JP7077439B1 - Vehicle type discrimination system and vehicle type discrimination program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a vehicle type discrimination system for discriminating a vehicle type classification with high accuracy. A vehicle type determination system 1000 includes a category determination unit 250 and a vehicle type classification determination unit 290. The category determination unit 250 determines the category based on the side image captured from the side of the approaching vehicle entering the toll road. The vehicle type classification determination unit 290 determines the vehicle type classification of the approaching vehicle based on the category. The vehicle type classification represents the classification for determining the toll of the toll road. The category is determined according to the vehicle type classification and the shape of the vehicle belonging to the vehicle type classification. [Selection diagram] Fig. 3

Description

The present invention relates to a vehicle type discrimination system and a vehicle type discrimination program.

At the entrance of a toll road such as an expressway, the vehicle type classification of the vehicle entering the toll road is determined in order to determine the toll of the vehicle traveling on the toll road.

For example, Patent Document 1 includes a camera that captures a vehicle entering a toll road from diagonally above and a laser scanner that scans the vehicle from the side, an image captured by the camera, and a detection point detected by the laser scanner. Based on the above, a device for determining the vehicle type classification of a vehicle is disclosed. This device uses a laser scanner to determine the vehicle type classification based on the number of axles and the distance between the axles.

Japanese Unexamined Patent Publication No. 2017-220076

Vehicles traveling on toll roads are classified into different vehicle types according to the presence or absence of a tolled vehicle, but it is difficult to determine the presence or absence of a towed vehicle based only on the number of axles and the distance between the axles. In general, the license plate is imaged and the vehicle type classification is determined based on the vehicle model number of the license plate, but the vehicle model number of the license plate may not be recognized due to snow or dirt.

In view of the above situation, one of the purposes is to provide a vehicle type discrimination system that discriminates vehicle type classification with high accuracy. Other purposes can be understood from the following description and description of embodiments.

Hereinafter, means for solving the problem will be described using the numbers and codes used in the embodiment of the invention. These numbers and codes are added in parentheses for reference in order to show an example of the correspondence between the description of the claims and the embodiment for carrying out the invention. Therefore, the scope of claims should not be construed in a limited manner by the description in parentheses.

The vehicle type determination system (1000) according to the embodiment for achieving the above object includes a category determination unit (250) and a vehicle type classification determination unit (290). The category determination unit (250) determines the category based on the side image captured from the side of the approaching vehicle entering the toll road. The vehicle type classification determination unit (290) determines the vehicle type classification of the approaching vehicle based on the category. The vehicle type classification represents the classification for determining the toll of the toll road. The category is determined according to the vehicle type classification and the shape of the vehicle belonging to the vehicle type classification.

The vehicle type determination program (500) according to the embodiment for achieving the above object tells the arithmetic unit (220) to determine the category based on the side image captured from the side of the approaching vehicle entering the toll road. Let it run. Further, the vehicle type determination program (500) causes the arithmetic unit (220) to determine the vehicle type classification of the approaching vehicle based on the category. The vehicle type classification represents the classification for determining the toll of the toll road. The category is determined according to the vehicle type classification and the shape of the vehicle belonging to the vehicle type classification.

According to the above embodiment, the vehicle type discrimination system can discriminate the vehicle type classification with high accuracy.

It is a schematic diagram of the vehicle type discrimination system in one embodiment. It is a block diagram of the processing apparatus of the vehicle type discrimination system in one Embodiment. It is a functional block diagram realized by the processing apparatus in one Embodiment. It is a flowchart which shows the processing by the processing apparatus in one Embodiment. It is a flowchart which shows the process by the vehicle type classification determination part in one Embodiment. It is a flowchart which shows the process by the vehicle type classification determination part in one Embodiment. It is a flowchart which shows the process by the vehicle type classification determination part in one Embodiment. It is a flowchart which shows the process by the vehicle type classification determination part in one Embodiment. It is a flowchart which shows the process by the vehicle type classification determination part in one Embodiment. It is a flowchart which shows the process by the vehicle type classification determination part in one Embodiment.

(Embodiment)
In the present embodiment, as shown in FIG. 1, a vehicle type discrimination system 1000 installed at the entrance of a toll road such as an expressway will be described as an example. At the entrance of the toll road, a right island 20R is provided on the right side and a left island 20L is provided on the left side along a traveling lane 10 through which a vehicle, for example, a car or a two-wheeled vehicle passes. The right island 20R is provided with a toll ticket issuing machine 300 and a start controller 450. The left island 20L is provided with a number imager 100, a vehicle side imager 110, a vehicle front imager 120, a processing device 200, and a starting vehicle detector 400.

The vehicle type determination system 1000 uses the vehicle side imager 110 to classify approaching vehicles entering the toll road into categories such as passenger cars, freight vehicles, tractor heads, buses, motorcycles, and large special vehicles. The category is determined according to the vehicle type classification and the shape of the vehicle belonging to the vehicle type classification. The vehicle type discrimination system 1000 classifies approaching vehicles into categories and uses the classified categories to determine the vehicle type classification with high accuracy. Here, the vehicle type classification represents the classification for determining the toll of the toll road.

