JPH06307827A - Car type judging device - Google Patents

Abstract

PURPOSE:To make the discrimination of car type, particularly between a large- sized special car, large size car, and a bus. CONSTITUTION:When a car passes between a group 12 of car front position sensors and a group 13 of silhoet sensors, the output signals are emitted from these sensor groups 12, 13 and received, and the overhang amount (h) of the car is determined from an overhang amount sensing circuit 17 while the total number (k) of shafts is determined by a shafts counting circuit 18, and the car type is judged by a microcomputer 19 on the basis of combination of these overhang amount (h) and total number (k) of shafts.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vehicle type discriminating apparatus for discriminating a vehicle type such as an automobile entering an expressway or a vehicle passing through a toll collection point.

[0002]

2. Description of the Related Art An automatic entrance system for discriminating a vehicle type such as an automobile entering an expressway, coding the vehicle type, recording the same on a pass ticket, and issuing a ticket is disclosed in, for example, Japanese Unexamined Patent Application Publication No. 5 (1999) -58242.
6-79397.

At the core of such a system is a tread device that measures the number of axles and the passing position of the wheels of the vehicle. In addition to this, a height sensor that measures the height of the vehicle in three stages is provided. Has been. Therefore, this system
The vehicle is classified into light vehicle, ordinary vehicle, medium-sized vehicle, large-sized vehicle, and large-sized special vehicle based on the measurement results of the tread device and the height sensor.

However, in such a system, it is difficult to discriminate a vehicle having a special shape and specification, and in order to cope with this, an auxiliary sensor, for example, a sensor for measuring only the length of a light vehicle is added. ing.

Further, for example, among large-sized transport vehicles having four or more axes, a single-body vehicle may not be determined to be a large special vehicle, but may be simply determined to be large. This vehicle body has, for example, a trailer that is extremely similar in shape and size, and it is required to discriminate them mechanically.

Under the present circumstances of such vehicle type discrimination, as a technique for discriminating between a large special vehicle and a large vehicle, Japanese Unexamined Patent Publication No. 59-59
As described in Japanese Unexamined Patent Publication No. 160299, the continuity of a predetermined height position in the vehicle is detected, and the large special vehicle and the large vehicle are discriminated based on the combination of the tread width and the number of axes. Specifically, for example, by arranging an optical sensor at a height of 1500 mm above the ground and checking the intermittentness of the optical sensor between the first axis and the second axis in the vehicle, for example, with the trailer shown in FIG. 14 (b), If there is an interruption, there is continuity, and if it is a single car body shown in Fig. 6 (c), there is continuity, so it can be determined.

In such a vehicle type discrimination, a trailer is a single-body vehicle, which is one of the only exceptional vehicles in a vehicle body with four or more axles, which is one of the conditions for large special vehicles. Can be distinguished from.

However, in the four-axis single vehicle body shown in FIG. 1 (a), the driver's cab and the cargo bed are separated from each other, and the trailer is not always distinguishable. Another problem with large vehicles is the bus vehicle. That is, it is necessary to distinguish a route bus as a large vehicle and a tourist bus as a large special vehicle, but in reality it is difficult to make this distinction mechanically, and the buses are collectively considered as unknown vehicles. It is desired to process. However, it is difficult to distinguish between a bus and a truck by the technique for distinguishing between a large special vehicle and a large vehicle based on the combination of the tread width and the number of axes as described above.

On the other hand, as a technique which has already been put into practical use for a vehicle type discriminating device, as shown in FIG. 15, a tread plate device 3 is provided in a passage 2 of a vehicle 1 to measure an axle distance and an axle number, and a vehicle separator. 4a and 4b are arranged on both sides of the aisle 2 to separate the vehicles 1 one by one, and vehicle height detectors 5a and 5b are provided above the vehicle separators 4a and 4b to detect the vehicle height. The length detectors 6a and 6b are arranged on both sides of the passage 2 to detect the length of the vehicle 1, and the axle distance, the number of axles,
In some cases, the vehicle type of the vehicle 1 is determined based on data such as vehicle height and vehicle length.

The tread device 3 is provided with a large number of pressure sensors to detect the weight of the vehicle 1 as the amount of mechanical displacement, and measures the number of axles, the width of tires, and the distance between them. Further, the vehicle separators 4a and 4b have a plurality of optical sensors arranged in a line in the vertical direction.

However, in such an apparatus, since the vehicle type is determined based on rough data such as the axle spacing and the number of axles, and the spatial shape data of the vehicle 1 is not used, the vehicle type is often erroneously determined. .

Further, in the mechanical measurement by the tread device 1,
In addition to taking up space for the discriminating device, the tread device 1 is solidly buried, resulting in high cost, and moreover, failure is likely to occur and reliability is low. The tread device 1 is easily damaged due to mechanical fatigue, vibration, dust, moisture, and high temperature.

[0013]

As described above, it is difficult to distinguish between a large-sized special vehicle, a large-sized vehicle, and a bus by the conventional method when determining the vehicle type. In addition, the vehicle type is often erroneously discriminated, and in the measurement of the tread device 1, the space is increased and the cost is increased, and further, the breakdown easily occurs and the reliability is low.

Therefore, it is an object of the present invention to provide a vehicle type identification device capable of identifying a large special vehicle, a large vehicle, and a bus in particular in vehicle type identification. Another object of the present invention is to provide a vehicle type discriminating device which is highly reliable and is capable of stably and accurately discriminating a vehicle type. It is another object of the present invention to provide a vehicle type identification device that can improve the accuracy of the determination even if the vehicle type identification standard is changed.

