Classifications

• • G—PHYSICS
  • G08—SIGNALLING
  • G08G—TRAFFIC CONTROL SYSTEMS
  • G08G1/00—Traffic control systems for road vehicles
  • G08G1/01—Detecting movement of traffic to be counted or controlled
  • G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles

Definitions

• the present invention relates to a method for detecting vehicles in motion, in particular trucks on motorways, with the aid of at least one camera and / or other devices for detecting the contour or spatial structure of a vehicle and possibly its speed.
• the present invention also relates to a device for detecting vehicles in motion, in particular trucks on motorways, which has at least one camera and / or other device for detecting the contour or spatial structure of the vehicle and possibly its speed.
• Toll monitoring devices are also already known. Similar to the devices at traffic lights or at speed measuring points, these devices generally have a camera which detects a vehicle and, in particular, its number plate when it passes a toll station. In addition, however, such toll devices often also have communication devices which connect to corresponding communication devices which are arranged on a vehicle and in the process exchange all the data relevant for the toll calculation. This can also be linked to automatic booking and communication to a central facility.
• US Patents 5,757,286 and 5,809,161 as well as the corresponding international publications WO 94/28377 and WO 93/19441 could possibly be closest to the present invention, the first-mentioned publication in each case providing a device for speed measurement with the aid of two video cameras with the production of Has stereo images at predetermined time intervals, which results in the location, path and speed of a vehicle.
• the second-mentioned publication describes a camera that is capable of moving objects to be distinguished and tracked from other moving and static objects, this being done essentially by an appropriate image processing system.
• corresponding highway sections are generally limited or interrupted by large stops, which have a number of payment counters at which the vehicles stop briefly so that the driver pays a corresponding fee or at which payment of the fee is briefly checked using a corresponding receipt or stamp becomes.
• systems have already been proposed for this purpose, e.g. Exchange information with a toll station via transponders or other communication devices, the vehicle being identified and a corresponding debit being carried out on an account assigned to the vehicle owner.
• the toll booths with stops and payment stations or with a check have the disadvantage that they are generally quite large structures, since they usually have to have far more checkout or payment points than the motorway in question has lanes in order to avoid significant traffic jams on the To create a freeway.
• these toll booths have the disadvantage that they either have to be additionally provided at every exit or driveway of a motorway, or that no further driveway or exit may exist between two corresponding toll booths. This means that the user may have to accept long distances or detours to reach the next freeway entrance or exit, which significantly reduces the acceptance of such a system and at least partially nullifies the time saved.
• such a system is completely impractical in cases in which only a certain part, even if under certain circumstances a considerable part, of the vehicles using a motorway is subject to charges. This applies, for example, to a motorway fee limited to trucks, as is currently being considered for the Federal Republic of Germany.
• a modern toll system should therefore limit the stopping of vehicles to the absolutely unavoidable level, for example to the payment of a toll fee at most once a day, insofar as journeys can be predicted. Otherwise, tolls can also be paid for longer periods, as a package for certain routes or for any number of kilometers, and it should also be possible to register the vehicle with an automatic debit system.
• there is no need to set up toll booths in such systems if there are opportunities to reliably identify any toll bumpers.
• a corresponding system must therefore be secure enough to effectively prevent any attempt at manipulation and any attempt to travel a certain distance without paying the fees due.
• the toll monitoring points must therefore be able to clearly identify each vehicle that is subject to toll as such, and the error rate should be as low as possible. This means that, on the one hand, vehicles which are not subject to the toll requirement and, on the other hand, on vehicles subject to tolls for which a toll fee has actually been paid should not be given the error message that the fee has not been paid.
• all toll vehicles should actually be recorded as such.
• induction loops as a trigger for a camera, vehicles that are subject to tolls and those that are not subject to tolls are recorded indiscriminately and the evaluation of corresponding image recordings would require an immense effort, which in none would stand in relation to the achievable toll fees.
• the present invention has for its object to provide a method and an apparatus for detecting vehicles in motion, which are able to make an automatic distinction between toll and non-toll vehicles, so that the unnecessary detection of large Avoid amounts of data.
• This system can preferably also be expanded in such a way that it only registers vehicles that are subject to tolls and where the proper payment of the tolls is in doubt.
• the object on which the invention is based is achieved in that, prior to the detection of the contour and / or spatial structure, the detection and tracking of the vehicle is carried out by a LIDAR system, the path and speed of the vehicle being determined from the LIDAR data which are assigned to the contour and structure data determined subsequently.
• an independent speed and path estimate is carried out by a LIDAR system, the path and speed of the vehicle being able to be estimated with sufficient accuracy from the LIDAR data in order to clearly identify the independently measured structural data, that is to say in particular height and / or width of the vehicle, the number of axles, the vehicle type, the license plate etc., to be able to assign to the LIDAR data.