The configuration of the vehicle type discrimination system 1000 will be described. The vehicle type discrimination system 1000 includes a number imager 100, a vehicle side imager 110, a vehicle front imager 120, and a processing device 200. The toll ticket issuing machine 300 issues a toll ticket according to the vehicle type classification determined by the vehicle type determination system 1000. The driver of the approaching vehicle receives the toll ticket issued by the toll ticket issuing machine 300. When the toll ticket is taken from the toll ticket issuing machine 300, the start controller 450 drives the actuator 460 to raise the start control rod 470. When the start control rod 470 is raised, the driver of the approaching vehicle advances the approaching vehicle. The starting vehicle detector 400 detects the advance of the approaching vehicle and determines that the approaching vehicle has left the traveling lane 10. When the approaching vehicle exits the traveling lane 10, the start controller 450 drives the actuator 460 to lower the start control rod 470.

The number imager 100 is configured to image the license plate of the approaching vehicle that has entered the traveling lane 10. The license plate image of the license plate captured is converted into an electric signal and transmitted to the processing device 200. The number imager 100 is, for example, a camera having high sensitivity and a high shutter speed capable of photographing the license plate of an approaching vehicle moving at a high speed, and may be provided with a strobe.

The vehicle side imager 110 is configured to image the side surface of the approaching vehicle. The side image captured on the side surface of the vehicle is converted into an electric signal and transmitted to the processing device 200. The vehicle side imager 110 is, for example, a camera provided with a camera capable of capturing a range of 180 degrees in a horizontal direction, for example, in the front-rear direction of the approaching vehicle, for example, in order to reliably image the side surface of the approaching vehicle, for example, a camera equipped with a fisheye lens.

The vehicle front imager 120 is configured to image the front of the approaching vehicle. The vehicle front imager 120 images the vehicle from above the front of the approaching vehicle. The captured front image is converted into an electric signal and transmitted to the processing device 200. The vehicle front imager 120 is, for example, a wide-angle camera that captures the front surface of an approaching vehicle.

As shown in FIG. 2, the processing device 200 includes an input / output device 210, a communication device 230, an arithmetic unit 220, and a storage device 240. The processing device 200 is, for example, a computer. Information for the arithmetic unit 220 to execute the process is input to the input / output device 210. Further, the input / output device 210 outputs the result of the processing executed by the arithmetic unit 220. The input / output device 210 includes various input devices and output devices, and includes, for example, a keyboard, a mouse, a microphone, a display, a speaker, a touch panel, and the like. The input / output device 210 may be omitted.

The communication device 230 is electrically connected to each device and communicates with each device. The communication device 230 acquires images captured by each of the number imager 100, the vehicle side imager 110, and the vehicle front imager 120, and transfers them to the arithmetic unit 220. Further, the signal generated by the arithmetic unit 220 is transferred to the ticket issuing machine 300. The communication device 230 includes various interfaces such as NIC (Network Interface Card) and USB (Universal Serial Bus).

The storage device 240 stores various data for discriminating the vehicle type classification of the approaching vehicle that has entered the traveling lane 10, for example, the vehicle type discriminating program 500. The storage device 240 is used as a non-transitory tangible storage medium for storing the vehicle type discrimination program 500. The vehicle type determination program 500 may be provided as a computer program product recorded on a computer-readable storage medium 1, or may be provided as a computer program product that can be downloaded from a server.

The arithmetic unit 220 performs various data processing for determining the vehicle type classification of the approaching vehicle. For example, the arithmetic unit 220 includes a central processing unit (CPU) and the like.

By reading and executing the vehicle type determination program 500, the arithmetic unit 220 reads and executes the category determination unit 250, the axle number determination unit 260, the vehicle length determination unit 270, the license plate determination unit 280, and the license plate determination unit 280. The vehicle type classification determination unit 290 is realized. The category determination unit 250 determines the category of the approaching vehicle based on the side image captured by the vehicle side imager 110. The axle number determination unit 260 determines the number of axles of the approaching vehicle based on the side captured image. The vehicle length determination unit 270 determines the number of axles of the approaching vehicle based on the side captured image. The license plate determination unit 280 determines the vehicle model number attached to the license plate of the approaching vehicle based on the license plate image captured by the license plate imager 100. The vehicle type classification determination unit 290 determines the vehicle type classification of the approaching vehicle based on each of the determined information.

Further, as shown in FIG. 2, the communication device 230 of the processing device 200 is communicably connected to the toll ticket issuing machine 300, the starting vehicle detector 400, and the starting controller 450. The arithmetic unit 220 receives signals from the toll ticket issuing machine 300 and the starting vehicle detector 400 via the communication device 230 to control the starting controller 450.

(Operation of vehicle type discrimination system)
When the approaching vehicle enters the traveling lane 10, the arithmetic unit 220 of the processing device 200 executes the process shown in FIG. 4 to determine the vehicle type classification of the approaching vehicle. For example, when the arithmetic unit 220 detects an approaching vehicle in the front image captured from the vehicle front imager 120, it determines that the approaching vehicle has entered the traveling lane 10 and executes the process shown in FIG. The vehicle type determination system 1000 may be provided with an approach detection sensor for detecting the approach of an approaching vehicle at the entrance of the traveling lane 10 shown in FIG. The approach detection sensor may include, for example, a laser transceiver that transmits a laser from the right island 20R to the left island 20L, and may detect the approach of the approaching vehicle by blocking the laser from the approaching vehicle.

