JP2011133989A - Window detection device, vehicle type discrimination device, window detection method, and vehicle type discrimination method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To detect a window of a vehicle and to discriminate a type of the vehicle based on a detection result. <P>SOLUTION: A radiation part 111 radiates radiation light 751 to the vehicle. A light receiving part 115 receives the reflection light 754 which is the radiation light 751 hitting the vehicle and being reflected. A reflection position computing part 122 computes a reflection position where the radiation light 751 hits the vehicle 850 and is reflected based on time until the light receiving part 115 receives the reflection light 754 after the radiation part 111 radiates the radiation light 751. A window detection part 130 detects the position of the window of the vehicle based on the reflection position which the reflection position computing part 122 has computed. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a window detection device that detects the position of a window of a vehicle and a vehicle type determination device that determines the vehicle type of a vehicle using a detection result of the window detection device.

As a method for discriminating the vehicle type, for example, there is a method of photographing a license plate, reading characters, and discriminating the vehicle type based on the read characters.
There is also a method of measuring the cross-sectional shape of a vehicle and determining the vehicle type based on the measured cross-sectional shape.

JP 2000-20877 A JP-A-11-191196

In the method of reading the license plate, the license plate may be dirty or bent, and characters may not be recognized. In addition, when the vehicle interval is narrow due to traffic congestion or the like, the license plate is hidden behind the front and rear vehicles, and the license plate cannot be photographed in the first place.
The method of discriminating the vehicle type based on the cross-sectional shape of the vehicle may have a similar cross-sectional shape, such as a large truck and a large bus, and may cause a determination error.
The present invention has been made to solve the above-described problems, for example, and has an object to detect a vehicle window and correctly determine the vehicle type based on the detection result.

The window detection device according to the present invention includes a radiation unit, a light receiving unit, a reflection position calculation unit, and a window detection unit,
The radiating part emits radiant light to the vehicle,
The light receiving unit receives reflected light reflected by the radiated light hitting the vehicle,
The reflection position calculation unit calculates a reflection position where the radiated light is reflected by the vehicle based on a time from when the radiating unit emits radiated light until the light receiving unit receives the reflected light.
The window detection unit detects the position of the window of the vehicle based on the reflection position calculated by the reflection position calculation unit.

  According to the window detection device of the present invention, since the position of the vehicle window can be known, it is possible to easily determine a vehicle type that is difficult to determine only by the shape of the vehicle.

FIG. 3 is a perspective view showing an example of a vehicle type identification system 800 in the first embodiment. FIG. 2 is a plan view showing an example of a vehicle type identification system 800 in the first embodiment. FIG. 6 is a front view illustrating an example of a radiation range 703 in Embodiment 1; FIG. 3 illustrates a principle of window detection in the first embodiment. FIG. 3 illustrates a principle of window detection in the first embodiment. FIG. 5 shows an example of a point group 710 in Embodiment 1; FIG. 2 is a block configuration diagram showing an example of a hardware configuration of a control device 200 in the first embodiment. FIG. 3 is a block configuration diagram showing an example of a functional block configuration of a vehicle type identification system 800 in the first embodiment. FIG. 5 is a flowchart showing an example of a flow of a reflection position calculation process S510 in the first embodiment. FIG. 6 is a flowchart showing an example of a flow of window detection processing S520 in the first embodiment. The flowchart figure which shows an example of the flow of the axle shaft detection process S530 in Embodiment 1. FIG. FIG. 5 shows an example of a point group 710 in Embodiment 1; The flowchart figure which shows an example of the flow of vehicle width calculation process S540 in Embodiment 1. FIG. The flowchart figure which shows an example of the flow of the number mark recognition process S550 in Embodiment 1. FIG. FIG. 6 is a diagram showing an example of an image that is a target of number mark recognition processing S550 in the first embodiment. FIG. 3 is a diagram illustrating an example of a positional relationship between a window and an axle detected by the vehicle type identification system 800 in Embodiment 1 for a vehicle 850. FIG. 10 is a diagram showing another example of the positional relationship between the window and the axle detected by the vehicle type identification system 800 in the first embodiment for the vehicle 850. The figure which shows another example of the positional relationship of the window and axle which the vehicle type identification system 800 in Embodiment 1 detects about the vehicle 850. FIG. The figure which shows an example of the charge classification in a toll road. The flowchart figure which shows an example of the whole flow of the vehicle type discrimination | determination process S560 in Embodiment 1. FIG. FIG. 6 is a flowchart showing an example of a flow of number mark determination processing S570 in the first embodiment. FIG. 6 is a flowchart showing an example of a flow of window determination processing S580 in the first embodiment. FIG. 6 is a front view showing an example of a vehicle type identification system 800 according to Embodiment 2.

Embodiment 1 FIG.
The first embodiment will be described with reference to FIGS.

FIG. 1 is a perspective view showing an example of a vehicle type discrimination system 800 in this embodiment.
The vehicle type identification system 800 is installed, for example, at a toll gate on a toll road. The vehicle type discrimination system 800 discriminates the vehicle type of the vehicle 850 that passes therethrough. The vehicle type is a type of the vehicle 850 such as a passenger car, a freight vehicle, a large vehicle, and a light vehicle. In the vehicle type discriminating system 800 installed at a toll booth or the like, it is important to discriminate which charge category the vehicle 850 corresponds to, so the vehicle type is particularly the charge category applicable to the vehicle 850. That's it. The vehicle type determination system 800 collects information for determining the vehicle type of the vehicle 850 by various methods such as window detection, axle detection, vehicle width measurement, and license plate reading. The vehicle type discrimination system 800 discriminates the vehicle type of the vehicle 850 based on the collected information.

  The vehicle type identification system 800 includes two laser sensors 110, an infrared imaging device 180 (number plate imaging unit), and a control device 200. In order to distinguish the two laser sensors 110, they may be referred to as a laser sensor 110a and a laser sensor 110b with a lower case alphabet. The same applies to a plurality of other elements. The infrared imaging device 180 continuously photographs the vehicle 850 that passes therethrough using infrared rays.

  The vehicle type identification system 800 is installed on two islands 810 across a road 820 through which the vehicle 850 passes. The laser sensor 110a is installed on an island 810a located on the left side when viewed from the vehicle 850. The laser sensor 110b is installed on an island 810b located on the right side when viewed from the vehicle 850. The infrared imaging device 180 and the control device 200 may be installed on the left island 810a, or may be installed on the right island 810b. The infrared imaging device 180 is installed at a position where a license plate of a vehicle 850 passing through the road 820 enters the imaging range 701.

FIG. 2 is a plan view showing an example of a vehicle type discrimination system 800 in this embodiment.
The laser sensor 110 is a reflection type. The laser sensor 110 emits laser light (radiated light) and receives the reflected light that is reflected by the vehicle 850 and measures the distance to the reflection position. The laser sensor 110 measures the shape of the vehicle 850 by scanning (scanning) while changing the direction in which the laser light is emitted up and down.
A radiation range 703 a in which the laser sensor 110 a emits laser light is substantially perpendicular to the vehicle traveling direction 702. The laser sensor 110a emits laser light from the left side to the vehicle 850. The laser sensor 110a mainly measures the shape of the left side surface of the vehicle 850.
Similarly, a radiation range 703 b in which the laser sensor 110 b emits laser light is substantially perpendicular to the vehicle traveling direction 702. The laser sensor 110b emits laser light from the right side to the vehicle 850. Laser sensor 110b mainly measures the shape of the right side surface of vehicle 850.
In the vehicle traveling direction 702, the laser sensor 110b is installed at a position shifted forward (back) from the laser sensor 110a. This has two meanings. One is to prevent the other laser sensors 110 from receiving the laser light emitted from the laser sensor 110 directly or the reflected light reflected from the vehicle 850 or the road 820. The other is to make it possible to calculate the traveling direction of the vehicle 850 and the speed of the vehicle 850 from the difference in time when the vehicle 850 passes through the radiation ranges 703a of the two laser sensors 110. The laser sensor 110b may be installed at a position shifted rearward (near) from the laser sensor 110a.

FIG. 3 is a front view showing an example of the radiation range 703 in this embodiment.
The radiation range 703a of the laser light emitted by the laser sensor 110a is set so as to cover the entire vehicle 850 from the top to the bottom. Similarly, the radiation range 703b of the laser light emitted by the laser sensor 110b is set so as to cover the entire vehicle from the top to the bottom. For example, the laser sensors 110a and 110b are installed at a height of 1.5 meters [m] from the road 820, and the radiation ranges 703a and 703b are set to a range of 90 degrees from the upper 45 degrees to the lower 45 degrees.

FIG. 4 is a diagram for explaining the principle of window detection in this embodiment.
This figure shows a case where the radiated light 751 emitted by the laser sensor 110 hits the vehicle body 851 at a portion that is not a window of the vehicle 850. The light reflected from the vehicle body 851 is divided into specularly reflected regular light 752 and diffusely reflected diffusely reflected light 753. Since the reflection angle of the regular reflection light 752 is equal to the incident angle of the radiation light 751, the regular reflection light 752 has a direction different from the direction of the laser sensor 110 unless the radiation light 751 hits the vehicle body 851 perpendicularly. Head to. On the other hand, the diffuse reflected light 753 travels in various directions, and includes reflected light 754 that returns to the direction of the laser sensor 110. The laser sensor 110 measures the distance to the reflection position by detecting the reflected light 754.

