic accidents, road obstacles, etc.), which may lead to misjudgment.

[0004] The monitoring method based on GPS positioning can estimate the traffic density and congestion level of roads by collecting the GPS tracks of vehicles, combining map data and traffic flow analysis; this method is usually suitable for large-scale road monitoring and has good spatial coverage. However, in this method, the accuracy of GPS data is limited by the vehicle's positioning equipment, signal shielding and data update frequency. In addition, GPS positioning cannot directly reflect the specific traffic conditions on the road section, and in areas with densely populated high-rise buildings in the city, positioning errors may lead to data deviations, thereby affecting the accuracy of congestion judgment.

[0005] The above two methods have certain limitations in practical applications, especially they are not suitable for sections where vehicles travel slowly, and they are prone to misjudgment, which reduces the accuracy of traffic congestion monitoring. Summary of the invention

[0006] The purpose of this application is to provide a method, device, electronic device and storage medium for identifying traffic congestion from a low-altitude perspective, which can be widely used in traffic road congestion monitoring, especially suitable for sections of roads where vehicles travel slower, and the judgment is relatively accurate, greatly improving the accuracy of traffic road congestion monitoring.

[0007] In order to achieve the above objectives, in a first aspect, the present application provides a method for identifying traffic congestion from a low-altitude perspective, comprising:

[0008] Acquire a road image of a road section to be identified;

[0009] Performing target detection on vehicles in the road image using a preset deep learning model to obtain bounding boxes of each vehicle in the road image;

[0010] Determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image;

[0011] According to the density between the vehicles and the number of the densely packed vehicles, a traffic congestion result of the road section to be identified is identified;

[0012] Determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image includes:

[0013] Obtaining image pixel coordinates of the frame of each vehicle in the road image according to the frame of each vehicle in the road image;

[0014] The density of vehicles in the road image and the number of densely packed vehicles are determined according to the borders of each vehicle in the road image and the corresponding image pixel coordinates.

[0015] In a preferred embodiment of the present application, the step of determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates includes:

[0016] Acquire the road planning line of the road section to be identified;

[0017] The density of vehicles in the road image and the number of densely packed vehicles are determined according to the road planning line of the road section to be identified, the borders of each vehicle in the road image and the corresponding image pixel coordinates.

[0018] In a preferred embodiment of the present application, the step of obtaining the road planning line of the road section to be identified includes:

[0019] Extracting the exposed road planning line of the road section to be identified from the road image;

[0020] Performing a first compensation for the blocked road planning line of the road section to be identified in the road image according to the exposed road planning line, the frame of each vehicle in the road image, and the distance between lanes;

[0021] Judging whether there are parking lanes on both sides of the road section to be identified according to the status of vehicle lights on both sides in the road image; when there are parking lanes on both sides of the road section to be identified, performing a second compensation for the blocked road planning lines of the road section to be identified in the road image according to the exposed road planning lines, the borders of each vehicle in the road image, and the size of the parking lanes;

[0022] The road planning line of the road section to be identified is obtained according to the exposed road planning line of the road section to be identified and the blocked road planning line of the road section to be identified.

[0023] In a preferred embodiment of the present application, the step of determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates includes:

[0024] Determining the density of vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates;

[0025] According to the density of vehicles in the road image and a preset density threshold, a dense vehicle group is obtained;

[0026] The number of densely packed vehicles between the vehicles in the road image is determined based on the vehicles in the densely packed group of vehicles.

[0027] In a preferred embodiment of the present application, the step of determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates includes:

[0028] Determining an average density between vehicles in each column of the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates;

[0029] Determine whether there are parked vehicles and sub-roads where parked vehicles are distributed in the road section to be identified according to the average density between vehicles in each column in the road image and the headlight status of each column of vehicles in the road image;

[0030] When there are parked vehicles on the road section to be identified, differentiating between moving vehicles and parked vehicles in the road image by using frames of different colors;

[0031] The density of vehicles and the number of densely packed vehicles in the road image are determined according to the frames of the running vehicles in the road image and the corresponding image pixel coordinates.

[0032] In a preferred embodiment of the present application, the identification of the traffic congestion result of the road section to be identified according to the density between the vehicles and the number of the densely packed vehicles includes:

[0033] Obtaining a target density corresponding to densely packed vehicles from the density between the vehicles;

[0034] The traffic congestion result of the road section to be identified is obtained according to the number of densely packed vehicles, the target density corresponding to the densely packed vehicles, the preset densely packed vehicle threshold and the preset target density threshold.

[0035] In a preferred embodiment of the present application, after identifying the traffic congestion result of the road section to be identified according to the density between the vehicles and the number of densely packed vehicles, the method further includes:

[0036] When the traffic congestion result indicates that there is congestion in the road section to be identified, the position information of the dense area in the road section to be identified is obtained according to the position information of the shooting device and the position information of the dense center of the dense area in the road image.

