JP6889288B2 - Traffic measurement system - Google Patents

Description

The present invention relates to a system for measuring traffic volume using a surveillance camera in a tunnel on a highway or a general road, and particularly relates to a traffic volume measurement system capable of improving measurement accuracy.

[Explanation of prior art]
Conventionally, there is a system for measuring traffic volume by installing a surveillance camera in a tunnel of a highway or a general national highway.
The tunnel where the traffic volume is measured is long and has a plurality of surveillance cameras, and a channel (CH) is assigned to each surveillance camera to perform image processing on the plurality of channels.
However, the traffic volume measurement has come to be performed by the image processing of 1CH assigned corresponding to each surveillance camera.

[Related technology]
As a conventional technique for measuring traffic volume, there is Japanese Patent Application Laid-Open No. 2009-138466 “Tunnel Monitoring Device” (Patent Document 1).
Patent Document 1 describes an in-tunnel monitoring device that analyzes images from a surveillance camera and measures the type, number, and average speed of vehicles.

Japanese Unexamined Patent Publication No. 2009-138466

However, in the conventional traffic measurement system, if the road to be monitored has two or more lanes, a plurality of vehicles may overlap in the field of view of the surveillance camera. In that case, the plurality of vehicles overlapping on the image may be overlapped. Since it is regarded as one unit, there is a problem that accurate counting cannot be performed and the measurement accuracy is lowered.

It should be noted that Patent Document 1 does not describe individually detecting a plurality of overlapping vehicles on an image.

The present invention has been made in view of the above circumstances, and is a traffic volume measurement system capable of distinguishing and detecting a plurality of vehicles that overlap in an image of one surveillance camera and improving the measurement accuracy of the traffic volume. The purpose is to provide.

The present invention for solving the problems of the above-mentioned conventional example is a traffic volume measuring system for measuring traffic volume, and corresponds to a plurality of cameras arranged continuously and images from the cameras to the cameras. An image processing device that analyzes the set measurement period and outputs the number of vehicles and the running speed of the vehicle during the measurement period with the camera, and stores a plurality of cameras as a group for image processing. When the number of vehicles and the running speed are input from the device, the number of vehicles and the running speed corresponding to a plurality of cameras stored as the same group are averaged, and the number of vehicles and the running speed corresponding to the group are calculated. , The totaling device is equipped with a totaling device that outputs as traffic volume, and the totaling device has a common length of measurement period corresponding to cameras belonging to the same group, and the start timing of the measurement period is sequentially delayed according to the arrangement order of the cameras. It is controlled , and when the traveling speed becomes faster, the length of the measurement period is shortened and the time difference of the start timing is reduced, and when the traveling speed is slower, the length of the measurement period is lengthened and the time difference of the start timing is increased. The time difference between the length of the measurement period and the start timing is stored in advance according to the running speed, and based on the time difference between the length of the measurement period and the start timing stored corresponding to the running speed calculated by the group. It is characterized by adjusting the measurement period and start timing of the cameras belonging to the group.

Further, the present invention is characterized in that the traffic volume measuring system includes a plurality of groups, and the combination of cameras belonging to the plurality of groups can be changed.

According to the present invention, it is a traffic volume measuring system that measures traffic volume, and analyzes a plurality of cameras arranged continuously and images from the cameras for a measurement period set corresponding to the cameras. An image processing device that outputs the number of vehicles and the running speed of the vehicle during the measurement period by the camera, and a plurality of cameras are stored as a group, and the number of vehicles and the running speed are input from the image processing device. Then, the number and running speed corresponding to the plurality of cameras stored as the same group are averaged, the number of vehicles corresponding to the group and the running speed are calculated, and the totaling device is output as the traffic volume. with a collection apparatus has a length of the measurement period corresponding to a camera belonging to the same group as a common, the start timing of the measurement period, which controls so as to sequentially delayed in accordance with the arrangement order of the camera, aggregation device When the running speed becomes faster, the length of the measurement period is shortened and the time difference of the start timing is reduced, and when the running speed is slower, the length of the measurement period is lengthened and the time difference of the start timing is increased. The time difference between the length of the period and the start timing is stored in advance, and the measurement period of the cameras belonging to the group is based on the time difference between the length of the measurement period and the start timing that are stored corresponding to the running speed calculated by the group. Since it is a traffic volume measurement system that adjusts the start timing, multiple vehicles that overlap with the image of one camera can be discriminated by different cameras in the group, and the measurement accuracy of traffic volume can be improved. It can be improved, and it is possible to increase the possibility that the same vehicle is photographed by a plurality of cameras in the group according to the actual traveling speed, and it is possible to further improve the measurement accuracy of the traffic volume .