FIG. 5 is a diagram for explaining the principle of window detection in this embodiment.
This figure shows a case where the radiated light 751 emitted by the laser sensor 110 hits the glass 852 which is a window portion of the vehicle 850. The emitted light 751 is divided into transmitted light that enters the glass 852 and reflected light that is reflected by the surface of the glass 852. The reflected light is divided into specularly reflected light 752 that is specularly reflected and diffusely reflected light that is diffusely reflected. The transmitted light is divided into regular transmitted light 755 refracted according to Snell's law and diffuse transmitted light diffused and transmitted. Since the surface of the glass 852 is relatively smooth, regular reflection light 752 and regular transmission light 755 are strong, and diffuse reflection light and diffuse transmission light are weak. For this reason, the reflected light reflected in the direction of the laser sensor 110 is weak, and the laser sensor 110 does not detect the reflected light.

ただし、ｘは、レーザセンサ１１０と反射位置との間の水平距離を表わす。ｙは、レーザセンサ１１０と反射位置との間の高さの差を表わす。ｃは、空気中における光速を表わす。φは、レーザセンサ１１０がレーザ光を放射した放射方向と水平方向とがなす角を表わす。ｔは、レーザセンサ１１０がレーザ光を放射してから反射光を受光するまでの遅延時間を表わす。 FIG. 6 is a diagram showing an example of the point group 710 in this embodiment.
If the angle in the radiation direction in which the laser sensor 110 radiates the laser light and the delay time from when the laser sensor 110 radiates the laser light to receiving the reflected light are known, the reflection position can be obtained by the following equation. .

Here, x represents the horizontal distance between the laser sensor 110 and the reflection position. y represents a height difference between the laser sensor 110 and the reflection position. c represents the speed of light in the air. φ represents an angle formed between the radiation direction in which the laser sensor 110 radiates laser light and the horizontal direction. t represents a delay time from when the laser sensor 110 emits laser light until it receives reflected light.

ただし、Ｗは、車幅を表わす。Ｒは、センサ間隔７２１を表わす。ｒ_ａは、レーザセンサ１１０ａと車両８５０との間の水平距離を表わす。ｒ_ｂは、レーザセンサ１１０ｂと車両８５０との間の水平距離を表わす。 The point group 710a is a plot of the reflection position calculated based on the observation result of the laser sensor 110a. The point group 710b is a plot of the reflection position calculated based on the observation result of the laser sensor 110b.
As described above, since the reflected light from the window 854 of the vehicle 850 is weak, no reflecting object is observed in that direction, and data is lost. For this reason, the point group 710a is divided into a group corresponding to the roof 853 of the vehicle 850 and a group corresponding to the main body 855 of the vehicle 850 and the like. Therefore, by analyzing the point group 710a, the window upper end height 711a, the window lower end height 712a, the sensor vehicle body distance 713a, and the like are known. The window upper end height 711a is the difference between the height of the portion that hits the upper end of the window 854a and the height of the road 820. The window lower end height 712a is the difference between the height of the portion that hits the lower end of the window 854a and the height of the road 820. The sensor inter-vehicle distance 713a is a distance from the laser sensor 110a to the main body 855 of the vehicle 850. Similarly, the point group 710b is divided into a group corresponding to the roof 853 of the vehicle 850 and a group corresponding to the main body 855 of the vehicle 850. Therefore, by analyzing the point group 710b, the window upper end height 711b, the window lower end height 712b, the sensor vehicle body distance 713b, and the like are known. Of the horizontal distance between the laser sensor 110a and the laser sensor 110b, a component (hereinafter referred to as “sensor interval 721”) in a direction perpendicular to the vehicle traveling direction 702 (a direction crossing the road 820) is known in advance. Then, the vehicle width 722 can be obtained using the following equation.

However, W represents a vehicle width. R represents the sensor interval 721. r _a represents the horizontal distance between the laser sensor 110a and the vehicle 850. r _b represents the horizontal distance between the laser sensor 110b and vehicle 850.

FIG. 7 is a block configuration diagram showing an example of a hardware configuration of the control device 200 in this embodiment.
The control device 200 includes, for example, a storage device 920, a processing device 911, an input device 902, an output device 901, and the like.
The storage device 920 stores a program executed by the processing device 911, data processed by the processing device 911, and the like. Examples of the storage device 920 include a nonvolatile storage device (hereinafter referred to as “ROM”) and a volatile storage device (hereinafter referred to as “RAM”). The ROM stores a program executed by the processing device 911, for example. The RAM temporarily stores, for example, data being processed by the processing device 911.
The processing device 911 processes data by executing the program stored in the storage device 920 and controls the entire control device 200 such as the processing device 911 itself, the storage device 920, the input device 902, and the output device 901.
The input device 902 inputs data and signals processed by the processing device 911 from the outside of the control device 200. The input device 902 converts input data and signals into a data format that can be processed by the processing device 911. The data converted by the input device 902 may be directly processed by the processing device 911 or may be temporarily stored by the storage device 920. The input device 902 includes, for example, an analog-digital converter (hereinafter referred to as “ADC”) that converts an analog signal into digital data, a latch circuit that holds the digital signal, and a receiving device that receives data transmitted by another device. There are operation input devices such as buttons and keyboards for inputting the operation of the operator.
The output device 901 converts the data processed by the processing device 911 or the data stored in the storage device 920 and outputs the converted data to the outside of the control device 200. The output device 901 includes, for example, a digital-analog converter (hereinafter referred to as “DAC”) that converts digital data into an analog signal, a latch circuit that holds the digital signal, and a transmission device that transmits data to other devices. There are display devices that display data in a form visible to the operator, such as light emitting diodes and liquid crystal display devices.

  The functional blocks of the control device 200 described below are realized by the processing device 911 executing the program stored in the storage device 920. However, this is merely an example, and the functional block of the control device 200 is not limited to a program, and may be realized by an electronic circuit such as an analog circuit or a digital circuit, or may be realized by a configuration other than the electronic circuit such as a mechanical configuration. May be.

FIG. 8 is a block configuration diagram showing an example of a functional block configuration of the vehicle type identification system 800 in this embodiment.
The laser sensor 110 includes a radiation unit 111 and a light receiving unit 115.
The control device 200 includes a laser sensor control unit 120, a window detection unit 130, a traveling direction determination unit 210, an axle detection unit 220, a vehicle width calculation unit 230, a number mark recognition unit 280, and a vehicle type determination unit 290.
Among these, the laser sensor 110, the laser sensor control unit 120, and the window detection unit 130 constitute the window detection device 100. The window detection device 100 is a part that detects the window of the vehicle 850 in the vehicle type identification system 800.

The radiating unit 111 emits radiated light 751 under the control of the laser sensor control unit 120. The radiation unit 111 includes a laser oscillation unit 112 and a laser scanning unit 113.
The laser oscillation unit 112 emits laser light in accordance with a control signal from the laser sensor control unit 120.
The laser scanning unit 113 bends the traveling direction of the laser light emitted from the laser oscillation unit 112 in accordance with a control signal from the laser sensor control unit 120, and sets the radiation light 751 to scan the radiation range 703.

The light receiving unit 115 receives the reflected light 754 and generates a signal indicating the intensity of the received reflected light 754. The light receiving unit 115 includes a laser light receiving unit 116 and an amplification unit 117.
The laser light receiving unit 116 receives the reflected light 754 reflected by the radiated light 751 when it hits the vehicle 850 or the like, and converts it into an electrical signal.
The amplifying unit 117 amplifies the electric signal converted by the laser light receiving unit 116 to obtain a signal that can be input by the control device 200.

The laser sensor control unit 120 controls the laser sensor 110 to calculate the reflection position. Note that one laser sensor control unit 120 may control the two laser sensors 110a and 110b. Further, two laser sensor control units 120a and 120b are provided corresponding to the two laser sensors 110a and 110b, and the respective laser sensor control units 120a and 120b control the corresponding laser sensors 110a and 110b. May be. The laser sensor control unit 120 includes a scanning control unit 121 and a reflection position calculation unit 122.
The scanning control unit 121 generates a control signal for controlling the laser oscillation unit 112 and the laser scanning unit 113 when the processing device 911 processes the data. For example, the scanning control unit 121 generates a control signal for controlling the laser oscillation unit 112 so that the laser light emitted from the laser oscillation unit 112 is pulsed at a predetermined interval. For example, the scanning control unit 121 generates a control signal for controlling the laser scanning unit 113 so that the radiation direction of the emitted light 751 bent by the laser scanning unit 113 changes at a predetermined angular velocity.

