Method and traffic monitoring system for monitoring a traffic area

Abstract

The invention relates to a method for monitoring a traffic area (12), comprising a traffic monitoring system (10), and to such a traffic monitoring system (10) for monitoring a traffic area (12). The traffic monitoring system (10) has at least one sensor system, wherein the sensor system comprises at least one sensor (20) for detecting the traffic area (12), and a controller (26) for processing the sensor information detected by the sensors (20). At least two processing algorithms (28, 30, 32, 34) are provided for processing the sensor information and for outputting object information (42, 44, 46) to at least one output (36, 38, 40). At least one criterion is stored for each processing algorithm (28, 30, 32, 34), wherein in each case an output of the object information (42, 44, 46) takes place at an output (36, 38, 40) with the processing algorithm (28, 30, 32, 34) which is assigned to a criterion.

Description

The invention relates to a method for monitoring a traffic area with a traffic monitoring system and to such a traffic monitoring system.Various traffic monitoring systems for monitoring a traffic area are known from the prior art. The traffic monitoring systems have a plurality of sensors for detecting the entire traffic area. The sensor information detected by the sensors is sent to a controller and further processed by this controller in order to obtain an image of the entire traffic area. Usually, the processing is further processed using one or more processing algorithms. In particular, such processing algorithms are used for object recognition or for object tracking, wherein an optimization to a property or to a use purpose is carried out by the respective processing algorithm. This optimization of the sensor information for a purpose of use often means that the data processed with this processing algorithm is less usable with respect to other features or uses. For example, a processing algorithm that is optimized for the highest possible recognition rate during object recognition can result in more erroneous detections, i.e., erroneous object recognitions.In order to have information with sufficiently high quality available for all purposes, controllers known hitherto therefore have a plurality of processing algorithms, a separate output being provided for each processing algorithm. As a result, the construction of such a controller or such a traffic monitoring system is very complicated. In addition, the output defines the processing algorithm and thus the intended use for which the sensor information is optimized.It is an object of the invention to provide a method for traffic monitoring and a traffic monitoring system which has a simpler structure and can be used flexibly.To achieve the object, a method for monitoring a traffic area is provided, having a traffic monitoring system, wherein the traffic monitoring system has at least one sensor system having at least one sensor for detecting the traffic area and a controller for processing the sensor information detected by the sensors, wherein at least two processing algorithms for processing the sensor information and for outputting object information are provided in the controller. For each processing algorithm, at least one criterion is stored in the controller, wherein the object information is output in each case using the processing algorithm which is assigned to a criterion.The invention is based on the consideration that object information created with each of the processing algorithms can be provided at each output of the controller. The selection of the respective processing algorithm takes place depending on the respective required purpose. That is to say that not all object information generated with all processing algorithms is output at a plurality of outputs, but rather exactly the object information or the object information generated with the processing algorithm which is currently required is output at each output. For example, the object information can be optimized with respect to object recognition, i.e., a better recognition rate, or with respect to object tracking. For example, the object information can be provided depending on the situation or depending on the position, so that the optimized object information required for the respective intended use is always output.For example, the criterion is position-dependent. Within a traffic area, there are different areas on which different requirements are made of the object information. For example, a very high recognition rate may be required in a specific area, for example in the vicinity of pedestrian areas or on a stop line for vehicles, in order reliably not to overlook a road user in such areas. In other areas, on the other hand, a small number of erroneous object recognitions may be required, for example, in order not to have a negative influence on the traffic flow. In particular, exactly one processing algorithm is stored for each position, wherein the object information for objects at this position is output using this processing algorithm. Object information is therefore determined for a moving object using different processing algorithms, depending on the position at which the object is located.The traffic area can be divided into a plurality of zones, for example, wherein the criterion is a profile, in particular a safety profile of the zone, and wherein each processing algorithm is assigned at least one profile, in particular a safety profile, and for each zone the object information of the objects which are located in this zone is carried out using the processing algorithm assigned to the respective profile of the zone. For example, the same processing algorithm is not permanently assigned to a zone or a position, but rather is assigned to the zone or the position depending on the situation. The situation or the safety profile can be established, for example, depending on the