JP2001216598A - Traffic flow measuring instrument - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve a problem that a conventional traffic flow measuring instrument sometimes recognizes front and rear vehicles erroneously since it is not capable of separating them when a distance between traveling vehicles is short because the conventional instrument usually photographs a road with an ITV camera from the rear side of the vehicle, and that the judging precision of the kind of a vehicle receives influence from the measuring error of a vehicle length which is large because the vehicle is photographed from its rear side. SOLUTION: By installing a fisheye lens camera at an angle nearly vertical to a road surface or the side surface of the traveling vehicle, the vehicle is separated surely on video to measure the traffic flow. In addition, the vehicle length is measured to be used for deciding the kind of the vehicle.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for performing image processing on a road image to measure the speed, type, and traffic volume of a traveling vehicle, and to detect events such as a high-speed vehicle, a stopped vehicle, a low-speed vehicle, and congestion. The present invention relates to a flow measuring device.

[0002]

2. Description of the Related Art FIG. 6 shows a conventional traffic flow measuring device. In FIG.
Is a vehicle traveling on the road 1, 2b is a vehicle traveling in front of 2a, 3 is a column mounted on the road side, 4 is a column 3 at a height to monitor from a diagonal rear of the vehicle, and a height 3 m from the road surface.
An industrial television camera (IT
V camera), 5 is a cable for transmitting an image signal from the ITV camera 4, 6 is an image processing device, and 17 is an example of a road monitoring range of the ITV camera 4.

FIG. 7 shows a processing configuration of the image processing apparatus 6. In the figure, 7 is an A / D converter for converting an analog image signal into digital luminance, 8 is an image storage for storing digital luminance for each image pickup screen, 9 is a vehicle detector for detecting a vehicle from an image, and 10 is a detector. A vehicle type determining unit for determining the vehicle type of the vehicle; 11, a trajectory creating unit for creating a running locus of the vehicle; 12, an event determining unit for detecting events such as a high-speed vehicle, a low-speed vehicle, a stopped vehicle, and congestion based on the created vehicle locus It is.

Next, the operation will be described. In FIG. 7, a road image taken by the ITV camera 4 is
Is input to the A / D conversion section 7 in FIG. The A / D converter 7 converts the image analog signal into, for example, a 256-level digital luminance value for each pixel constituting the imaging screen. The image storage unit 8 stores the digital luminance value for each imaging screen, and inputs image data to the vehicle detection unit 9. The vehicle detection unit 9 extracts a moving object by calculating a difference for each pixel of the input image data and image data that is temporally preceding and following, and background image data having no moving object such as a vehicle. The vehicle is detected from the shape and size, and the position of the vehicle is specified. The vehicle type determination unit 10 measures the vehicle width from the extraction area, for example, and determines the vehicle type as a large vehicle if the vehicle width is equal to or larger than a preset threshold, and a normal vehicle if it is smaller than the threshold. The trajectory creating unit 11 creates a vehicle trajectory by associating the vehicle detection positions with temporally continuous images. The event determination unit 12 measures the traveling speed based on the created vehicle trajectory, and determines a high speed, a low speed, and a stopped vehicle. In addition, the traffic volume and average speed of passing vehicles are totaled for a certain period of time, and congestion is determined. For example, when the traffic volume in the past 5 minutes is 20 or more and their average speed is 40 km / h or less, it is determined that there is a traffic jam.

FIG. 8 shows an example of a road image taken by the ITV camera 4. In the figure, 1 is a road, 2a is a road 1
The vehicle traveling above 2b is a vehicle traveling ahead of 2a.

[0006]

As described above, in the conventional traffic flow measuring device, the camera is usually installed at an angle for monitoring the vehicle from obliquely behind the vehicle, so that the distance between the two vehicles is reduced. When these are not enough, they could not be separated on the image, and erroneous recognition sometimes occurred. For example, in the image shown in FIG. 8, two vehicles may be erroneously recognized as one. In particular, in a place where the installation height of the camera is limited, such as in a tunnel, this misrecognition is remarkable. Further, when an event such as a low-speed vehicle, a stopped vehicle, or a traffic jam occurs, the inter-vehicle distance of the traveling vehicle may be particularly short, and at that time, it may not be possible to detect these events. . Furthermore, since the vehicle is monitored from diagonally behind the vehicle, the measurement error of the vehicle length of the vehicle is large, and the vehicle type determination is performed without using the vehicle length information, thereby affecting the accuracy of the vehicle type determination.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problem, and it is possible to accurately separate a plurality of traveling vehicles on an image and to determine a vehicle type by using a measured value of a vehicle length, thereby achieving high-precision traffic. It is intended to measure flow.

[0008]

A traffic flow measuring device according to a first aspect of the present invention is a traffic flow measuring device which detects a traveling vehicle from an image captured by capturing a road and measures the traffic flow. The camera is provided with a camera installed at a position where the traveling vehicle is viewed from a close angle and equipped with an ultra-wide-angle lens, and an image processing device that measures a traffic flow from an image captured by the camera.