  The reflection position calculation unit 122 inputs the signal amplified by the amplification unit 117 when the processing device 911 controls the input device 902. The reflection position calculation unit 122 processes the data, and the laser oscillation unit 112 emits the radiated light 751 in accordance with the control signal generated by the scanning control unit 121, and then the laser light receiving unit 116 reflects the reflected light. The time t required to receive 754 is calculated. The reflection position calculation unit 122 processes the data by the processing device 911 and calculates the time t and the direction φ in which the laser scanning unit 113 radiates the radiated light 751 in accordance with the control signal generated by the scanning control unit 121. Based on this, the relative coordinates (x, y) of the reflection position with respect to the position of the laser sensor 110 are calculated. Note that the laser sensor control unit 120 may be configured to calculate the absolute coordinates of the reflection position. For example, the storage device 920 stores the coordinates of the laser sensor 110 in a predetermined coordinate system in advance. The laser sensor control unit 120 performs coordinate conversion using the coordinates of the laser sensor 110 stored in the storage device 920, and calculates the absolute coordinates of the reflection position in the coordinate system.

Window detection unit 130 (window glass detection unit) detects a window of vehicle 850 based on the reflection position calculated by laser sensor control unit 120. The window detection unit 130 includes a discontinuous position calculation unit 131 and a window determination unit 132.
The discontinuous position calculating unit 131 processes the data so that the position where the reflection position is not continuous from the series of reflection positions calculated by the reflection position calculating unit 122 (hereinafter referred to as “discontinuous position”). Called.)
As shown in FIG. 6, when the radiated light 751 hits the roof 853, the main body 855, or the road 820 of the vehicle 850 and is reflected, the laser sensor 110 detects the reflected light 754, and the reflection position calculation unit 122 determines the reflection position. calculate. On the other hand, when the radiated light 751 hits the window 854 of the vehicle 850 or when there is nothing in the radiating direction, the laser sensor 110 does not detect the reflected light 754, so the reflection position calculation unit 122 calculates the reflection position. Without it, it will be missing. Therefore, if the vehicle 850 has the roof 853, the reflection position calculated by the reflection position calculation unit 122 is divided into two groups with a portion corresponding to the window 854 interposed therebetween.
As described above, when the reflection position in a certain direction is missing, and there is a reflection position that is not missing in both the upper and lower directions, the discontinuous position calculation unit 131 performs discontinuity. Calculate the position. For example, the discontinuous position calculation unit 131 calculates the reflection position located at the bottom of the reflection positions upward from the missing measurement direction as the upper discontinuity position. In addition, the discontinuous position calculation unit 131 calculates the reflection position located at the top of the reflection positions below the missing measurement direction as the lower discontinuity position. Thus, the discontinuous position calculation unit 131 calculates a pair of an upper discontinuous position and a lower discontinuous position.

The window determination unit 132 determines whether or not the discontinuity position calculated by the discontinuity position calculation unit 131 corresponds to a window by processing the data by the processing device 911. For example, the window determination unit 132 compares the height of the upper discontinuity position calculated by the discontinuity position calculation unit 131 with a predetermined threshold value, and determines that the window is not a window if the height is lower than the threshold value. In addition, for example, the window determination unit 132 compares the discontinuous position calculated by the discontinuous position calculation unit 131 with the discontinuous position calculated by the discontinuous position calculation unit 131 at the time of the previous scan. When the variation amount of the discontinuous position is larger than a predetermined threshold, it is determined that the window is not a window.
In this way, the window determination unit 132 removes the pair determined not to be a window from the pair of discontinuous positions calculated by the discontinuous position calculation unit 131. For the remaining pairs, the window determination unit 132 sets the upper discontinuity position as the upper end position of the window and the lower discontinuity position as the lower end position of the window.

  The reflection position calculated by the reflection position calculation unit 122 can be used for various purposes other than detection of a window.

The traveling direction determination unit 210 (traveling direction determination processing unit) determines the traveling direction of the vehicle 850 based on the reflection position calculated by the reflection position calculation unit 122 when the processing device 911 processes the data. For example, the traveling direction determination unit 210 determines whether the vehicle 850 is present in the radiation range 703 of the laser sensor 110 from the reflection position calculated by the reflection position calculation unit 122. If the vehicle 850 is not present, the series of reflection positions form the shape of the road 820. Therefore, when a series of reflection positions have a shape different from the shape of the road 820, it means that some object other than the road 820 exists in the radiation range 703. Therefore, the traveling direction determination unit 210 determines whether or not the series of reflection positions form the shape of the road 820, and determines that the vehicle 850 exists when the shape is different from the shape of the road 820. The traveling direction determination unit 210 may be configured to analyze in more detail the shape formed by the series of reflection positions and determine whether or not the vehicle 850 is present. That is, the traveling direction determination unit 210 determines whether the shape formed by the series of reflection positions is the shape of the vehicle 850 or the shape of an object other than the vehicle 850 (for example, a person such as a staff at a toll gate) Only when it is determined that the shape of the vehicle 850 is present, it is determined that the vehicle 850 exists.
Based on the determination result, the traveling direction determination unit 210 determines the time when the vehicle 850 starts to exist in the radiation ranges 703a and 703b for each of the two laser sensors 110a and 110b (hereinafter referred to as “vehicle detection start time”) and the like. , And the time when it ended (hereinafter referred to as “vehicle detection end time”) is calculated. Since the two laser sensors 110a and 110b are installed at positions shifted in the vehicle traveling direction 702, if the vehicle 850 is traveling in the vehicle traveling direction 702, the vehicle detection start time and the vehicle detection end time in the laser sensor 110a. Instead, the vehicle detection start time and the vehicle detection end time of the laser sensor 110b are later. Conversely, when the vehicle 850 runs backward in the direction opposite to the vehicle traveling direction 702, the vehicle detection start time or vehicle detection end time in the laser sensor 110a is greater than the vehicle detection start time or vehicle detection end time in the laser sensor 110b. Is later. The traveling direction determination unit 210 determines the traveling direction of the vehicle 850 by comparing the vehicle detection start times or the vehicle detection end times of the two laser sensors 110a and 110b.

The axle detector 220 (axle detector) detects the axle of the vehicle 850 when the processing device 911 processes the data. The axle is a unit that counts tires arranged in the lateral direction of the vehicle 850 as one axis. For example, in a normal four-wheel vehicle, two front wheels are one axle, two rear wheels are another axle, and the number of axles is two. On the other hand, some large trucks have a large number of axles such as three axles and four axles.
In the example shown in FIG. 6, the tire 856 is just reaching the radiation range 703 of the laser sensor 110. Therefore, the shape of the tire 856 appears in the shape formed by the series of reflection positions calculated by the reflection position calculation unit 122.
As described above, the axle detection unit 220 determines that the axle has been detected when the shape of the tire 856 appears in the shape formed by the series of reflection positions. Further, the axle detection unit 220 calculates the number of axles by counting the number of detected axles for one vehicle 850. It should be noted that axle detection unit 220 adds 1 to the number of axles when it detects an axle when vehicle 850 is traveling in vehicle traveling direction 702 based on the determination result of traveling direction determination unit 210. Conversely, when the axle is detected when the vehicle 850 is running backward, 1 is subtracted from the number of axles. Thereby, even when the vehicle 850 repeats forward and backward movement, the number of axles can be correctly counted without erroneously counting.

  The vehicle width calculation unit 230 (vehicle width measurement unit) calculates the vehicle width 722 of the vehicle 850 when the processing device 911 processes the data. For example, as described above, the vehicle width calculation unit 230 calculates the horizontal distance between the laser sensor 110a and the vehicle 850 (sensor vehicle body distance 713a). The vehicle width calculation unit 230 calculates the horizontal distance between the laser sensor 110b and the vehicle 850 (sensor body distance 713b). The vehicle width calculation unit 230 calculates the sum of the sensor body distance 713a and the sensor body distance 713b. The vehicle width calculation unit 230 calculates a difference obtained by subtracting the calculated sum from the sensor interval 721 (inter-installation distance in the road horizontal direction) to obtain a vehicle width 722.

  The number mark recognition unit 280 (number plate recognition unit) inputs an infrared image captured by the infrared imaging device 180 when the processing device 911 controls the input device 902. The number mark recognizing unit 280 performs image processing on the input infrared image by processing the data by the processing device 911, analyzes the license plate of the vehicle 850, and analyzes the size, The registration number of the vehicle 850 described on the naver plate is read.

  The vehicle type determination unit 290 (window glass vehicle type determination unit / vehicle type determination unit) is a window detected by the window detection unit 130, an axle detected by the axle detection unit 220, and a vehicle width calculation unit when the processing device 911 processes the data. Based on the information such as the vehicle width 722 calculated by 230 and the registration number read by the number mark recognition unit 280, the vehicle type of the vehicle 850 is determined. The vehicle type determination unit 290 outputs the determined vehicle type by the processing device 911 controlling the output device 901.