[0037] In a preferred embodiment of the present application, after obtaining the position information of the dense area in the road section to be identified according to the position information of the shooting device and the position information of the dense center of the dense area in the road image, the method further includes:

[0038] Displaying the location information of the densely populated area in the road section to be identified, the time of occurrence of congestion and the traffic congestion result on a predetermined map interface;

[0039] Alternatively, the location information of the densely populated area in the road section to be identified, the time of occurrence of congestion and the traffic congestion result are displayed in a predetermined graphical form.

[0040] In a second aspect, the present application provides a traffic congestion identification device from a low-altitude perspective, comprising:

[0041] An acquisition module, used for acquiring a road image of a road section to be identified;

[0042] A target detection module, used to perform target detection on vehicles in the road image using a preset deep learning model to obtain a bounding box of each vehicle in the road image;

[0043] A density analysis module, used to determine the density of vehicles in the road image and the number of densely packed vehicles according to the borders of each vehicle in the road image;

[0044] An identification module, used for identifying and obtaining the traffic congestion result of the road section to be identified according to the density between the vehicles and the number of the densely packed vehicles;

[0045] The intensive analysis module is specifically used for:

[0046] Obtaining image pixel coordinates of the frame of each vehicle in the road image according to the frame of each vehicle in the road image;

[0047] The density of vehicles in the road image and the number of densely packed vehicles are determined according to the borders of each vehicle in the road image and the corresponding image pixel coordinates.

[0048] In a third aspect, the present application provides an electronic device, including a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the above-mentioned traffic congestion identification method from a low-altitude perspective.

[0049] In a fourth aspect, the present application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned method for identifying traffic congestion from a low-altitude perspective.

[0050] The present application provides a method, device, electronic device and storage medium for identifying traffic congestion from a low-altitude perspective. Compared with the prior art, it has at least the following beneficial effects:

[0051] The road image of the road section to be identified used in the present application is obtained by taking pictures at a low-altitude perspective. The road image can clearly show the vehicles in the road section to be identified. Since it is taken at a low-altitude perspective, the present application is particularly suitable for sections where vehicles travel slowly. At the same time, the present application can also be applied to other sections and is widely used in traffic congestion monitoring. The present application detects the vehicles in the road image through a preset deep learning model to obtain the borders of each vehicle in the road image, and then determines the density of the vehicles in the road image and the number of dense vehicles. The density of the vehicles in the road image and the number of dense vehicles can clearly and accurately reflect the traffic congestion situation of the road section to be identified, so that the traffic congestion result of the road section to be identified can be accurately identified, so that the accuracy of traffic congestion monitoring can be greatly improved. In addition, the borders of each vehicle are calibrated by using image pixel coordinates, so that the distribution and calibration of the borders of each vehicle are more systematic and standardized, the calculation standard of the spacing between the rectangular borders of the vehicles is unified, and the accuracy of the calculation of the spacing between the rectangular borders of the vehicles is improved, so that the accuracy of the calculation of the distance between the vehicles in the road image can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0053] Figure 1 It is a flow chart of a method for identifying traffic congestion from a low-altitude perspective provided in an embodiment of the present application;

[0054] Figure 2 is a flowchart of step S130 provided in an embodiment of the present application;

[0055] Figure 3 is a flowchart of step S140 provided in an embodiment of the present application;

[0056] Figure 4 It is a structural block diagram of a traffic congestion identification device under a low-altitude perspective provided in an embodiment of the present application;

[0057] Figure 5 It is a schematic diagram of the internal structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0058] The specific implementation methods of the present application are further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present application but are not intended to limit the scope of the present application.

[0059] At present, the methods of monitoring whether there is traffic congestion based on vehicle speed and GPS positioning have certain limitations in practical applications, and are especially not suitable for sections where vehicles travel slowly. Both methods are prone to misjudgment in such sections of traffic roads, reducing the accuracy of traffic congestion monitoring.

[0060] In response to the above-mentioned problems in the prior art, the embodiments of the present application provide a method, device, electronic device and storage medium for identifying traffic congestion from a low-altitude perspective, which can be widely used in traffic road congestion monitoring, especially suitable for sections of roads where vehicles travel slower, and the judgment is relatively accurate, greatly improving the accuracy of traffic road congestion monitoring.

[0061] See also Figure 1 , Figure 1 It is a flow chart of a method for identifying traffic congestion from a low-altitude perspective provided in an embodiment of the present application.

[0062] The traffic congestion identification method under a low-altitude perspective described below in the embodiment of the present application can be applied to computer equipment such as servers.

[0063] The present application embodiment provides a method for identifying traffic congestion from a low-altitude perspective, comprising the following steps:

[0064] Step S110, obtaining a road image of a road section to be identified.

