JP2002260163A - Image processing area determining method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processing area determining method capable of determining an image processing area for processing image without any direct human works on the road. SOLUTION: Deviation of a vehicle 1 traveling on the road 4 in the road crossing direction from a magnetic nail 5 installed on the road is detected, the image of the vehicle traveling on the road is picked up, the reference position of the vehicle is calculated by processing the picked-up image, the position information of the magnetic nail which is stored in advance and the information on the road size in the road crossing direction are read, and the image processing area of the picked-up image is determined based on the deviation, the reference position, and the position information of the magnetic nail, and information on the road size in the road crossing direction.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used for detecting road conditions such as traffic volume, speed, and sudden events by capturing an image of a road with a camera installed on the side of the road and processing the captured image. And a method for determining an image processing area to be performed.

[0002]

2. Description of the Related Art On a road for automatic driving, a driving support road system (AHS system) supports driving to a vehicle by a magnetic nail installed on a road or a CCTV camera installed on a side of the road. A large number of magnetic nails are installed on the center of the lane so that the polarity is reversed at regular intervals in the vehicle traveling direction. The position of the vehicle with respect to the center of the lane can be measured by measuring the amount of deviation from the magnetic nail using a magnetic sensor mounted on a vehicle traveling on this road. The driving support road system guides the vehicle to the center of the lane based on the position information. Further, the CCTV camera plays an important role in detecting road conditions, such as detecting obstacles such as falling objects and understanding vehicle behavior.

[0003] Vehicle behavior grasp using a CCTV camera detects the behavior of a passing vehicle (position, speed, acceleration with respect to the road), obstacles such as falling objects, and the like by processing images from the CCTV. It can detect traffic abnormalities such as accidents, stops, and traffic jams.
Further, safe driving is realized by transmitting the information to an AHS vehicle (a vehicle capable of automatic driving in the AHS system) through road-to-vehicle communication or the like. Also, AHS
It also plays an important role in grasping the vehicle behavior status of AHS vehicles and non-AHS vehicles in the transition period of system introduction.
Therefore, when calculating the positions of vehicles and obstacles on a road by performing image processing on an image obtained from a CCTV camera, it is necessary to find the relationship between the coordinates on the image and the actual coordinates on the road. By obtaining this relationship, the driving support road system can know the road area on the image, that is, the image processing area, and the vehicle running on the road can detect the position, speed, acceleration, and the like on the lane. .

Since the CCTV camera is fixedly installed on the side of a road, its position with respect to the road (distance from the shoulder and installation height) and optical conditions (angle of view, distance of converging point, etc.) are known. However, in many cases, the imaging direction of a CCTV camera in a three-dimensional space can be obtained only after measurement. If the road to be image-processed is a straight road and there is no change in road shape such as a gradient change, white lines inside the road are detected by image processing, and the distance between the white lines (the distance between the white lines is known in advance) ) Can be calculated from the relationship.

However, in general, many roads are curved or their gradients are changed. Therefore, the image processing area cannot be calculated from the white line on the screen.
Therefore, in the case of such a road, a color cone or the like is installed as a mark for image processing along the actual road alignment (lane boundary line), and the position on the road and the position on the image are measured. The correspondence between the coordinates of the actual road on the image and the coordinates of the actual road is acquired, and an image processing area is set.

[0006]

However, when the road is bent or the gradient is changed, the above-mentioned work of installing the color cone can be easily performed by a person such as a road on which a vehicle is running. There was a problem that it was dangerous work that could not be approached. The present invention has been made in view of such a situation, and provides a method of determining an image processing area that can determine an image processing area for image processing without direct work by a human on a road. The purpose is to:

[0007]