  The vehicle type determined by the vehicle type determination system 800 is used by, for example, an automatic toll collection system (hereinafter referred to as “ETC”). For example, the ETC calculates a toll for the vehicle 850 based on the vehicle type determined by the vehicle type determination system 800. Alternatively, the ETC determines whether there is a contradiction between the information related to the vehicle type of the vehicle 850 acquired by another method such as information received from the ETC on-vehicle device and the vehicle type determined by the vehicle type determination system 800. Discover the vehicle.

FIG. 9 is a flowchart showing an example of the flow of the reflection position calculation process S510 in this embodiment.
In the reflection position calculation process S510, the reflection position calculation unit 122 calculates the reflection position. The reflection position calculation process S510 includes a scanning direction acquisition step S511, a delay time calculation step S512, a distance calculation step S513, and a coordinate conversion step S514.

In the scanning direction acquisition step S511, the reflection position calculation unit 122 causes the processing device 911 to process the data so that the scanning control unit 121 controls the laser scanning unit 113 to emit the emitted light 751. To get.
In the delay time calculation step S512, the reflection position calculation unit 122 causes the processing device 911 to process the data so that the laser oscillation unit 112 emits the emitted light 751 and the light reception unit 115 receives the reflected light 754. And a delay time t is calculated.
In the distance calculation step S513, the reflection position calculation unit 122 converts the delay time t calculated in the delay time calculation step S512 into a distance to the reflection position by the processing device 911 processing the data. Specifically, for example, the reflection position calculation unit 122 calculates the quotient obtained by dividing the delay time t by the speed of light c in the air, thereby calculating the length of the round-trip path through which the radiated light 751 and the reflected light 754 have passed. calculate. The reflection position calculation unit 122 calculates the length of the one-way path, that is, the distance to the reflection position, by calculating a quotient obtained by dividing the calculated path length by 2.
In the coordinate conversion step S514, the reflection position calculation unit 122 processes the data so that the angle φ acquired in the scanning direction acquisition step S511 and the distance to the reflection position calculated in the distance calculation step S513 are processed. Based on this, the relative coordinates (x, y) of the reflection position or the absolute coordinates of the reflection position are calculated. The reflection position calculation unit 122 converts the polar coordinate display based on the angle φ and the distance into an orthogonal coordinate display with the horizontal direction and the height direction as axes.
Thereafter, the reflection position calculation unit 122 returns to the scanning direction acquisition step S511 and calculates the next reflection position.

  In this way, the reflection position calculation unit 122 calculates a series of reflection positions.

FIG. 10 is a flowchart showing an example of the flow of the window detection processing S520 in this embodiment.
In the window detection process S520, the window detection unit 130 detects the window of the vehicle 850. Window detection processing S520 includes scanning direction selection step S521, missing measurement determination step S522, state determination step S523, state transition step S524, state determination step S525, window upper end determination step S526, state transition step S527, window lower end determination step S528, It has windowless determination process S529.

In the scanning direction selection step S <b> 521, the discontinuous position calculation unit 131 selects the radiation direction from the top in the radiation direction in which the laser scanning unit 113 radiates the radiation 751 by processing the data with the processing device 911. To do.
When the discontinuous position calculation unit 131 selects the radiation direction, the process proceeds to a missing measurement determination step S522.
When the processing is completed up to the lowest radiation direction and there is no radiation direction that can be selected, the discontinuous position calculation unit 131 proceeds to the windowless determination step S529.

In the missing measurement determination step S522, the discontinuous position calculation unit 131 has the reflection position calculated by the reflection position calculation unit 122 for the radiation direction selected in the scanning direction selection step S521 by processing the data by the processing device 911. It is determined whether or not.
When there is a reflection position calculated by the reflection position calculation unit 122, that is, when there is no missing measurement, the discontinuous position calculation unit 131 proceeds to the state determination step S523.
When there is no reflection position calculated by the reflection position calculation unit 122, that is, when there is a missing measurement, the discontinuity position calculation unit 131 proceeds to the state determination step S525.

  In the state determination step S523, the discontinuous position calculation unit 131 determines the detection state when the processing device 911 processes the data.

The detection state is a state where the currently selected radiation direction corresponds to the vehicle 850. In this example, since the radiation direction is viewed in order from the top, the process starts from the direction in which the vehicle 850 does not exist. This is called an “empty” state. In the “empty” state of the radiation direction, the reflection position calculated by the reflection position calculation unit 122 does not exist, that is, missing. Prior to the start of the window detection process S520, the discontinuous position calculation unit 131 stores in advance in the storage device 920 that the state is “empty”.
When the radiation direction decreases, the roof 853 of the vehicle 850 is first seen. That is, the state changes from the missing measurement to the state where the reflection position calculated by the reflection position calculation unit 122 exists. This is called the “roof” state.
If there is a window 854, then the window 854 is visible. That is, the measurement returns to the missing measurement again from the state where the reflection position calculated by the reflection position calculation unit 122 exists. This is called the “window” state.
When the window 854 is finished, the main body 855 can be seen next. That is, the state shifts from the missing measurement to the state where the reflection position calculated by the reflection position calculation unit 122 exists. After that, even if the reflective object changes to the tire 856, the road 820, or the like, the state where the reflection position calculated by the reflection position calculation unit 122 is not changed, and reaches the lowest radiation direction.
When there is no window 854, it reaches the lowest radial direction without transitioning from the “roof” state to the “window” state.

When the detection state is the “empty” state, it means that the state is changed from the missing measurement state to the state where the reflection position calculated by the reflection position calculation unit 122 exists. The discontinuous position calculation unit 131 proceeds to the state transition process S524.
Similarly, when the detection state is the “window” state, it means that the state is changed from the missing measurement state to the state where the reflection position calculated by the reflection position calculation unit 122 exists. The discontinuous position calculation unit 131 proceeds to the window lower end determination step S528.
When the detection state is other than “sky” and “window”, it means that the state where the reflection position calculated by the reflection position calculation unit 122 exists continues. The discontinuous position calculation unit 131 returns to the scanning direction selection step S521 and selects the next radiation direction.

In the state transition step S524, the discontinuous position calculation unit 131 causes the processing device 911 to process the data, thereby causing the detection state to transition from the “empty” state to the “roof” state. The storage device 920 stores that the current detection state is the “roof” state.
Thereafter, the discontinuous position calculation unit 131 returns to the scanning direction selection step S521 and selects the next radiation direction.

In the window lower end determination step S528, the window determination unit 132 processes the data so that the reflection position calculated by the reflection position calculation unit 122 in the radiation direction selected in the scanning direction selection step S521 is the lower end of the window. It is determined whether or not.
When it determines with it being the lower end of a window, the window determination part 132 makes the reflection position the position of the lower end of a window, and complete | finishes window detection process S520.
When the window determination unit 132 determines that the window is not the lower end of the window, the discontinuous position calculation unit 131 returns to the scanning direction selection step S521 and selects the next radiation direction.

In the state determination step S525, the discontinuous position calculation unit 131 determines the detection state when the processing device 911 processes the data.
When the detection state is the “roof” state, it means that the state where the reflection position calculated by the reflection position calculation unit 122 is changed to the missing state. The discontinuous position calculation unit 131 proceeds to the window upper end determination step S526.
If the detection state is other than the “roof” state, it means that the missing measurement state continues. The discontinuous position calculation unit 131 returns to the scanning direction selection step S521 and selects the next radiation direction.

In the window upper end determination step S526, the window determination unit 132 calculates the reflection position calculation unit 122 for the radiation direction one higher than the radiation direction selected in the scanning direction selection step S521 by processing the data by the processing device 911. It is determined whether or not the reflection position is the upper end of the window.
When it determines with it being the upper end of a window, the window determination part 132 makes the reflection position the position of the upper end of a window, and progresses to state transition process S527.
When it determines with it not being the upper end of a window, the window determination part 132 returns to scanning direction selection process S521, and selects the next radiation direction.

In the state transition step S527, the window determination unit 132 causes the processing device 911 to process the data, thereby causing the detection state to transition from the “roof state” to the “window” state. The storage device 920 stores that the current detection state is the “window” state.
Thereafter, the discontinuous position calculation unit 131 returns to the scanning direction selection step S521 and selects the next radiation direction.

  In the windowless determination step S529, the window determination unit 132 determines that no window has been detected by the processing device 911 processing the data, and ends the window detection process S520.

FIG. 11 is a flowchart showing an example of the flow of the axle detection process S530 in this embodiment.
In the axle detection process S530, the axle detection unit 220 detects the axle of the vehicle 850. The axle detection process S530 includes a scanning direction selection step S531, a height determination step S532, a continuity determination step S533, an axle presence determination step S534, an axle absence determination step S535, and a vehicle absence determination step S536.

In the scanning direction selection step S531, the axle detection unit 220 selects the radiation direction in order from the bottom among the radiation directions in which the laser scanning unit 113 radiates the radiated light 751 by the processing device 911 processing the data.
When the axle detection unit 220 selects the radial direction, the process proceeds to a height determination step S532.
If the processing is completed up to the top radial direction and there is no radial direction that can be selected, the axle detector 220 proceeds to the vehicle absence determination step S536.