[0065] The road image of the road section to be identified is obtained by taking pictures at a low altitude perspective by a shooting device, which may be a drone or a camera set up on the road section to be identified. In this embodiment, a drone is used as an example of a shooting device to illustrate and describe the corresponding contents of the embodiment.

[0066] In one embodiment, the road image of the section to be identified can be a single image taken by a drone at a low altitude perspective, or it can be a composite image; when the road image of the section to be identified is a composite image, it can be obtained by synthesizing multiple images of the section to be identified taken by a drone, or it can be obtained by obtaining a corresponding image frame from a video taken by a drone, and the road image of the section to be identified is obtained as a composite image by image synthesis. The road image of the section to be identified using a composite image can more comprehensively and clearly reflect the road conditions and vehicles of the section to be identified, which is conducive to congestion identification and judgment of the section to be identified.

[0067] Step S120, performing target detection on vehicles in the road image using a preset deep learning model to obtain the bounding box of each vehicle in the road image.

[0068] In one embodiment, the preset deep learning model is a trained deep learning model. Specifically, a trained target detection model can be used, which has the function of performing target detection on vehicles on a traffic road section.

[0069] The border of each vehicle in the road image can be a rectangular border or a frame of other shapes, for example, an elliptical border. In this embodiment, a rectangular border is used as the border of each vehicle in the road image. Preferably, the border of each vehicle is a border that surrounds / circumscribes the vehicle.

[0070] Step S130, determining the density of vehicles in the road image and the number of densely packed vehicles according to the borders of each vehicle in the road image.

[0071] In one embodiment, the density between vehicles may be the distance between vehicles, that is, the distance between vehicles is used as the density, and the value of the distance between vehicles may be a percentage of the vehicle length (rounded), 1 to 100%; specifically, when determining the distance between vehicles in a road image, the distance between vehicles in the road image may be calculated by the spacing between the rectangular frames of the vehicles and the distance between the drone and the ground of the road section to be identified.

[0072] After determining the distance between vehicles in the road image, the number of densely packed vehicles in the road image can be determined by determining the distance between vehicles in the road image. The number of densely packed vehicles is the number of densely packed vehicles, and the number of densely packed vehicles can be 4 to 1000.

[0073] In other embodiments, the density between vehicles may also be measured by other parameters, for example, it may be the spacing between the rectangular frames of the vehicles. Correspondingly, the number of densely packed vehicles in the road image may be determined by the spacing between the rectangular frames of the vehicles. It can be understood that if the density between vehicles is measured by the spacing between the rectangular frames of the vehicles, the distance between the drone and the ground of the same road section to be identified may be fixed, thereby ensuring the accuracy of the spacing between the rectangular frames of the vehicles as a measure of the density between the vehicles.

[0074] Step S140, identifying and obtaining the traffic congestion result of the road section to be identified according to the density between vehicles and the number of densely packed vehicles.

[0075] In one embodiment, a corresponding threshold interval can be set according to the density between vehicles to match different levels of density. Similarly, a corresponding threshold interval can be set according to the number of dense vehicles to match different levels of dense vehicles. Furthermore, the traffic congestion results of the road section to be identified can be determined by the different levels of density and different levels of dense vehicles. It can be understood that the traffic congestion results of the road section to be identified also correspond to different congestion levels, for example, light congestion, moderate congestion, medium-to-high congestion and high congestion.

[0076] The traffic congestion identification method under low-altitude perspective of the embodiment of the present application adopts a road image of the section to be identified that is taken under low-altitude perspective. The road image can clearly show the vehicles in the section to be identified. Since it is taken under low-altitude perspective, the present application is particularly suitable for sections where vehicles travel slowly. At the same time, the present application can also be applied to other sections and is widely used in traffic road congestion monitoring. The present application performs target detection on vehicles in the road image through a preset deep learning model, obtains the bounding box of each vehicle in the road image, and then determines the density of vehicles and the number of densely packed vehicles in the road image. Among them, the density of vehicles and the number of densely packed vehicles in the road image can clearly and accurately reflect the traffic road congestion situation of the section to be identified, so that the traffic congestion result of the section to be identified can be accurately identified, so that the accuracy of traffic road congestion monitoring can be greatly improved.

[0077] See also Figure 2 , Figure 2 It is a flowchart of step S130 provided in an embodiment of the present application.

[0078] In one embodiment, step S130, determining the density of vehicles in the road image and the number of densely packed vehicles according to the borders of each vehicle in the road image, may include:

[0079] Step S131, obtaining the image pixel coordinates of the frame of each vehicle in the road image according to the frame of each vehicle in the road image;

[0080] Step S132, determining the density of vehicles in the road image and the number of densely packed vehicles based on the borders of each vehicle in the road image and the corresponding image pixel coordinates.