• These data together result in e.g. the approximate length, width and / or height as well as other parameters of the vehicle and the fact whether the vehicle may have a trailer.
• This information is sufficient to classify the vehicle with a high degree of accuracy, whereby it is important to differentiate trucks with a certain size and weight class with sufficient accuracy, so that false messages from the system are largely avoided.
• the corresponding device is characterized in that it has a LIDAR system which, together with an evaluation device for the detection and tracking of a vehicle at a distance before reaching a measuring point or recording position of a camera, the camera and / or the other devices for recording structural data being arranged and designed in such a way that they are at least one of the dimensions height or width of the one previously recorded by the LIDAR and for that Record the pre-calculated vehicle when it reaches the measuring point.
• the at least one profile parameter (height or width) of the vehicle is detected by laser distance sensors which are oriented essentially perpendicular to the direction of travel of the vehicle, so that any perspective distortions are minimal and can be corrected easily. It is expedient if the laser distance sensors assigned to the monitoring of a lane are arranged above and laterally offset from the lane, and if in particular a pair of corresponding laser distance sensors is used. These then simultaneously capture a side front and the roof of the vehicle and the comparison of the data determined by the two laser sensors then enables a precise, stereoscopic determination of the cross section. In order to further improve the accuracy of the height and width dimensions, a method is preferred in which the data from the laser distance sensors are statistically evaluated or averaged.
• the LIDAR is a laser or maser measuring system, which generally consists of a larger number of individual beams which span a plane in the manner of a fan, the opening angle of this fan preferably not being significantly greater than 60 for the purposes of the present invention ° is.
• Each lane is expediently assigned its own LIDAR system and a corresponding LIDAR camera is preferably arranged above and next to the lane to be monitored by it.
• the plane spanned by the individual beams of the LIDAR is inclined both with respect to the street plane and with respect to a plane perpendicular to the street plane. In the case of a three-axis coordinate system, of which two axes span the street plane and the third the direction perpendicular thereto, the plane spanned by the LIDAR is rotated about all three axes.
• a line of intersection of the LIDAR plane with the plane of the road runs along a line which intersects the direction of travel at an acute angle and, for a plane parallel to the street plane, this line of intersection accordingly moves closer to the LIDAR system.
• Each of the n lasers that span the LIDAR plane is able to carry out a distance measurement over the transit time of a reflected laser beam.
• a vehicle In the described alignment of the LIDAR, a vehicle must necessarily break through the LIDAR plane and thereby generate corresponding distance data which are only evaluated in such a way that the path, the speed and the acceleration of the vehicle or object can be estimated therefrom In particular, whether the vehicle remains on the lane on which it was recorded or whether it is currently changing lanes, the LIDAR system can then use this path and speed data to determine the location and time of the structure measurement or height and width measurement, ie he can clearly assign the distance and speed data to the measurement data of the corresponding sensors, and a suitable point in time can be determined on the basis of the LIDAR data, at which a license plate camera is triggered, which essentially detects the front of a vehicle.
• the optical axis of the camera is preferably inclined at an angle between 15 and 70 °, preferably at an angle between 25 and 40 °, with respect to the road surface and the camera itself is, if possible, arranged directly above the lane on which it detects the vehicles.
• An infrared camera with an infrared flash which does not dazzle the drivers of the vehicles and which generally also has a long service life, is expediently used for this purpose.
• the flash is preferably arranged in the vicinity of the camera, so that retroreflective sections, as are widespread with license plates, stand out clearly as a lighter image area.
• communication devices are also provided which allow communication with a corresponding communication device provided in a vehicle. It would basically be possible, for example, to use the lidar system at the same time as a communication device, for which purpose coding and decoding devices could possibly also be used in order to form one or more of the lidar beams as a communication beam and to obtain the essential vehicle information.
• At least one communication device and preferably even two or more communication devices per lane are provided for each lane of a roadway, although each of these communication devices only has a relatively narrowly limited detection area.
• an approximately conical detection area with an elliptical cross section can be produced, the longitudinal axis of the elliptical cross section running in the direction of the lane and the short semi-axis of the elliptical cross section perpendicular this runs.
• the detection areas can be less than the width of a lane and the length can be, for example, two to three times the width, but is usually limited to a maximum of 20 m.
• This communication device should preferably contain the data about the type and type and size class of the vehicle, its identity (license plate, including country code), and toll fee data, so that the registration system can dispense with the collection or storage of data that has already been recorded if the toll is properly paid is determined by the communication device.
• an embodiment is preferred in which exactly one communication device, preferably based on infrared, is used per lane, wherein the detection areas may well overlap.
• the double detection of vehicles that is possible when changing lanes due to the overlap of the detection areas can be corrected later by data comparison.