In the height determination step S532, the axle detection unit 220 determines that the height of the reflection position calculated by the reflection position calculation unit 122 in the radial direction selected in the scanning direction selection step S531 by processing the data by the processing device 911. It is determined whether or not the height of the road 820 is higher.
When the height of the reflection position is higher than the height of the road 820, the axle detection unit 220 proceeds to the continuity determination step S533.
When the height of the reflection position is the height of the road 820, the axle detection unit 220 returns to the scanning direction selection step S531, and selects the next radiation direction.

In the continuity determination step S533, the axle detection unit 220 processes the data by the processing device 911, thereby reducing the reflection position calculated by the reflection position calculation unit 122 in the radial direction selected in the scanning direction selection step S531. The distance from the reflection position calculated by the reflection position calculation unit 122 in the radiation direction is calculated. The axle detection unit 220 determines whether the calculated distance is larger or smaller than a predetermined threshold.
If the calculated distance is smaller than the threshold value, it means that a portion in contact with the road 820 has been detected. The axle detector 220 proceeds to the axle presence determination step S534.
If the calculated distance is larger than the threshold value, it means that a part away from the road 820 has been detected. The axle detector 220 proceeds to the axle absence determination step S535.

  In the axle presence determination step S534, the axle detection unit 220 determines that there is an axle, and ends the axle detection process S530.

  In the axle absence determination step S535, the axle detector 220 determines that there is no axle, and ends the axle detection process S530.

FIG. 12 is a diagram showing an example of the point group 710 in this embodiment.
When there is no axle within the radiation range 703 of the laser sensor 110, the vehicle 850 is detected as if it floats in the air, as shown in this figure. In that case, the distance between the reflection position (the measurement point on the tire surface) detected at a position higher than the road 820 and the reflection position (the measurement point on the road surface) detected on the road 820 becomes long, and the measurement distance Is greatly reduced.
On the other hand, when the axle is within the radiation range 703 of the laser sensor 110, as shown in FIG. 6, since the tire 856 in contact with the road 820 is detected, the reflection detected at a position higher than the road 820 is detected. The distance between the position (measurement point on the vehicle body) and the reflection position (measurement point on the road surface) detected on the road 820 becomes close, and the measurement distance does not change greatly.
The axle detection unit 220 uses this difference to detect the presence or absence of the axle.

FIG. 13 is a flowchart showing an example of the flow of the vehicle width calculation step S540 in this embodiment.
In the vehicle width calculation step S540, the vehicle width calculation unit 230 calculates the vehicle width 722 of the vehicle 850. The vehicle width calculation step S540 includes a sensor selection step S541, a minimum distance initialization step S542, a scanning direction selection step S543, a height determination step S544, a horizontal distance comparison step S545, a minimum distance update step S546, and a vehicle width calculation step S547. .

In the sensor selection step S541, the vehicle width calculation unit 230 selects one of the laser sensors 110 in order to process the two laser sensors 110a and 110b one by one as the processing device 911 processes the data. For example, the vehicle width calculation unit 230 first selects the laser sensor 110a. When the processing for the laser sensor 110a is completed, the vehicle width calculation unit 230 next selects the laser sensor 110b.
When any one of the laser sensors 110 is selected, the vehicle width calculation unit 230 proceeds to the minimum distance initialization step S542.
When the processing for the two laser sensors 110a and 110b ends and there is no laser sensor 110 to be selected, the vehicle width calculation unit 230 proceeds to the vehicle width calculation step S547.

  In the minimum distance initialization process S542, the vehicle width calculation unit 230 processes the data so that the horizontal distance between the laser sensor 110 selected in the sensor selection process S541 and the vehicle 850 is a sufficiently large value. (Eg, sensor interval 721). The storage device 920 stores the initialized horizontal distance.

In the scanning direction selection step S543, the vehicle width calculation unit 230 selects a radiation direction from among the radiation directions in which the laser scanning unit 113 radiates the radiated light 751 when the processing device 911 processes the data. In this case, the selection order may be from the top or from the bottom.
When the radial direction is selected, the vehicle width calculation unit 230 proceeds to the height determination step S544.
When the processing for all the radiation directions is completed and there is no radiation direction to be selected, the vehicle width calculation unit 230 returns to the sensor selection step S541 and selects the next laser sensor 110.

In the height determination step S544, the vehicle width calculation unit 230 processes the data so that the height of the reflection position calculated by the reflection position calculation unit 122 in the radial direction selected in the scanning direction selection step S543 is calculated. It is determined whether or not the height of the road 820 is higher.
When the height of the reflection position is higher than the height of the road 820, the vehicle width calculation unit 230 proceeds to the horizontal distance comparison step S545.
When the height of the reflection position is the height of the road 820, the vehicle width calculation unit 230 returns to the scanning direction selection step S543 and selects the next radiation direction.

In the horizontal distance comparison step S545, the vehicle width calculation unit 230 processes the data by the processing device 911 so that the reflection position calculated by the reflection position calculation unit 122 and the laser sensor 110 with respect to the radiation direction selected in the scanning direction selection step S543. And the horizontal distance stored in the storage device 920 are compared.
When the horizontal distance calculated by the reflection position calculation unit 122 is smaller, the vehicle width calculation unit 230 proceeds to the minimum distance update step S546.
When the horizontal distance calculated by the reflection position calculation unit 122 is larger, the vehicle width calculation unit 230 returns to the scanning direction selection step S543 and selects the next radiation direction.

In the minimum distance update step S546, the vehicle width calculation unit 230 updates the horizontal distance stored in the storage device 920 when the processing device 911 processes the data. The storage device 920 stores the distance between the reflection position calculated by the reflection position calculation unit 122 and the laser sensor 110 in the radiation direction selected in the scanning direction selection step S543 as a horizontal distance.
Thereafter, the vehicle width calculation unit 230 returns to the scanning direction selection step S543 and selects the next radiation direction.

  In the vehicle width calculation step S547, the vehicle width calculation unit 230 adds up the horizontal distances stored in the storage device 920 for the two laser sensors 110a and 110b as the processing device 911 processes the data. The vehicle width calculation unit 230 calculates the vehicle width 722 by subtracting the sum of the horizontal distances from the sensor interval 721 when the processing device 911 processes the data.

FIG. 14 is a flowchart showing an example of the flow of the number mark recognition process S550 in this embodiment.
In the number mark recognition process S550, the number mark recognition unit 280 reads the number plate. The number mark recognition process S550 includes an image input step S551, an edge extraction step S552, a block detection step S553, a number mark extraction step S554, a binarization step S555, and a character recognition step S556.

  In the image input step S551, the number mark recognition unit 280 inputs an image captured by the infrared imaging device 180 by the processing device 911 controlling the input device 902.

  In the edge extraction step S552, the number mark recognition unit 280 processes the data by the processing device 911, and performs image processing on the image input in the image input step S551 to detect a contour line. For example, the number mark recognition unit 280 differentiates an input image to generate a differential image.

  In the block detection step S553, the number mark recognition unit 280 processes the data by the processing device 911 to divide the image processed in the edge extraction step S552 into strip-shaped sectors, and to generate a plurality of images elongated in the vertical direction. Is generated. The number mark recognition unit 280 detects a block for each of a plurality of images generated by the processing device 911 processing data.

In the number mark extraction step S554, the number mark recognition unit 280 processes the data, and the number is recognized among the images input in the image input step S551 based on the blocks detected in the block detection step S553. Determine the position where the plate is reflected. The number mark recognition unit 280 processes the data, and cuts out an image of a portion in which the number plate appears from the image input in the image input step S551 based on the determined position.
The number mark recognition unit 280 determines whether the number plate of the vehicle 850 is a large plate or a middle plate (ordinary plate) based on the size of the clipped image.

  In the binarization step S555, the number mark recognition unit 280 binarizes the image cut out in the number mark cutout step S554 by the processing device 911 processing the data.

  In the character recognition step S556, the number mark recognition unit 280 recognizes characters in the image binarized in the binarization step S555 as the processing device 911 processes the data.

FIG. 15 is a diagram showing an example of an image that is a target of the number mark recognition process S550 in this embodiment.
The input image 731 is an original image captured by the infrared imaging device 180 and input by the number mark recognition unit 280. The input image 731 is a monochrome image for infrared imaging, but includes various intermediate tones and is not uniform.
The differential image 732 is an image in which the number mark recognition unit 280 has extracted the contour line in the edge extraction step S552.
The plurality of sector images 733 are images divided by the number mark recognition unit 280 in the block detection step S553.
The number mark recognition unit 280 projects each sector image 733 to generate a histogram 734. The number mark recognition unit 280 generates a detection waveform 735 binarized by applying a predetermined threshold to the generated histogram 734. The number mark recognition unit 280 creates a run length based on the generated detection waveform 735. The number mark recognition unit 280 detects a block based on the created run length and identifies the position of the license plate.
Based on the position of the license plate specified in this way, the number mark recognition unit 280 calculates the cutout range 736.
The number mark recognition unit 280 cuts out a portion corresponding to the cutout range 736 from the input image 731 and generates a cutout image 737.
The number mark recognition unit 280 recognizes characters (for example, “Shonan”, “55”, “A”, “42”, “49”, etc.) that appear in the cut out image 737 that has been cut out. Various characters are written on the license plate, but what is important in distinguishing the vehicle type is the portion of the classification number (in this example, “55”).