The category determination unit 250 realized by the arithmetic unit 220 detects the area of the approaching vehicle from the side image captured by the vehicle side imager 110 in step S110 shown in FIG. For example, the category determination unit 250 may use an artificial intelligence program for detecting the area of the approaching vehicle from the side captured image. Artificial intelligence programs are learned to detect various vehicle areas from images. Further, when the vehicle side imager 110 includes a fisheye lens, the category determination unit 250 extracts a predetermined determination region including the center of the side image, for example, a determination region within a predetermined angle from the optical axis of the vehicle side imager 110. Then, the area of the approaching vehicle may be detected from the extracted determination area. For example, the category determination unit 250 extracts a determination region within 45 degrees from the optical axis of the vehicle side imager 110 from the side image, and detects the region of the approaching vehicle from the extracted determination region. As a result, when the vehicle side imager 110 includes a fisheye lens, it is possible to suppress the influence of distortion of the captured image that occurs in the vicinity of the outer edge of the side captured image.

In step S120, the category determination unit 250 determines whether or not the approaching vehicle captured in the side image is a connected vehicle. A connected vehicle represents a vehicle that is towing another vehicle. For example, the category determination unit 250 detects both the towed vehicle and the towed vehicle as different areas of the approaching vehicle as the area of the approaching vehicle from the side captured image. When the distance between the areas of the two approaching vehicles is shorter than the predetermined distance, the category determination unit 250 determines that the vehicle is a connected vehicle. The category determination unit 250 executes the process of step S150 when the approaching vehicle is not a connected vehicle. When the approaching vehicle is a connected vehicle, the category determination unit 250 executes the process of step S130.

In step S130, the category determination unit 250 determines the region of the towed vehicle and the region of the towed vehicle from the side captured image. For example, the category determination unit 250 determines that, of the regions of the two approaching vehicles determined to be connected vehicles, the traveling direction of the approaching vehicle, for example, the region existing in front is the region of the towing vehicle, and the reverse of the traveling direction of the approaching vehicle. The direction, for example, the area existing behind is determined to be the area of the towed vehicle.

In step S140, the axle number determination unit 260 determines the number of axles of the towed vehicle based on the side captured image. The axle number determination unit 260 detects an axle region from the side captured image. For example, the axle number determination unit 260 may use an artificial intelligence program for detecting an axle region from a side captured image. Artificial intelligence programs are trained to detect axle areas from images. Further, when the vehicle side imager 110 includes a fisheye lens, the axle number determination unit 260 extracts a predetermined region including the center of the side image, for example, a determination region within a predetermined angle from the optical axis of the vehicle side imager 110. Then, the area of the axle may be detected from the extracted determination area.

The axle number determination unit 260 determines the number of axles of the towed vehicle based on the detected axle region and the towed vehicle region. The axle number determination unit 260 determines the axle area of the towed vehicle based on the axle area and the towed vehicle area, and counts the determined axle area of the towed vehicle to count the towed vehicle. Determine the number of axles. For example, the axle number determination unit 260 determines that the area of the axle included in the area of the towed vehicle is the area of the axle of the towed vehicle. Further, the axle number determination unit 260 may determine the region of the axle that overlaps the region of the towed vehicle as the region of the axle of the towed vehicle in the traveling direction of the approaching vehicle, for example, the horizontal direction.

In step S150, the category determination unit 250 determines the category of the towing vehicle of the approaching vehicle, for example, the connected vehicle, based on the side captured image. The category determination unit 250 extracts the vehicle area of the detected approaching vehicle, and determines the category of the approaching vehicle based on the extracted vehicle area from the side captured image. When the approaching vehicle is a connected vehicle, the category determination unit 250 determines the category of the towing vehicle with the towing vehicle as the approaching vehicle. For example, the category determination unit 250 extracts a region of the towing vehicle and determines the category of the towing vehicle based on the extracted region of the side captured image.

The category determination unit 250 determines the category of the approaching vehicle by using, for example, an artificial intelligence program. Artificial intelligence programs are trained to determine different vehicle categories from images. The artificial intelligence program is learned to distinguish between a plurality of vehicle types included in one category, for example, a light vehicle included in a light vehicle, a two-wheeled vehicle, and a three-wheeled vehicle. When the artificial intelligence program determines that the approaching vehicle is one of a light vehicle, a two-wheeled vehicle, and a three-wheeled vehicle, the category determination unit 250 determines that the approaching vehicle is included in the light vehicle category.

Further, the category determination unit 250 determines that the ordinary vehicle includes, for example, a sedan excluding a light vehicle, an RV (Recreational Vehicle) vehicle, a sports car, a minivan, a wagon, a hatchback, and the like. In this case, the artificial intelligence program is learned to distinguish and determine sedans, RVs, sports cars, minivans, wagons, hatchbacks, etc., excluding light vehicles. When the artificial intelligence program determines that the vehicle type is included in the ordinary vehicle, the category determination unit 250 determines that the approaching vehicle is included in the ordinary vehicle.

Similarly, the artificial intelligence program included in the category determination unit 250 distinguishes and determines a plurality of vehicle types included in one category for other categories such as large special vehicles, trucks, tractor heads, and buses. Learned to do. The category determination unit 250 determines the category of the approaching vehicle based on the vehicle type determined by the artificial intelligence program.