In order to achieve the above-mentioned object, the invention according to claim 1 is to reduce the amount of deviation of a vehicle traveling on a road from a magnetic nail installed on the road in a cross-road direction. Detect and image vehicles traveling on the road,
The reference position of the vehicle is calculated by performing image processing on the captured image, and the position information of the magnetic nail and the road size information in the cross-road direction stored in advance are read, and the deviation amount,
The image processing area of the captured image is determined based on the reference position, the position information of the magnetic nail, and the road size information in the cross road direction. That is, on the road, a plurality of magnetic nails are installed at predetermined intervals along the longitudinal direction of the road. A vehicle travels on the vehicle, and the vehicle detects an amount of displacement in a cross-road direction with respect to the magnetic nail. On the other hand, an image of the road on which the vehicle is traveling is taken. The reference position of the vehicle on the image is calculated based on the captured image. Here, the road size information and the installation position of the magnetic nail are known in advance. Then, the image processing area of the captured image is determined based on the displacement amount, the reference position, the position information of the magnetic nail, and the size information of the road in the cross road direction. By determining the image processing area, for example, it is possible to know where the position and speed of the vehicle obtained by performing predetermined image processing on the captured image are located on the actual road and the speed.

According to a second aspect of the present invention, in the first aspect, the image processing area is a lane area. Thereby, it is possible to know how much the vehicle running in the lane deviates from the lane.

[0009]

An embodiment of the present invention will be described with reference to the accompanying drawings. FIG. 1 shows the appearance of a road or a camera to which the present invention is applied. The CCTV camera 2 captures an image of the vehicle 1 traveling on a road 4. The position, speed, acceleration, and the like of the vehicle are detected from the captured image in the image processing device 3, and the information is transmitted to the road management center and the captured vehicle 1. A system composed of these components is called a driving support road system. At the center of each lane of the road 4, a large number of magnetic nails 5 are provided at predetermined intervals in the vehicle traveling direction. In this embodiment, the left lane in the drawing will be described. The automatically traveling vehicle travels on the road along the magnetic nail 5. Even at a corner, by traveling along the magnetic nail 5, the vehicle does not run off the lane.
The principle by which the vehicle 1 detects the magnetic nail 5 will be described later. Note that, in the drawing, white arrows indicate the traveling direction of the vehicle 1. The white line 10 separates each lane. Road 4 has two lanes. In FIG. 1, the magnetic nail 5 is clearly and clearly shown, but this is for the purpose of explanation. Magnetic nails actually installed on roads are difficult to visually distinguish from roads on the captured image, and when a captured image as shown in FIG. 1 is obtained, the magnetic nail is visible on the captured image. Not a target.

FIG. 2 shows the relationship between the image coordinate system 12 and the road coordinate system 13. The road 4 is imaged in the acquired image 11. The origin 0 _IJ of the image coordinate system 12 is set at the upper left of the acquired image 11. An I coordinate system is provided rightward from the origin 0 _IJ ,
A J coordinate system is provided downward. On the other hand, the road coordinate system
13 XY are provided along the road 4. An X coordinate system is provided in the traveling direction of the road, and Y is a direction orthogonal to the X coordinate system.
A coordinate system is provided. The origin 0 _XY of the road coordinate system 13 is provided substantially at the lower center of FIG. In the case as shown in FIG. 2, the image processing area determination method aims to determine the area of the road 4 in the image coordinate system 12 as the image processing area.

FIG. 3 shows details of the road coordinate system 13 of FIG.
The X axis of the road coordinate system 13 is set along the vehicle traveling direction indicated by a white line 10 in FIG. The Y axis is set so as to be orthogonal to the X axis. That is, the Y axis is set in the normal direction of the X axis at an arbitrary coordinate position of the X axis.

