JP2851355B2 - Vehicle type identification device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vehicle type discrimination device applied to a vehicle type discrimination in a toll collection device for a toll road.

[Prior Art] FIG. 2 shows a configuration of a conventional vehicle type discriminating apparatus. In FIG. 2, a vehicle separator 01 detects entry of a vehicle, an imaging device 05 and an image processor 06 recognize a license plate of a vehicle, a tread plate 03 measures the number of axes and a tread of the vehicle, and a vehicle height detector 02 detects a vehicle height. The overhang detector 04 measures the distance from the tip of the vehicle to the first axis, and comprehensively determines the recognition result and the measurement result to determine the vehicle type.

[Problems to be Solved by the Invention] As shown in FIG. 2, in the conventional vehicle type discrimination, a tread is required for measuring the number of axes of the vehicle and measuring the tread. Since the treads must be buried in the road, when installing the equipment, the vehicle must be closed for a long time, and it has been considered difficult to introduce a vehicle type discriminating device on an existing road with heavy traffic. The measurement on the tread is a contact method performed by pressing the tire. Therefore, the service life is remarkably short as compared with other non-contact type sensors. Therefore, conventionally, it was necessary to replace the treads at regular intervals during the operation period.

The present invention conventionally performs the number of axes and tread measurement performed by a tread plate by image processing, and configures a vehicle type determination device that does not use a tread plate, thereby facilitating choreography of the equipment and reducing equipment costs, The purpose is to improve reliability.

[Means for Solving the Problems] 1. From the front image of the vehicle taken at the start of the approach, the left and right tire positions of the first axis are detected, and the tread is obtained.

2.During the vortex of the vehicle, repeatedly image the side of the vehicle,
By detecting tires from the image and evaluating the history of tire position changes, forward / reverse is determined, and the number of passing axes until the end of vehicle passing is determined.

By providing these two functions, it is possible to configure a vehicle type determination device that does not use a tread.

[Action] By performing tread measurement and counting the number of axes conventionally performed using a tread using these two means,
A vehicle type discrimination device that does not use a tread can be configured.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a configuration diagram of a vehicle type identification device according to the present invention. In the figure, 1 is a vehicle separator, 2 is a vehicle height detector, 4 is an overhang detector, 5 is a front image pickup device, 6 is an image processor, and 7 is a side image pickup device.

The vehicle separator 1 is an optical sensor for detecting a vehicle passing through a lane. A plurality of light beams pass through the vehicle in the axial direction of the lane. It detects and outputs a vehicle detection signal all the time during passing. At the time of vehicle intrusion, the image processor 6 issues an imaging command to the front image capturing device 5 at the same time when the vehicle detection signal is turned on, and captures an image of the front of the vehicle. The front image pickup device 5 outputs a still image video signal, and the image processor 6 takes in the signal, performs A / D conversion, and processes the digital image.
Perform license plate recognition and tread measurement.

Also, while the vehicle is passing (while the vehicle detection signal is on)
, The side of the vehicle is imaged and output every 1/30 second by the side image pickup device 7. The image processor 6 extracts the tire portion by capturing and processing the image. By performing this processing every time (every 1/30 second), the movement of the tire is detected and the number of axes is counted.

In addition, the image processor 6 can perform the process of performing license plate recognition and tread measurement from the vehicle front image and the process of sequentially extracting the tire portion from the vehicle side image and counting the number of axes. .

At the end of the passage of the vehicle, the first axis tread, the Nahan-plate size (large plate / medium plate), the vehicle type classification number, two digits obtained by the processing of the image processor 6, the number of axes, and the vehicle height obtained by other sensors The vehicle type is determined by comprehensively determining the measurement results such as the vehicle length and the vehicle length.

Next, an example of a method for realizing each of (1) tread measurement by image processing and (2) axis number measurement by image processing, which are features of the present invention, will be described.

(1) Tread Measurement by Image Processing Tread measurement by image processing is performed by processing the front image of the vehicle captured by the front image capturing device 5 at the moment when the vehicle is detected by the vehicle separator 1.

FIG. 3A shows an imaging range of a front image of the vehicle.
This imaging range is set so that both tires are reflected in the range.