Further, according to the present invention, the traffic volume measuring system includes a plurality of groups and can change the combination of cameras belonging to the plurality of groups. Therefore, an appropriate group is used according to the road condition and the driving tendency. It can be organized, and has the effect of being able to accurately determine the traffic volume of the entire section for which traffic volume is to be measured.

It is explanatory drawing which shows the outline of photography of this system. It is a schematic block diagram which shows the schematic structure of this system. It is a schematic explanatory drawing which shows the outline of the measurement method of this system. It is explanatory drawing which shows the adjustment of the measurement unit time and the measurement start time difference. It is a block diagram of the traffic volume measurement totaling apparatus 3. It is explanatory drawing of the group management table. It is explanatory drawing of the measurement period table.

Embodiments of the present invention will be described with reference to the drawings.
[Outline of Embodiment]
The traffic volume measurement system (the present system) according to the embodiment of the present invention is a system that measures traffic volume based on images from a plurality of surveillance cameras, and an image processing device uses images from a plurality of surveillance cameras. Is analyzed for each preset measurement period, the number of vehicles and the running speed are detected, and the traffic volume measurement totaling device is registered as the same group with a plurality of surveillance cameras continuously arranged along the road. The number of detected vehicles and the traveling speed are averaged to calculate the section traffic volume. In particular, the measurement period start timing is sequentially delayed according to the order of arrangement of the surveillance cameras in the group, and one surveillance camera Even when a plurality of vehicles overlap in the field of view, another surveillance camera can discriminate without overlapping, and the measurement accuracy of traffic volume can be improved.

[Outline of shooting with this system: Fig. 1]
Before explaining the configuration of this system, the outline of shooting of this system will be briefly described with reference to FIG. FIG. 1 is an explanatory diagram showing an outline of shooting of this system.
As shown in FIG. 1, in this system, a plurality of surveillance cameras (hereinafter referred to as “cameras”) (1) to (5) are continuously arranged along the road, and the cameras (1) to (5) are arranged in succession. By processing the traffic volume obtained from the image of 5) as one group, it is possible to identify a plurality of vehicles that overlap each other with one camera.
Here, the road to be monitored is inside the tunnel, but it may be outside (light part).

In the example of FIG. 1, in the field of view of the camera (1), the vehicle B cannot be measured because it overlaps with the vehicle A, but the cameras (3) and (4) do not overlap with the vehicle A and can be measured. ..
In this way, in this system, the traffic volume (number of vehicles and running speed) measured based on the images from multiple cameras is averaged within the group and output as the traffic volume of the entire section of the group. is there.
Here, the traveling speed is obtained by obtaining the speed of each vehicle within the measurement period and averaging the measured number of vehicles.

Therefore, the traffic volume is measured from the images of each camera according to the vehicle traveling speed in the section so that the vehicle photographed by the upstream camera (1) is photographed by the downstream cameras (2) to (5) as much as possible. The measurement period to be measured and the timing of the start of the measurement period are adjusted, which is a feature of this system. These will be described later.

[Configuration of this system: Fig. 2]
Next, the configuration of this system will be described with reference to FIG. FIG. 2 is a schematic configuration diagram showing a schematic configuration of this system.
As shown in FIG. 2, this system includes a plurality of cameras 1 (cameras (1) to (5)), an image processing device 2, and a traffic volume measurement / aggregation device 3.
The camera 1 captures the traffic situation of the vehicle on the road and outputs a continuous image (image) without a time break to the image processing device 2. Here, the number of cameras 1 is five, but as long as there are two or more cameras, the number may be more or less.