number and type of road users or the traffic situation within the zone and / or in a defined radius around the position. For example, in the presence of pedestrians or a vehicle from which persons might exit, a processing algorithm with a higher recognition rate may be used temporarily.Furthermore, zones can be created temporarily, for example depending on traffic information. For example, a stop line may only appear temporarily, for example, to allow pedestrians to cross a roadway. In such a situation, a very high recognition rate is required. Furthermore, the traffic information can be hazard zones or the speeds, the number, the distances or the type of road users. Furthermore, the traffic information can be information about the traffic area, for example the number of lanes or the state or the display of traffic signs, such as traffic lights or guidance systems.In addition, the criterion can be object-dependent. For example, a higher recognition rate may be desired or necessary for certain objects or in the vicinity of these objects. In the immediate vicinity of facilities for local area traffic or public transportation, for example, the most reliable possible pedestrian detection may be required.The criterion can also be movement information of at least one object, wherein the movement information is detected in particular with the sensor system. For example, the movement information can be the falling below a defined speed in order to be able to reliably detect exiting persons.The processing algorithms typically require a certain settling time. However, a rapid change between the various processing algorithms, i.e. a change between various object information, is often required, for example when a vehicle is moving rapidly within the traffic area or between various zones. The object information is therefore preferably created permanently using all processing algorithms, wherein only an output of the object information is carried out, the creation of which is carried out using a processing algorithm which is assigned to a criterion. Since all processing algorithms are executed in parallel, it is possible to switch rapidly between the different object information items, i.e. object information items generated alternately with different processing algorithms can be output at an output without delay as soon as the respective criterion is fulfilled.To achieve the object, a traffic monitoring system for monitoring a traffic area is furthermore provided, having at least one sensor system, the sensor system having at least one sensor for detecting the traffic area, and a controller for processing the data detected by the sensors, at least two processing algorithms being provided for processing the data and for outputting object information, at least one criterion being stored for each processing algorithm, the object information being output in each case by the processing algorithm which is associated with a criterion, the traffic area being monitored in particular by a method according to one of the preceding claims.Further advantages and features will become apparent from the following description in conjunction with the accompanying drawings. In these show: FIG. 1 is a schematic illustration of a traffic area monitored with a traffic monitoring system according to the invention; and FIG. 2 shows a schematic illustration of a traffic monitoring system for monitoring the traffic area from FIG. 1.FIG. 2 shows a traffic monitoring system 10 for monitoring a traffic area 12 illustrated in FIG. 1. The traffic area 12 has at least one traffic area 14, on which road users 16, 18 can move. The road users 16 can be pedestrians or cyclists, for example, the road users 18 can be motorized road users, for example vehicles. Different traffic areas 14 can be provided for the road users 16, 18, wherein the traffic areas 14 can overlap each other, for example in the area of a crosswalk across a road.The traffic monitoring system 10 has a plurality of sensors 20 for detecting the entire traffic area 12. Furthermore, the traffic monitoring system has at least one communication device 22 for receiving further information and / or for communicating with one or more road users 16, 18.For example, the traffic monitoring system 10 can establish a temporary or constant communication connection 24 to one or more road users 16, 18. Traffic instructions, warnings or driving instructions can be sent to the road user 16, 18 via this communication link 24. Furthermore, the road users 16, 18 can send traffic information to the traffic monitoring system 10, for example sensor information of the sensors installed in a vehicle.The sensors 20 are positioned in such a way that the entire traffic area 12 can be detected, in particular redundantly. The sensors may use, for example, various sensor technologies. For example, the sensors are radar or lidar sensors, cameras or other sensors which are suitable for detecting the traffic area or the road users 16, 18 or objects located on the traffic area.The sensors 20 and the communication device 22 are connected to a controller 26. In the controller 26, the raw data that have been transmitted from the sensors 20 to the controller 26 are processed. Furthermore, traffic information received via the communication device 22 may also be processed.For example, object detection and object tracking of the road users 16, 18 located within the traffic area 12 take place in the controller 26; this object detection and object tracking can be used, for example, to generate traffic information, warning messages or driving instructions and to transmit them to different road users 16, 18. For this object detection and object tracking, various processing algorithms 28, 30, 32, 34 for the sensor information are available in the controller 26.The processing algorithms 28, 30, 32, 34 differ with regard to the processing of the sensor information, wherein the processing algorithms 28, 30, 32, 34 are optimized with