[0081] The image pixel coordinates of the frame of each vehicle in the road image can reflect the distribution of each vehicle in the road image, and are conducive to calculating the spacing between the rectangular frames of the vehicles, thereby improving the calculation speed of the vehicle distance between the vehicles in the road image. In addition, the frame of each vehicle is calibrated using the image pixel coordinates, so that the distribution and calibration of the frame of each vehicle are more systematic and standardized, the calculation standard of the spacing between the rectangular frames of the vehicles is unified, and the accuracy of the calculation of the spacing between the rectangular frames of the vehicles is improved, thereby improving the accuracy of the calculation of the vehicle distance between the vehicles in the road image.

[0082] In one embodiment, step S132, determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates, may include:

[0083] Obtain the road planning line of the road section to be identified;

[0084] The density of vehicles in the road image and the number of densely packed vehicles are determined according to the road planning lines of the road section to be identified, the borders of each vehicle in the road image and the corresponding image pixel coordinates.

[0085] In this embodiment, the road planning line of the road section to be identified may reflect the specific road planning of the road section to be identified, for example, the number of lanes, the driving direction of vehicles on the lanes, and the arrangement of parking lanes (whether there are parking lanes, the planned number of parking lanes, and the length of parking lanes), etc.;

[0086] When determining the density between vehicles in a road image based on the road planning lines of the road section to be identified, the bounding boxes of each vehicle in the road image, and the corresponding image pixel coordinates, the bounding boxes of each vehicle in the road image can be first divided using the road planning lines of the road section to be identified and the bounding boxes of each vehicle in the road image, and the bounding boxes of vehicles in different driving directions and the bounding boxes of vehicles in parking lanes can be distinguished. Then, based on the bounding boxes of vehicles in different driving directions and the bounding boxes of vehicles in parking lanes, as well as the corresponding image pixel coordinates, the density between vehicles in different driving directions in the road image can be determined, wherein the bounding boxes of vehicles in parking lanes do not need to be used to determine the density between vehicles in different driving directions in the road image. After determining the density between vehicles in different driving directions in the road image, the number of densely packed vehicles in different driving directions in the road image can be determined.

[0087] The road planning lines of the sections to be identified can be used to accurately know the specific road planning of the sections to be identified. Combined with the specific road planning of the sections to be identified, their road distribution can be determined. Then, when determining the density of vehicles and the number of densely packed vehicles in the road image, a clear division is made accordingly according to the road distribution, thereby improving the accuracy of the density of vehicles and the number of densely packed vehicles in the sections to be identified, thereby improving the accuracy of the traffic congestion results for the sections to be identified. At the same time, vehicles in parking lanes are not taken into consideration, which can reduce interference and recognition speed in traffic congestion recognition of the sections to be identified.

[0088] In this embodiment, optionally, obtaining a road planning line of a road section to be identified may include:

[0089] Extracting the exposed road planning line of the road section to be identified from the road image;

[0090] Based on the exposed road planning lines, the borders of each vehicle in the road image and the distance between lanes, a first compensation is performed on the blocked road planning lines of the road section to be identified in the road image;

[0091] According to the status of the lights of the vehicles on both sides of the road image, it is determined whether there are parking lanes on both sides of the road section to be identified; when there are parking lanes on both sides of the road section to be identified, a second compensation is performed on the blocked road planning lines of the road section to be identified in the road image according to the exposed road planning lines, the borders of each vehicle in the road image and the size of the parking lanes;

[0092] The road planning line of the road section to be identified is obtained according to the exposed road planning line of the road section to be identified and the blocked road planning line of the road section to be identified.

[0093] The exposed road planning lines of the road section to be identified are road planning lines that are not blocked by objects such as vehicles in the road image and are exposed in the road image; when performing the first compensation for the blocked road planning lines of the road section to be identified in the road image, the borders of each vehicle in the road image are used to determine the number of lanes of the road section to be identified, and the spacing between the lanes is used to perform the first compensation for the blocked road planning lines in combination with the exposed road planning lines after determining the number of lanes of the road section to be identified;

[0094] The status of the headlights of vehicles on both sides of the road image can be photographed by a drone at a lower viewing angle and obtained through image recognition, wherein the drone can be different from the drone used to obtain the road image, that is, the shooting is performed by another drone; when judging whether there are parking lanes on both sides of the road section to be identified, the proportion of the headlights of vehicles on both sides being on / off can be identified to determine whether there are parking lanes on both sides of the road section to be identified. It can be understood that the headlights of vehicles in parking lanes are usually in the off state; when there are parking lanes on both sides of the road section to be identified, the parking lane planning lines are preliminarily supplemented by revealing the road planning lines and the sizes of the parking lanes. After the preliminarily supplemented, the parking lane planning lines are corrected in combination with the borders of each vehicle in the parking lane, so as to complete the second compensation for the blocked road planning lines of the road section to be identified in the road image.