  Next, the principle of vehicle type discrimination based on window detection will be described.

FIG. 16 is a diagram showing an example of a positional relationship between a window and an axle detected by the vehicle type identification system 800 in this embodiment for the vehicle 850.
The horizontal axis represents time. If the vehicle 850 is moving at a substantially constant speed in the vehicle traveling direction 702, the time can be regarded as a distance in the traveling direction of the vehicle 850.
As in this example, when the vehicle 850 is a passenger car, the ratio of the windows detected to the entire length of the vehicle 850 is high.

FIG. 17 is a diagram showing another example of the positional relationship between the window and the axle detected by the vehicle type identification system 800 in this embodiment for the vehicle 850. In FIG.
As in this example, when the vehicle 850 is a truck, the ratio of detecting windows to the entire length of the vehicle 850 is low.

FIG. 18 is a diagram showing still another example of the positional relationship between the window and the axle detected by the vehicle type identification system 800 in this embodiment for the vehicle 850. In FIG.
As in this example, when the vehicle 850 is a bus, as in the case of a passenger car, the ratio of the windows detected to the entire length of the vehicle 850 is high.

As is clear from these three examples, the ratio of windows differs greatly between freight vehicles such as trucks and passenger cars and buses. This is because a passenger seat is generally provided with a window but a cargo bed is not provided with a window. Therefore, it is possible to distinguish between a passenger car or a bus and a truck based on the ratio of windows.
Also, the bonnet-type vehicle 850, like the passenger car in the example shown in FIG. 16, is characterized in that no window is detected before the first axle (front wheel). Therefore, if the ratio of the window to the total length of the vehicle 850 is calculated, the ratio of the window of the passenger car is reduced, and the difference from the truck may be reduced. Therefore, the ratio of windows between the first axle (front wheel) and the second axle (rear wheel) is compared. Thereby, a truck and a passenger car or a bus can be correctly identified.

FIG. 19 is a diagram illustrating an example of charge classification on a toll road.
In this example, there are five fee categories of “light”, “normal”, “medium”, “large”, and “extra large”. Of these, “light” includes motorcycles and light vehicles. “Normal” includes small cars other than “light” and ordinary passenger cars with a capacity of 10 or less. “Medium-sized” vehicles have a capacity of 11 to 29 passengers and a regular passenger car with a total weight of less than 8 tons [t], or a regular lorry with a maximum of 3 axles or less, a total weight of less than 8 tons, and a maximum loading capacity of less than 5 tons. Is included. “Large” includes regular passenger cars with a capacity of 11 to 29 people, total weight of 8t or more and vehicle length of less than 9m, and regular lorries with 3 to 4 axles and total weight within a specified range. . “Extra large” includes ordinary passenger cars with a capacity of 11 to 29 people, total weight of 8 tons or more, vehicle length of 9 meters or more, ordinary passenger cars with a capacity of 30 people or more, ordinary trucks with 4 or more axles within the specified range Is included. A route bus is treated as “large” if it falls under “extra large” in the above category.

  The vehicle type discrimination system 800 installed at the toll gate of the toll road having such a toll system has the goal of correctly discriminating five vehicle types that match the above-mentioned toll category.

FIG. 20 is a flowchart showing an example of the overall flow of the vehicle type discrimination process S560 in this embodiment.
In the vehicle type determination process S560, the vehicle type determination unit 290 determines the vehicle type of the vehicle 850. The vehicle type determination process S560 includes an axle number determination process S561, a multi-axis determination process S562, a vehicle width determination process S563, a number mark determination process S570, a window determination process S580, and a vehicle width determination process S566.

In the number-of-axes determination step S561, the vehicle type determination unit 290 determines whether the number of axles of the vehicle 850 is two or three or more based on the axle detected by the axle detection unit 220 when the processing device 911 processes the data. It is determined whether it is.
If the number of axles is three or more, the vehicle type determination unit 290 proceeds to the multi-axis determination step S562.
When the number of axles is two, the vehicle type determination unit 290 proceeds to the vehicle width determination step S563.

In the multi-axis determination step S562, the vehicle type determination unit 290 determines whether the vehicle type of the vehicle 850 is “medium”, “large”, or “extra large” when the processing device 911 processes the data. . It is assumed that a known method is used for the determination, and detailed description thereof is omitted here.
Thereafter, the vehicle type determination unit 290 ends the vehicle type determination process S560.

In the vehicle width determination step S563, the vehicle type determination unit 290 determines whether the vehicle type of the vehicle 850 is “light” based on the vehicle width 722 calculated by the vehicle width calculation unit 230 when the processing device 911 processes the data. Judge whether it is other than “light”. Specifically, for example, the vehicle type determination unit 290 compares the vehicle width 722 with a predetermined threshold value W ₁ (eg, 1.48 m), and is “light” if the vehicle width 722 is equal to or less than the threshold value W _1. judge.
When it is determined that the vehicle type of the vehicle 850 is “light”, the vehicle type determination unit 290 ends the vehicle type determination process S560.
When it is determined that the vehicle type of the vehicle 850 is other than “light”, the vehicle type determination unit 290 proceeds to the number mark determination process S570.

In the number mark determination processing S570, the vehicle type determination unit 290 processes the data so that the vehicle type of the vehicle 850 is “ordinary” based on the number plate size and the classification number read by the number mark recognition unit 280. ”,“ Medium ”or“ Large ”. In addition, since the information from the license plate cannot distinguish between the route bus and the sightseeing bus, in that case, the vehicle type determination unit 290 determines that the vehicle 850 is a “large bus”. Therefore, the “large bus” includes a case of “large” and a case of “extra large”. In addition, if the characters cannot be read due to dirt or bending of the license plate, or if the license plate is hidden behind the previous vehicle due to traffic jams or the like, the vehicle type determination unit 290 determines “unknown”. .
If it is determined as “large bus” or “unknown”, the vehicle type determination unit 290 proceeds to window determination processing S580.
When it determines with it being other than that, the vehicle type discrimination | determination part 290 complete | finishes vehicle type discrimination | determination process S560.

In the window determination process S580, the vehicle type determination unit 290 determines whether the vehicle 850 is a lorry or a passenger car based on the window detected by the window detection unit 130 by the processing device 911 processing the data. To do. If it is determined that the vehicle is a passenger car, the vehicle type determination unit 290 determines whether the vehicle type of the vehicle 850 is “normal”, “medium”, “large”, or “extra large”.
If it is determined that the item is “cargo”, the vehicle type determination unit 290 proceeds to the vehicle width determination step S566.
If it is determined that the vehicle type is not “cargo”, the vehicle type determination unit 290 ends the vehicle type determination process S560.

In the vehicle width determination step S566, the vehicle type determination unit 290 determines whether the vehicle type of the vehicle 850 is “normal” based on the vehicle width 722 calculated by the vehicle width calculation unit 230 when the processing device 911 processes the data. It is determined whether it is “medium” or “large”. Specifically, for example, the vehicle type determination unit 290 compares the vehicle width 722 with predetermined threshold values W ₂ (for example, 1.7 m) and W ₃ (for example, 2.2 m), and the vehicle width 722 is less than the threshold value W ₂ . if any, determined as "normal", if the threshold value W ₂ ultra-W ₃ equal to or less than the "medium-sized", if the threshold W ₃ than "large".
Thereafter, the vehicle type determination unit 290 ends the vehicle type determination process S560.

FIG. 21 is a flowchart showing an example of the flow of the number mark determination process S570 in this embodiment.
The number mark determination process S570 includes a classification number determination step S571 and two size determination steps S572 and S573.

In the classification number determination step S571, the vehicle type determination unit 290 determines whether the vehicle 850 is a “1” number based on the classification number read by the number mark recognition unit 280 by the processing device 911 processing the data. It is determined whether the number is “3” or “3” to “7”.
If the classification number is unreadable, the vehicle type determination unit 290 determines “unknown” and ends the number mark determination process S570.
When the numbers are “3” to “7”, the vehicle 850 is any one of a small passenger car, a small truck, and a normal passenger car (capacity of 10 or less). The vehicle type determination unit 290 determines that the vehicle type of the vehicle 850 is “normal”, and ends the number mark determination process S570.
If the number is “1”, the vehicle 850 is a regular truck. The vehicle type determination unit 290 proceeds to the size determination step S572.
When the number is “2”, the vehicle 850 is a normal passenger car (capacity of 11 or more). The vehicle type determination unit 290 proceeds to the size determination step S573.

