WO1995017740A1

WO1995017740A1 - Automatic vehicle identification toll collection system with optical vehicle recognition

Info

Publication number: WO1995017740A1
Application number: PCT/US1994/004541
Authority: WO
Inventors: Kenneth R. Aubrey
Original assignee: Cubic Automatic Revenue Collection Group
Priority date: 1993-12-20
Filing date: 1994-04-25
Publication date: 1995-06-29
Also published as: AU7353894A

Abstract

The improved AVI toll collection system verifies the identity of a tag in an AVI toll collection system by optically inspecting vehicles proceeding through a toll plaza (14). An identifying tag is located in the vehicle (12) which communicates with a toll collection unit (16) located at the toll plaza. In addition to communication means, the toll collection unit includes one or more optical sensors such as video cameras (32 - 42), and a means for extracting and recognizing (44) one or more optically perceptible features from the images produced by the optical sensors. The information obtained by the optical sensors is used to verify or demonstrate a reasonable probability of correctness of the information obtained from the tag and to provide the location of the vehicle within the toll plaza. If an error is detected in the information from the tag, or if no information is received from the tag, the optically obtained information is stored in the toll collection unit's memory to provide a record of incomplete or erroneous toll transactions for later action.

Description

AUTOMATIC VEHICLE IDENTIFICATION TOLL COLLECTION SYSTEM WITH OPTICAL VEHICLE RECOGNITION

BACKGROUND OF THE INVENTION

The present invention relates generally to automatic toll collection systems in which the toll transaction is automatically performed with a device located aboard a vehicle as the vehicle approaches a toll plaza and, more specifically, to an improvement for such a system that verifies the identity of the device participating in the transaction by optically inspecting the vehicle.

Automatic revenue collection systems are used in toll collection systems on thoroughfares such as highways and bridges. Such systems generally comprise a collection unit located at a point of entry or exit. A patron's vehicle carries an electronic device, sometimes called a tag, that automatically communicates with the toll collection unit to effect a toll or fare transaction when the tag approaches the collection unit. The communication link commonly uses radio frequency transmissions, although some systems may use optical or acoustical transmissions. The tag may include a microprocessor and memory, which is used for storing information identifying the tag. This information may include an account number or electronic serial number.

The patron may establish an account for payment of the vehicle's tolls with the authority conducting toll collections. The balance of the account is adjusted each time the tag effects a transaction. The balance may be stored in the tag memory or in another memory accessible to the collection unit. The patron may pre-purchase either a specific number of tolls or a specific monetary amount that is then credited to his account. The toll amount involved in the transaction may be different for different classes of vehicles, and vehicles may be classified by axle count, weight or other criteria.

.. Each time a vehicle carrying a tag approaches the toll collection unit, the tag identifies itself by transmitting its serial number or other identifying designation. The tag may identify itself in response to an interrogation from the collection unit or other condition. The collection unit may verify the serial number by matching it to a list of valid serial numbers. In response to receiving a valid serial number, the collection unit can reduce the account balance to effect the toll transaction. Systems of this type are sometimes called Automatic Vehicle Identification (AVI) systems. The effects of crosstalk, reflections, antennae lobe distortion, and moving metallic objects reduce the reliability of AVI toll collection systems in toll plazas that have multiple lanes. The problems are most evident when several tags attempt to communicate with the collection unit at approximately the same time. Toll transactions intended for one tag and its associated account may be incorrectly attributed to a different tag. Such an occurrence may cause incorrect toll amounts to be debited from account balances if the vehicles are not of the same class.