Then, in this system, a group is set in advance in order to aggregate (average) the traffic volume based on the images of a plurality of cameras.
It should be noted that the group is composed of a plurality of cameras provided in a series of sections having similar driving conditions such as relatively smooth flow and easy congestion. For example, a straight road section (FIG. 1). Example), curved section, slope section, near tunnel entrance, near tunnel exit, etc.
The group information is stored in the traffic volume measurement / aggregation device 3.

The image processing device 2 inputs an image taken by one camera as one channel, performs image processing, and outputs traffic volume information for each channel.
Specifically, the image processing device 2 includes channel processing units (CH (1) processing unit to CH (5) processing unit) that perform image processing for each channel, and each channel processing unit measures a predetermined image. It is divided into periods and image processing is performed for each measurement period.

Each channel processing unit detects a vehicle based on its size, shape, etc. by image processing, obtains the number of vehicles that have passed during the measurement period, and calculates the vehicle traveling speed of the measurement period in the shooting range of the camera. ..
Specifically, the channel processing unit calculates the traveling speed of each vehicle from the time it takes for the vehicle to travel a certain section (distance), totals the traveling speed of each vehicle that has passed during the measurement period, and calculates the number of vehicles. Divide to obtain the running speed for the measurement period.
Then, the number of vehicles and the traveling speed are output to the traffic volume measurement / aggregation device 3 as traffic volume information corresponding to the shooting range of each camera. Channel processing units are provided according to the number of cameras 1.

Here, as a feature of this system, in the measurement of traffic volume, the image processing device 2 performs measurement according to the timing of measurement start and the unit time of measurement (measurement unit time) instructed by the traffic volume measurement totaling device 3. This will be described later.

The traffic volume measurement / aggregation device 3 is a characteristic part of this system, and the traffic volume for each channel input from the image processing device 2 is aggregated and averaged for each preset group, and a camera belonging to the group is installed. It is calculated as the traffic volume (section traffic volume) of the entire section.
When there are multiple groups in the tunnel, the traffic volume in those sections is totaled and averaged, the traffic volume of the entire tunnel is calculated, and the traffic volume is output to the outside.
The aggregated traffic volume (the number of vehicles averaged in the section traffic volume and the traveling speed) is provided on a display board showing road information and media such as a radio.

At the same time, the traffic volume measurement / aggregation device 3 of this system determines an appropriate measurement period and measurement start timing of each camera for each group based on the aggregation result, and outputs an instruction to the image processing device 2. The measurement period and measurement start timing will be described later.

[Measurement method in this system: Fig. 3]
Next, the traffic volume measurement method of this system will be described with reference to FIG. FIG. 3 is a schematic explanatory view showing an outline of the measurement method of this system.
FIG. 3 shows a case where the cameras (1) to (5) shown in FIG. 1 are grouped into one group for measurement.
As shown in FIG. 1, the channel processing unit corresponding to each camera of the image processing device 2 cuts out an image of a set measurement unit time from the continuous video captured by each camera at a set measurement start timing. (Divided) and measure the traffic volume. The “band” shown in FIG. 1 indicates the measurement period corresponding to each camera.
The period of measurement unit time starting from the measurement start timing is called the measurement period corresponding to each camera.

Here, as a feature of this system, the length of the measurement period (measurement unit time) of a plurality of cameras belonging to the same group are all the same, and the image processing unit 2 describes the images of the cameras (1) to (5). Measure the traffic volume for each common unit time.
Further, the measurement periods of the cameras (1) to (5) are set to be staggered in time, and the time difference of the start timing of the measurement periods between adjacent cameras is Δt.

In the example of FIG. 3, the measurement start timing of the camera (1) is time t1, the measurement period is time t1 to t6, and the length of the measurement period (measurement unit time) is t6-t1 = T1. That is, in the image processing device 2, the channel (1) processing unit counts the number of vehicles and calculates the traveling speed for the image of the time T1 from the time t1 to the time t6.
The image of the time T1 starting from the time t6 of the camera (1) is used for the aggregation process of the next time.