In step S160, the vehicle length determination unit 270 determines the vehicle length of the approaching vehicle based on the side captured image. The vehicle length determination unit 270 determines the vehicle length of the approaching vehicle based on the extracted area of the approaching vehicle. For example, the vehicle length determination unit 270 calculates the length of the region of the approaching vehicle extracted in the side captured image in the traveling direction of the approaching vehicle, for example, the horizontal direction, and the vehicle of the approaching vehicle is based on the calculated length. Determine the length. The vehicle length determination unit 270 is preset with an arithmetic expression for calculating the vehicle length of the approaching vehicle from the length of the region of the approaching vehicle in the side captured image. As shown in FIG. 1, the approaching vehicle is imaged by the vehicle side imager 110 while traveling in the traveling lane 10 sandwiched between the right island 20R and the left island 20L. Therefore, when the vehicle side imager 110 images an approaching vehicle, the distance from the vehicle side imager 110 to the approaching vehicle is included in a limited predetermined range. As a result, the vehicle length determination unit 270 can calculate the vehicle length of the approaching vehicle from the length of the approaching vehicle captured in the side image.

When the approaching vehicle is a connected vehicle, the vehicle length determination unit 270 determines the length of the approaching vehicle based on the area of the towing vehicle. For example, the vehicle length determination unit 270 calculates the length of the region of the towing vehicle in the side captured image in the traveling direction of the approaching vehicle, for example, the horizontal direction, and enters the vehicle length of the towing vehicle based on the calculated length. Determined as the length of the vehicle.

In step S170 shown in FIG. 4, the axle number determination unit 260 determines the number of axles of the approaching vehicle based on the side captured image. The axle number determination unit 260 detects an axle region from the side captured image. The axle number determination unit 260 determines the number of axles of the approaching vehicle based on the detected axle area and the approaching vehicle area. The axle number determination unit 260 determines the area of the axle of the approaching vehicle based on the area of the axle and the area of the approaching vehicle, and counts the area of the determined axle of the approaching vehicle to determine the number of axles of the approaching vehicle. judge. For example, the axle number determination unit 260 determines that the area of the axle included in the area of the approaching vehicle is the area of the axle of the approaching vehicle. Further, the axle number determination unit 260 may determine the region of the axle that overlaps the region of the approaching vehicle as the region of the axle of the approaching vehicle in the traveling direction of the approaching vehicle, for example, the horizontal direction.

When the approaching vehicle is a connected vehicle, the axle number determination unit 260 determines the number of axles of the towing vehicle based on the side captured image. For example, the axle number determination unit 260 determines the number of axles of the towing vehicle as the number of axles of the approaching vehicle based on the detected axle region and the towing vehicle region.

In step S180, the license plate determination unit 280 determines the vehicle model number of the license plate based on the license plate image captured by the license plate imager 100. The license plate determination unit 280 extracts a license plate area from the license plate captured image, and determines a vehicle model number included in the extracted license plate area. For example, the license plate determination unit 280 determines the vehicle model number by performing character recognition in the area of the extracted license plate. Further, the license plate determination unit 280 determines whether the license plate is a medium plate (medium-sized plate) or a large plate (large plate) based on the size of the area of the extracted license plate.

In step S190, the vehicle type classification determination unit 290 determines the vehicle type classification of the approaching vehicle based on each determined information, specifically, the category of the approaching vehicle, the length of the vehicle, the number of axles, and the like. For example, the vehicle type classification determination unit 290 executes the processes shown in FIGS. 5A to 5F to determine the vehicle type classification of the approaching vehicle. When the approaching vehicle is a connected vehicle, first, the vehicle type classification determination unit 290 executes the processes shown in FIGS. 5A to 5F to determine the vehicle type classification of the towing vehicle. Next, the vehicle type classification determination unit 290 determines the vehicle type classification of the approaching vehicle based on the number of axles of the towed vehicle and the vehicle type classification of the towed vehicle.

The vehicle type classification determination unit 290 determines whether the license plate can be recognized by the license plate imager 100 in step S210 shown in FIG. 5A. Specifically, the vehicle type classification determination unit 290 determines whether the license plate determination unit 280 can determine the vehicle type number of the license plate. When the license plate determination unit 280 can determine the license plate vehicle type number, the vehicle type classification determination unit 290 executes the process of step S220 in order to determine the vehicle type classification using the license plate number. When the license plate determination unit 280 cannot determine the vehicle type number of the license plate, the vehicle type classification determination unit 290 executes the process of step S510 shown in FIG. 5D in order to determine the vehicle type classification without using the vehicle type number.

In step S220 shown in FIG. 5A, the vehicle type classification determination unit 290 performs different processing depending on the first digit of the vehicle type number determined by the license plate determination unit 280, specifically, the leftmost number of the vehicle type number. When the first digit of the vehicle type number is "8", the vehicle type classification determination unit 290 executes the process of step S230 and determines that the vehicle type classification of the approaching vehicle is a special purpose vehicle. In step S230, when the approaching vehicle is a connected vehicle, the vehicle type classification determination unit 290 determines that the vehicle type classification of the towing vehicle is a special purpose vehicle.

In step S220, the vehicle type classification determination unit 290 executes the process of step S240 when the first digit of the vehicle type number is any of "3" to "7". In step S240, the vehicle type classification determination unit 290 determines whether the approaching vehicle is a light vehicle based on the front image captured by the vehicle front imager 120 shown in FIG. For example, the vehicle type classification determination unit 290 extracts a region of the license plate of the approaching vehicle from the front captured image, and when the extracted license plate region is yellow, determines that the approaching vehicle is a light vehicle. Further, the vehicle type classification determination unit 290 may determine that the approaching vehicle is a light vehicle when the category determination unit 250 determines that the category of the approaching vehicle is a light vehicle.