[0015]

According to a first aspect of the present invention, a first sensor group is formed by arranging a plurality of sensors in a vertical direction with respect to a vehicle, and a plurality of sensors is arranged in a traveling direction of the vehicle. A second sensor group, a vehicle analysis means for obtaining the length and the number of axes from the front of the vehicle to a predetermined position based on the detection signals of the first and second sensor groups, and the length obtained by the vehicle analysis means. A vehicle type discriminating apparatus, which is provided with a discriminating means for discriminating a vehicle type of a vehicle based on a combination of the number of axes, and which is intended to achieve the above object.

According to a second aspect of the present invention, there is provided a silhouette creating means for sequentially receiving signals from the first and second sensor groups and creating silhouette data for the vehicle. According to claim 3, a sensor group in which a plurality of sensors are arranged vertically with respect to the vehicle, a silhouette creating means for sequentially receiving detection signals of the sensor group and creating silhouette data for the vehicle, and the silhouette Window processing means for obtaining the sum of the light and dark values in each window set for the silhouette data created by the creating means, and the vehicle type of the vehicle based on the light and dark values in each window found by this window processing means A vehicle type discriminating device that includes a discriminating means for discriminating and tries to achieve the above object.

According to claim 4, a first sensor group formed by arranging a plurality of sensors in a vertical direction with respect to the vehicle, and a second sensor group formed by arranging a plurality of sensors in a traveling direction of the vehicle, A silhouette creating unit that sequentially receives the detection signals of the first and second sensor groups to create silhouette data for a vehicle, and the silhouette data created by the silhouette creating unit and the reference silhouette data are pattern-matched to obtain a vehicle silhouette. Based on the discrimination means for discriminating the vehicle type, if the vehicle type is the same, silhouette data is added to the reference silhouette data, and if it is a different vehicle type, the silhouette data is subtracted from the reference silhouette data to obtain the reference silhouette data. And a learning means for performing learning and updating of the vehicle type discriminating apparatus for achieving the above object.

According to the present invention, the silhouette creating means for creating silhouette data for a vehicle sequentially receives the detection signals of the sensor group output according to the traveling speed of the vehicle and holds the state of the detection signal. It has a holding circuit and a silhouette memory that sequentially stores the detection signals held by the information holding circuit to create silhouette data for the vehicle.

[0019]

According to the present invention, when the vehicle passes through the first and second sensor groups, the vehicle analysis means receives the detection signals of these sensor groups at this time and the length from the front of the vehicle to the predetermined position is determined by the vehicle analysis means. And the number of axes are obtained, and the vehicle type of the vehicle is determined by the determining means based on the combination of the length and the number of axes.

In this case, according to the second aspect, the signals of the first and second sensor groups are sequentially received and the silhouette data for the vehicle is created. According to the third aspect, when the vehicle passes through the sensor group, the silhouette generation means generates the silhouette data for the vehicle by sequentially receiving the detection signals of the sensor group at this time. For this silhouette data,
Each window is set by the window processing means, the sum of the light and dark values in the window is obtained, and the vehicle type of the vehicle is judged by the judging means based on the light and dark values.

According to the fourth aspect, when the vehicle passes through the first and second sensor groups, the silhouette creating means creates the silhouette data for the vehicle by sequentially receiving the detection signals from these sensor groups. Then, the silhouette data and the reference silhouette data are pattern-matched by the discriminating means to discriminate the vehicle type of the vehicle. Based on the discrimination result, the learning means adds the silhouette data to the reference silhouette data if the vehicle type is the same. If the vehicle type is different, the silhouette data is subtracted from the reference silhouette data to learn and update the reference silhouette data.

According to the fifth aspect of the invention, the silhouette creating means sequentially receives the detection signals of the first sensor group, holds the state of the detection signals, and sequentially stores the held detection signals in the silhouette memory. The silhouette data for is created.

[0023]

【Example】

(1) Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of a vehicle type identification device. A tire passage sensor 11, a vehicle front position sensor group 12, and a silhouette sensor group 13 are arranged in a passage 10 through which a vehicle passes.

As shown in FIG. 2, the tire passage sensor 11 is composed of an optical sensor in which a light projector 11a and a light receiver 11b are combined on both sides of the passage 10 and is arranged at a position corresponding to the tire position of the vehicle, and its light shielding is performed. Is to detect the passage of tires.

The vehicle front position sensor group 12 obtains data for measuring the position of the front of the vehicle. Each optical sensor is a combination of a plurality of light projectors 121a-12na and light receivers 121b-12nb on both sides of the passage 10. Made of. The sensor group 12 is arranged along the traveling direction of the vehicle at a position approximately corresponding to the number plate mounting position of the vehicle.

The vehicle front position sensor group 12 includes, for example, a tire passage sensor 11 and a sensor (121) in FIG.
a, 121b), the distance Lo between the sensors can be set to 500 mm, the distance between the sensors ΔL can be set to 50 mm, and the number of sensors n can be set to 40.
From b) to the sensors (12na, 12nb), the light shielding state is sequentially changed.

The silhouette sensor group 13 is used for entering a vehicle,
Detecting exit and separation of vehicles, a plurality of light emitters 131a to 13na and light receivers 131b to both sides of the passage 10 are detected.
It consists of each sensor that is a combination of 13 nb. The sensor group 13 is arranged vertically with respect to the vehicle.

A vehicle separation circuit 14 is connected to the output terminal of the silhouette sensor group 13, and is used for evaluating the continuity of the vehicle from the vehicle separation circuit 14 and outputs a vehicle passage signal T for notifying that the vehicle is passing. It has become a thing.

A signal conversion circuit 15 is connected to the output terminal of the tire passage sensor 11. The signal conversion circuit 15 receives the output signal of the tire passage sensor 11, converts it into a front edge edge passage signal P1 of the tire when the light passing is changed to the light blocking, and outputs it when the light passing is changed to the light passing. It has a function of converting and outputting to the trailing edge passing signal P2.