The measurement unit time of the camera (2) to the camera (5) is T1, which is the same as that of the camera (1).
However, the start timings (measurement start timings) of the measurement periods of the cameras (2) to (5) are sequentially delayed, with the camera (2) at time t2, the camera (3) at time t3, and the camera (4). Is time t4, and the camera (5) is time t5.

The time difference (measurement start time difference) (t2-t1, t3-t2, t4-t2, t5-t4) of the start timing of the measurement period between adjacent cameras is Δt.
That is, t2 = t1 + Δt, t3 = t1 + 2Δt, t4 = t1 + 3Δt, t5 = t1 + 4Δt.

When the image processing results of the cameras (1) to (5) stored in advance as a group are input from the image processing device 2, the traffic volume measurement / aggregation device 3 calculates an average value for the number of vehicles and the traveling speed. , With time information, it is output as the number of section vehicles and section running speed in the section where the cameras of the group are installed. The time information can be set arbitrarily, but for example, the time is t1. In the above example, the time t1 is the earliest measurement start timing.

In this way, by sequentially delaying the image cropping timing (measurement start timing) according to the camera arrangement order, it is possible to detect vehicles that could not be photographed by one camera, and improve the traffic volume measurement accuracy. It is something that can be made to do.

[Adjustment of measurement period: Fig. 4]
Further, as a feature of this system, the traffic volume measurement / aggregation device 3 adjusts the measurement period corresponding to each camera for each group according to the measured traveling speed.
Specifically, when the traveling speed is high, the measurement unit time and the measurement start time difference are shortened, and when the traveling speed is slow, the measurement unit time and the measurement start time difference are lengthened.

FIG. 4 is an explanatory diagram showing adjustment of the measurement unit time and the measurement start time difference.
For example, at a certain traveling speed, as shown in FIG. 4A, it is assumed that the number of vehicles and the traveling speed are measured with the measurement unit time T1 and the measurement start time difference Δt.

When the vehicle traveling speed decreases from that state, as shown in FIG. 4B, the traffic volume measurement / aggregation device 3 sets the measurement unit time to a time longer than T1 (T2) and the measurement start time difference. Is a time longer than Δt (Δt'), and a new measurement unit time and measurement start timing are instructed to the image processing apparatus 2.

This is because when the traveling speed is high, the vehicle may be overlooked if the image processing is not performed frequently in a short measurement unit time, but when the running speed is slow, the change in the image is small and even if the measurement unit time is lengthened. This is because the effect is small.
Further, when the traveling speed is high, it is necessary to shorten the measurement start time difference in order to capture the same vehicle with the adjacent cameras, but when the traveling speed is slow, there is no problem even if the measurement start time difference is increased. ..

When the traveling speed is extremely slow as in a traffic jam, it is difficult to photograph the same vehicle with the cameras (1) to (5), but the average value of the number of vehicles and the traveling speed should be calculated. There is no problem with this, and it can be expected that the measurement accuracy will be improved by using the images from multiple cameras.

[Structure of Traffic Measurement and Aggregation Device 3: FIG. 5]
Next, the configuration of the traffic volume measurement / aggregation device 3 will be described with reference to FIG. FIG. 5 is a block diagram of the traffic volume measurement / aggregation device 3.
As shown in FIG. 5, the traffic volume measurement / aggregation device 3 of this system basically includes a control unit 31, a storage unit 32, and an interface unit 33.

The control unit 31 aggregates for each group based on the number of vehicles and the traveling speed corresponding to each camera (each CH) input from the image processing device 2, and calculates the average number of vehicles and the average traveling speed in the section. Then, it is output as the section traffic volume.
When there are a plurality of sections (multiple groups) in the tunnel, the average value obtained by averaging the traffic volumes of the plurality of sections is output as the traffic volume of the entire measurement target section (entire tunnel).

