Methods for estimating the traffic density of road users on a given road section

Abstract

The invention relates to a method for estimating the traffic density of road users on a given road section, in which a travel speed of the vehicles (3, 5) is determined on the basis of data sent to an evaluation device (9) by a large number of road users (3, 5), by means of which the traffic density is inferred.In a method based on continuous data collection, but limiting the computational effort, the image data transmitted by road users (3, 5), which are created by the road users (3, 5) while driving on a given road section (A, B, C) from the oncoming road users (3, 5), are processed by a computer vision algorithm, whereby object data is created for each oncoming road user (5, 3) which records the crossing time of each oncoming road user (3, 5) in the given road section (A, B, C) and the number of oncoming vehicles (3, 5) that have the same crossing time, whereby the traffic density for a given period is evaluated as a function of the number of oncoming road users detected in this period and their crossing times.

Description

[0001] The invention relates to a method for estimating a traffic density of road users on a given traffic section, in which a travel speed of the vehicles is determined on the basis of data sent from a large number of road users to an evaluation device external to the vehicle, by means of which the traffic density is deduced.

[0002] DE 100 22 812 A1 discloses a method for determining the traffic situation based on reporting vehicle data for a traffic network with traffic-controlled network nodes. This method involves obtaining data from reporting vehicles moving in traffic along specified route sections. This data is used to determine their travel times, which are used to determine the number of vehicles in the queue and the average vehicle density outside the queue on the route section under consideration.

[0003] US 2019 / 0220679 A1 discloses a method for localizing a traffic situation in cooperation with a plurality of vehicles. Situation data containing images of vehicles located in a geographical region are inserted into an image database. Related situation objects are located in images of the image database. The images of different vehicles are compared in the image database, with the compared images exhibiting corresponding features of the related situation objects.

[0004] The object of the invention is to provide a method for estimating a traffic density of road users on a given traffic section, which is based on continuous data collection but limits the computational effort.

[0005] The invention is based on the features of the independent claims. Advantageous developments and refinements are the subject of the dependent claims. Further features, possible applications, and advantages of the invention will become apparent from the following description and the explanation of exemplary embodiments of the invention, which are illustrated in the figure.

[0006] The problem is solved by the subject matter of patent claim 1.

[0007] In the method explained at the outset for estimating a traffic density of road users on a given road section, in which a travel speed of the vehicles is determined on the basis of data sent by a large number of road users to an evaluation device external to the vehicle, by means of which the traffic density is inferred, the image data sent by road users, which are created by the road users while driving on a given road section from the road users in oncoming traffic, are processed by a computer vision algorithm, whereby object data is created for each road user in oncoming traffic, which record the crossing time of each road user in oncoming traffic in the given road section and the number of vehicles in oncoming traffic that have the same crossing time,The traffic density is evaluated for a given period of time depending on the number of oncoming traffic users detected during this period and their crossing times. The continuous collection of object data by all vehicles traveling on the given route section eliminates the need for reporting vehicles, which can only be deployed on selected sections. This allows for real-time information on traffic density on any given section of road at any time. The large number of evaluated image data increases the accuracy of the determined traffic density, as each individual crossing of a route section by a road user is taken into account in the evaluation. The aggregation of individual crossings over each route section includes the recording of the compass direction.

[0008] In one embodiment, the speed of the road users capturing the image data and the speed of oncoming road users are considered to be approximately identical. This reduces the computational effort and provides the results in a timely manner.

[0009] In a further embodiment, an average value is calculated from the recorded ratios of the number of vehicles per crossing time for the crossing times recorded within the specified period, which reflects the traffic density on the specified traffic section during the specified period. This averages out inaccuracies caused by the large number of vehicles considered.

[0010] In a further embodiment, the recorded object data, which relate to the direction of travel and the start time of travel on the specified traffic section, as well as the number of road users for each crossing time, are compared and stored in an object list. This object list supports the targeted estimation of different groups of road users, such as passenger cars, commercial vehicles, buses, cyclists, and pedestrians.

[0011] In a further embodiment, the object data of road users in a vehicle fleet is collected and sent to a vehicle backend for analysis. This can involve using services from a vehicle manufacturer that wirelessly connect to and exchange data with all vehicles in their global customer fleet via a vehicle backend.

[0012] Further advantages, features, and details will become apparent from the following description, which describes at least one exemplary embodiment in detail. Described features may form the subject matter of the invention alone or in any meaningful combination, possibly even independently of the claims, and may, in particular, also be the subject matter of one or more separate applications.

[0013] It shows: Fig. 1 an embodiment of the method according to the invention.