[0095] Through the above method, the actual road planning line of the road section to be identified can be obtained, avoiding the situation that the road planning line is not known after adjustment; at the same time, through the secondary compensation of the obstructed road planning line, the accuracy of the obstructed road planning line is highly guaranteed, and in the secondary compensation processing, the spacing between lanes, the status of the headlights of vehicles on both sides and the size of the parking lane are fully combined to fully ensure the accuracy of the compensated obstructed road planning line, so as to better ensure the accuracy of the road planning line of the road section to be identified.

[0096] In one embodiment, step S132, determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates, may include:

[0097] Determine the average density between vehicles in each column of the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates;

[0098] According to the average density of vehicles in each column of the road image and the status of the lights of vehicles in each column of the road image, it is determined whether there are parked vehicles and sub-roads where parked vehicles are distributed in the road section to be identified;

[0099] When there are parked vehicles on the road section to be identified, different colored frames are used to distinguish between moving vehicles and parked vehicles in the road image;

[0100] The density of vehicles and the number of densely packed vehicles in the road image are determined according to the frames of the running vehicles in the road image and the corresponding image pixel coordinates.

[0101] When determining the average density between vehicles in each column in the road image, firstly divide the vehicles into multiple columns and the frames of the vehicles according to the frames of the vehicles in the road image, and calculate the average density between vehicles in each column according to the frames of the vehicles in each column and the corresponding image pixel coordinates;

[0102] The average density between vehicles in each column can be used to determine whether there are parked vehicles on the road section to be identified, and then determine whether there are parking lanes and sub-roads where parked vehicles are distributed on the road section to be identified. Specifically, the judgment can be made based on the numerical value and difference of the average density between vehicles in each column; and the status of the lights of each column of vehicles in the road image can also be used to determine whether there are parked vehicles on the road section to be identified. Specifically, the judgment can be made based on the proportion of the lights of each column of vehicles in the on / off state. Optionally, a preliminary judgment can be made based on the average density between vehicles in each column, and then a secondary judgment can be made based on the status of the lights of each column of vehicles in the road image, and then the preliminary judgment result can be corrected based on the secondary judgment result to obtain the final judgment result.

[0103] This dual determination and correction method greatly ensures the accuracy of determining whether there is a parking lane on the road section to be identified and the sub-roads where parked vehicles are distributed. Different colored borders are used to distinguish between moving vehicles and parked vehicles in the road image, and then the borders of the moving vehicles and the corresponding image pixel coordinates are used to determine the density of vehicles and the number of densely packed vehicles in the road image, which greatly improves the calculation efficiency and eliminates the interference of parked vehicles, making the density of vehicles and the number of densely packed vehicles in the road image more reasonable, thereby improving the accuracy of identifying the traffic congestion results of the road section to be identified.

[0104] In another embodiment, step S132, determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates, may include:

[0105] Determine the density of vehicles in the road image based on the borders of each vehicle in the road image and the corresponding image pixel coordinates;

[0106] According to the density of vehicles in the road image and a preset density threshold, a dense vehicle group is obtained;

[0107] According to the vehicles in the dense group of vehicles, the number of dense vehicles between the vehicles in the road image is determined.

[0108] In this embodiment, the preset density threshold is used to determine whether vehicles in the road image need to be divided into dense vehicles, and to divide dense vehicle groups; it can be understood that the number of dense vehicle groups can be one or more, especially in complex traffic conditions (such as traffic accidents, road obstacles, etc.), the number of dense vehicle groups is likely to be multiple.

[0109] By dividing dense vehicles and dense vehicle groups, and then determining the number of dense vehicles between vehicles in the road image, the accuracy of the determined number of dense vehicles can be improved. In addition, when there are multiple dense vehicle groups, the accuracy of determining the number of dense vehicles is greatly improved, and the distribution of dense vehicle groups on the road section to be identified can be clearly reflected, which is conducive to the driver obtaining more information about the road conditions of the road section to be identified, thereby facilitating the driver's decision-making.

[0110] See also Figure 3 , Figure 3 It is a flowchart of step S140 provided in an embodiment of the present application.

[0111] In one embodiment, step S140, based on the density of vehicles and the number of densely packed vehicles, identifying the traffic congestion result of the road section to be identified includes:

[0112] Step S141, obtaining a target density corresponding to densely packed vehicles from the density between vehicles;

[0113] Step S142, identifying and obtaining the traffic congestion result of the road section to be identified according to the number of densely packed vehicles, the target density corresponding to the densely packed vehicles, the preset densely packed vehicle threshold and the preset target density threshold.

[0114] In this embodiment, the dense vehicles correspond to the vehicles in the determined dense vehicle number, and the obtained target density corresponding to the dense vehicles corresponds to the specific density between the vehicles in the determined dense vehicle number.