In a traffic flow measuring device according to a second aspect of the present invention, the camera is provided for each lane of a road.

A traffic flow measuring device according to a third aspect of the present invention is a vehicle detecting unit for detecting a traveling vehicle from an image taken by the camera as the image processing device, and a vehicle for measuring the length of the detected traveling vehicle. The vehicle includes a length measuring unit and a vehicle type determining unit that determines a vehicle type based on the measured vehicle length information.

[0011]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiment 1 FIG. 1 is a diagram showing a first embodiment of the present invention. In the figure, 1 is a road composed of a plurality of lanes in a tunnel or outdoors,
2a is a vehicle traveling on the road, 2b is a vehicle traveling ahead of the vehicle 2a, and the inter-vehicle distance to 2a is not sufficient. 1
Reference numeral 3 denotes a pillar which is composed of a pillar perpendicular to the road surface and a pillar (arm) parallel to the road surface, and is an F-shaped support installed so that a camera can be mounted above the road. Reference numeral 14 denotes an optical axis perpendicular to the road surface. A fish-eye lens camera, which is installed on the arm of the F-shaped column 13 so as to look down from below at an angle close to that of the F-shaped support 13 and is equipped with a fish-eye lens (ultra-wide-angle lens) as an objective lens, and 5 transmits an analog image signal from the fish-eye lens camera 14. The cable 6 is an image processing device for processing the received analog image signal. The fisheye lens camera 14 is installed at a position where a plurality of lanes can be monitored. Reference numeral 18 denotes an example of a road monitoring range of the fisheye lens camera 14. An image signal of the captured image is input to the image processing device 6 via the cable 5.

FIG. 2 is a diagram showing a processing configuration of the image processing device 6. In the figure, reference numeral 7 denotes an A / D converter for converting an analog image signal into digital brightness, 8 an image storage unit for storing digital brightness for each image pickup screen, 9 a vehicle detection unit for detecting a vehicle from an image, and 15 a detection unit. A vehicle length measuring unit that measures the vehicle length of the vehicle that has been performed, 16 is a vehicle type determining unit that uses a vehicle length that determines the vehicle type of the detected vehicle using vehicle length information, 11 is a trajectory creating unit that creates a traveling locus of the vehicle, Reference numeral 12 denotes an event determination unit that detects an event such as a high-speed vehicle, a low-speed vehicle, a stopped vehicle, or traffic jam based on the created vehicle trajectory.

Next, the operation will be described. In FIG. 2, the road image taken by the fisheye lens camera 4 is a cable 5
Is input to the A / D converter 7 in the image processing device 6 via the. The A / D converter 7 converts the image analog signal into, for example, a 256-level digital luminance value for each pixel constituting the imaging screen. The image storage unit 8 stores the digital luminance value for each imaging screen, and inputs the stored image data and the current image data output from the A / D conversion unit 7 to the vehicle detection unit 9. In the vehicle detection unit 9, each of the image data stored in the image storage unit 8 before and after in time, the background image data without a moving object such as a vehicle, and the current image data output from the A / D conversion unit 7. By taking a difference for each pixel, a moving object is extracted, a vehicle is detected from the shape and size of the extraction area, and the position of the vehicle is specified. Vehicle length measurement unit 1
5 measures the length of a portion corresponding to the vehicle length of the extraction area when the vehicle 2 passes directly below the fisheye lens camera 14.
The vehicle type determining unit 16 using the vehicle length determines the vehicle type by comparing the vehicle length information with a preset threshold value in addition to the vehicle width information. For example, if the vehicle width is 20 pixels or more and the vehicle length is 50 pixels or more, the vehicle is a large vehicle; The trajectory creating unit 11 creates a vehicle trajectory by associating the vehicle detection positions with temporally continuous images. The event determination unit 12 measures the traveling speed based on the created vehicle trajectory, and determines a high speed, a low speed, and a stopped vehicle. In addition, the traffic volume and average speed of passing vehicles are totaled for a certain period of time, and congestion is determined. For example, when the traffic volume in the past 5 minutes is 20 or more and their average speed is 40 km / h or less, it is determined that there is a traffic jam.

FIG. 3 is an example of a road image to be processed according to the first embodiment of the present invention. In the figure, 1 is a road,
2a is a vehicle traveling on the road, 2b is a vehicle traveling ahead of the vehicle 2a, and the inter-vehicle distance to 2a is not sufficient. If the fisheye lens camera 14 is installed at an angle perpendicular to the road surface,
Even vehicles with an insufficient inter-vehicle distance can be separated on the image, and these vehicles can be individually detected.
For this reason, the accuracy of traffic flow measurement is improved, and even in a traffic jam where the inter-vehicle distance is particularly short, the vehicles can be reliably separated on the image, so that the accuracy of the traffic jam detection is improved. In particular, even in a place where the height is limited, such as in a tunnel, the vehicle can be reliably separated on the image. Further, by determining the vehicle type using the vehicle length and the vehicle width, the accuracy of traffic flow measurement for each vehicle type is improved.