A front position detection circuit 16 is connected to the output terminal of the vehicle whole surface position sensor group 12. The front position detection circuit 16 receives the output signal of the vehicle front position sensor group 12 to detect the front position of the vehicle and outputs the front position signal F to the overhang amount detection circuit 17.
It has a function to send to.

The overhang amount detecting circuit 17 receives the vehicle passage signal T to judge the start of the vehicle entry, and then the value f1 of the front position signal F when the first tire leading edge passage signal P1 is received. , A function of calculating the overhang amount hh = (f1 + f2) / 2 (1) by the value f2 of the front position signal F when the first tire trailing edge passing signal P2 is received after the vehicle enters is doing.

The axis number counting circuit 18 has a function of receiving the tire front edge edge passage signal P1 and the tire trailing edge edge passage signal P2 and obtaining the total number of axes k until the vehicle departs.
The microcomputer 19 has a function of discriminating a vehicle as a large vehicle, a large special vehicle, or an unknown vehicle based on the combination of the overhang amount h and the total number of axes k. The conditions for the discrimination are as follows. .

Condition 1 h <hth and k <4 Large vehicle Condition 2 h <hth and k ≧ 4 Large special vehicle Condition 3 h ≧ hth and k <4 Unknown vehicle Condition 4 h ≧ hth and k = 4 Large vehicle Condition 5 h ≧ hth and k> 4 Large special vehicle Note that hth is set to 1500 mm, for example.

Here, the condition 1 corresponds to the case where the single body truck shown in FIG. 8 (c) has three or less axes. For trailers and triaxial vehicles, this condition 1 is also satisfied, and most large vehicles correspond to this. Condition 2 corresponds to most of trailers having four or more axles or a single vehicle having a towing vehicle.

Condition 3 corresponds to the tank truck shown in FIG. 8 (d) and the bus shown in FIG. 8 (e). Condition 4 corresponds to a 4-axis single-body truck. The condition 5 corresponds to a large special vehicle other than the above conditions 1 to 4.

By the way, to explain the grounds for these conditions 1 to 5, the trailer is characterized in that its driver's cab separates the cargo bed and moves independently. For this reason, the first axis supports the driver's cab and front It is not far away.

On the other hand, the four-axis single-body truck has a shape similar to that of a tank truck, and the first axis exists at a position distant from the front. Therefore, the overhang amount h is longer than that of a trailer and a general three-axis single vehicle body.

Since the bus has an entrance / exit near the front, the overhang amount h from the front to the first axis is longer than that of a general truck or trailer. In addition, four-axle vehicles such as buses and tank trucks are rare. Based on the above grounds.

Next, the operation of the apparatus configured as described above will be described. When the vehicle passes through the tire passage sensor 11 and the vehicle front position sensor group 12 from the silhouette sensor group 13, the silhouette sensor group 13 determines the sensor groups (131a to 13na, 131b to 13n) according to the height of the vehicle.
b) is in a light-shielded state, and its output signal is the vehicle separation circuit 1
4 and the circuit 14 outputs a vehicle passage signal T indicating that the vehicle is passing. This vehicle passage signal T is sent to the signal conversion circuit 15, the front position detection circuit 16, the overhang amount detection circuit 17, the axis number counting circuit 18, and the microcomputer 19.

Further, the tire passage sensor 11 detects the passage of the tire due to the light shielding due to the passage of the vehicle and outputs the detection signal. This detection signal is sent to the signal conversion circuit 15, so that the signal conversion circuit 15 outputs the front edge passing signal P1 of the tire when the light passing is changed to the light blocking, and the light blocking is changed to the light passing. Then, the trailing edge passing signal P2 of the tire is output.

On the other hand, the vehicle front position sensor group 12 sequentially changes to the light-shielding state from the sensors (121a, 121b) to the sensors (12na, 12nb) as the vehicle passes. The output signal of the vehicle front position sensor group 12 at this time is sent to the front position detection circuit 16,
The detection circuit 16 detects the front position of the vehicle and sends the front position signal F to the overhang amount detection circuit 17.

The overhang amount detecting circuit 17 receives the vehicle passage signal T to judge the start of the vehicle entry, and then the value f1 of the front position signal F when receiving the first tire leading edge edge passage signal P1. , The above equation (1) is calculated from the value f2 of the front position signal F when the first tire trailing edge passing signal P2 is received after the vehicle enters, and the overhang amount h is obtained.

On the other hand, the axis number counting circuit 18 receives the tire front edge edge passage signal P1 and the tire trailing edge edge passage signal P2 to obtain the total number of axes k until the vehicle exits. Thus,
The microcomputer 19 receives the overhang amount h and the total number of axes k, and determines the vehicle as a large vehicle, a large special vehicle, or an unknown vehicle based on the above conditions 1 to 5.

As described above, in the first embodiment, when the vehicle passes through the vehicle front position sensor group 12 and the silhouette sensor group 13, the sensor groups 12, 1 at this time are passed.
Since the vehicle overhang amount h and the total number of axes k are obtained by receiving the output signals of No. 3 and the vehicle type is determined based on the combination of the overhang amount h and the total number of axes k, the vehicle type is particularly large. Can be distinguished from special vehicles, large vehicles, and unknown vehicles such as buses.

The first embodiment may be modified as follows. For example, the above hth is not one, but hth2 = 2
It is also possible to set a plurality of size determination criteria such as 000 mm and perform more detailed vehicle type classification.

The tire passage sensor 11 may use any one of the sensors in the silhouette sensor group 13. Further, the overhang amount h may be set between the tire front edge position and the front.