The image processor 6 obtains the outline of the vehicle entering the lane by processing the front image of the vehicle. Normally, this can be easily performed by storing an image when the vehicle is not passing through and performing a subtraction process with the vehicle front image.
FIG. 3 (b) shows the result of extracting the outline of the vehicle, and the inside of the detected outline is indicated by oblique lines.

Next, as shown in FIG.
It is divided into II and III areas. The region II is a range where the tire portion cannot exist in the front image of the vehicle, and the division points a and b can be obtained by statistical processing or the like, and may be fixed positions as long as the visual field and the position of the imaging device are the same. . Next area
From “I” and “III”, “the lowest point on the outline” is obtained and defined as P1 and P2. Since these are almost equal to the points in contact with the road surface, as shown in FIG. 3 (c), they are regarded as points P1 'and P2' on the road surface, and the distance between P1 'and P2' is obtained.
Let it be the tread.

(2) Measurement of the number of axes by image processing The image processing device 6 takes in the vehicle side image that is imaged and output every 1/30 second by the side image imaging device 7 every time, and performs the processing. Find the position of the tire. The imaging range of the vehicle side image is a range in which two-axis tires are not captured at the same time.
Either there is no tire or only one axis. The image processor 6 recognizes the start of passing through the axle by changing the state of “no tire” from the state of “no tire” to the state of “with tire” in the vehicle side image, and changes the state of “with tire” to “no tire”. The end of the passing of the axle is recognized by having become. At the end of the passage of the axle, the change in the tire position from the start of the passage is examined to determine whether the axle is moving forward or backward. At the start of the passage of the vehicle, the number of measurement axes is set to 0, and the number of axles of the vehicle can be obtained by setting +1 when passing through the forward axis and −1 when passing through the reverse axis. FIG. 4 shows an example of shaft passage detection. The figure shows time t0, t1, t2, t3, where the interval is 1/30 second.
At t4, t0, t4 are in the “no tire” state, and t1, t2, t3
Shows an example in which the vehicle has passed through the forward axis at time t4 when the vehicle is in the “with tire” state. FIGS. 7A and 7B show a vehicle side image at t1 to t3, a contour line extraction result, and a detected tire portion. FIG. 7C shows that the position of the tire changes from P1 to P2 to P3 corresponding to t1 to t3. At the end of the axis passage (t4), it is determined that the vehicle is moving forward when the position of the tire is moving from left to right with time (in a case where the right side surface in the traveling direction of the vehicle is imaged).

[Effects of the Invention] As described above, according to the present invention, since a tread plate is not required unlike the conventional vehicle type discriminating apparatus, the installation of the equipment becomes easy and the equipment cost can be reduced. In addition, reliability is improved by eliminating the necessity of replacing a tread.

[Brief description of the drawings]

FIG. 1 is a device configuration diagram of a vehicle type identification device according to a first embodiment of the present invention, FIG. 2 is a device configuration diagram of a conventional vehicle type identification device, FIG. 3 is an example of tread measurement by image processing, and FIG. It is an example of axis passage detection by image processing. DESCRIPTION OF SYMBOLS 1 ... Vehicle separator, 2 ... Vehicle height detector, 4 ... Overhang detector, 5 ... Front image pickup machine, 6 ... Image processor, 7 ... Side image pickup machine.

Claims (1)

(57) [Claims] 1. A vehicle separator for detecting a vehicle passing a toll gate at a specific position, and a front image for capturing a front surface of the vehicle when a vehicle detection signal output from the vehicle separator is turned on. An image capturing device, a side image capturing device that continuously captures an image of a side surface of a vehicle while the vehicle detection signal is on, a vehicle height detector that detects a vehicle height of the passing vehicle, and an overhang of the passing vehicle. An overhang detector for detecting, and a license plate portion is extracted from a front image of the vehicle captured by the front image capturing device to determine a license plate size and recognize a vehicle type classification number, and to perform left and right left and right movements on the first axis. First image processing means for detecting a tire position to obtain a tread; detecting a tire position from a side image of the vehicle taken by the side image pickup device; And the second image processing means for determining the number of passing axes until the end of vehicle passing, the determination result of the license plate size by the first image processing means, the recognition result of the vehicle type classification number, The vehicle type is comprehensively determined from a tread measurement result, an axis number measurement result by the second image processing means, an overhang measurement result by the overhang detector, and a vehicle height measurement result by the vehicle height detector. Vehicle type discriminating device.