In step S240, when the vehicle type classification determination unit 290 determines that the approaching vehicle is a light vehicle, the process of step S250 is executed to determine that the vehicle type classification of the approaching vehicle is a light vehicle. If it is determined that the approaching vehicle is not a light vehicle, the vehicle type classification determination unit 290 determines that the vehicle type classification of the approaching vehicle is a normal vehicle.

In step S220, the vehicle type classification determination unit 290 executes the process of step S310 shown in FIG. 5B when the first digit of the vehicle type number is "1". In step S310, the vehicle type classification determination unit 290 determines whether or not the size of the license plate is the middle plate. The size of the license plate is determined by the license plate determination unit 280 in step S180 shown in FIG. 4 whether or not it is a middle plate. The vehicle type classification determination unit 290 acquires whether or not the size of the license plate of the approaching vehicle is the middle plate based on the determination by the license plate determination unit 280. When the size of the license plate of the approaching vehicle is the middle plate, the vehicle type classification determination unit 290 executes the process of step S320 shown in FIG. 5B to determine that the vehicle type classification of the approaching vehicle is a medium-sized vehicle. The vehicle type classification determination unit 290 executes the process of step S330 when the size of the license plate of the approaching vehicle is not the middle plate.

In step S330, the vehicle type classification determination unit 290 performs different processing depending on the number of axles of the approaching vehicle. When the number of axles of the approaching vehicle is "2", the vehicle type classification determination unit 290 executes the process of step S340. In step S340, the vehicle type classification determination unit 290 determines whether the category of the approaching vehicle determined by the category determination unit 250 is a tractor head. When the category of the approaching vehicle is the tractor head, the vehicle type classification determination unit 290 executes the process of step S350 and determines that the vehicle type classification of the approaching vehicle is a medium-sized vehicle. When the category of the approaching vehicle is not the tractor head, the vehicle type classification determination unit 290 executes the process of step S360 and determines that the vehicle type classification of the approaching vehicle is a large vehicle.

In step S330, when the number of axles of the approaching vehicle is "3", the vehicle type classification determination unit 290 executes the process of step S360 and determines that the vehicle type classification of the approaching vehicle is a large vehicle.

In step S330, when the number of axles of the approaching vehicle is "4", the vehicle type classification determination unit 290 executes the process of step S370. In step S370, the vehicle type classification determination unit 290 determines whether the approaching vehicle is equal to or less than the limit of the vehicle restriction order. For example, the vehicle type classification determination unit 290 determines whether the vehicle length and the vehicle height of the approaching vehicle are equal to or less than the limit of the vehicle restriction order. The vehicle length is determined by the vehicle length determination unit 270 in step S160 shown in FIG. The vehicle type classification determination unit 290 acquires the vehicle length determined by the vehicle length determination unit 270. Further, the vehicle type classification determination unit 290 determines the vehicle height of the approaching vehicle based on the side captured image. For example, the vehicle type classification determination unit 290 determines the vehicle height of the approaching vehicle based on the length of the region of the approaching vehicle in the direction perpendicular to the traveling direction of the approaching vehicle, for example, the vertical direction in the side captured image. When the vehicle length and the vehicle height of the approaching vehicle are equal to or less than the limit of the vehicle restriction order, the vehicle type classification determination unit 290 executes the process of step S380 shown in FIG. 5B, and the vehicle type classification of the approaching vehicle is a large vehicle. Decide that there is. When the length and height of the approaching vehicle are not less than or equal to the limit of the vehicle restriction order, the vehicle type classification determination unit 290 determines that the vehicle type classification of the approaching vehicle is an oversized vehicle.

In step S330, when the number of axles of the approaching vehicle is "5" or more, the vehicle type classification determination unit 290 executes the process of step S390 and determines that the vehicle type classification of the approaching vehicle is an oversized vehicle.

In step S220 shown in FIG. 5A, when the first digit of the vehicle model number is "2", the vehicle type classification determination unit 290 executes the process of step S410 shown in FIG. 5C. In step S410, the vehicle type classification determination unit 290 determines whether or not the size of the license plate is the middle plate. The size of the license plate is determined by the license plate determination unit 280 in step S180 shown in FIG. 4 whether or not it is a middle plate. The vehicle type classification determination unit 290 acquires whether or not the size of the license plate of the approaching vehicle is the middle plate based on the determination by the license plate determination unit 280. When the size of the license plate of the approaching vehicle is the middle plate, the vehicle type classification determination unit 290 executes the process of step S420 shown in FIG. 5C to determine that the vehicle type classification of the approaching vehicle is a medium-sized vehicle. The vehicle type classification determination unit 290 executes the process of step S430 when the size of the license plate of the approaching vehicle is not the middle plate.

In step S430, the vehicle type classification determination unit 290 determines whether or not the approaching vehicle is a route bus based on the side captured image or the front captured image. For example, the vehicle type classification determination unit 290 determines that the approaching vehicle is a route bus when the side image or the front image includes an area representing the destination of the route bus. When the vehicle type classification determination unit 290 determines that the approaching vehicle is not a fixed-route bus, the vehicle type classification determination unit 290 executes the process of step S440. When the approaching vehicle is a fixed-route bus, the vehicle type classification determination unit 290 executes the process of step S460 to determine that the vehicle type classification of the approaching vehicle is a large vehicle.