In addition, by adding a silhouette creating unit in another embodiment to be described later, inputting silhouette data from the silhouette creating unit to the microcomputer 19 enables comprehensive vehicle type discrimination. (2) Next, a second embodiment of the present invention will be described.

FIG. 3 is a block diagram of the vehicle type identification device. A silhouette sensor group 21 is provided in the passage 20. The silhouette sensor group 21 is provided with a projector group 21a and a light receiver group 21b on both sides of the passage 20, respectively.

Of these, the projector group 21a is formed by arranging a plurality of projectors 211a to 21na in the vertical direction of the vehicle, and the projector group 21b is composed of these projectors 211a to 21na.
A plurality of light receivers 211 are provided at positions corresponding to 21 na.
b to 21 nb are arranged. Each sensor is formed by a combination of the light projectors 211a to 21na and the light receivers 211b to 21nb.

The silhouette creating unit 2 is connected to the output terminal of the silhouette sensor group 21 via the silhouette input unit 22.
3 is connected. A vehicle speed sensor 24 is provided on the side of the passage 20, and an image of a traveling vehicle is picked up by the vehicle speed sensor 24 and a video signal thereof is sent to a vehicle speed detector 25.

The vehicle speed detecting section 25 has a function of obtaining the amount of movement of the vehicle in the left-right direction from the video signal from the vehicle speed sensor 24 and obtaining the speed of the vehicle from the amount of movement. The silhouette creating unit 23 is the silhouette input unit 2
2 has a function of receiving detection signals from the silhouette sensor group 21 sequentially input through 2 and creating silhouette data of the entire vehicle viewed from the lateral direction. In creating this silhouette data, the vehicle speed from the vehicle speed detection unit 25 It has a function of correcting the expansion and contraction of the silhouette based on the vehicle speed.

The window processing unit 26 receives the silhouette data created by the silhouette creating unit 23, and with respect to this silhouette data, as shown in FIG. It has a function of generating W5 and obtaining total sums N1 to N5 of the number of pixels of silhouette portions in these windows W1 to W5. Note that FIG. 4 shows an example of composition of silhouette data A, B, and C for three types of vehicles.

The fuzzy logic unit 27 uses each window W1.
It has a function of receiving the sum total N1 to N5 of the numbers of pixels in W5 to W5 and executing a fuzzy calculation to determine the vehicle type of the vehicle. Specifically, the fuzzy logic unit 27 divides the total sums N1 to N5 of the numbers of pixels into a plurality of stages of "small, small, slightly small, medium, slightly large, large, very large". Evaluate by membership function. And
As shown in FIG. 5, the rules expressed by words such as “minor, few, ...” in these membership functions are shown in FIG.
It is evaluated by the size of ~ n. That is, if the rule 1 total pixel number N1 is large, N2 is small or small, N3 is very large, and N5 is small, the vehicle type is A.

Rule 2 If the total number of pixels N1 is large, N2 is small, N4 is very large, and N5 is large, the vehicle type is B. Rule 3 Total number of pixels N1 is small, N2 is large, N3
If there are many, N4 is many, and N5 is few, it is the vehicle model C. Are examples of rules.

Next, the operation of the apparatus configured as described above will be described. When the vehicle travels in the passage 20 and passes between the light projector group 21a and the light receiver group 21b of the silhouette sensor group 21, the light to the light receiver group 21b is blocked. At this time, the silhouette sensor group 21 includes the light receivers 2 that are shielded from light.
A detection signal corresponding to 11b to 21nb is output. This detection signal is sent to the silhouette creating unit 23 through the silhouette input unit 22.

Since the vehicle passes in front of the vehicle speed sensor 24, the vehicle speed sensor 24 picks up an image of the traveling vehicle and outputs a video signal thereof. This video signal is sent to the vehicle speed detection unit 25, so that the vehicle speed detection unit 25
Calculates the amount of movement of the vehicle in the left-right direction from the video signal and determines the speed of the vehicle from this amount of movement.

The silhouette creating section 23 sequentially receives the detection signals from the silhouette sensor group 21 to create the silhouette data of the entire vehicle viewed from the lateral direction, and the silhouette data is sent from the vehicle speed detecting section 25 to the vehicle speed. The expansion and contraction of the silhouette is corrected based on.

When the silhouette data is created in this way, this silhouette data is passed to the window processing unit 26. The window processing unit 26 receives the silhouette data and generates the windows W1 to W5 shown in FIG. 4 for the silhouette data. Next, the window processing unit 26 obtains the sum total N1 to N5 of the number of pixels of the silhouette portion in each of the windows W1 to W5. The sums N1 to N5 of the numbers of pixels are passed to the fuzzy logic unit 27.

When the fuzzy logic unit 27 receives the sums N1 to N5 of these pixel numbers, the sum N1 of these pixel numbers is received.
Fuzzy calculation is performed based on N5 to determine the vehicle type of the vehicle. For example, if the total number N1 of pixels in the window W1 is No shown in FIG. 5, mo% is obtained for “many N1” in rules 1 and 2 and 0% for “n1 is small” in rule 3. Will be acquired. After obtaining the score of each input condition in this way,
The fuzzy logic unit 27 uses the minimum value of the score in each rule as the accuracy of the result of that rule.

Next, the fuzzy logic unit 27 compares the magnitudes of the respective accuracies obtained for each rule, that is, for each vehicle type, and if one of the accuracies is larger than the other accuracies by a predetermined value or more, the accuracy of the accuracies is determined. The vehicle type is determined to be the vehicle type of the vehicle. Note that, for example, when the two vehicle model accuracies do not fall within a predetermined value, it is determined that it is difficult to determine the vehicle model.

