EP1361488B1 - Method and apparatus for classifying vehicles equipped with wheels - Google Patents

Abstract

The method involves acquiring (4) a view (A) of the wheels (R1-3) associated with one side of the vehicle (1) to be classified, determining at least one characteristic property (m1,m2,DR,DF) of the wheels by evaluating the view and allocating the determined characteristic property to a vehicle class. The determined characteristic property can be the diameter or rim of the tire of at least one wheel. AN Independent claim is also included for the following: (a) an arrangement for implementing the inventive method.

Description

The invention relates to a method for automatically classifying wheeled vehicles. Furthermore, the invention relates to a corresponding device for carrying out the method.

The ever-increasing load on the road network by lorries and the associated costs of maintaining motorways and expressways in particular have led to a toll on the use of the roads concerned. Usually this toll is paid when entering the street in question at an arranged in the area of the junction tollbooth. For this purpose, the respective vehicle must stop or at least drive so slowly that the payment of the toll can be handled properly and the access to the toll road can be released.

In the known way, although an exact control of the payment of the toll and access to the toll road network can be accomplished. However, the payment process and the release of access only after payment has resulted in a significant delay in the flow of traffic. This can lead to the formation of large traffic jams in front of the toll booths, especially in the case of heavy traffic.

Therefore, from a traffic planning point of view, there is the requirement to pay the toll in such a way that the traffic flow remains largely undisturbed. This requirement can be fulfilled by giving vehicle drivers or drivers the opportunity of a flat-rate payment in advance of a road user fee. It is also conceivable to subsequently charge the use of a route.

Regardless of whether the payment of the user fee is made in advance or is billed subsequently, an effective control of the road use is required. On the one hand, this control is complicated by the variety of vehicle types. On the other hand, the control of the vehicles is particularly complex in that the toll to be paid usually should not only be made dependent on the distance traveled, but also on the load caused by the use of the infrastructure by the respective vehicle. For example, heavy goods vehicles should pay higher fees than light ones, as the heavier trucks, for example, cause higher road wear than lighter vehicles. Therefore, the road charges to be paid have been classified into load classes depending on the respective weight of the vehicle.

It has been proposed to automatically detect the respective load class of a vehicle by an automated detection of its geometric dimensions, such as overall width, total height, total length, etc., perform. The disadvantage of this procedure, however, is that the results of an automatically occurring Image evaluation greatly influenced by the vehicle respectively transported loads or other structures that do not lead to a significant change in the weight of the vehicle itself, however, change its appearance compared to the standard type of each vehicle but significantly. Due to these influences, reliable automated detection of vehicles characterized by their external appearance can not be accomplished with the required safety.

Attempts have been made to circumvent the problems inherent in image-based detection and classification of vehicles by detecting magnetic characteristics of the vehicles instead of geometrical characteristics. For this purpose, magnetic field sensors are arranged in the road surface, which detect the change in an electromagnetic field occurring as a function of the mass of magnetizable material of the vehicle and assign it to a specific weight class. In this way, it is true that the accuracy of the weight-dependent classification of the respective detected vehicle can be increased. The disadvantage, however, is that the known magnetic field sensors must be firmly arranged in the road surface and this extensive structural measures on the respective monitored road are required, which lead to a disability of current traffic and considerable costs.

For automatic classification of wheeled vehicles is from the US 4,813,004 and the JP 2001-283375 A a method is known in which a view of the vehicle is detected, which images the wheels. In this process become characteristic Characteristics of the wheels, namely their width and number and the outer tire diameter and rim diameter determined. To determine the characteristic features of the wheels, the vehicle is laterally detected by a camera. Finally, the determined characteristic features of the wheels are assigned to a vehicle class.

A disadvantage of this prior art is that the detection of characteristic features of the wheels can lead to misinterpretations in the assignment of a vehicle to a particular class, since it is now possible for every driver of a vehicle easily, the tire or rim size individually after to change his wishes. This is especially true for small or light vehicles. So it is conceivable that, for example, cars or SUVs have wheels with extra-large diameter, which would actually be typical of a different vehicle class. In this case, therefore, cars or SUVs would be mistakenly classified, for example, in the category of lorries and the toll system would impose an incorrect charge.

It is therefore an object of the present invention to provide an automatically feasible method for classifying vehicles, in particular cars or off-road vehicles, which leads to reliable results with reliable results. Likewise, a device should be created with which this method can be carried out in a particularly simple manner.