In step S440, the vehicle type classification determination unit 290 determines whether the vehicle length is less than 9 m. The vehicle length is determined by the vehicle length determination unit 270 in step S160 shown in FIG. The vehicle type classification determination unit 290 acquires the vehicle length determined by the vehicle length determination unit 270. When the acquired vehicle length is not less than 9 m, the vehicle type classification determination unit 290 executes the process of step S450 shown in FIG. 5C to determine that the vehicle type classification of the approaching vehicle is an oversized vehicle. When the acquired vehicle length is less than 9 m, the vehicle type classification determination unit 290 executes the process of step S460 and determines that the vehicle type classification of the approaching vehicle is a large vehicle.

In step S210 shown in FIG. 5A, when the vehicle type classification determination unit 290 determines that the license plate cannot be recognized, the process of step S510 shown in FIG. 5D is executed. In step S510, the vehicle type classification determination unit 290 executes different processing depending on the category of the approaching vehicle determined by the category determination unit 250. When the category of the approaching vehicle is "light vehicle", the vehicle type classification determination unit 290 executes the process of step S520 and determines that the vehicle type classification of the approaching vehicle is a light vehicle.

In step S510, when the category of the approaching vehicle is "ordinary vehicle", the vehicle type classification determination unit 290 executes the process of step S530 and determines that the vehicle type classification of the approaching vehicle is the ordinary vehicle.

In step S510, when the category of the approaching vehicle is "large special vehicle", the vehicle type classification determination unit 290 executes the process of step S540 and determines that the vehicle type classification of the approaching vehicle is an oversized vehicle.

In step S510, when the category of the approaching vehicle is "tractor head", the vehicle type classification determination unit 290 executes the process of step S550. In step S550, the vehicle type classification determination unit 290 determines whether the number of axles of the approaching vehicle is "2". When the number of axles of the approaching vehicle is "2", the vehicle type classification determination unit 290 executes the process of step S560 and determines that the vehicle type classification of the approaching vehicle is a medium-sized vehicle. When the number of axles of the approaching vehicle is not "2", for example, when the number of axles of the approaching vehicle is "3", the vehicle type classification determination unit 290 executes the process of step S570, and the vehicle type classification of the approaching vehicle is a large vehicle. Is determined to be.

In step S510, when the category of the approaching vehicle is "truck", the vehicle type classification determination unit 290 executes the process of step S610 shown in FIG. 5E. In step S610, the vehicle type classification determination unit 290 executes different processing depending on the number of axles of the approaching vehicle. When the number of axles of the approaching vehicle is "2" or "3", the vehicle type classification determination unit 290 executes the process of step S620.

In step S620, the vehicle type classification determination unit 290 determines whether or not the size of the license plate is the middle plate. When the size of the license plate of the approaching vehicle is not the middle plate, the vehicle type classification determination unit 290 executes the process of step S630 and determines that the vehicle type classification of the approaching vehicle is a large vehicle. The vehicle type classification determination unit 290 executes the process of step S640 when the size of the license plate of the approaching vehicle is the middle plate.

In step S640, the vehicle type classification determination unit 290 determines whether the approaching vehicle is a small freight vehicle (also referred to as a small freight vehicle). For example, the vehicle type classification determination unit 290 determines whether the approaching vehicle is a small freight vehicle based on the length of the approaching vehicle and the vehicle height. The vehicle length is determined by the vehicle length determination unit 270 in step S160 shown in FIG. The vehicle type classification determination unit 290 acquires the vehicle length determined by the vehicle length determination unit 270. Further, the vehicle type classification determination unit 290 determines the vehicle height of the approaching vehicle based on the side captured image. For example, the vehicle type classification determination unit 290 determines the vehicle height of the approaching vehicle based on the length of the region of the approaching vehicle in the direction perpendicular to the traveling direction of the approaching vehicle, for example, the vertical direction in the side captured image. When the vehicle length and the vehicle height of the approaching vehicle are equal to or less than the regulation of the small freight vehicle, the vehicle type classification determination unit 290 executes the process of step S650 shown in FIG. 5E, and the vehicle type classification of the approaching vehicle is a normal vehicle. Is determined to be. When the length and height of the approaching vehicle are larger than the regulation of the small freight vehicle, the vehicle type classification determination unit 290 determines that the vehicle type classification of the approaching vehicle is a medium-sized vehicle.

In step S610, when the number of axles of the approaching vehicle is "4", the vehicle type classification determination unit 290 executes the process of step S670. In step S670, the vehicle type classification determination unit 290 determines whether the approaching vehicle is equal to or less than the limit of the vehicle restriction order. For example, the vehicle type classification determination unit 290 determines whether the vehicle length and the vehicle height of the approaching vehicle are equal to or less than the limit of the vehicle restriction order. When the vehicle length and the vehicle height of the approaching vehicle are equal to or less than the limit of the vehicle restriction order, the vehicle type classification determination unit 290 executes the process of step S680 shown in FIG. 5E, and the vehicle type classification of the approaching vehicle is a large vehicle. Decide that there is. When the length and height of the approaching vehicle are not less than or equal to the limit of the vehicle restriction order, the vehicle type classification determination unit 290 determines that the vehicle type classification of the approaching vehicle is an oversized vehicle.