As described above, in the second embodiment, the detection signals of the silhouette sensor group 21 are sequentially received to create the silhouette data for the vehicle, and the sum of the brightness values in the windows W1 to W5 is added to the silhouette data. Since the vehicle type of the vehicle is discriminated based on these brightness values, it is possible to discriminate each vehicle type with high reliability and stability.

The second embodiment may be modified as follows. For example, the silhouette sensor group 21 may use a line sensor or a video camera. Further, the silhouette data may be obtained from the difference image between the background image captured by the video camera and the image showing the vehicle.
Further, as the vehicle speed sensor 24, an ultrasonic sensor or a microwave sensor may be used. On the other hand, the fuzzy logic unit 29 may use a decision tree. (3) Next, a third embodiment of the present invention will be described.

The third embodiment is an improvement of the vehicle silhouette data creation, and can be applied to the silhouette data creation of the first and second embodiments, and can also be applied to various silhouette data creation. Is.

FIG. 6 is a block diagram of a silhouette data creation system in the vehicle type identification device. Passage 20 through which vehicle A travels
Is provided with a silhouette sensor Q as a first sensor group.

This silhouette sensor Q is a combination of a light projector Qa and a light receiver Qb, and has a passage 20.
Are erected on each side of. Of these, the projector Qa
Is a one-dimensional array of a plurality of light emitting elements in the vertical direction, and the light receiving device Qb is a one-dimensional array of a plurality of light receiving elements in the vertical direction at a position facing each light emitting element. ing.

As a result, the silhouette sensor Q includes the sensors q1 to qt in which the light projecting elements and the light receiving elements are combined,
It is an array of Q1 to Qn, and is divided into a lower sensor group composed of the sensors q1 to qt and an upper sensor group composed of the respective sensors Q1 to Qn above it from the bottom.

Each of the sensors q1 to qt and Q1 to Qn of the silhouette sensor Q outputs an H level (high level) signal in the light-shielded state and an L level (low level) signal in the light-transmitting state. Is output.

A vehicle passage position sensor E is provided as a second sensor group. This vehicle passage position sensor E is
A combination of the light projector Ea and the light receiver Eb is provided in a direction perpendicular to the light projector Qa and the light receiver Qb of the silhouette sensor Q, that is, in the traveling direction of the vehicle A and at the height position of the vehicle body.

Of these, the light emitter Ea is a plurality of light emitting elements arranged one-dimensionally in the horizontal direction, and the light receiver Eb has a plurality of light receiving elements one-dimensionally in the horizontal direction at positions facing each light emitting element. It has been arranged in a regular manner.

As a result, the vehicle passage position sensor E has the respective sensors E1 to Em in which the light emitting elements and the light receiving elements are combined.
It is an array of. Therefore, the vehicle passage position sensor E shields light in the order of the sensors E1 to Em when the vehicle A travels in the direction of the arrow A in the passage 20, and outputs the silhouette data sampling signal SA.

In FIG. 7, the sensors q1 to qt, Q1 in the silhouette sensor Q and the vehicle passage position sensor E are shown.
˜Qn and E1˜Em are shown. On the other hand, the light shielding information holding circuit 20a is connected to the lower sensor groups q1 to qt of the silhouette sensor Q, and the light transmission information holding circuit 20b is connected to the upper sensor groups Q1 to Qn.

The light shielding information holding circuit 20a takes in the output PA1 of the upper sensor groups Q1 to Qn in accordance with the sampling signal SA from the vehicle passage position sensor E, and outputs this output PA1.
It has a function of holding and outputting a signal level state of 1, and is connected to each of the sensors Q1 to Qn.

As shown in FIG. 8, the specific configuration is such that flip-flops F1 and F2 are connected in two stages, of which the OR gate a1 is connected to the D input terminal of the first-stage flip-flop F1 and the inverter is connected to the CLR terminal. Connect a2,
Moreover, the Qo output terminal of the flip-flop F1 is connected to the input terminal of the OR gate a1.

Then, the sampling signal SA is passed through the inverter a2 to the CL of the first-stage flip-flop F1.
It is input to the R terminal and supplied to the second-stage flip-flop F2.

The lower sensor group Q of the silhouette sensor Q
One of the outputs PA1 of 1 to Qn is supplied to the input terminal of the OR gate a1. The flip-flop F1 of the first stage includes the sensors q1 to q of the silhouette sensor Q.
A sampling trigger CLKi synchronized with the sensor timing signals of t, Q1 to Qn is input. Where C
LKi is a signal that is sufficiently faster than the sampling signal SA.

The light transmission information holding circuit 20b has a function of taking in the output PA2 of the lower sensor group q1 to qt in accordance with the sampling signal SA from the vehicle passage position sensor E and holding and outputting the signal level state of this output PA2. And is connected to each of the sensors q1 to qt.

As shown in FIG. 9, the concrete configuration is such that flip-flops F3 and F4 are connected in two stages, of which the OR gate a3 is connected to the D input terminal of the first-stage flip-flop F3 and the inverter is connected to the CLR terminal. Connect a4,
Further, the Qo output terminal of the flip-flop F1 is connected to one input terminal of the OR gate a3, and the inverter a5 is connected to the other input terminal of the OR gate a3.

Then, the sampling signal SA is supplied to the CL of the first-stage flip-flop F3 through the inverter a4.
It is input to the R terminal and is also supplied to the second-stage flip-flop F4.

Further, the lower sensor group q of the silhouette sensor Q
The output PA2 of 1 to qt is supplied to the other input terminal of the OR gate a3 through the inverter a5. The sampling trigger CLKi is input to the first-stage flip-flop F3.