Embodiment 2 FIG. 4 shows a second embodiment of the present invention. In the figure, 1a is a driving lane, 2
b is an overtaking lane or oncoming lane, 2 is a vehicle running in lane 1a, 13 is an F-shaped support installed on the roadside, 14a is a position monitoring lane 1a, and the optical axis is directly below the road surface at an angle almost perpendicular to the road surface. , A fish-eye lens camera 14b is installed at the position where the lane 1b is monitored, and a fish-eye lens camera 14a is installed at the same angle as 4a, and 5a is an analog image signal acquired by the fish-eye lens camera 14a. A cable for transmission, 5b is a cable for transmitting an analog image signal acquired by the fish-eye lens camera 14b, 6a is an image processing device for processing an analog image signal received by the cable 5a, and 6b is a device for processing an analog image signal received by the cable 5b. An image processing device. 18
a is an example of the road monitoring range of the fisheye lens camera 14a, and 18b is an example of the road monitoring range of the fisheye lens camera 14b.
The configurations of the image processing devices 6a and 6b are the same as in the first embodiment. The fish-eye lens cameras 14a and 14b capture road images of each lane, and are input to the image processing devices 6a and 6b via cables connected to the cameras.

Next, the operation will be described. The fisheye lens cameras 14a and 14b each take an image of one lane,
The image processing device 6 is connected via cables 5a and 5b, respectively.
a, 6b.

FIG. 5 is an example of a road image taken by the fish-eye lens camera 14a. In the figure, 1a is a driving lane, 2
Is a vehicle running on the road. By using a fish-eye lens camera, vehicles with short inter-vehicle distances in the same lane can be reliably separated on an image. In addition, by installing the camera for each lane, the vehicle passes near the center of the image, so that the vehicle can be photographed in a large size. Therefore, errors due to pixel resolution are reduced, and detection accuracy is improved.

[0018]

According to the present invention, by installing the fisheye lens camera at an angle nearly perpendicular to the road surface,
There is an effect that the accuracy of the traffic flow measurement is improved when the distance between the running vehicles is short and the conventional installation method cannot separate the vehicles from each other on the image.

In addition, even when an event such as traffic congestion that shortens the distance between vehicles occurs, the vehicles can be individually detected, and the accuracy of event determination is improved.

Further, by installing the fisheye lens camera at an angle nearly perpendicular to the road surface, the length of the vehicle can be measured with high accuracy, and there is an effect that the accuracy of measuring the traffic volume for each vehicle is improved.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a first embodiment of the present invention.

FIG. 2 is a diagram illustrating a configuration of an image processing apparatus according to Embodiment 1 of the present invention;

FIG. 3 is an example of a road image to be processed according to the first embodiment of the present invention;

FIG. 4 is a diagram showing a configuration of a second embodiment of the present invention.

FIG. 5 is an example of a road image to be processed according to the second embodiment of the present invention;

FIG. 6 is a diagram showing a configuration of a conventional traffic flow measuring device.

FIG. 7 is a diagram illustrating a configuration of an image processing device of a conventional traffic flow measuring device.

FIG. 8 is an example of a road image to be processed by the conventional traffic flow measuring device.

[Explanation of symbols]

1 Road consisting of multiple lanes, 1a Running lane, 1b
Overtaking or oncoming lane, 2 running vehicle, 2a
Vehicle traveling in the traveling lane, 2b vehicle traveling in front of vehicle 2a, 3 pillars, 4 ITV cameras, 5 video cable, 5a video cable transmitting lane 1a video,
5b, an image cable for transmitting an image of lane 1b, 6 an image processing device, 6a an image processing device for processing an image of lane 1a, 6b an image processing device for processing an image of lane 1b,
7 A / D conversion unit, 8 image storage unit, 9 vehicle detection unit,
Reference Signs List 10 vehicle type judgment unit, 11 trajectory creation unit, 12 event judgment unit, 13 F-shaped support, 14 fisheye lens camera, 14a
A fisheye lens camera for photographing the lane 1a, a fisheye lens camera for photographing the lane 1b, a vehicle length measuring unit,
16 vehicle type determination unit using the commander, 17 road monitoring range of ITV camera, 18 road monitoring range of fisheye lens camera 14, 18a road monitoring range of fisheye lens camera 14a,
18b Road monitoring range of fisheye lens camera 14b.

Claims (3)

[Claims] 1. A traffic flow measuring device that detects a traveling vehicle from an image captured by capturing a road and measures a traffic flow.
A traffic flow measurement device equipped with a camera installed at a position overlooking the traveling vehicle at an angle close to perpendicular to the road surface and equipped with an ultra-wide-angle lens, and an image processing device that measures traffic flow from images taken by this camera . 2. The traffic flow measuring device according to claim 1, wherein the camera is installed for each lane of a road. 3. The image processing apparatus, comprising: a vehicle detection unit that detects a traveling vehicle from an image captured by the camera; a vehicle length measurement unit that measures a vehicle length of the detected traveling vehicle; The traffic flow measuring device according to claim 1, further comprising a vehicle type determining unit that determines a vehicle type based on vehicle length information.