[0014] In Fig.1 shows an exemplary embodiment of the method according to the invention, in which a traffic section 1 is shown in the form of a road junction, which consists of three road sections A, B, C. On road sections A and C, two vehicles 3, 5 are approaching each other, for example. The first vehicle 3 is traveling on road section A and will then turn onto road section B. The second vehicle 5 is traveling on road section C in the direction of road section A. Both vehicles 3, 5 are each equipped with an environmental camera 7, with which they record the respective oncoming vehicle 5, 3. The image data recorded by the environmental cameras 7 are processed in the respective vehicle 3, 5 by a computer vision algorithm and various objects, including their positions, are recognized and stored. For example, the object with position x, y is recognized at time t.This object data is transmitted wirelessly to a vehicle backend 9, which creates an object list from the object data of the two oncoming vehicles 3, 5. Alternatively, the object list for each vehicle 3, 5 can also be created in the respective vehicle 3, 5, and the respective object list can be sent to the vehicle backend 9, which combines the object lists of both vehicles 3, 5.

[0015] An example of such an object list is shown in Table 1. While vehicle 3 is traveling on section A, every oncoming vehicle 5, commercial vehicle, cyclist, or pedestrian, is detected, stored, and counted. A driver assistance system automatically differentiates between active and inactive road users, such as parked cars. At the same time, the time t F known which vehicle 3 needs to cross section A. Table 1 section of route Direction of travel Start time Duration of crossing Counted objects A South 01.08.2024, 16:15 85s 2 cars, 1 truck A South 01.08.2024, 16:30 45s 2 cars, 1 cyclist A South 01.08.2024 , 16:25 70s 2 cars

[0016] Using a special aggregation process, the vehicle backend 9 estimates traffic densities for various traffic sections from thousands of data points underlying the object lists, taking into account that object data for the respective traffic section is not available for every minute. The aggregation process will be explained in more detail using the recorded cars.

[0017] For the object list presented in Table 1, the traffic density for cars on northbound section A is 2 cars, each with a crossing time of 85 seconds, corresponding to a measurement of oncoming traffic of 170 seconds. This estimate is only valid if the oncoming traffic has the same speed as the oncoming measuring vehicle 3. During this time, 2 cars are detected, and the oncoming traffic is measured in approximately 170 seconds. The ratio of 2 cars to 170 seconds results in a ratio of 1 / 85 per car / s.

[0018] For the second crossing, two cars were detected, each with a crossing time of 45 seconds. Thus, two cars were detected in 90 seconds of oncoming traffic. The ratio of two cars / 90 seconds results in a ratio of 1 / 45 per car / second per car.

[0019] During the third crossing, two cars were detected with a crossing time of 35 seconds, meaning that two cars were detected in 70 seconds of oncoming traffic. The ratio of two cars per 70 seconds results in a ratio of 1 / 35 per car per second.

[0020] From the three crossings, an average ratio of approximately 1 / 48 of a car per second is determined. Extrapolated over an hour, this results in a traffic density of 75 cars. For the period between 4:00 p.m. and 5:00 p.m., the traffic density is thus given as 75 cars. QUOTES CONTAINED IN THE DESCRIPTION

[0000] This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Cited patent literature

[0000] DE 100 22 812 A1

[0002] US 2019 / 0220679 A1

[0003]

Claims

[1] Method for estimating a traffic density of road users on a given traffic section, in which a travel speed of the vehicles (3, 5) is determined on the basis of data sent from a plurality of road users (3, 5) to an evaluation device (9) external to the vehicle, by means of which the traffic density is deduced, characterized byin that the image data sent by road users (3, 5), which are created by the road users (3, 5) while driving on a predetermined traffic section (A, B, C) of the road users (3, 5) in oncoming traffic, are processed by a computer vision algorithm, wherein object data are created for each road user (5, 3) of the oncoming traffic, which record the crossing time of each road user (3, 5) of the oncoming traffic in the predetermined section of road (A, B, C) and the number of vehicles (3, 5) of the oncoming traffic which have the same crossing time, wherein the traffic density for a predetermined period of time is evaluated as a function of the number of road users of the oncoming traffic detected in this period and their crossing times. [2] Method according to claim 1, characterized bythat the speed of the road users (3, 5) recording the image data and the speeds of the oncoming road users (3, 5) are considered to be approximately identical. [3] Method according to claim 1 or 2, characterized by that for the crossing times recorded within the specified period, an average value is formed from the recorded ratios of the number of road users (3, 5) per crossing time, which reflects the traffic density in the specified period on the specified traffic section (A, B, C). [4] Method according to claim 1, 2 or 3, characterized by that the recorded object data, which relate to the direction of travel and the start time of driving on the specified traffic section (A, B, C), and a number of road users (3, 5) are related to each crossing time and stored in an object list. [5] Method according to at least one of the preceding claims, characterized by that the object data of road users (3, 5) of a vehicle fleet are recorded and sent to a vehicle backend (9) for evaluation.

Images