These light-shielding and light-transmitting information holding circuits 20a,
The silhouette memory 20c is connected to the output terminal of 20b, and the holding outputs of the holding circuits 20a and 20b are sequentially stored in the silhouette memory 20c to create the silhouette data of the vehicle A.

Next, the operation of creating silhouette data in the apparatus configured as described above will be described. When the vehicle A travels on the passage A and enters between the vehicle passage position sensors E,
The sensors E1 to Em of the vehicle passage position sensor E are shielded from light according to the traveling direction of the vehicle A, that is, in the order of the sensors E1 to Em. As a result, the vehicle passage position sensor E is
The H-level sampling signal SA is output every time each of the sensors E1 to Em is shielded from light.

When the vehicle A enters between the silhouette sensors Q, the sensors q1 to q of the silhouette sensor Q
Of t and Q1 to Qn, the H-level signal is output from the light-shielded sensor.

On the other hand, the light shielding information holding circuit 20a takes in the output PA1 of the upper sensor groups Q1 to Qn in accordance with the sampling signal SA and holds and outputs the signal level state of this output PA1.

That is, the first-stage flip-flop F
1 is cleared by the sampling signal SA, and the signal level of the output terminal Qo is returned to L level. After that, the input to the flip-flop F1 is applied to each sensor Q1.
Since each output of ~ Qn and the output terminal Qo passes through the OR gate a1 and becomes a logical sum, each sensor Q1
When the output of ~ Qn becomes H level, the flip-flop F
The output of the No. 1 output terminal Qo is held at the H level.

Output terminal Qo of this flip-flop F1
Is sent to the second-stage flip-flop F2,
It is taken in at the rising edge of the sampling signal SA, and the output of its output terminal Qo is sent to the silhouette memory 20c.

At the same time, the light transmission information holding circuit 20b takes in the output PA2 of the upper sensor groups Q1 to Qn in accordance with the sampling signal SA and holds and outputs the signal level state of this output PA2. The operation is performed by the lower sensor group q
The outputs PA2 of 1 to qt are inverted by the inverter a5 and input to the OR gate a3, whereby the sensors q1 to q1 are output.
When the output of t becomes the L level, the output of the output terminal Qo of the flip-flop F3 is held at the L level, and the signal of the output of the flip-flop F3, which is not the output terminal Qo of the flip-flop F3, is taken into the flip-flop F4 at each rising edge of the sampling signal SA. Then, the output is supplied to the silhouette memory 20c. Therefore, the output of the OR gate 3 is inverted and operated.

In this way, the holding circuits 20a and 20b are provided.
The silhouette data of the vehicle A is created by sending the holding output of the above to the silhouette memory 20c. Next, the effect of this circuit will be described.

Here, as shown in FIG. 10, the silhouette sensor Q includes the upper sensor group Q1 to Q3 and the lower sensor group q1.
It is assumed that the vehicle passage position sensor E is formed by each sensor E1 to E9. When the vehicle A enters the sensor E1, each sensor q1, q2, q3, Q of the silhouette sensor Q is shown in FIG.
The outputs "011000" of 1, Q2 and Q3 are stored in the silhouette memory 20c, and when the vehicle A next enters the sensor E2, the same sensor output "011000" is stored in the silhouette memory 20c. Similarly, each time the vehicle A enters each of the sensors E3, E4, ..., Each sensor output is stored in the silhouette memory 20c.

In this example, since the outputs of the sensors Q1 to Q3 of the silhouette sensor Q come through the light shielding information circuit 20a, the sampling times Ti and Ti + 1 (i =
If the light is blocked even once during 1, 2, ...), the information is retained. Therefore, regardless of the distance between the vehicle passage position sensors E1 to E9, the silhouette data can be obtained even in a narrow portion such as the window frame of the vehicle A without losing the data.

Further, the sensors q1 to q1 of the silhouette sensor Q are
At q3, since it passes through the light-shielding information holding circuit 20b, it avoids mud smaller than the interval of the vehicle passage position sensor E,
It is possible to exclude grounding etc. as noise and leave only a silhouette with a certain width such as a tire.

Thus, in the silhouette sensor Q,
By selecting the information to be left as a silhouette by each position in the vertical direction of the silhouette sensor Q, only the sheet image necessary for recognition as shown in FIG. 11 can be input.

Therefore, if such a method of creating silhouette data is applied to the second embodiment, for the silhouette data of the vehicle A, the sum of the light and dark values in each of the windows W1 to W5 is calculated to obtain these light and dark values. When the vehicle type of the vehicle is determined based on the above, the reliability of each vehicle type determination can be further increased, and the accuracy thereof can be stably increased. (4) Next, a fourth embodiment of the present invention will be described.

FIG. 12 is a block diagram of a vehicle type identification device. In the passage 31 on which the vehicle 30 travels, the silhouette sensor group 3
Two are provided. This silhouette sensor group 32 is
The projector group 32a and the light receiver group 32b are arranged on both sides of the passage. Among them, the projector group 32a is formed by arranging a plurality of projectors 321a to 32na in the vertical direction with respect to the vehicle 30, and the light receiver group 32b is a plurality of light receivers at positions corresponding to the respective projectors 321a to 32na. Is an array of.

Further, a passage sensor group 33 is provided.
The passage sensor group 33 is provided on both sides of the passage with the projector groups 33a.
And a light receiver group 33b are arranged. Of these, the light projector group 33a is provided integrally with the light projector group 32a of the silhouette sensor group 32, and the light receiver group 33b is provided in the light receiver group 32b of the silhouette sensor group 32. It is provided integrally. The projector group 32a includes a plurality of projectors 331a to 331a-3.
3na are arranged along the traveling direction of the vehicle 30, and the light receiver group 33b includes the light projectors 331a to 33n.
A plurality of light receivers are arranged at a position corresponding to a.