• Erfassen einer die einer Fahrzeugseite zugeordneten Räder abbildenden Ansicht des jeweils zu klassifizierenden Fahrzeugs,
• Ermitteln mindestens eines charakteristischen Merkmals der Räder durch Auswerten der Ansicht,
• Erfassen der die gesamte Form des Fahrzeugs kennzeichnenden Merkmale wie Gesamtbreite und - höhe, und
• Zuordnen des ermittelten charakteristischen Merkmals der Räder zu einer Fahrzeugklasse.

The previously derived and indicated object is achieved by a method for automatic Classifying wheeled vehicles, comprising the following steps:

• Detecting a view of each vehicle to be classified, which images the wheels associated with a vehicle side,
• Determining at least one characteristic feature of the wheels by evaluating the view,
• Detecting the features characterizing the entire shape of the vehicle such as overall width and height, and
• Assigning the determined characteristic feature of the wheels to a vehicle class.

• mit einer Bilderfassungseinrichtung zum Erfassen eines Bildes der Ansicht des Fahrzeugs,
• mit einer Einrichtung zum Erfassen der die gesamte Form des Fahrzeugs kennzeichnenden Merkmale wie Gesamtbreite und -höhe, und
• mit einer Auswerteinrichtung zum Ermitteln des charakteristischen Merkmals mindestens eines der Räder des Fahrzeugs durch Auswerten des von der Bilderfassungseinrichtung jeweils erfassten, an die Auswerteinrichtung gelieferten Bildes und zum Zuordnen des ermittelten charakteristischen Merkmals zu einer Fahrzeugklasse.

Furthermore, the object is achieved by a device for carrying out this method:

• with an image capture device for capturing an image of the view of the vehicle,
• with a device for detecting the overall shape of the vehicle characterizing features such as overall width and height, and
• with an evaluation device for determining the characteristic feature of at least one of the wheels of the vehicle by evaluating the image detected by the image capture device respectively, supplied to the evaluation device and for assigning the determined characteristic feature to a vehicle class.

Advantageous embodiments of the method and the device according to the invention are the subject of the dependent claims.

In other words, according to the method and the device according to the invention, characteristic features of the wheels, for example the number of wheels, the distance between the axles and / or the diameter of the wheels, are still used to assign the vehicle to a vehicle class. In addition to the characteristic features of the wheels, the overall shape of the vehicle is detected characteristics such as overall width and height.

In this way, before the final assignment is made on the basis of the determined characteristic of the wheels to a vehicle class, it can be checked whether the determined value of the characteristic of the wheels, for example the value determined for the wheel diameter, also typically corresponds to the vehicle it is to be classified.

If, for example, the diameter is used and the value determined for the diameter is particularly large, it is still checked before the final assignment to a vehicle class, whether the vehicle is also a vehicle due to its width or height, which typically wheels with such a large diameter has. In this way, the accuracy of the classification of the vehicle is significantly increased and reduces the likelihood of misinterpretation.

The captured view may be any type of image that is suitable for uniquely determining the location and size of the wheels of the respective detected vehicle. Thus, the recording may be a representation of a longitudinal side of the vehicle detected by a camera oriented transversely to the direction of travel of the vehicle. Likewise, it is possible, using appropriately trained evaluation systems, to use images which are detected at a certain angle. In addition, it may be appropriate to use stereoscopic images of the relevant side surface. It is essential only that the position and size of the wheels can be determined from the detected view with sufficient accuracy.

The respectively determined characteristic size of the wheels may be, for example, a diameter, for example the outer diameter, at least one tire, the diameter of at least one rim, the difference between the outer diameter of the tire and the rim of at least one of the wheels, or the number and distribution the number of axles of the vehicle assigned to each side of the vehicle associated, visible on the view wheels. In this case, the reliability of the detection can be increased by not only determining a respective geometric characteristic of the wheels, but also determining two and more such characteristics and taking them into account in the allocation of the respective load class. Particularly reliable results are obtained when the number of axles of each vehicle is combined with a characteristic characterizing the tire or rim size of the wheels.

The invention is based on the idea that the tire and rim size can be used together with the number of axles as characteristic parameters that stand for the respective permissible total weight. Thus, in particular for lorries, the respective load class can be assigned with high reliability on the basis of easily recognizable geometric variables of the vehicle to be classified in each case. The clearly defined form of vehicle wheels, which is particularly unmistakeable in the view, also permits an unambiguous, unmistakable determination of the features which give the assigned classification, even in the case of automatic image acquisition and evaluation.