In the size determination step S572, the vehicle type determination unit 290 determines whether the size of the license plate is a middle plate or a large plate by the processing device 911 processing data.
When the size of the license plate is the middle plate, the vehicle type determination unit 290 determines that the vehicle type of the vehicle 850 is “medium”, and ends the number mark determination process S570.
When the size of the license plate is a large plate, the vehicle type determination unit 290 determines that the vehicle type of the vehicle 850 is “large”, and ends the number mark determination process S570.

In the size determination step S573, the vehicle type determination unit 290 determines whether the size of the license plate is a middle plate or a large plate by the processing device 911 processing data.
When the size of the license plate is the middle plate, the vehicle type determination unit 290 determines that the vehicle type of the vehicle 850 is “medium”, and ends the number mark determination process S570.
When the size of the license plate is a large plate, the vehicle type determination unit 290 determines “large bus”, and ends the number mark determination processing S570.

FIG. 22 is a flowchart showing an example of the flow of the window determination processing S580 in this embodiment.
The window determination process S580 includes a presence / absence determination step S581, a ratio determination step S582, an upper end determination step S583, and a lower end determination step S584.

In the presence / absence determination step S581, the vehicle type determination unit 290 determines whether the window 854 is detected at any position of the vehicle 850 based on the detection result detected by the window detection unit 130 by the processing device 911 processing the data. Determine whether or not.
If no window 854 is detected, the vehicle 850 is considered to have no roof 853. Since the open car is a passenger car, the vehicle type determination unit 290 determines that the type of the vehicle 850 is “normal” and ends the window determination process S580.
When the window 854 is detected at any position, the vehicle type determination unit 290 proceeds to the ratio determination step S582.

In the ratio determination step S582, the vehicle type determination unit 290 performs processing based on the detection result detected by the window detection unit 130 and the detection result detected by the axle detection unit 220 when the processing device 911 processes the data. A ratio at which a window is detected between the eye axle and the second axle is calculated. The vehicle type determination unit 290 compares the calculated ratio with a predetermined threshold value a by the processing device 911 processing the data.
If the window detection ratio is less than or equal to the threshold a, the vehicle 850 is considered to be a truck. The vehicle type determination unit 290 determines that the item is “cargo”, and ends the window determination process S580.
If the window detection ratio is greater than the threshold value a, the vehicle 850 is considered to be a passenger car. The vehicle type determination unit 290 proceeds to the upper end determination step S583.

In the upper end determination step S583, the vehicle type determination unit 290 calculates the average height of the upper ends of the windows by the processing device 911 processing the data. The vehicle type determination unit 290 compares the calculated average with predetermined threshold values h ₁ and h _{2 as} the processing device 911 processes the data.
If the average height of the window top edge is the threshold value h ₁ below, the vehicle 850 is considered not to be a bus. The vehicle type determination unit 290 determines that the vehicle type of the vehicle 850 is “normal”, and ends the window determination process S580.
If the average height of the window top edge is below the threshold value h ₁ super h _2, it is considered that the vehicle 850 is mini bus. The vehicle type determination unit 290 determines that the vehicle type of the vehicle 850 is “medium”, and ends the window determination process S580.
If the average height of the window top edge is threshold h ₂ greater, the vehicle 850 is considered to be large buses. The vehicle type determination unit 290 proceeds to the lower end determination step S584.

In the lower end determination step S584, the vehicle type determination unit 290 calculates the average height of the lower ends of the windows by the processing device 911 processing the data. Vehicle type identification unit 290, by the processing device 911 processes the data, comparing the calculated average with a predetermined threshold value h _3.
If the average height of the window lower is the threshold value h ₃ below, the vehicle 850 is considered to be local bus. The vehicle type determination unit 290 determines that the vehicle type of the vehicle 850 is “large”, and ends the window determination process S580.
If the average height of the window bottom is threshold h ₃ than is believed that the vehicle 850 is sightseeing bus. This is because a sightseeing bus generally has a space for loading passengers' luggage at the lower part of the vehicle, and therefore has a higher window position than a route bus. The vehicle type determination unit 290 determines that the vehicle type of the vehicle 850 is “extra large” and ends the window determination process S580.

In this way, the vehicle type determination system 800 determines the vehicle type of the vehicle 850.
Discrimination by window detection is useful as it is often used as an aid when the license plate cannot be read, but the license plate cannot be read because it is hidden by the vehicle in front.

The window detection apparatus 100 in this embodiment includes a radiation unit 111, a light receiving unit 115, a reflection position calculation unit 122, and a window detection unit 130.
The radiating unit 111 emits radiated light 751 to the vehicle 850.
The light receiving unit 115 receives the reflected light 754 that is reflected when the radiated light 751 hits the vehicle 850.
The reflection position calculation unit 122 reflects the reflected light 751 on the vehicle 850 based on the time from when the radiation unit 111 emits the emitted light 751 to when the light receiving unit 115 receives the reflected light 754. The reflection position is calculated.
The window detection unit 130 detects the position of the window 854 of the vehicle 850 based on the reflection position calculated by the reflection position calculation unit 122.
Since the position of the window 854 of the vehicle 850 is known, the vehicle type of the vehicle 850 can be easily determined.

The radiating unit 111 scans the radiated light 751 in the vertical direction.
The reflection position calculation unit 122 includes a time from when the radiating unit 111 radiates the radiated light 751 to when the light receiving unit 115 receives the reflected light 754, and a radiation direction in which the radiating unit 111 radiates the radiated light 751. Based on the above, a plurality of reflection positions in the vertical direction where the radiated light 751 is reflected by the vehicle 850 are calculated.
The window detection unit 130 calculates a position where the reflection position is discontinuous based on the plurality of reflection positions consecutive in the vertical direction calculated by the reflection position calculation unit 122, and the calculated position is the window 854 of the vehicle 850. It is determined that it is the upper end or the lower end of.
Since the position where the reflection position is discontinuous is determined as the upper end or the lower end of the window 854, the position of the window 854 can be easily detected.

  When the window 854 is open, the light receiving unit 115 receives the reflected light 754 of the radiated light 751 hitting a person or an object in the vehicle, so that the reflection position in the direction corresponding to the window 854 is not missing. There is. In that case, the reflection position where the radiated light 751 hits a person or an object in the vehicle is farther than the reflection position where the reflected light hits the roof 853 or the main body 855. Therefore, the window detection unit 130 determines that the direction is a window not only when the reflection position is missing, but also when the distance from the laser sensor 110 to the reflection position becomes discontinuous (spot-like). It is good also as a structure to determine. Then, even when the window 854 is opened, the position of the window 854 can be detected correctly.

The vehicle type discriminating device (vehicle type discriminating system 800) in this embodiment includes a window detecting device 100 and a vehicle type discriminating device (vehicle type discriminating unit 290).
The vehicle type determination device determines the vehicle type of the vehicle 850 based on the position of the window 854 detected by the window detection device 100.
Since the vehicle type of the vehicle 850 is determined based on the position of the window 854, the vehicle type of the vehicle 850 can be easily and correctly determined even when the license plate cannot be read.

The vehicle type discriminating apparatus, based on the position of the window detected the window detector 100, determines the passenger car and the bus, when the height of the upper end of the window 854 is lower than the first upper threshold value h ₁ of the predetermined Further, it is determined that the vehicle 850 is a normal passenger car.
Since the normal passenger car and the bus are discriminated based on the height of the upper end of the window 854, the vehicle type of the vehicle 850 can be easily discriminated correctly.

The vehicle type discriminating apparatus, based on the position of the window 854 in which the window detector 100 has detected, to determine the microbus and large bus, when the height of the upper end of the window is higher than a predetermined second upper threshold value h ₂ Then, it is determined that the vehicle is a large bus.
Since the microbus and the large bus are discriminated based on the height of the upper end of the window 854, the vehicle type of the vehicle 850 can be easily discriminated correctly.

The vehicle type discriminating apparatus, based on the position of the window 854 that detected the window detector 100, determines the local bus and tourist bus, when the height of the lower end of the window 854 is higher than a predetermined lower threshold value h ₃ Then, it is determined that the vehicle 850 is a sightseeing bus.
Since the route bus and the sightseeing bus are discriminated based on the height of the lower end of the window 854, the vehicle type of the vehicle 850 can be easily discriminated correctly.

The vehicle type discrimination device further includes an axle detection device (axle detection unit 220).
The axle detection device detects the position of the axle of the vehicle 850.
The vehicle type determination device determines the vehicle type of the vehicle based on the position of the window 854 detected by the window detection device 100 and the position of the axle detected by the axle detection device.
By combining the position of the window 854 and the position of the axle, the vehicle type of the vehicle 850 can be easily determined correctly.

When the axle detection device detects two axles, the vehicle type discriminating device calculates a ratio in which a window 854 exists between the two axles, and the calculated ratio is smaller than a predetermined ratio threshold a. It is determined that vehicle 850 is a truck.
Since the passenger car and the cargo car are discriminated based on the ratio of the window 854 between the two axles, the vehicle type of the vehicle 850 can be easily discriminated correctly.