Practitioners have attempted to alleviate these problems by performing triangulation calculations and amplitude-comparisons on the various signals to determine their point of origin. However, none of these methods is sufficiently accurate. It would be highly desirable to provide a method for verifying the identity of the tag versus its relative position in the toll lane with which the collection unit is communicating without relying on the use of signals susceptible to the effects described above. Furthermore, unlike conventional toll systems, vehicles in an AVI system do not stop at toll plazas. Tolls are paid "on the fly" as the vehicles pass through, thereby improving traffic flow. In conventional toll systems, vehicles that fail to pay the toll can be easily identified and detained by enforcement personnel. However, in an AVI system the advantage of improved traffic flow is lost if vehicles are detained at the toll plaza in order to remedy non-payment of a toll. On the other hand, toll collection systems cannot ignore non-payment and mispayment of tolls. It would be desirable to provide a method for remedying non-payment of a toll without obstructing traffic flow through the toll plaza in an AVI system. Optical recognition systems have also been used for AVI to detect certain visual characteristics of a vehicle, for example, license plate numbers. However, inherent in a optical recognition system of this type is a large universe of possibilities. In the simple case of a license plate, there may be on the order of eight or nine alphanumeric characters. The process becomes even more complicated if detection of other visible characteristics is required, such as vehicle type. In order to distinguish one vehicle from another using exclusively optical data requires complete and accurate registration of the information, e.g., the entire license number. Also of concern is the frequent occurrence of partial obstruction of characters on a license plate by the vehicle's bumper, a trailer hitch, or by dirt. With such a critical need for accuracy, detection on-the-fly is virtually impossible, so that vehicles must proceed at a sufficiently slow speed to assure full registration. Thus, many of the advantages of AVI may be lost in an optical system.

These problems and deficiencies are clearly felt in the art and are solved by the present invention in the manner described below.

SUMMARY OF THE INVENTION

The present invention verifies the identity of a tag in an AVI toll collection system by optically inspecting vehicles proceeding through a toll plaza. The invention comprises a tag located in the vehicle, a toll collection unit located at the toll plaza, one or more optical sensors such as video cameras, and a means for extracting and recognizing one or more optically perceptible features from the images produced by the optical sensors.

The optical sensors monitor vehicles passing through the toll plaza and may operate continuously or in response to detection of an approaching vehicle. In a toll plaza that has multiple lanes, each lane should be monitored . Although an exemplary embodiment monitors and communicates with approaching vehicles, it should be noted that the optical monitoring and communication may occur not only when the vehicle is approaching but also when the vehicle is in other positions with respect to the toll plaza, such as when the vehicle is in or moving away from the toll plaza, the only limitation being that the vehicle must be within optical range of the optical sensor. The feature recognition means produces information that uniquely identifies the vehicle or identifies it as belonging to a class of vehicles. Any image recognition hardware or software known in the art may be used to perform this process, including those using fuzzy logic. The identifying information may correspond to any suitable optically perceptible feature that is unique to a vehicle or class of vehicles, including a license plate number or portion thereof, vehicle body type, vehicle axle count, and vehicle color. The toll collection unit transmits an interrogation signal. The tag responds to the interrogation signal by identifying information comprising information representative of an optically perceptible feature unique to that vehicle or class. The collection unit receives this identifying information and compares it to the information extracted by the feature recognition means. If a match, or the reasonable probability of a match, is found, the toll transaction is permitted to proceed. If they do not correspond, or the probability of match is low, it is likely that the communication signals have been corrupted or that the tag transmitting the information is in a different lane from that which the optical sensor is monitoring. In such an error situation, the collection unit may repeat the interrogation and/or may store the identifying information for later action by the collection authority.

The toll transaction comprises altering the patron's account. In embodiments in which the account balance is stored in memory in the tag, the tag transmits the account balance to the collection unit, which debits the toll amount and transmits the new balance back to the tag. The tag then stores the debited balance in its memory. Alternatively, a processor in the tag may debit the account balance. In still other embodiments, the account balance is not stored in the tag but rather is stored in a computer memory connected to the collection unit.

If the account balance is insufficient to pay for a toll, the collection unit stores the extracted information in a memory for later analysis by enforcement personnel. In addition, if a vehicle approaches the toll plaza and the collection unit receives no response to its interrogation signal, the collection unit stores the extracted information in the memory for later analysis. Similarly, if the tag responds with an invalid serial number or other identifying information, the collection unit may store the extracted information.

The foregoing, together with other features and advantages of the present invention, will become more apparent when referring to the following specification, claims, and accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

For a more complete understanding of the present invention, reference is now made to the following detailed description of the embodiments illustrated in the accompanying drawings, wherein:

Figure 1 illustrates a tag located in a vehicle;

Figure 2 is a schematic block diagram of a vehicle approaching a toll plaza, showing the operation of the present invention.