[0115] By selecting the target density corresponding to dense vehicles, the interference and calculation amount of non-dense vehicles are removed, and the recognition efficiency of traffic congestion results of the road section to be identified is greatly improved.

[0116] In one embodiment, in step S140, after the traffic congestion result of the road section to be identified is obtained according to the density between vehicles and the number of densely packed vehicles, the method for identifying traffic congestion from a low-altitude perspective of the embodiment of the present application may further include:

[0117] When the traffic congestion result indicates that there is congestion in the road section to be identified, the position information of the dense area in the road section to be identified is obtained according to the position information of the shooting device and the position information of the dense center of the dense area in the road image.

[0118] The traffic congestion result indicates that there is congestion on the road section to be identified, which may be a situation where the traffic congestion result of the road section to be identified is one of light congestion, moderate congestion, medium-high congestion and high congestion.

[0119] In this embodiment, the location information of the shooting device may be the longitude and latitude of the shooting device; the dense area in the road image may be determined based on the corresponding vehicle in the dense vehicle number, and it can be understood that the dense area in the road image may be one or more places; the dense center is the center point of the dense area, and the location information of the dense center may be the image pixel coordinates of the dense center; in this embodiment, the location information of the dense area in the road section to be identified is calculated by the longitude and latitude of the shooting device and the image pixel coordinates of the dense center of the dense area in the road image.

[0120] By combining the location information of the shooting equipment with the location information of the dense center of the dense area in the road image, the location information of the dense area in the road section to be identified can be accurately obtained, and the road section to be identified can be precisely located; the location information of the dense area in the road section to be identified can be used by the server, for example, it can be reflected in the corresponding map software, or corresponding road congestion information can be generated and transmitted to the corresponding road processing personnel.

[0121] In this embodiment, the method for identifying traffic congestion from a low-altitude perspective in the embodiment of the present application may also include:

[0122] Displaying the location information, congestion occurrence time and traffic congestion results of the densely populated area in the road section to be identified on a predetermined map interface;

[0123] Alternatively, the location information of the densely populated area in the road section to be identified, the time of occurrence of congestion and the result of traffic congestion are displayed in a predetermined graphical form.

[0124] In this embodiment, optionally, the server can be a road monitoring background server for performing local road monitoring, which displays a map interface of the monitored local road; the location information of dense areas in the road section to be identified, the time when congestion occurs and the results of traffic congestion can be displayed on this predetermined map interface; similarly, the road monitoring background server can also display the location information of dense areas in the road section to be identified, the time when congestion occurs and the results of traffic congestion in a predetermined graphical form to facilitate monitoring by relevant personnel.

[0125] In order to execute the methods corresponding to the above embodiments and achieve corresponding functions and technical effects, a traffic congestion identification device under a low-altitude perspective is provided below.

[0126] See also Figure 4 , Figure 4 It is a structural block diagram of a traffic congestion identification device from a low-altitude perspective provided in an embodiment of the present application.

[0127] The traffic congestion identification device under low-altitude viewing angle provided in the embodiment of the present application includes:

[0128] An acquisition module 410 is used to acquire a road image of a road section to be identified;

[0129] The target detection module 420 is used to perform target detection on vehicles in the road image using a preset deep learning model to obtain the bounding box of each vehicle in the road image;

[0130] A density analysis module 430, for determining the density of vehicles in the road image and the number of densely packed vehicles according to the borders of each vehicle in the road image;

[0131] The identification module 440 is used to identify the traffic congestion result of the road section to be identified according to the density between vehicles and the number of densely packed vehicles.

[0132] The traffic congestion identification device under low-altitude perspective of the embodiment of the present application adopts a road image of the section to be identified that is taken under a low-altitude perspective. The road image can clearly show the vehicles in the section to be identified. Since it is taken under a low-altitude perspective, the present application is particularly suitable for sections where vehicles travel slowly. At the same time, the present application can also be applied to other sections and is widely used in traffic road congestion monitoring. The present application performs target detection on vehicles in the road image through a preset deep learning model, obtains the bounding box of each vehicle in the road image, and then determines the density of vehicles and the number of densely packed vehicles in the road image. Among them, the density of vehicles and the number of densely packed vehicles in the road image can clearly and accurately reflect the traffic road congestion situation of the section to be identified, so that the traffic congestion result of the section to be identified can be accurately identified, so that the accuracy of traffic road congestion monitoring can be greatly improved.

[0133] In one embodiment, the dense analysis module 430 may be specifically used to:

[0134] According to the frame of each vehicle in the road image, the image pixel coordinates of the frame of each vehicle in the road image are obtained;

[0135] According to the borders of each vehicle in the road image and the corresponding image pixel coordinates, the density of vehicles in the road image and the number of densely packed vehicles are determined.