Further, the control unit 31 adjusts the measurement period in the image processing device 2 according to the calculated average traveling speed for each group.
Specifically, the control unit 31 adjusts the measurement unit time according to the traveling speed of each group, outputs an instruction to the corresponding channel processing unit, adjusts the measurement start time difference, and starts the measurement. Instruct each channel processing unit of the measurement start timing based on the time difference.

The instruction (notification) of the measurement unit time and the measurement start timing may be periodically given from the traffic volume measurement totaling device 3 to the image processing device 2, or when the current measurement unit time and the measurement start timing are changed. You may try to do it.
The operation of the control unit 31 will be described later.

The storage unit 32 stores the program executed by the control unit 31, and has a group management table for each group that stores the information of the cameras belonging to the group, the set measurement unit time, and the measurement start time difference. ..
Further, the storage unit 32 stores a measurement period table that stores an appropriate measurement unit time and measurement start time difference corresponding to various traveling speeds as a reference for the measurement unit time and the measurement start timing. These will be described later.
The interface unit 33 is connected to the image processing device 2 and the aggregation result output unit.

[Group management table: Fig. 6]
Next, the group management table stored in the storage unit 32 of the traffic volume measurement / aggregation device 3 will be described with reference to FIG. FIG. 6 is an explanatory diagram of the group management table.
As shown in FIG. 6, the group management table stores information on cameras belonging to the group, measurement unit time, and measurement start time difference corresponding to the group number (group No.).

Here, the camera information is stored in the order of arrangement from the upstream to the downstream of the road.
Specifically, in group 1, cameras (1), cameras (2) ... (5) are arranged in this order, and in group 2, cameras (6), cameras (7), and cameras (8) are arranged in this order. They are arranged in order.
The camera arrangement order is information required when specifying the measurement start timing.
Further, new establishment or abolition of the group, change of the combination of cameras belonging to the group, etc. can be performed at any time from the traffic volume measurement / aggregation device 3.

[Measurement period table: Fig. 7]
Next, the measurement period table stored in the storage unit 32 of the traffic volume measurement totaling device 3 will be described with reference to FIG. 7. FIG. 7 is an explanatory diagram of the measurement period table.
As shown in FIG. 7, the measurement period table stores an appropriate measurement unit time and a measurement start time difference obtained in advance in accordance with the traveling speed.
As described above, the appropriate measurement unit time and measurement start time difference are set so that the vehicle photographed by one camera is photographed by another camera in the group as much as possible according to the traveling speed. It is a thing.

[Operation of traffic volume measurement / aggregation device 3]
The operation of the traffic volume measurement / aggregation device 3 will be described.
When the traffic volume measurement / aggregation device 3 inputs the number of vehicles and the traveling speed for each camera as the processing result from the image processing device 2, the number of vehicles and the traveling speed (section traffic volume) for each group are referred to with reference to the group management table. Is calculated.
When there are a plurality of groups, the traffic volume measurement / aggregation device 3 further averages them, calculates the traffic volume (number of vehicles, traveling speed) of the entire measurement target section, and calculates the traffic volume of the measurement target section. Output to the outside.

Further, the traffic volume measurement / aggregation device 3 reads an appropriate measurement unit time and measurement start time difference corresponding to the traveling speed calculated for each group with reference to the measurement period table shown in FIG. 7, and is shown in FIG. Update the measurement period and measurement start time difference in the group management table.
In addition, instead of reading the appropriate measurement unit time and measurement start time difference corresponding to the calculated running speed from the table, the reference running speed and the corresponding measurement unit time and measurement start time difference, calculation formula, parameters, etc. May be stored and the measurement unit time and the measurement start time difference may be obtained by calculation based on the calculated traveling speed.

Then, when the measurement unit time and the measurement start time difference are changed, the traffic volume measurement totaling device 3 notifies the image processing device 2 of the new measurement period and the measurement start time difference. At that time, the traffic volume measurement / aggregation device 3 notifies the image processing device 2 by designating the channel processing device corresponding to each camera belonging to the group.