Experiments have shown that by integrating tire and rim size registration, the number of trucks properly classified could be significantly increased. While the known methods each only provided recognition quotas in the range of 70%, misrecognitions and classifications constitute the exception in the method according to the invention. Therefore, the procedure according to the invention is suitable, for example, for reliable automatic differentiation of tolled and toll-free vehicles when entering roads, the for vehicles with a gross vehicle weight exceeding a certain weight.

It may be convenient, carried out according to the invention recognition of characteristic features of the wheels themselves or their geometric arrangement on the vehicle with a conventional detection of the entire shape of the vehicle characterizing features, such as Total width, height, etc., to further increase the accuracy of the classification.

An apparatus which is particularly suitable for carrying out the method according to the invention is equipped with an image capture device for detecting the view of the vehicle and with an evaluation device which evaluates the respective characteristic feature of at least one of the wheels of the vehicle by evaluating the image acquisition device respectively acquired by the image acquisition device Determined image and the determined characteristic feature assigns a vehicle class. Such a device can be easily operated with conventional image capture devices, such as a commercially available portable digital video system, which are available at low cost and can be easily assembled and transported.

In this case, the degree of automation of the classification process can be further enhanced by the image capture device comprising a device for automatically detecting the approach of a moving vehicle.

The invention will be explained in more detail with reference to an embodiment schematically illustrating drawing.

The sole FIGURE shows a lorry 1, which enters a toll road 2 for lorries with a certain permissible gross weight. In the area of the entrance is a device 3 for automatically determining the weight category of the trucks hauling into the road 2.

The device 3 comprises an image capture device 4, which is formed by a conventional digitally operating, portable video system that is set up on a roadside tripod not shown here. In addition, the image capture device 4 is equipped with a device 4a for detecting the approach of a motor vehicle.

The image capture device 4 is connected via a data line 4b to an evaluation device 5 formed by a conventional data processing device. The evaluation device 5 is arranged in a likewise transportable housing 5a.

If an on-site inspection is to be carried out, a display device 6 connected to the evaluation device 5 in the form of a conventional screen can also be accommodated in the housing 5a.

The evaluation device 5 is equipped with a memory device in which the result of the evaluation carried out by the evaluation device 5 is stored. In this case, the respectively evaluated captured views A can also be stored for securing evidence. Likewise, the evaluation device 5 via a suitable, not shown remote data transmission device with a central monitoring point, also not shown, for example, a central data acquisition device or a traffic control center. Each of these embodiments enables a fully automatic classification operation that requires no operator intervention.

The device 3 can also be part of a monitoring system which not only carries out the classification of the respective vehicle according to the load class, but also other measures required for detecting the vehicle, such as automatic recognition of the vehicle registration number, automatic recognition of identification means carried by the vehicle, etc ,

In addition, the device may be connected to a signal device which emits a signal for discharging this vehicle from the current traffic upon detection of a suspected of improper payment of fees vehicle.

The truck 1 has three axles A1, A2, A3 for each pair of wheels, wherein the first axis A1 in the area of the driver's cab 1a and the two other axles A2, A3 are arranged closely adjacent as twin axles in the area of the rear 1b of the truck 1. Accordingly, on the right in the direction of travel F, visible in the figure side of the truck 1 three wheels R1, R2, R3 can be seen.

If the device 4a detects the approach of the truck 1, the image capture device 4 makes a receptacle A of the side view of the vehicle visible in the direction of travel F. Truck 1 This illustration A of the side view of the truck 1 is supplied to the evaluation device 5 as a digital data record via the data line 4b.

The evaluation device 5 first determines the position of the wheels R1, R2, R3 in the illustration A by applying a suitable evaluation algorithm. Subsequently, the spacings m1, m2 of the center points of the wheels R1 and R2 or R1 and R3 corresponding to the distances of the axles A1, A2 and A1, A3 are determined in order to make a clear statement about the number of axles of the truck 1. In addition, for the wheel R1, the outer diameter D _{R of} the tire and the outer diameter D _{F of} the rim are detected.

The number of axes A1, A2, A3 and the diameters D _R , D _{F of} the tire _R 1 are then compared as geometrical features characterizing the wheels _{R 1} , _{R 2} , _R 3 by the evaluation device 5 with data records which are stored in a memory of the evaluation device 5 and indicate which axle numbers combined with typical tire and rim diameters are assigned to which load class. This comparison gives the load class LK, which is assigned to the truck 1.