The output terminals of the silhouette sensor group 32 and the passage sensor group 33 are connected to the silhouette creating section 34. The silhouette creating unit 34 sequentially receives the detection signals from the silhouette sensor group 32 to create two-dimensional binarized silhouette data of the vehicle 30 viewed from the lateral direction, and the silhouette data is obtained from the passage sensor group 33. It has a function of correcting the expansion and contraction of the silhouette based on the detection signal indicating the vehicle speed. The silhouette creating unit 34 creates silhouette data for the first axis of the vehicle 30.

Further, the reference data storage unit 35 stores each reference silhouette data for each vehicle type. The determination unit 36 receives the silhouette data created by the silhouette creation unit 34, and performs pattern matching between the silhouette data and the reference silhouette data to determine the vehicle type of the vehicle 30, and as a result of this determination, the reference silhouette of the same vehicle type. It has a function as a learning unit 37 that adds silhouette data to data and subtracts silhouette data from reference silhouette data of different vehicle types to learn and update the reference silhouette data.

When the vehicle type of the vehicle 30 is discriminated by pattern matching, the discriminating section 36 obtains the degree of similarity by pattern matching, and discriminates the vehicle type of the reference silhouette data having a high degree of similarity as the vehicle type of the vehicle 30. There is.

This similarity is calculated by the correlation coefficient value ρoo ρoo = {n ² Σf · g− (Σf) (Σg)} ÷ [{n ² Σf ² − (Σf) ² } · {n ² Σg ² − (Σg) ² }] ^1/2 (2) is used. Here, f is two-dimensional (n × n) silhouette reference data (multivalued data), and g is two-dimensional silhouette data (binary data).

Next, the operation of the apparatus configured as described above will be described. When the vehicle 30 passes between the silhouette sensor group 32 and the passage sensor group 33, the respective light receivers of the silhouette sensor group 32 are in a light-shielded state according to the height of the vehicle 30, and the output signal thereof is the silhouette data creation unit 34.
To send to.

Further, in the passage sensor group 33, each light receiver sequentially changes to the light-shielded state as the vehicle 30 passes, and the detection signals are sequentially sent to the silhouette forming section 34. The silhouette creation unit 34 receives a detection signal from the silhouette sensor group 32 in sequence and receives a two-dimensional two-dimensional view of the vehicle 30 from the lateral direction.
Create valued silhouette data. In this case, the silhouette creation unit 34 corrects the silhouette data for the expansion and contraction of the silhouette of the vehicle speed based on the detection signal indicating the vehicle speed from the passage sensor group 33.

Next, the discriminator 36 determines the silhouette generator 34.
The reference silhouette data stored in the reference data storage unit 35 is read out while receiving the silhouette data created by.

Then, the discriminating section 36 discriminates the vehicle type of the vehicle 30 by pattern matching the silhouette data and the reference silhouette data. In this case, the discriminating unit 36 obtains the similarity shown in the above equation (2) by pattern matching, and determines that the vehicle type of the reference silhouette data having the high similarity is the vehicle type of the vehicle 30.

Further, as a result of the discrimination of the vehicle type, the learning section 37 of the discrimination section 36 adds the silhouette data to the reference silhouette data of the same vehicle type and subtracts the silhouette data from the reference silhouette data of the different vehicle type. Learn and update the reference silhouette data.

For example, as shown in FIG. 13 (a), when the reference silhouette data is the vehicle model of vehicle Q1 and the silhouette data created by the silhouette creating unit 34 is vehicle Q2, the vehicle model of this vehicle Q2 is vehicle Q1. If it is the same type, the silhouette data of the vehicle Q2 is added to the reference silhouette data of the vehicle Q1 as shown in FIG. The silhouette data in FIG. 6B shows an example of a 6 × 6 matrix structure.

Next, the silhouette creating unit 34 causes the vehicle Q3
If the vehicle type of this vehicle Q2 is different from that of the vehicle Q1, the silhouette data of the vehicle Q2 is generated as shown in FIG.
The silhouette data of the vehicle Q3 is subtracted from the reference silhouette data created by adding 1 and Q2.

As described above, the learning section 37 adds the silhouette data of the same vehicle type to the reference silhouette data and subtracts the silhouette data of the different vehicle type to learn and update the reference silhouette data.

As a result, reference silhouette data for each vehicle type is created by learning and stored in the reference data storage unit 35. As described above, in the third embodiment, the silhouette data is created by receiving the detection signals of the silhouette sensor group 32 and the passage sensor group 33 and compared with the reference silhouette data. The silhouette data is added and the silhouette data is subtracted from the reference silhouette data to learn and update the reference silhouette data in the case of a different vehicle type. Therefore, the silhouette data of each passing vehicle 30 is added for learning. Therefore, it is possible to obtain the optimum reference silhouette data for discriminating each vehicle type. As a result, even if the vehicle type determination standard is changed, the determination can be performed and the accuracy of the vehicle type determination can be increased.

Since the silhouette sensor group 32 and the passage sensor group 33 are non-contact type measurements, they are less likely to cause a failure than the mechanical type such as a tread device, can be used stably for a long period of time, and are reliable. You can improve the property.

The fourth embodiment may be modified as follows. For example, the range in which the degree of similarity is obtained by pattern matching is not limited to the first axis to the first axis of the vehicle 30, and the degree of similarity does not depend on the correlation coefficient value ρoo.

Further, in learning the reference silhouette data, each pixel of the silhouette data is not processed one-to-one with respect to the reference silhouette data, but for example, the silhouette data of the input vehicle is calculated according to the similarity to the silhouette of the reference vehicle. A method may be used in which the value is changed and the magnitude of addition and subtraction in the learning unit is changed.