When the image processing device 2 receives the number of the channel processing device corresponding to the group, the measurement unit time, and the measurement start time difference from the traffic volume measurement totaling device 3, the image processing device 2 sets them in each channel processing device.

As described above, the measurement unit time is a time common to all channel processing devices in the group.
Further, when the instructed measurement start time difference is Δt, the image processing device 2 uses the measurement start timing (arbitrary timing) of the channel processing device corresponding to the most upstream camera as a reference, and the second channel processing device sets the second channel processing device. The measurement start timing is set so that the measurement period is started later by the time Δt and the third channel processing device is started later by Δt.
Hereinafter, in the same manner, the measurement start timing is set by delaying the measurement start timing by Δt in sequence according to the camera arrangement order for all the channel processing devices in the group.
As a result, as shown in FIGS. 3 and 4, image processing is performed by sequentially delaying the measurement period according to the order in which the cameras are arranged.

[Effect of Embodiment]
According to the traffic volume measuring system according to the embodiment of the present invention, it is a system that measures the traffic volume based on the images from a plurality of surveillance cameras, and the image processing device 2 obtains images from the plurality of surveillance cameras. The number of vehicles and the traveling speed are detected by analyzing for each preset measurement period, and the traffic volume measurement totaling device 3 stores a plurality of continuously arranged surveillance cameras as one group and is the same. The section traffic volume is calculated by averaging the number of vehicles and the traveling speed detected by multiple surveillance cameras registered as a group. In particular, the measurement period start timing of the surveillance cameras is determined according to the arrangement order of the surveillance cameras in the group. Even if multiple vehicles overlap in the field of view of one surveillance camera, they can be discriminated without overlapping by another surveillance camera, and the measurement accuracy of traffic volume can be improved. effective.

In addition, according to this system, the measurement unit time of the measurement period and the measurement start timing are appropriately adjusted according to the measured traveling speed, so that the detection accuracy of the vehicle is improved and the traffic volume is reduced. It has the effect of further improving the measurement accuracy.

In the above-mentioned example, the image processing device 2 and the traffic volume measurement / aggregation device 3 are described as different configurations, but they may be formed as an integrated device.

Further, in the examples shown in FIGS. 3 and 4, the detection periods corresponding to the cameras are overlapped in time, but they do not have to overlap.

The present invention is suitable for a traffic volume measurement system capable of distinguishing and detecting a plurality of vehicles that overlap in an image of one surveillance camera and improving the measurement accuracy of the traffic volume.

1 ... Surveillance camera, 2 ... Image processing device, 3 ... Traffic measurement and totaling device, 31 ... Control unit,
32 ... Storage unit, 33 ... Interface unit

Claims (2)

It is a traffic volume measurement system that measures traffic volume.
With multiple cameras arranged in succession,
An image processing device that analyzes an image from the camera for a measurement period set corresponding to the camera and outputs the number of vehicles in the measurement period of the camera and the traveling speed of the vehicle. ,
When the plurality of cameras are stored as a group and the number of vehicles and the traveling speed are input from the image processing device, the number of vehicles and the traveling speed corresponding to the plurality of cameras stored as the same group are stored. It is equipped with a counting device that averages, calculates the number of vehicles and traveling speed corresponding to the group, and outputs it as traffic volume.
The tabulation device has a common length of measurement period corresponding to cameras belonging to the same group, and controls the start timing of the measurement period so as to be sequentially delayed according to the arrangement order of the cameras .
When the traveling speed is increased, the totaling device shortens the length of the measurement period and reduces the time difference of the start timing, and when the traveling speed is slow, the length of the measurement period is increased and the time difference of the start timing is increased. , The time difference between the length of the measurement period and the start timing is stored in advance according to the traveling speed.
Traffic characterized by adjusting the measurement period and start timing of cameras belonging to the group based on the time difference between the length of the measurement period and the start timing stored corresponding to the traveling speed calculated by the group. Quantity measurement system. The traffic volume measuring system according to claim 1, further comprising a plurality of groups and making it possible to change the combination of cameras belonging to the plurality of groups.

Images