FIG. 4 is a block diagram of the vehicle-mounted device for detecting and recording the amount of deviation from the magnetic nail 5 and an overview of the vehicle 1. The vehicle 1 has an on-board unit 14 mounted thereon. OBE 14
Is a magnetic sensor 16, a vehicle position measurement (calculation) unit 17, a clock 18,
It comprises a recording unit 19 and a road reference position detection unit 13.
The magnetic sensor 16 outputs an electric signal based on the presence or absence of the magnetic nail 5 installed on the road. Based on the electric signal, the vehicle position measuring unit 17 calculates a deviation amount indicating how much the vehicle is deviated from the magnetic nail 5 with respect to the road crossing direction which is the normal direction of the vehicle traveling direction. The principle of detecting the shift amount will be described later.

On the other hand, a road reference position detector 13 detects the presence or absence of a reference tag 12 installed on a road. For example, by transmitting an electromagnetic wave of a predetermined frequency toward the road surface from the road reference position detecting unit 13, the presence or absence of the reference tag 12 is determined by the presence or absence of a reflected wave of the reference tag 12 generated by resonance with the electromagnetic wave. Is detected. Since the magnetic nails 5 are installed at predetermined intervals on the road, the vehicle position measuring unit 17
Calculates the travel distance from the reference tag 12 to the magnetic nail 5 detected by the magnetic sensor 16 by integrating the number of detected magnetic nails 5 after detecting the reference tag 12. Here, the vehicle travel distance and the amount of deviation calculated by the vehicle position measuring unit 17 are collectively referred to as a vehicle position. The vehicle position calculated every time the vehicle position is calculated is recorded in the recording unit 19 together with the detection time output from the clock 18. An image processing sign 15 is attached to the rear of the vehicle 1. CCT
When the V camera 2 acquires an image including the sign 15, the reference position of the vehicle 1 in the acquired image 11 acquired by the CCTV camera 2 is calculated. The distance between the marker 15 and the magnetic sensor 16 is known, and the distance is referred to as a vehicle length correction amount 1c.

FIG. 5 shows the vehicle 1 when the vehicle 1 is viewed from the front.
FIG. 3 is a diagram showing a positional relationship between a magnetic sensor 16 mounted on the vehicle and a magnetic nail 5 installed on the road 4. Magnetic thin film which is a detecting element of the magnetic nail 5 in the magnetic sensor 16
As shown in the figure, a plurality of 42 are provided on the four road surfaces in the cross direction of the road. The magnetic sensor 16 is a magnetic nail 5
It outputs the amount of change in the electric signal generated from each magnetic thin film 42 when passing through the top. Then, the amount of deviation from the magnetic nail 5 in the cross-road direction is calculated according to the amount of change.

FIG. 6 shows a block diagram of the magnetic sensor 16.
The magnetic sensor 16 includes a plurality of magnetic thin films across the width of the vehicle 1.
It has 42. A high frequency is supplied from the high frequency circuit 41 to each magnetic thin film. Each voltmeter 43 measures the voltage of each magnetic thin film. When the magnetic thin film 42 passes near the magnetic nail 5, the voltage of each magnetic thin film 42 changes. Each change amount extraction circuit 44 extracts the change from each voltmeter 43. The extracted amounts of change are output to the vehicle position measuring unit 17. The vehicle position measuring unit 17 calculates the amount of deviation of the magnetic sensor 16 from the magnetic nail 5 based on the magnitude of these changes.

FIG. 7 shows a specific example of the detection result stored in the recording unit 19. For each detection time (unit: second), the traveling position (unit: m) of the vehicle 1 and the position in the road crossing direction (unit :)
m) is recorded. At a position crossing the road, a positive value indicates that the vehicle is shifted leftward in the vehicle traveling direction, and a negative value indicates that the vehicle is shifted leftward in the vehicle traveling direction. The reference position detected by the road reference position detection unit 13 is a traveling distance of 0 m at time 0 s (second).
It is. The position after time 1 s is detected by the magnetic sensor 16.