• the simultaneous detection of a vehicle by two communication devices also provides additional information about the current position of this vehicle.
• All or most of the sensors of the system according to the invention are preferably arranged on the cross member of a bridge spanning the lane or lanes, a part of the sensors being able to be arranged on the vertical pillars of such a bridge if required.
• Figure 1 is a schematic view of a road bridge with different types of
• FIG. 2 shows a roadway view near a bridge from above, detection areas for different sensors being marked differently,
• Figure 3 is a top view of a camera / lighting system for a front view of a truck and its number plate
• Figure 4 is a data processing and communication scheme in the detection of vehicles.
• a highway bridge 10 can be seen in FIG. 1, the cross member 12 of which extends over three parallel carriageways and a standing lane, the cross member 12 resting on two pillars 11.
• the bridge can be a normal road bridge, but it will usually be a separately erected bridge in the manner of the known gantry bridges or the bridges for the accommodation of speed measuring devices and cameras.
• the bridge 10 can also be designed so that it can be walked on, safety devices on the pillars 11, not shown here, ensuring that climbing onto the bridge 10 by unauthorized persons is generally prevented.
• the direction of view in Figure 1 is the direction of travel and you can see on the right a standing lane of width b and then on the left a total of three lanes of width S, with the contour of two trucks 8 being recognizable in a rear view on the right and in the middle lane.
• 1 denotes a license plate camera with integrated lighting, which preferably works on an infrared basis.
• a license plate camera 1 is arranged approximately in the middle above the right and the left lane, whereas an overview camera 6 with LED flash lighting 5 is provided above the left lane.
• the cameras 1, like the overview camera 6, preferably operate on an infrared basis, which has the advantage, among other things, that the drivers of the motor vehicles are not dazzled by the flashes of light which are required anyway in poor visibility conditions, with corresponding infrared flash units also having a very long operating time.
• the LIDAR camera 2 is attached to the right pillar 11 and is provided for monitoring the right lane, while that on the right lane on the cross member 12 of the LIDAR camera 2 'is provided for the detection of the middle lane. Furthermore, laser distance sensors 3, 3' can be seen, which are provided in pairs for each lane.
• the pair 3 of laser distance sensors each consist of a line camera and one over one Angle of a little more than 90 ° scanned laser beam, whereby one of the laser distance sensors 3 is mounted on the right pillar 11 and the other laser distance sensor 3 is mounted above the left lane, but should detect vehicles in the right lane and only for the detection of the Vehicles in the right lane is evaluated by the simultaneous detection of a truck 8 in the right lane by the b With the laser distance sensors 3, a stereoscopic image or a real cross section of the truck 8 located in the right lane can be generated.
• a corresponding pair of laser sensors 3 ' is provided in order to detect trucks 8 in the middle lane.
• This pair of sensors is mounted above the left and right lanes, so that each of these laser sensors 3 'detects both the roof or the top of a truck and one of the sides, that is to say either the right or the left side of the truck.
• the transition along the right and left roof edges is very easy to detect on the basis of the abruptly changing distance data, and by comparing the data recorded by the two line cameras or laser distance sensors 3 ', a stereoscopic image or an exact cross-section can be generated, during the passage of a truck through the bridge 10, if possible, several measurement series are recorded with each of the laser distance sensors 3, 3 ', so that, in addition to the comparison of the data of the pairs of sensors, each of which has detected the same vehicle from different positions, to determine the true dimensions ( Height and width), statistical averaging over several measurements can also be carried out.
• the dimensions in the longitudinal direction of the trucks 8 can also be detected under certain circumstances solely by the laser distance sensors 3, 3 ', since the wheels or axles of the trucks which can be seen from the side are also detected during the passage of the truck.
• the length of a truck which is obtained by the laser distance sensors 3, 3 'in a large number of successive cross-sectional measurements can then be e.g. be determined under the condition that the wheels with these measurements with detected wheels must have a substantially circular shape.
• This boundary condition can also be introduced as a control condition if the length of the truck is determined solely from the driving time (total detection time of the distance sensors 3, 3 ') and the independently determined speed of the truck 8.
• the LIDAR systems 2, 2 ' are used to determine the speed. These are explained in more detail in connection with FIG. 2.
• the bridge 10 also shows what are known as DSRC units 4, which are used for communication with corresponding communication units installed in trucks.
• each of these communication units 4 has a detection area 4 ', which has the shape of an elongated ellipse, with the major axis of the ellipse approximately in the direction of travel and the minor axis perpendicular to it.
• Each lane is covered by two of these communication units 4, the overlap with the communication units of adjacent lanes being kept as small as possible.
• the length of the elliptical detection areas 4 ' is in any case less than 20 m and is approximately 8 m in the specifically illustrated case. This means that about 0.35 seconds are available for communication between a communication unit 4 and the corresponding devices in a truck when the truck 8 moves through the area 4 ′ at a speed of 80 km / h.
• the communication system must exchange all relevant data, that is, record the identity of the vehicle, preferably via its electronically transmitted license plate number, the distance traveled or the intended route and the confirmation of a toll fee or the arrival hand over to a toll payment agency where a corresponding toll has been paid or will be paid.