[0136] Optionally, when the density analysis module 430 determines the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates, it can be specifically used to:

[0137] Obtain the road planning line of the road section to be identified;

[0138] The density of vehicles in the road image and the number of densely packed vehicles are determined according to the road planning lines of the road section to be identified, the borders of each vehicle in the road image and the corresponding image pixel coordinates.

[0139] Furthermore, when acquiring the road planning line of the road section to be identified, the dense analysis module 430 can be specifically used to:

[0140] Extracting the exposed road planning line of the road section to be identified from the road image;

[0141] According to the exposed road planning lines, the borders of each vehicle in the road image and the distance between lanes, the first compensation is performed on the blocked road planning lines of the road section to be identified in the road image;

[0142] According to the status of the lights of the vehicles on both sides of the road image, it is determined whether there are parking lanes on both sides of the road section to be identified; when there are parking lanes on both sides of the road section to be identified, a second compensation is performed on the blocked road planning lines of the road section to be identified in the road image according to the exposed road planning lines, the borders of each vehicle in the road image and the size of the parking lanes;

[0143] The road planning line of the road section to be identified is obtained according to the exposed road planning line of the road section to be identified and the blocked road planning line of the road section to be identified.

[0144] Optionally, when the density analysis module 430 determines the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates, it can be specifically used to:

[0145] Determine the average density between vehicles in each column of the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates;

[0146] According to the average density of vehicles in each column of the road image and the status of the lights of vehicles in each column of the road image, it is determined whether there are parked vehicles and sub-roads where parked vehicles are distributed in the road section to be identified;

[0147] When there are parked vehicles on the road section to be identified, different colored frames are used to distinguish between moving vehicles and parked vehicles in the road image;

[0148] The density of vehicles and the number of densely packed vehicles in the road image are determined according to the frames of the running vehicles in the road image and the corresponding image pixel coordinates.

[0149] Optionally, when the density analysis module 430 determines the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates, it can be specifically used to:

[0150] Determine the density of vehicles in the road image based on the borders of each vehicle in the road image and the corresponding image pixel coordinates;

[0151] According to the density of vehicles in the road image and a preset density threshold, a dense vehicle group is obtained;

[0152] According to the vehicles in the dense group of vehicles, the number of dense vehicles between the vehicles in the road image is determined.

[0153] In one embodiment, the identification module 440 may be specifically configured to:

[0154] Obtain target density corresponding to dense vehicles from the density between vehicles;

[0155] According to the number of densely packed vehicles, the target density corresponding to the densely packed vehicles, the preset densely packed vehicle threshold and the preset target density threshold, the traffic congestion result of the road section to be identified is obtained.

[0156] As an optional implementation, the traffic congestion identification device under the low-altitude perspective of the present application further includes a positioning module, which is used to:

[0157] When the traffic congestion result indicates that there is congestion in the road section to be identified, the position information of the dense area in the road section to be identified is obtained according to the position information of the shooting device and the position information of the dense center of the dense area in the road image.

[0158] Furthermore, the traffic congestion identification device under the low-altitude perspective of the present application also includes a display module for:

[0159] Displaying the location information, congestion occurrence time and traffic congestion results of the densely populated area in the road section to be identified on a predetermined map interface;

[0160] Alternatively, the location information of the densely populated area in the road section to be identified, the time of occurrence of congestion and the result of traffic congestion are displayed in a predetermined graphical form.

[0161] The above-mentioned low-altitude traffic congestion recognition device can implement the above-mentioned low-altitude traffic congestion recognition method. The specific limitations and other contents of the above-mentioned low-altitude traffic congestion recognition device embodiment can be found in the content of the low-altitude traffic congestion recognition method, which will not be repeated in the embodiment.

[0162] An embodiment of the present application also provides an electronic device, including a memory and a processor, wherein the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the above-mentioned traffic congestion identification method under a low-altitude perspective.

[0163] Optionally, the electronic device mentioned above may be a computer device such as a server.

[0164] In one embodiment, the internal structure of the electronic device of the present application can be as follows: Figure 5 shown.

[0165] An embodiment of the present application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned method for identifying traffic congestion from a low-altitude perspective.

[0166] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0167] In addition, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0168] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0169] The above description is only an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings.

[0170] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0171] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations; at the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "including one..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

Claims

1. A method for identifying traffic congestion from a low-altitude perspective, characterized in that: include: Acquire a road image of a road section to be identified; Performing target detection on vehicles in the road image using a preset deep learning model to obtain bounding boxes of each vehicle in the road image; Determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image; According to the density between the vehicles and the number of the densely packed vehicles, a traffic congestion result of the road section to be identified is identified; Determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image includes: Obtaining image pixel coordinates of the frame of each vehicle in the road image according to the frame of each vehicle in the road image; The density of vehicles in the road image and the number of densely packed vehicles are determined according to the borders of each vehicle in the road image and the corresponding image pixel coordinates.