Figure 3 is a schematic block diagram of a typical tag; Figure 4 is a schematic block diagram of the collection unit of the present invention; and

Figure 5 is a flow diagram of a method for verifying the identity of a tag.

DESCRIPTION OF THE PREFERRED EMBODIMENT

In Figures 1 and 2, a tag 10 is disposed in a vehicle 12. The vehicle user may use tag 10 in an AVI toll collection system to pay tolls for vehicle 12. A toll collection unit 16 continuously transmits an interrogation signal 17. When vehicle 12 approaches a toll plaza 14 to the position shown in broken line, tag 10 is activated by interrogation signal 17. If tag 10 is valid and functioning properly, it responds by transmitting identifying information 18 to collection unit 16. Identifying information 18 comprises information representative of a set of one or more optically detectable characteristics or features (not shown) of vehicle 12. For example, identifying information 18 may comprise several or all of the alphanumeric characters on the license plate (not shown) of vehicle 12. Information 18 may also include a serial number and any other information commonly transmitted by tags in AVI systems.

Toll plaza 14 has multiple lanes 20, 22, 24, 26, and 28, and 30, which are observed by multiple video cameras 32, 34, 36, 38, 40, and 42, respectively. Although video cameras 32-42 are shown mounted on toll plaza 14 above the center of each lane, they may be mounted in any other suitable location such as on the apron strips between adjacent lanes. As illustrated, video camera 32 captures video images of vehicle 12 as it approaches toll plaza 14 in lane 20. It should be noted that optical verification by the video cameras can also occur while the vehicle is in the toll plaza or as it leaves the toll plaza, the important criteria being that the vehicle is within optical range of the video cameras. This versatility allows the inventive system to be adapted to variations in license plate placement. For example, some states do not require a license plate on the front of the vehicle.

An identifier circuit 44 receives the video images and extracts information corresponding to one or more of the features discussed above. Collection unit 16 compares the portion of the extracted information corresponding to a feature to each portion of information 18 that corresponds to a feature. A transaction may occur only if the comparison results in a reasonable probability of a match.

If none of the features demonstrate a probable match, it is likely that signal reflections, fading, or crosstalk from similar signals transmitted by tags in vehicles (not shown) in other lanes would impair the accuracy of a subsequent toll transaction. Thus, collection unit 16 does not attempt to perform a toll transaction. Collection unit 16 may transmit a new request to tag 10 to which tag 10 would respond with identifying information 18 in the same manner as it should to the initial interrogation signal 17. Alternatively, or if after transmitting one or more new requests, collection unit 16 does not detect a probable match, it may discontinue the attempts and store the extracted information in a memory for later analysis by enforcement personnel. Enforcement personnel can display the information on a computer monitor (not shown) or otherwise use the information to locate the vehicle and its owner. In addition, the file that is created provides evidence that may be used when the individual whose vehicle has committed the violation is brought before the court or other enforcement authority, as may be needed for ultimate collection of the unpaid tolls. If information 18 is a likely match to the extracted information, a toll transaction occurs. Tag 10 transmits a signal 46 value equal to its stored account balance to collection unit 16. Collection unit 16 debits the account balance by the, amount of a toll and transmits a signal 48 corresponding to the debited balance to tag 10, which stores the debited balance. If the stored account balance is insufficient to cover a toll, collection unit 16 stores the extracted information in a memory for later analysis by -1-

enforcement personnel. Similarly, if collection unit 16 receives no response to interrogation signal 17 or an invalid response, such as an unrecognizable serial number, it may store the extracted information in a memory for later analysis. In Figure 3, tag 10 is illustrated in further detail. Tag 10 may comprise a tag unit radio frequency (RF) transceiver 50 having an antenna 52. A microprocessor 54 and associated memory 56 are connected to transceiver 50.