regard to specific criteria, that is to say the object information 44, 46, 48, 50 processed with a processing algorithm and output at an output 36, 38, 40, 42 are optimized for a specific purpose.Such an optimization can be effected, for example, with regard to better pedestrian recognition, reliable recognition of existing objects, recognition of speeds of moving objects or other criteria for stationary or moving objects.However, optimizing the object information with respect to a purpose of use frequently leads to this object information being less well suited for other purposes. Hitherto, therefore, the sensor information has been processed simultaneously with all processing algorithms 28, 30, 32, 34, wherein an output was provided at a controller 26 for each processing algorithm 28, 30, 32, 34 or for the object information with this processing algorithm 28, 30, 32, 34. The various downstream applications were each connected to one of these outputs, at which output only the object formations created with the respective processing algorithm 28, 30, 32, 34 were output.As can be seen in FIG. 1, all processing algorithms 28, 30, 32, 34 are connected to all outputs 36, 38, 40, 42 so that all object information 44, 46, 48, 50 created with the different processing algorithms 28, 30, 32, 34 can be output at all outputs 36, 38, 40, 42.Different criteria are present for the selection of a processing algorithm 28, 30, 32, 34 for creating and outputting the object information 44, 46, 48, 50 at one of the outputs 36, 38, 40, 42.These criteria can be, for example, position information within the traffic area 12. For example, the traffic area 12 is divided into different zones 52, 54, 56. Within these zones 52, 54, 56, there are different requirements for the object information made available. For example, a stop line 58 may be provided in close proximity to a zone 56 that includes an intersection region. In such a zone 52, a particularly reliable object detection may be required in order to avoid road users 16, 18 not being detected at stop line 58. Furthermore, particularly good position detection is required in order to ensure that a vehicle 18 comes to a standstill exactly on the stop line. In a zone 50 that is further away from this stop line 56, for example, speed detection may be more important, whereas in the zone 52 that comprises the intersection region, particularly reliable position detection is important, in particular if a plurality of vehicles are located within the intersection region.If a road user 16, 18 moves within the traffic area 12, its position and the zone 52, 54, 56 in which the road user 16, 18 is located are detected permanently. Depending on the position of the road user 16, 18, the object information 44, 46, 48, 50 relating to this road user 16, 18 is ascertained using the processing algorithm 28, 30, 32, 34 assigned to this zone 52, 54, 56 and is output at the output 36, 38, 40, 42 selected for this purpose. If the road user 16, 18 changes the zone 52, 54, 56, the processing algorithm 28, 30, 32, 34 used for creating the object information 44, 46, 48, 50 also changes, so that the object information 44, 46, 48, 50 output at the same output 36, 38, 40, 42 is now optimized with respect to a different purpose.Therefore, only the object information 44, 46, 48, 50 that is actually required is output at the outputs 36, 38, 40, 42.In order to avoid changeover times between the processing algorithms 28, 30, 32, 34, the sensor information items of the sensors 20 can also be processed in parallel in the background with all the processing algorithms 28, 30, 32, 34, wherein, however, only an output of the object information items takes place at one of the outputs 36, 38, 40, 42 using the respectively selected processing algorithm 28, 30, 32, 34.In addition to or instead of a position-dependent criterion, additional criteria can be defined. For example, criteria can comprise safety profiles for the zones 52, 54, 56 or the road users 16, 18. For example, a certain safety profile can be assigned to a zone 52, 54, 56 depending on the traffic situation. For example, a holding area for a public means of transport can be located in a zone 52, 54, 56, wherein particularly reliable pedestrian detection is required when a public means of transport is stopped. In such a case, a high safety profile can be assigned to the zone 52, 54, 56 and a corresponding processing algorithm 28, 30, 32, 34 can be used.If there is no public means of transport in the zone 52, 54, 56 and / or does not stop, a lower safety profile can be assigned to the zone 52, 54, 56 and another processing algorithm 28, 30, 32, 34 can be used.Furthermore, the processing algorithm 28, 30, 32, 34 can be selected depending on traffic information, in particular the traffic situation within a zone 52, 54, 56. For example, in the case of dense traffic and / or low speeds, i.e. shorter distances between the road users 16, 18, a particularly reliable object identification of the individual road users 16, 18 and an accurate position and distance detection between the vehicles may be required.In addition, the processing algorithms 28, 30, 32, 34 can also be selected depending on the object and / or depending on movement information. For example, in the environment of specific road users 16, 18, for example public means of transport, a particular person recognition may be required. However, this may additionally be dependent on the speed of the public transportation means.If additional information is required for the selection of a processing algorithm 28, 30, 32, 34, this information can be determined, for example, with the sensors 20 or received via the communication device 22. For example, the vehicles can identify themselves when entering the traffic area 12, wherein the vehicle type can also be detected. Furthermore, the road users 16, 18 can have identification features which make it possible to assign a road user 16, 18 to an object group.