In step S610, when the number of axles of the approaching vehicle is "5" or more, the vehicle type classification determination unit 290 executes step S690 to determine that the vehicle type classification of the approaching vehicle is an oversized vehicle.

In step S510 shown in FIG. 5D, when the category of the approaching vehicle is "bus", the vehicle type classification determination unit 290 executes the process of step S710 shown in FIG. 5F. In step S710, the vehicle type classification determination unit 290 determines whether or not the size of the license plate is the middle plate. When the size of the license plate of the approaching vehicle is the middle plate, the vehicle type classification determination unit 290 executes the process of step S720 and determines that the vehicle type classification of the approaching vehicle is a medium-sized vehicle. The vehicle type classification determination unit 290 executes the process of step S730 when the size of the license plate of the approaching vehicle is not the middle plate.

In step S730, the vehicle type classification determination unit 290 determines whether or not the approaching vehicle is a route bus based on the side captured image or the front captured image. For example, the vehicle type classification determination unit 290 determines that the approaching vehicle is a route bus when the side image or the front image includes an area representing the destination of the route bus. When the vehicle type classification determination unit 290 determines that the approaching vehicle is not a fixed-route bus, the vehicle type classification determination unit 290 executes the process of step S740. When the approaching vehicle is a fixed-route bus, the vehicle type classification determination unit 290 executes the process of step S760 and determines that the vehicle type classification of the approaching vehicle is a large vehicle.

In step S740, the vehicle type classification determination unit 290 determines whether the vehicle length is less than 9 m. When the vehicle length is not less than 9 m, the vehicle type classification determination unit 290 executes the process of step S750 and determines that the vehicle type classification of the approaching vehicle is an oversized vehicle. When the acquired vehicle length is less than 9 m, the vehicle type classification determination unit 290 executes the process of step S760 and determines that the vehicle type classification of the approaching vehicle is a large vehicle.

In this way, the vehicle type classification determination unit 290 can determine the vehicle type classification of the approaching vehicle. Further, when the approaching vehicle is a connected vehicle, the vehicle type classification determination unit 290 changes the vehicle type classification of the approaching vehicle based on the number of axles of the towed vehicle and the vehicle type classification of the towing vehicle. When the vehicle type classification of the towed vehicle is any of a light vehicle, an ordinary vehicle, and a medium-sized vehicle, the vehicle type classification determination unit 290 determines whether the number of axles of the towed vehicle is "1". When the number of axles of the towed vehicle is "1", the vehicle type classification determination unit 290 changes the vehicle type classification of the towed vehicle from a light vehicle to an ordinary vehicle, from an ordinary vehicle to a medium-sized vehicle, and from a medium-sized vehicle to a large-sized vehicle. When the number of axles of the towed vehicle is "2" or more, the vehicle type classification determination unit 290 changes the vehicle type classification of the towed vehicle from a light vehicle to a medium-sized vehicle, from a normal vehicle to a large vehicle, and from a medium-sized vehicle to an oversized vehicle. ..

When the vehicle type classification of the towing vehicle is a large vehicle, the vehicle type classification determining unit 290 determines whether the number of axles of the towing vehicle is "2". When the number of axles of the towing vehicle is "3" or more, the vehicle type classification determination unit 290 changes the vehicle type classification of the towing vehicle from a large vehicle to an oversized vehicle. When the number of axles of the towed vehicle is "2", the vehicle type classification determination unit 290 further determines whether the number of axles of the towed vehicle is "2" or more, and the number of axles of the towed vehicle is "2" or more. When, the vehicle type classification of the towing vehicle is changed from a large vehicle to an oversized vehicle.

In this way, the vehicle type classification determination unit 290 can determine the vehicle type classification of the approaching vehicle even when the vehicle type number of the license plate cannot be determined by using the category of the approaching vehicle determined by the category determination unit 250. can. Further, even when the vehicle type number of the license plate can be determined, the vehicle type classification determination unit 290 can determine the vehicle type classification of the approaching vehicle with high accuracy by using the category of the approaching vehicle determined by the category determination unit 250. Can be done. Further, by capturing the side surface of the approaching vehicle with the vehicle side imager 110, for example, a camera, the distance of the traveling lane 10 can be shortened as compared with the case where the number of axles is determined by using the laser sensor. This is because when a laser sensor is used, the vehicle needs to pass in front of the laser sensor. In particular, when the vehicle side imager 110 is a camera capable of capturing a range of 180 degrees, the distance of the traveling lane 10 can be further shortened.

(Modification example)
In step S120 shown in FIG. 4, the category determination unit 250 may determine whether or not the approaching vehicle is a connected vehicle by any method. For example, the category determination unit 250 extracts the area of the approaching vehicle from the side image captured by the vehicle side imager 110. By determining whether the towed vehicle is imaged in the area of the approaching vehicle, the category determination unit 250 may determine whether or not the approaching vehicle is a connected vehicle. In this case, when the towed vehicle is imaged in the area of the approaching vehicle, the category determination unit 250 determines that the approaching vehicle is a connected vehicle.

The towed vehicle may be included in the category determined by the category determination unit 250. In this case, the category determination unit 250 extracts the area of the approaching vehicle in step S110 shown in FIG. 4, and determines the category of the approaching vehicle based on the extracted area. When the category of the approaching vehicle is determined to be a towed vehicle, the category determination unit 250 determines that the approaching vehicle in front of the towed vehicle is a towed vehicle, and the approaching vehicle is a connected vehicle consisting of the towed vehicle and the towed vehicle. Is determined to be.