This load class LK is displayed by the display device 6, for example to indicate to a monitoring person that the lorry 1 is a toll vehicle.

In addition, the result delivered by the evaluation device 5, for example, trigger an automatic payment or posting process, by a Account of the vehicle owner or driver is charged with the due toll.

The classification of the lorry 1 automatically carried out by the device 3 can be supplemented by further recognition methods, with which not only the load class can be classified, but also an exact determination of the identity of the respective vehicle can be carried out. The advantage of the operation of the device 3 is that it enables a safe selection of vehicles passing by according to their permissible total weight in real time in the inventive mode, so that the optionally complementary carried out recognition or payment procedures are actually subject to the toll Vehicles can concentrate.

REFERENCE NUMBERS

1

Lorry

1a

Cab of the lorry 1

1b

Rear of the truck 1

2

toll road

3

Device for automatically determining the weight category of trucks hauling into road 2

4

Image capture device

4a

Device for detecting the approach of a motor vehicle

4b

data line

5

evaluation

5a

Portable housing of the evaluation device 5

6

display

A

Recording the visible side view of the truck 1

A1, A2, A3

Axles of the truck 1

D _R

Outer diameter of the tire of the wheel R1

D _F

Outer diameter of the rim of the wheel R1

F

direction of travel

LK

load class

m1, m2

Distances of the axles of the wheels R1, R2 or R1, R3

R1, R2, R3

Wheels of the truck 1

Claims (15)

1. A process which can be performed automatically for the classification of vehicles (1) fitted with wheels (R1, R2, R3) involving the following steps:

   - Acquisition of an image (A) of the wheels (R1, R2, R3) arranged on the side of a vehicle (1) which requires classification,

   - Determination of at least one characteristic feature of the wheels (R1, R2, R3) by evaluating the image (1),

   - Acquisition of the features characterising the overall shape of the vehicle, such as overall width and height, and

   - Assignment of the characteristic features (m1, m2, D_R, D_F) of the wheels (R1, R2, R3) so obtained to a vehicle class (LK).
2. A process according to claim 1, characterised in that a diameter (D_R) of the tyre on at least one of the wheels (R1) is determined as a characteristic feature of the wheels (R1, R2, R3).
3. A process according to one of the foregoing claims, characterised in that the diameter (D_F) of the rim on at least one of the wheels (R1) is determined as a characteristic feature of the wheels (R1, R2, R3).
4. A process according to one of the foregoing claims, characterised in that the difference of the outer diameter (D_R, D_F) of the tyre and rim on at least one of the wheels (R1) is determined as a characteristic feature of the wheels (R1, R2, R3).
5. A process according to one of the foregoing claims, characterised in that the number of axles (A1, A2, A3) of the vehicle is determined as a characteristic feature of the wheels (R1, R2, R3)
6. A process according to one of the foregoing claims, characterised in that the image (A) is acquired using a digital image acquisition device (4).
7. A process according to claim 6, characterised in that the image acquisition device comprises at least two cameras.
8. A process according to one of the foregoing claims, characterised in that the operating step in acquiring the image (A) comprises the self-actuated recognition of the approach of a vehicle (1).
9. A process according to one of the foregoing claims, characterised in that the vehicles concerned are trucks (1).
10. A device to perform the process designed according to one of the claims 1 to 9 to classify motor vehicles (1) with:

    - an image acquisition device (4) to capture an image (A) of the vehicle (1),

    - a device to capture the features characterising the overall shape of the vehicle, such as overall width and height, and

    - an evaluation device (5) to determine the characteristic features of at least one of the wheels (R1, R2, R3) of the vehicle (1) by evaluating the image (A) captured by the image acquisition device (4) and transmitted to the evaluation device (5), and to assign the characteristic features so obtained to a vehicle class (LK).
11. A device according to claim 10, characterised in that the image acquisition device (4) comprises a device (4a) for the automatic detection of the approach of a vehicle (1) in motion.
12. A device according to one of claims 10 or 11, characterised in that it comprises a display device (6) to show the vehicle class (LK) assigned to the acquired vehicle (1) in question.
13. A device according to one of claims 10 to 12, characterised in that it is portable.
14. A device according to one of claims 10 to 13, characterised in that the image acquisition device (4) comprises at least two cameras.
15. A device according to one of claims 10 to 14, characterised in that the evaluation device (5) and the image acquisition device (4) form a structural module.

Images