Claims

Method for monitoring a traffic area (12), having a traffic monitoring system (10) for monitoring a traffic area (12), having at least one sensor system, wherein the sensor system has at least one sensor (20) for detecting the traffic area (12), and a controller (26) for processing the sensor information detected with the sensors (20), wherein at least two processing algorithms (28, 30, 32, 34) are provided for processing the sensor information and for outputting object information (42, 44, 46) at at least one output (36, 38, 40), wherein at least one criterion is stored for each processing algorithm (28, 30, 32, 34), wherein in each case an output of the object information (42, 44, 46) is effected at an output (36, 38, 40) with the processing algorithm (28, 30, 32, 34) which is assigned to a criterion.Method according to Claim 1, characterized in that the object information (42, 44, 46) is created permanently using all processing algorithms (28, 30, 32, 34), wherein only an output of that object information (42, 44, 46) is carried out, the creation of which has been carried out using a processing algorithm (28, 30, 32, 34) which is associated with a criterionMethod according to one of claims 1 and 2, characterised in that the criterion is position-dependent.Method according to one of Claims 1 to 3, characterized in that the traffic area (12) is divided into a plurality of zones (48, 50, 52), wherein the criterion is a profile, in particular a safety profile of the zone (48, 50, 52), and wherein each processing algorithm (28, 30, 32, 34) is assigned at least one profile, in particular a safety profile, and for each zone the object information (42, 44, 46) of the objects which are located in this zone is carried out using the processing algorithm (28, 30, 32, 34) assigned to the respective profile of the zone.Method according to Claim 4, characterized in that the zones (48, 50, 52) are created temporarily, in particular as a function of traffic information.Method according to one of the preceding claims, characterized in that the criterion is object-dependent.Method according to one of the preceding claims, characterized in that the criterion is movement information of at least one object, wherein the movement information is detected in particular by means of the sensor system.Traffic monitoring system (10) for monitoring a traffic area (12), having at least one sensor system, wherein the sensor system has at least one sensor (20) for detecting the traffic area (12), and a controller (26) for processing the sensor information detected with the sensors (20), wherein at least two processing algorithms (28, 30, 32, 34) are provided for processing the sensor information and for outputting object information (42, 44, 46) at at least one output (36, 38, 40), wherein at least one criterion is stored for each processing algorithm (28, 30, 32, 34), wherein in each case an output of the object information (42, 44, 46) is carried out at an output (36, 38, 40) with the processing algorithm (28, 30, 32, 34) which is assigned to a criterion, wherein the traffic area (12) is monitored in particular using a method according to one of the preceding claims.

Images