2. The method for identifying traffic congestion from a low-altitude perspective according to claim 1, characterized in that: Determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates includes: Acquire the road planning line of the road section to be identified; The density of vehicles in the road image and the number of densely packed vehicles are determined according to the road planning line of the road section to be identified, the borders of each vehicle in the road image and the corresponding image pixel coordinates.

3. The method for identifying traffic congestion from a low-altitude perspective according to claim 2, characterized in that: The step of obtaining the road planning line of the road section to be identified comprises: Extracting the exposed road planning line of the road section to be identified from the road image; Performing a first compensation for the blocked road planning line of the road section to be identified in the road image according to the exposed road planning line, the frame of each vehicle in the road image, and the distance between lanes; Judging whether there are parking lanes on both sides of the road section to be identified according to the status of vehicle lights on both sides in the road image; when there are parking lanes on both sides of the road section to be identified, performing a second compensation for the blocked road planning lines of the road section to be identified in the road image according to the exposed road planning lines, the borders of each vehicle in the road image, and the size of the parking lanes; The road planning line of the road section to be identified is obtained according to the exposed road planning line of the road section to be identified and the blocked road planning line of the road section to be identified.

4. The method for identifying traffic congestion from a low-altitude perspective according to claim 1, characterized in that: Determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates includes: Determining the density of vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates; According to the density of vehicles in the road image and a preset density threshold, a dense vehicle group is obtained; The number of densely packed vehicles between the vehicles in the road image is determined based on the vehicles in the densely packed group of vehicles.

5. The method for identifying traffic congestion from a low-altitude perspective according to claim 1, characterized in that: Determining the density of vehicles and the number of densely packed vehicles in the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates includes: Determining an average density between vehicles in each column of the road image according to the borders of each vehicle in the road image and the corresponding image pixel coordinates; Determine whether there are parked vehicles and sub-roads where parked vehicles are distributed in the road section to be identified according to the average density between vehicles in each column in the road image and the headlight status of each column of vehicles in the road image; When there are parked vehicles on the road section to be identified, differentiating between moving vehicles and parked vehicles in the road image by using frames of different colors; The density of vehicles and the number of densely packed vehicles in the road image are determined according to the frames of the running vehicles in the road image and the corresponding image pixel coordinates.

6. The method for identifying traffic congestion from a low-altitude perspective according to claim 1, characterized in that: The identifying and obtaining the traffic congestion result of the road section to be identified according to the density between the vehicles and the number of the densely packed vehicles includes: Obtaining a target density corresponding to densely packed vehicles from the density between the vehicles; The traffic congestion result of the road section to be identified is obtained according to the number of densely packed vehicles, the target density corresponding to the densely packed vehicles, the preset densely packed vehicle threshold and the preset target density threshold.

7. The method for identifying traffic congestion from a low-altitude perspective according to claim 1, characterized in that: After identifying the traffic congestion result of the road section to be identified according to the density between the vehicles and the number of densely packed vehicles, the method further includes: When the traffic congestion result indicates that there is congestion in the road section to be identified, the position information of the dense area in the road section to be identified is obtained according to the position information of the shooting device and the position information of the dense center of the dense area in the road image.

8. The method for identifying traffic congestion from a low-altitude perspective according to claim 7, characterized in that: After obtaining the position information of the dense area in the road section to be identified according to the position information of the shooting device and the position information of the dense center of the dense area in the road image, the method further includes: Displaying the location information of the densely populated area in the road section to be identified, the time of occurrence of congestion and the traffic congestion result on a predetermined map interface; Alternatively, the location information of the densely populated area in the road section to be identified, the time of occurrence of congestion and the traffic congestion result are displayed in a predetermined graphical form.

9. A traffic congestion identification device from a low-altitude perspective, characterized in that: include: An acquisition module, used for acquiring a road image of a road section to be identified; A target detection module, used to perform target detection on vehicles in the road image using a preset deep learning model to obtain a bounding box of each vehicle in the road image; A density analysis module, used to determine the density of vehicles in the road image and the number of densely packed vehicles according to the borders of each vehicle in the road image; An identification module, used for identifying and obtaining the traffic congestion result of the road section to be identified according to the density between the vehicles and the number of the densely packed vehicles; The intensive analysis module is specifically used for: Obtaining image pixel coordinates of the frame of each vehicle in the road image according to the frame of each vehicle in the road image; The density of vehicles in the road image and the number of densely packed vehicles are determined according to the borders of each vehicle in the road image and the corresponding image pixel coordinates.

10. An electronic device, characterized in that: It includes a memory and a processor, the memory is used to store a computer program, and the processor runs the computer program to enable the electronic device to execute the traffic congestion identification method under a low-altitude perspective according to any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that: It stores a computer program, which, when executed by a processor, implements the method for identifying traffic congestion from a low-altitude perspective as described in any one of claims 1 to 8.

Images