• a variant of the communication devices is currently preferred, in which only a single but wider detection area of a single communication device per lane is provided in each case compared to the illustration in FIG. 2, the detection areas also being able to overlap, since double detections can be made later Data synchronization can be corrected.
• the communication system can decide whether the remaining data to be recorded or still to be recorded for the vehicle in question must be saved or whether data storage is obsolete because the vehicle has been proven to pay the required toll has been.
• the detection areas 9 and 9 'of the two LIDAR systems 2 and 2' can also be seen in FIG.
• the LIDAR system detects corresponding laser echoes, generally in parallel on several channels, according to the number of laser beams used. In the present case, a number of 16 individual laser beams has proven to be sufficient for the LIDAR 2 or 2 '.
• the LIDAR data is processed automatically, i.e. the echoes of the different laser beams are assigned to the front or side surfaces of a truck due to their only slightly or more deviating values.
• This assignment of the individual laser channels to the front or side surface of a truck enables the current speed and also the direction of travel of a vehicle or truck to be determined on the basis of the continuously continued measurement, and in particular lane changes can also be detected and predicted.
• the recording can also be controlled with the aid of the license plate cameras 1 and possibly also the laser distance sensors 3, 3 '.
• the license plate cameras 1 are each arranged approximately centrally over the right and middle lanes and are inclined with their optical axis approximately 30 ° downwards in the direction of the road surface. They are triggered exactly at the point in time at which the front of the vehicle is essentially completely in the field of view of the camera 1, this point in time likewise being determined from the data of the associated LIDAR 2 or 2 '.
• the laser distance sensors 3, 3 ' permanently scan the empty street space, they can also be switched on if necessary before a truck 8 detected and tracked by the LIDAR system has reached the detection range of the laser distance sensors 3, 3' , They could be switched off again after a time interval which, given the detected speed of the truck and a theoretical maximum length of a truck and trailer (with trailer), is sufficient to drive through the bridge, this time in turn being calculated from the speed data recorded using the LIDAR can.
• All images and data are preferably recorded and stored digitally and, if necessary, forwarded to a central office. Further communication devices (not shown) are provided for this. However, it is expedient if the local system mounted on a bridge 10 is sufficiently self-sufficient that it itself decides which data are to be transmitted to the control center and which do not need to be transmitted. It goes without saying that both the remote transmission of the data and the data exchange between the communication devices 4 and corresponding counterparts in the truck should, if possible, be secured cryptographically and made tamper-proof.
• the license plate cameras 1 are primarily used for this purpose. In principle, these record a complete front image of a truck, but this is recorded with a sufficiently high resolution so that, in particular, the area of the license plate can be enlarged or evaluated. Numerous systems for the automatic reading of motor vehicle license plates are now known, so that these systems need not be discussed in more detail here.
• the communication device can establish a connection with a central point in order to check whether there are data on the payment of the toll fees for the license plate in question.
• the recorded data of the vehicle cross-section, the number of its axles and the determination of whether a trailer is present or not can be compared with the data stored centrally for the license plate, in order, for example, to prevent manipulation through the use of incorrect number plates.
• all recorded data including the frontal view of a truck saved and kept in order to be able to carry out a corresponding check. This is particularly true of course for the data of those vehicles that have clearly not paid the toll due.
• the data of the vehicles for which the fees have been clearly paid either due to direct communication or due to a request from a head office can be deleted immediately, unless you want them for statistical purposes or to compile a complete fee statement save and process anyway.
• a surveillance camera 7 can also be seen in FIG. 1, which is intended to prevent possible manipulation attempts on the bridge or the sensors mounted thereon.
• Figure 3 is shown schematically how the license plate cameras 1 work by an infrared flash light 13 is aligned and mounted at a small angle of 1 ° to 10 ° relative to the axis of the camera 1, so that the license plate retroreflective surfaces of trucks as well as infrared light particularly bright picture elements emerge in the picture taken by the camera 1.
• the overview image generated with a similar camera 6 is also only required in cases of doubt.
• it serves as a classification aid for a human observer who has received the corresponding data at a central office in order to decide whether a given vehicle is a toll vehicle.
• FIG. 4 shows the schematic sequence of the communication process, as well as the cryptographic process (encryption / decryption, digital signature, etc.). This preferably takes place on a plug-in card 32.
• the images digitized in 21 are digitally signed 32 before they are evaluated by various image processing methods 25. Together with the DSRC data, the data obtained from the image processing are fed to the decision process 26. If there are significant deviations between the data from communication and image processing, the corresponding vehicle is classified as a toll bouncer. The data from the communication are then digitally signed and stored together with the digitally signed images as evidence 30. Otherwise, the data is deleted at 31.