In Figure 4, toll collection unit 16 and identifier circuit 44 are shown in further detail. A video multiplexor or selector 58 receives multiple video inputs from video cameras 32-42. A suitable digitizer 60 converts the output of selector 58 to digital images. A feature recognizer 62 receives each image and extracts one or more features, which may be any characteristic of vehicle 12 that feature recognizer 62 is capable of identifying. For example, optical character recognition is well-known in the art, and a person skilled in the art can readily produce software for converting an image of a license plate into alphanumeric characters, e.g., encoded in ASCII format, that may be manipulated by a computer. Although it is preferred that feature recognizer 62 produce a complete license number, any portion of a license number that sufficiently distinguishes vehicle 12 from vehicles in other lanes is suitable.

According to the present invention, the use of video information to verify the position of a vehicle within the toll plaza is greatly simplified as compared to optical recognition systems of the prior art, such as license plate readers. The identifying information 18 which is transmitted by the tag 10 includes optically detectable characteristics, such as the type of vehicle, the state in which the vehicle is registered and the license plate number.and vehicle color, among others. With this preliminary information, the search by the optical system will limited at the outset to a much smaller universe. For example, if the tag transmits a signal that the vehicle corresponding to the transmitted information is a blue sedan with California license plates "123 ABC", the optical system will then be required only to extract sufficient information from the video signal to confirm a portion of the transmitted information. This can be as simple as registering a blue sedan or extracting a few characters, such as "23 A", from the license plate. The optical system will not be trying to determine the vehicle type, e.g., car, truck, recreational vehicle, nor will it be required to register the entire license plate number to determine the issuing state or identify the vehicle without prior knowledge of what the information should be. In fact, an absolute match of the extracted information is not required. The optical system need only demonstrate a reasonable probability of a match between the optical information and the transmitted information to provide the desired verification.

A microprocessor 64 and associated memory 66 are connected to feature recognizer 62 for controlling the fare transactions, communication, feature recognition, and video source selection. It should be recognized that several microprocessors or microcontrollers performing different ones of these tasks or even in parallel on the same task would also be suitable. A toll unit RF transceiver circuit 68 and associated antenna 70 are connected to microprocessor 64. Transceiver circuits 50 and 68 and respective microprocessors 54 and

64 allow bidirectional communication of information. Transceiver circuits

50 and 68 may use any suitable modulation method, such as amplitude modulation, frequency modulation, phase modulation, or impedance modulation, and the transmitted information may be encoded using any suitable method.

The operation of the present invention is illustrated in Figure 5. At step 72, toll collection unit 16 transmits interrogation signal 17. Tag 10 may respond by transmitting identifying information 18 at step 74. At step 76, collection unit 16 monitors for such a response. Whether or not a response is received, feature recognizer 62 extracts feature information from video images of vehicle 12 at steps 78 and 80, respectively. Microprocessor 64 controls video source selection via selector 58.

If no response is received or if the response does not comprise valid identifying information 18, as discussed above, microprocessor 64 stores the extracted feature information in memory 66 at step 82. If valid identifying information 18 is received in response to interrogation signal 17, microprocessor 64 compares the portion of identifying information 18 that corresponds to features with the extracted feature information at step 84.

At step 86, microprocessor 64 performs a toll transaction with tag 10 only if the comparison demonstrates a reasonable probability of a match.

If the extracted feature does not match any feature in the set of features represented by information 18, or shows a low probability of a match, microprocessor 64 stores the extracted feature information in memory 66 at step 82. As discussed above, the features may not match as a result of corrupted communications rather than a patron's willful non-payment of the toll. Thus, microprocessor 64 may attempt additional interrogations (not shown in Fig. 5) before storing the extracted feature information. If communications were only momentarily corrupted, additional interrogations may result in a sufficiently verifying match. This verification procedure thus not only records information that can be used to identify vehicles that have failed to pay the toll, but also improves the reliability of AVI systems by verifying with a reasonable probability that the tag from which the toll collection unit collects the toll is located in a vehicle that is approaching the toll plaza.

Obviously, other embodiments and modifications of the present invention will occur readily to those of ordinary skill in the art in view of these teachings. Therefore, this invention is to be limited only by the following claims, which include all such other embodiments and modifications when viewed in conjunction with the above specification and accompanying drawings.