Next, a description will be given of a roadside processing device whose outline is shown in FIG. The roadside processing device includes a CCTV camera 2 and an image processing device 3. The image processing device 3 includes an A / D conversion unit 21, a memory 22, a marker position extraction unit (target extraction processing unit) 24, a clock 23, and a recording unit 25. The CCTV camera 2 images a predetermined area of the road 4. The imaging signal from the CCTV camera 2 is A / D
It is output to the conversion unit 21 and converted from an analog signal to a digital signal,
That is, it is converted into image data. The converted image data is stored in the memory 22 together with the imaging time. Note that the imaging time is obtained from the clock 23. Then, the marker position extracting unit 24 extracts the position of the marker 15 in the image coordinate system 12 based on the stored image data. The clock 23 is the vehicle 1
The time is adjusted so that the time matches the time of the clock 18 mounted on the device.

FIG. 9 shows the appearance of the vehicle 1 to which the sign 15 is attached. This is an external view from behind the vehicle. Vehicle 1
Are provided with three signs 15 (15a, 15b, 15c). The marker position extracting unit 24 extracts the positions of the three markers 15 in the image coordinate system 12 by image processing.

FIG. 10 shows the result of the extraction. At each time (unit: second), each marker (15a, 15
b, 15c) are recorded. I in the table head indicates the position on the I coordinate, and J in the table head indicates the position on the J coordinate. Then, the marker position extracting unit 24 outputs the marker position extraction results of these three markers 15 and the imaging time to the recording unit 25. The recording unit 25 records the marker position extraction result and the imaging time. Next, a method of performing image processing on a position calculation result of the vehicle 1 based on the detection of the magnetic nail 5 and an image obtained by the CCTV camera 2 and determining an image processing area based on a result of calculating the position of the vehicle 1 explain.

FIG. 11 shows an outline of the image processing area determination processing. In step S1, image coordinate system vehicle position data and time data are read from the recording unit 25 in which the vehicle position based on the image captured by the CCTV camera 2 is recorded. In step S2, road coordinate system vehicle position data and time data are read from the recording unit 19 in which the vehicle position based on the detection result of the magnetic nail 5 is recorded. In step S3, the image coordinate system vehicle position data and the road coordinate system vehicle position data are associated with each other based on the read times.
In step S4, the position of the magnetic nail 5 in the image coordinate system 12 in the image coordinate system 12 is calculated from the associated image coordinate system vehicle position data and road coordinate system vehicle position data. A method for calculating this position will be described with reference to FIG.

FIG. 12 shows a road 4 and a magnetic nail 5 and a sign 15.
It shows the relationship with. Here, the first magnetic nail 5 detected by the magnetic sensor 16 is shown. The numbers in the subscripts of the symbols in the figure indicate the detection time or the imaging time. FIG. 12 shows the positional relationship detected and imaged at time 1s. The magnetic nail 5 exists on the road center line 33. Road center line 33 is X in road coordinate system 13
Corresponds to the axis. Position of magnetic nails 5 are indicated by point N _1. Central indicator 15b is indicated by a point P ₁ c. Left and right signs 1
5a and 15c are indicated by points P _1L and P _1R , respectively. The deviation amount in the road coordinate system 13 obtained from the magnetic sensor 16 is D
represented by c _1. The distance between the signs 15a and 15b in the road coordinate system 13 is indicated by D. The distance between the marker 15b and the marker 15c is also indicated by D. This distance D is known. Of course, since the position of the magnetic sensor 16 and the position of the sign 15 are different, the reference position of the vehicle 1 on the image and the magnetic sensor 16 With the reference position of.

On the other hand, the markers 15a and 15b extracted from the image
Distance in the image coordinate system 12 is a ₁Indicated by Ma
The distance between the sign 15b and the sign 15c is also a₁Indicated by
You. This is not known, and the sign position extraction unit 24
It is calculated based on the position of each extracted marker 15. Yo
The displacement e in the image coordinate system 12₁Is (a × Dc₁/
D). This shift amount e₁Only the image coordinate system 12
Of the marker 15b at₁By correcting c
The position Ic on the road in the image coordinate system 12₁Calculate
You. In this way, each time point in the image coordinate system 12
Find the road center position.