When the silhouette data is created in the fourth embodiment, the silhouette data creating technique of the third embodiment may be applied. By applying this silhouette data creation technology, it is possible to obtain optimum and highly accurate reference silhouette data for distinguishing each vehicle type,
Even if the vehicle type determination standard is changed, the determination can be made and the accuracy of the vehicle type determination can be increased.

[0112]

As described above in detail, according to the present invention, it is possible to provide a vehicle type identification device capable of particularly identifying a large special vehicle, a large vehicle and a bus in the vehicle type identification. Further, according to the present invention, it is possible to provide a vehicle type identification device which is highly reliable and is capable of stably and accurately identifying a vehicle type.

Further, according to the present invention, it is possible to provide a vehicle type discriminating apparatus capable of reducing the occurrence of a failure, improving the reliability, and improving the precision of the discrimination even if the vehicle type discrimination standard is changed. Further, according to the present invention, it is possible to provide a vehicle type identification device capable of accurately obtaining silhouette data of a vehicle without losing data even in a narrow portion of the vehicle.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing a first embodiment of a vehicle type identification device according to the present invention.

FIG. 2 is a specific configuration diagram of each sensor group in the same device.

FIG. 3 is a configuration diagram showing a second embodiment of a vehicle type identification device according to the present invention.

FIG. 4 is a view showing each window position according to each vehicle type in the same device.

FIG. 5 is a view showing a membership function of fuzzy inference in the device.

FIG. 6 is a configuration diagram showing a third embodiment of the vehicle type identification device according to the present invention.

FIG. 7 is a layout view of a silhouette sensor and a vehicle passage position sensor.

FIG. 8 is a configuration diagram of a light shielding information holding circuit.

FIG. 9 is a configuration diagram of a light transmission information holding circuit.

FIG. 10 is a schematic diagram showing silhouette data when a vehicle is passing through.

FIG. 11 is a diagram showing sensor data when a vehicle is passing through.

FIG. 12 is a configuration diagram showing a fourth embodiment of the vehicle type identification device according to the present invention.

FIG. 13 is a diagram showing an operation of a learning function of the device.

FIG. 14 is a diagram showing vehicle types of a large special vehicle, a large vehicle, and a bus in the vehicle.

FIG. 15 is a configuration diagram of a conventional device.

[Explanation of symbols]

10 ... Passage, 11 ... Tire passing sensor, 12 ... Vehicle front position sensor group, 13 ... Silhouette sensor, 14 ... Vehicle separation circuit, 15 ... Signal conversion circuit, 16 ... Front position detection circuit, 17 ... Overhang amount detection circuit, 18 ... Axis number counting circuit, 19 ... Microcomputer, 20 ... Passage, 21
... Silhouette sensor group, 22 ... Silhouette input section, 23
... Silhouette creation unit, 24 ... Vehicle speed sensor, 25 ... Vehicle speed detection unit, 26 ... Window processing unit, 27 ... Fuzzy logic unit, Q ... Silhouette sensor, E ... Vehicle passage position sensor,
20a ... Shading information holding circuit, 20b ... Light passing information holding circuit, 20c ... Silhouette memory, 30 ... Vehicle, 31 ... Passage, 32 ... Silhouette sensor group, 33 ... Passage sensor group,
34 ... Silhouette creation unit, 35 ... Reference data storage unit, 3
6 ... Discrimination unit, 37 ... Learning unit.

Claims (5)

[Claims] 1. A first sensor group in which a plurality of sensors are arranged in a vertical direction with respect to a vehicle, and a second sensor group in which a plurality of sensors are arranged in a traveling direction of the vehicle, and the first sensor group.
And a vehicle analysis means for obtaining the length and the number of axes from the front of the vehicle to a predetermined position from each detection signal of the second sensor group, and a combination of the length and the number of axes obtained by the vehicle analysis means. A vehicle type discriminating apparatus comprising: a discriminating means for discriminating a vehicle type of the vehicle. 2. The vehicle type identification device according to claim 1,
A vehicle type discriminating apparatus comprising a silhouette creating means for sequentially receiving signals from the first and second sensor groups and creating silhouette data for a vehicle. 3. A sensor group formed by arranging a plurality of sensors in a vertical direction with respect to a vehicle, a silhouette creating means for sequentially receiving detection signals of the sensor group and creating silhouette data for the vehicle, and the silhouette creating Window processing means for obtaining the sum of the light and dark values in each window set for the silhouette data created by the means, and the vehicle type of the vehicle based on the light and dark values in each window obtained by the window processing means A vehicle type discriminating device comprising: a discriminating means for discriminating. 4. A first sensor group formed by arranging a plurality of sensors in a vertical direction with respect to a vehicle, and a second sensor group formed by arranging a plurality of sensors in a traveling direction of the vehicle, and the first sensor group.
And a silhouette creating means for sequentially receiving respective detection signals of the second sensor group to create silhouette data for the vehicle, and pattern matching of the silhouette data created by the silhouette creating means and the reference silhouette data for the vehicle type of the vehicle. Based on the result of this determination, if the vehicle type is the same, the silhouette data is added to the reference silhouette data, and if the vehicle type is different, the silhouette data is subtracted from the reference silhouette data. A vehicle type discriminating apparatus comprising: a learning unit for learning and updating the reference silhouette data. 5. A vehicle type discriminating apparatus comprising a silhouette creating means for creating silhouette data for a vehicle, wherein the silhouette creating means sequentially receives detection signals of a sensor group output according to the traveling speed of the vehicle, and outputs the detection signals. And a silhouette memory that sequentially stores the detection signals held by the information holding circuit to create silhouette data for the vehicle.