I CLAIM:

Claims

1. A toll collection system, comprising: a tag disposed in a vehicle, said tag having memory for storing unique information representative of unique identifying characteristics of said vehicle and bidirectional communication means for transmitting said unique information, said unique identifying characteristics including a known optically perceptible characteristic; a collection unit having bidirectional communication means for communicating with said tag unit, said collection unit receiving said unique information transmitted by said tag; optical sensor means for producing optical information corresponding to said vehicle when said vehicle is within optical range of a toll plaza; feature recognition means for extracting a detected optically perceptible characteristic corresponding to said vehicle from said optical information; and said collection unit having processor means for performing a toll transaction with said tag and for comparing said detected optically perceptible characteristic with said known optically perceptible characteristic to verify an identity and position of said vehicle within said toll plaza.

2. The toll collection system described in Claim 1 , wherein said unique identifying characteristics include at least a portion of a license plate number.

3. The toll collection system described in Claim 1 , wherein said optical sensor means comprises a video camera.

4. The toll collection system described in Claim 1 , wherein: said toll transaction comprises performing an arithmetic operation on a stored numerical value associated with said tag; and said collection unit stores said detected optically perceptible characteristic when said numerical value is within a predetermined range after said toll transaction is performed.

5. The toll collection system described in Claim 4, wherein: said toll transaction comprises subtracting a monetary amount from said stored numerical value; and said collection unit stores said detected optically perceptible characteristic when said numerical value is less than zero after said toll transaction is performed.

6. The toll collection system described in Claim 1 , wherein said processor stores said detected optically perceptible characteristic when said collection unit does not receive a response to said interrogation from said tag.

7. The toll collection system described in Claim 1 , wherein said processor stores said detected optically perceptible characteristic when said collection unit does not receive valid unique information in response to said interrogation.

8. The toll collection system described in Claim 1 , wherein: said vehicle has one of a plurality of classes, and each class has a corresponding type of toll collection transaction; said unique information comprises said class of said vehicle; said detected optically perceptible characteristic comprises said class of said vehicle; and said processor means does not perform said toll transaction unless said class of said unique information is the same as said class of said detected optically perceptible characteristic.

9. The toll collection system described in Claim 8, wherein said processor stores said detected optically perceptible characteristic when said collection unit does not receive a response to said interrogation from said tag.

10. The toll collection system described in Claim 9, wherein said processor stores said detected optically perceptible characteristic when said collection unit does not receive valid unique information in response to said interrogation.

1 1. The toll collection system described in Claim 10, wherein: said toll transaction comprises performing an arithmetic operation on a stored numerical value associated with said tag; and said collection unit stores said detected optically perceptible characteristic when said numerical value is within a predetermined range after said transaction is performed.

12. A method for verifying the identity and position of a vehicle within optical range of a toll plaza, said vehicle having an electronic tag and said toll plaza having a collection unit, said method comprising the steps of: transmitting identifying information from said tag to said collection unit; gathering optically perceptible information from said vehicle when said vehicle is within optical range of said toll plaza; extracting at least one feature from said optically perceptible information; comparing said extracted feature to identifying information received by said collection unit; and performing a toll transaction if said identifying information corresponds to said extracted feature.

13. The method for verifying the identity of a vehicle within optical range of a toll plaza described in Claim 12, further comprising the step of storing said extracted feature if said collection unit does not receive said identifying information.

14. The method for verifying the identity of a vehicle within optical range of a toll plaza described in Claim 12, further comprising the step of storing said extracted feature if said collection unit receives invalid identifying information.

15. The method for verifying the identity of a vehicle within optical range of a toll plaza described in Claim 12, wherein: said toll transaction comprises the step of performing an arithmetic operation on a stored numerical value; said method further comprising the step of storing said extracted feature if said numerical value is within a predetermined range after said arithmetic operation is performed.

16. The method for verifying the identity of a vehicle within optical range of a toll plaza described in Claim 15, wherein said arithmetic operation is subtraction and said predetermined range is equal to the set of all negative numbers.

17. The method for verifying the identity of a vehicle within optical range of a toll plaza described in Claim 16, wherein said numerical value is stored in said tag.