Therefore, when the CCTV camera 2 captures an image of the vehicle 1, it is necessary to know which position of the road 4 the vehicle 1 is traveling on the basis of the road center position in the image coordinate system 12. Can be. Furthermore, the figure
In step S3 of step 11, the width information from the center of the lane to the lane boundary position is read from the road width database to calculate the area up to the lane boundary in step S4. Road width database 13
, The width W _1L from the lane center position in the road coordinate system 13 to the lane left boundary line 31 and the lane right boundary line 32 from the lane center position in the road coordinate system 13
The width W _{1R up to} is stored. Therefore, the image coordinate system 12
Width w _1L from the road center position I ₁ c to the left boundary of the lane in is calculated by W _1L × _{a 1} / D, the width w _1R from the road center position I ₁ c to the right boundary of the lane W _1R × a ₁ /
D. From the position information and the width information, the image processing area corresponding to the road area in the image coordinate system 12 in the captured image is determined.

In FIG. 13, the image processing area 35 determined in the acquired image 11 is indicated by a dotted line. By determining the image processing area, the vehicle 1 traveling on the road 4
Road 4 from information in the image obtained from CCTV camera 2
You can know where you are traveling. This image processing area coincides with the lane area of the road 4 shown in the captured image.

[0025]

According to the present invention, it is possible to know the position of a vehicle traveling on a road with respect to the reference position of the road on the image. An excellent effect that an image processing area for image processing can be determined can be expected.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating the appearance of a road, a camera, and the like according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating a relationship between an image coordinate system and a road coordinate system.

FIG. 3 is a diagram showing details of a road coordinate system in FIG. 2;

FIG. 4 is a block diagram of a vehicle-mounted device that detects and records a deviation amount from a magnetic nail.

FIG. 5 is a diagram showing a positional relationship between a magnetic sensor and a magnetic nail.

FIG. 6 is a block diagram of a magnetic sensor.

FIG. 7 is a diagram showing a specific example of data recorded in a recording unit of the vehicle-mounted device.

FIG. 8 is a block diagram of a roadside processing device.

FIG. 9 is a diagram showing the appearance of a sign attached to the rear of the vehicle.

FIG. 10 is a diagram illustrating a specific example of data recorded in a recording unit of the roadside processing device.

FIG. 11 is a schematic diagram of an image processing area determination process.

FIG. 12 is a diagram illustrating image processing area determination.

FIG. 13 is a diagram illustrating a determined image processing area.

[Explanation of symbols]

 Reference Signs List 1 vehicle 2 camera 3 image processing device 4 road 5 magnetic nail 15 sign 16 magnetic sensor 17 vehicle position measuring unit 18 clock 19 recording unit 21 A / D conversion unit 22 memory 23 clock 24 signage position extracting unit 25 recording unit

Continued on the front page F-term (reference) 2F063 AA02 AA03 BA15 BB03 BD15 CA21 DA01 DA05 DA06 DB04 DD06 GA52 LA29 LA30 ZA01 2F065 AA03 AA04 BB15 CC11 DD00 FF04 FF26 JJ03 JJ26 KK01 MM02 PP05 QQ03 QQ24 5H180 A01 DD01

Claims (2)

[Claims] A vehicle traveling on a road is detected on the basis of transmission / reception of electromagnetic waves to / from the magnetic nail, the amount of deviation of a vehicle traveling on the road from a magnetic nail installed on the road being detected in a direction transverse to the road. Calculating a reference position of the vehicle by imaging the vehicle and performing image processing on the captured image; reading out the pre-stored magnetic nail position information and road size information in a cross-road direction; An image processing area determining method for determining an image processing area of a picked-up image based on a reference position, magnetic nail position information, and road size information in a cross-road direction. 2. The method according to claim 1, wherein the image processing area is a lane area.