In step S160 shown in FIG. 4, the vehicle length determination unit 270 may calculate the vehicle height of the approaching vehicle based on the side captured image. For example, the vehicle length determination unit 270 determines the vehicle height of the approaching vehicle based on the length of the region of the approaching vehicle in the direction perpendicular to the traveling direction of the approaching vehicle, for example, the vertical direction in the side captured image. The determined vehicle height of the approaching vehicle is used in the process of determining the vehicle type classification by the vehicle type classification determination unit 290.

The axle number determination unit 260 may determine whether the detected axle is in contact with the road based on the side captured image. For example, the axle number determination unit 260 determines whether the lower end in the detected axle region, for example, the vertical lower end, is above the lower end in the detected other axle region in the vertical direction. When the lower end of the axle region is above the lower end of the other axle region, the axle number determination unit 260 determines that the axle corresponding to the upper region is not in contact with the road and is a lift axle. .. The axle number determination unit 260 determines the number of axles, excluding the axles determined to be lift axles. As a result, the axle number determination unit 260 can accurately determine the number of axles even if there is an axle of the lift axle. The axle number determination unit 260 may determine the axle of the lift axle by comparing the center positions of the axle regions in the vertical direction.

The embodiments and modifications described above are examples, and may be changed as long as the functions are not impaired. Further, the configurations described in the respective embodiments and modifications may be arbitrarily changed or / or arbitrarily combined as long as the functions are not impaired.

For example, the vehicle type determination system 1000 may omit the number imager 100 and determine the vehicle type classification of the approaching vehicle based on the side image and the front image. Further, even when the license plate determination unit 280 can determine the license plate, the vehicle type classification determination unit 290 executes the processes shown in FIGS. 5D to 5F and determines the vehicle type classification based on the category determined by the category determination unit 250. It may be determined. In this case, the vehicle type classification determination unit 290 uses the vehicle type classification determined by the processing shown in FIGS. 5A to 5C using the vehicle type number on the license plate, and the vehicle type classification determined using the category by the processing shown in FIGS. 5D to 5F. Based on the above, the vehicle type classification of the approaching vehicle may be determined.

1: Storage medium 10: Traveling lane 20R: Right island 20L: Left island 100: Number imager 110: Vehicle side imager 120: Vehicle front imager 200: Processing device 210: Input / output device 220: Arithmetic device 230: Communication device 240: Storage device 250: Category determination unit 260: Axle number determination unit 270: Vehicle length determination unit 280: License plate determination unit 290: Vehicle type classification determination unit 300: Passing ticket issuing machine 400: Start vehicle detector 450: Start controller 460: Actuator 470: Start control rod 500: Vehicle type discrimination program 1000: Vehicle type discrimination system

Claims (9)

A category determination unit that determines the category based on the side image captured from the side of the approaching vehicle entering the toll road.
Based on the category, the vehicle type classification determination unit that determines the vehicle type classification of the approaching vehicle, and the vehicle type classification determination unit.
Equipped with
The vehicle type classification represents a classification for determining the toll of the toll road.
The category is a vehicle type discrimination system determined according to the vehicle type classification and the shape of a vehicle belonging to the vehicle type classification. The category determination unit is equipped with an artificial intelligence program.
The vehicle type determination system according to claim 1, wherein the artificial intelligence program is learned to distinguish and determine a plurality of vehicle types included in at least one of the categories. A axle number determination unit for determining the number of axles of the approaching vehicle based on the side captured image is provided.
The vehicle type discrimination system according to claim 1 or 2, wherein the vehicle type classification determination unit determines the vehicle type classification of the approaching vehicle based on the number of axles of the approaching vehicle and the category. The axle number determination unit is
Based on the position of the axle of the approaching vehicle in the side image, it is determined whether the axle is a lift axle.
The vehicle type determination system according to claim 3, wherein when the axle is a lift axle, the number of axles is determined without counting the axles. The category determination unit determines whether the approaching vehicle is a connected vehicle towing the towed vehicle based on the side image captured image.
From claim 1, the vehicle type classification determination unit determines the vehicle type classification of the approaching vehicle based on the vehicle type classification of the towing vehicle and the number of axles of the towed vehicle when the approaching vehicle is the connected vehicle. The vehicle type discrimination system according to any one of 4. The vehicle type determination system according to claim 5, further comprising an axle number determination unit for determining the number of axles of the towed vehicle based on the side image. A vehicle side imager that captures the approaching vehicle from the side and acquires the side captured image is provided.
The vehicle type discrimination system according to any one of claims 1 to 6, wherein the vehicle side imager captures a range of 180 degrees in the traveling direction of the approaching vehicle. The category determination unit
In the traveling direction of the approaching vehicle, a determination region within a predetermined angle is extracted from the optical axis of the vehicle side imager from the side image captured.
The vehicle type determination system according to claim 7, wherein the category of the approaching vehicle is determined in the extracted determination area. Judging the category based on the side image taken from the side of the approaching vehicle entering the toll road,
To determine the vehicle type classification of the approaching vehicle based on the category,
To the arithmetic unit,
The vehicle type classification represents a classification for determining the toll of the toll road.
The category is a vehicle type discrimination program determined according to the vehicle type classification and the shape of a vehicle belonging to the vehicle type classification.

Images