Landscapes

• Physics & Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Traffic Control Systems (AREA)
• Image Analysis (AREA)
• Sorting Of Articles (AREA)
• Radar Systems Or Details Thereof (AREA)
• Optical Radar Systems And Details Thereof (AREA)
• Length Measuring Devices By Optical Means (AREA)
• Image Processing (AREA)

Applications Claiming Priority (3)

Publications (2)

Family

ID=7700900

Family Applications (1)

Country Status (5)

Cited By (1)

Families Citing this family (20)

Family Cites Families (10)

• 2001
  • 2001-09-29 DE DE10148289A patent/DE10148289A1/de not_active Withdrawn
• 2002
  • 2002-07-17 AT AT02754376T patent/ATE372523T1/de active
  • 2002-07-17 AU AU2002320979A patent/AU2002320979A1/en not_active Abandoned
  • 2002-07-17 WO PCT/DE2002/002621 patent/WO2003036319A2/fr active IP Right Grant
  • 2002-07-17 DE DE50210861T patent/DE50210861D1/de not_active Expired - Lifetime
  • 2002-07-17 EP EP02754376A patent/EP1446678B1/fr not_active Expired - Lifetime
  • 2002-07-20 AU AU2002320990A patent/AU2002320990A1/en not_active Abandoned
  • 2002-07-20 WO PCT/DE2002/002678 patent/WO2003036320A2/fr not_active Application Discontinuation

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events