The axle detection device detects the position of the axle of the vehicle 850 based on the reflection position calculated by the reflection position calculation unit 122 of the window detection device.
Since the axle is detected based on the reflection position calculated by the reflection position calculation unit 122, it is not necessary to provide a separate sensor for detecting the axle. Since the number of parts of the vehicle type identification system 800 is reduced, the manufacturing cost can be reduced.

  In the vehicle type identification device (vehicle type identification system 800) described above, the laser sensor control unit 120 collects distance information and angle information from the laser sensors 110a, 110b installed on the left and right of the toll road (road 820) to the vehicle 850. To do. The vehicle width measurement unit (vehicle width calculation unit 230) measures the vehicle width based on data from the laser sensors 110a and 110b. The window glass detector (window detector 130) detects the window glass position. The axle detection unit (axle detection unit 220) detects the axle position. The window glass vehicle type determination unit (vehicle type determination unit 290) determines the vehicle type from the position of the axle and the position of the window glass. The vehicle type determination unit (vehicle type determination unit 290) determines the vehicle type from the window glass position and vehicle width measurement results.

  The window glass vehicle type determination unit determines a passenger car and an ordinary freight car from the positional relationship between the window glass position and the axle position.

  The window glass vehicle type determination unit determines the vehicle type of the micro bus for medium-sized vehicles, the route bus for large vehicles, and the sightseeing bus for oversized vehicles from the upper edge position and the lower edge position of the window glass.

  The vehicle type identification device detects the position of the window glass of the passing vehicle using a laser sensor. The vehicle type discriminating apparatus collects the positional relationship between the tire position and the window glass by detecting the tire position of the passing vehicle by the laser sensor and detecting the window glass position by the laser sensor. The vehicle type identification device performs a vehicle type identification process together with vehicle width information measured by the laser sensor. Since a vehicle is generally provided with a glass window at a passenger's boarding position, it is possible to classify a passenger car in which a window glass exists to the rear part of the vehicle and a freight car in which the window part does not exist because the rear part of the vehicle is for luggage. . For passenger cars, it can also be used to determine the type of bus by determining the height of the window position. This makes it possible to determine the vehicle type even for a vehicle in which the license plate cannot be recognized and the vehicle type is difficult to determine.

Embodiment 2. FIG.
A second embodiment will be described with reference to FIG.
In addition, about the part which is common in Embodiment 1, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  In the first embodiment, the configuration in which the axle is detected using the detection results of the pair of laser sensors 110a and 110b provided to detect the window 854 of the vehicle 850 has been described. In this embodiment, a configuration in which an axle detection sensor is provided separately from the window detection sensor will be described.

FIG. 23 is a front view showing an example of a vehicle type identification system 800 according to this embodiment.
The vehicle type identification system 800 includes four laser sensors 110a, 110b, 110c, and 110d.
Laser sensors 110a and 110b are provided at higher positions than in the first embodiment. The laser sensors 110a and 110b specialize in window detection. The radiation range 703a of the laser sensor 110a and the radiation range 703b of the laser sensor 110b cover a range where the window 854 of the vehicle 850 can exist, but it is not always necessary to cover all of the lower end of the vehicle 850.
The laser sensors 110c and 110d are specialized in axle detection. The radiation range 703c of the laser sensor 110c and the radiation range 703d of the laser sensor 110d are limited below the horizontal direction.

In this way, the laser sensors 110a and 110b are specialized for window detection, and separately, by providing the axle sensor laser sensors 110c and 110d, the radiation range 703 of the emitted light 751 can be narrowed down. The resolution of position detection can be increased and the detection accuracy is increased.
Further, by providing the window detection laser sensors 110a and 110b close to the height of the window 854 and providing the axle detection laser sensors 110c and 110d close to the height of the axle, the vehicle 850 Even when traveling on the edge of the road 820, the window 854 and the axle can be detected correctly.

  The vehicle width calculation unit 230 may be configured to calculate the vehicle width based on detection results detected by the four laser sensors 110a to 110d. Alternatively, the vehicle width calculation unit 230 may be configured to calculate the vehicle width based on detection results detected by the axle detection laser sensors 110c and 110d. In this case, the window detection is performed by one of the two laser sensors 110a and 110b, and the other laser sensor 110a and 110b may not be provided.

  Further, the axle detection sensor is not limited to the laser sensors 110c and 110d, and may be another sensor such as a footboard type axle sensor.

  DESCRIPTION OF SYMBOLS 100 Window detection apparatus, 110 Laser sensor, 111 Radiation part, 112 Laser oscillation part, 113 Laser scanning part, 115 Light receiving part, 116 Laser light receiving part, 117 Amplifying part, 120 Laser sensor control part, 121 Scan control part, 122 Reflection position Calculation unit, 130 window detection unit, 131 discontinuous position calculation unit, 132 window determination unit, 180 infrared imaging device, 200 control device, 210 traveling direction determination unit, 220 axle detection unit, 230 vehicle width calculation unit, 280 number mark recognition , 290 vehicle type discrimination unit, 701 imaging range, 702 vehicle traveling direction, 703 radiation range, 710 point cloud, 711 window upper end height, 712 window lower end height, 713 sensor inter-vehicle distance, 721 sensor interval, 722 vehicle width, 751 synchrotron radiation, 752 specular reflection light, 753 diffuse reflection light, 754 reflection light, 755 positive Transmitted light, 800 vehicle type identification system, 810 island, 820 road, 850 vehicle, 851 body, 852 glass, 853 roof, 854 window, 855 body, 856 tire, 901 output device, 902 input device, 911 processing device, 920 storage device .

Claims (11)

A radiation unit, a light receiving unit, a reflection position calculation unit, and a window detection unit;
The radiating part emits radiant light to the vehicle,
The light receiving unit receives reflected light reflected by the radiated light hitting the vehicle,
The reflection position calculation unit calculates a reflection position where the radiated light is reflected by the vehicle based on a time from when the radiating unit emits radiated light until the light receiving unit receives the reflected light.
The window detection unit detects the position of the window of the vehicle based on the reflection position calculated by the reflection position calculation unit. The radiating section scans the radiated light in the vertical direction,
The reflection position calculation unit is configured to generate the radiation based on a time from when the radiating unit emits radiated light to when the light receiving unit receives reflected light and a radiation direction in which the radiating unit emits radiated light. Calculate a plurality of reflection positions in the vertical direction where the light is reflected by the vehicle,
The window detection unit calculates a position where the reflection position is discontinuous based on a plurality of reflection positions consecutive in the vertical direction calculated by the reflection position calculation unit, and the calculated position is an upper end or a lower end of the vehicle window. The window detection device according to claim 1, wherein it is determined that The window detection device according to claim 1 or claim 2 and a vehicle type identification device,
The vehicle type discriminating device discriminates the vehicle type of the vehicle based on the position of the window detected by the window detecting device.   The vehicle type discriminating device discriminates between a normal passenger car and a bus based on the position of the window detected by the window detecting device, and when the height of the upper end of the window is lower than a predetermined first upper end threshold value, the vehicle The vehicle type identification device according to claim 3, wherein the vehicle type is determined to be a normal passenger car.   The vehicle type discriminating device discriminates between a microbus and a large bus based on the position of the window detected by the window detecting device, and when the height of the upper end of the window is higher than a predetermined second upper end threshold, the vehicle 5. The vehicle type identification device according to claim 3, wherein the vehicle type is determined to be a large bus.   The vehicle type discriminating device discriminates a route bus and a sightseeing bus based on the position of the window detected by the window detecting device, and when the height of the lower end of the window is higher than a predetermined lower threshold, the vehicle is The vehicle type identification device according to any one of claims 3 to 5, wherein it is determined that the bus is a sightseeing bus. The vehicle type identification device further includes an axle detection device,
The axle detection device detects the position of the axle of the vehicle,
The vehicle type discriminating device discriminates the vehicle type of the vehicle based on the position of the window detected by the window detecting device and the position of the axle detected by the axle detecting device. Item 7. The vehicle type identification device according to any one of Items 6 to 7.   When the axle detection device detects two axles, the vehicle type discriminating device calculates a ratio in which a window exists between the two axles, and when the calculated ratio is smaller than a predetermined ratio threshold, the vehicle The vehicle type identification device according to any one of claims 3 to 7, wherein the vehicle type identification device determines that the vehicle is a truck.   9. The vehicle type determination according to claim 7, wherein the axle detection device detects the position of the axle of the vehicle based on the reflection position calculated by the reflection position calculation unit of the window detection device. apparatus. In a window detection method in which a window detection device having a radiation unit, a light receiving unit, a reflection position calculation unit, and a window detection unit detects a vehicle window,
The radiating part emits radiant light to the vehicle,
The light receiving unit receives reflected light reflected by the radiated light hitting the vehicle,
The reflection position calculation unit calculates a reflection position where the radiated light is reflected by the vehicle based on a time from when the radiating unit emits radiated light until the light receiving unit receives the reflected light.
The window detection unit detects the position of the window of the vehicle based on the reflection position calculated by the reflection position calculation unit.   A vehicle type determination method, wherein the vehicle type of the vehicle is determined based on the position of the window detected by the window detection method according to claim 10.

Images