JPH07266919A - Curved road alarm device - Google Patents

Abstract

PURPOSE:To realize a curved road alarm device giving an alarm to dangerous car speed by finding a radius of curvature of a curved road without finding the radius of curvature of the curved road from a white line coordinate in a two-dimensional picture image and without setting a transmitter on the curved road. CONSTITUTION:A radius of curvature of a curved road is decided by collating a letter pattern decided by applying binarization treatment and edge treatment to a picture image signal obtained from a two-dimensional camera picking up an image of a curve curvature display sign set on a curved road and a letter pattern of a previously memorized curve curvature, previously set allowed car speed is found from this radius of curvature, and when detected self car speed 2 is higher than this allowed car speed, it is judged as over speed and an alarm is energized for a specified period of time.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a curved road warning device, and more particularly to a device for issuing a warning when the vehicle speed is dangerous when a vehicle enters a curved road.

[0002]

2. Description of the Related Art In such a curved road warning device, it is necessary to first detect the radius of curvature of the curved road.
As such a conventional example, JP-A-3-186713 and
No. 3-139706 and the like, two white lines (guide lines) are detected from the output image of the two-dimensional camera, and the radius of curvature of the curved road is obtained from the coordinates of the white lines.

However, when the radius of curvature of the curved road is calculated from the white line coordinates in the two-dimensional image as in the conventional example, there are the following problems.

(1) At night, when the white line is not sufficiently illuminated, or when a soundproof wall is installed along the road, the white line in the distance cannot be detected and an accurate radius of curvature of the curved road is required. Absent.

(2) Since it is based on the "principle of monocular triangulation", the direction of the camera optical axis must be set strictly with respect to the road surface in calculating the distance of the white line, and when the vehicle pitches. Calculation error occurs in the radius of curvature of the curved road.

(3) In a curved road having a gradient or a bank, it is based on the above-mentioned principle of monocular triangulation.
There is a difference in the calculation of the radius of curvature of the curved road.

As described above, Japanese Patent Laid-Open No. 54-559 discloses a curved road warning device having no drawbacks based on the principle of monocular triangulation.
No. 23 publication is cited.

In this device, a transmitter is provided near the entrance of the curved road, the reference permissible vehicle speed information is transmitted from the transmitter to the receiver of the vehicle entering the curved road, and the received reference permissible vehicle speed is received. Based on the above, an alarm is issued based on the relationship with the current vehicle speed.

[0009]

However, in the case of such a conventional technique, it is necessary to newly install a transmitting device on each curved road, and it takes a long period to completely install the transmitting device and the transmitter is completely installed. Until it is done, it is not possible to exert sufficient effect.

In addition, there is a problem in that when such a transmitting device is newly installed on each curved road, a considerable cost is required and the maintenance thereof is also a heavy burden on the road installation / manager side.

Therefore, the present invention does not determine the radius of curvature of the curved road from the white line coordinates in the two-dimensional image, but also obtains the radius of curvature of the curved road without installing a transmitter on the curved road, which is dangerous vehicle speed. It is an object of the present invention to realize a curved road warning device that gives a warning to a driver.

[0012]

In order to achieve the above object, a curved road warning device according to the present invention is mounted on a vehicle, and is a two-dimensional camera for photographing a curve curvature radius display sign of a road, and a vehicle speed detection means. And an alarm device, and by comparing the character pattern detected by performing binarization processing and edge processing on the image signal obtained from the camera with the character pattern of the curve curvature radius stored in advance, the curve path Signal processing means for detecting the radius of curvature of the curve and obtaining an allowable vehicle speed stored in advance from the curve radius of curvature and energizing the alarm device for a certain period of time when the detected vehicle speed is higher than the allowable vehicle speed.
Is equipped with.

The signal processing means includes a memory for storing a character pattern of a curve radius of curvature and a memory for storing an allowable vehicle speed corresponding to the curve radius of curvature.

[0014]

[Function] A sign indicating the radius of curvature is installed at the entrance of the curved road of most highways, and the curvature radius is displayed on most of the sharp curved roads where attention must be paid to the vehicle speed when entering the curved road. The display value is also a value set in advance at the time of design and can be judged as a sufficiently accurate value.

Therefore, in the present invention, the signal processing means detects the character pattern by binarizing and edge-processing the image signal obtained from the two-dimensional camera mounted on the vehicle.

Then, by comparing and collating this character pattern with the character pattern having the curve curvature radius stored in advance in the memory, the curve curvature radius included in the curvature radius display mark of the curved road is recognized, and this curve curvature is recognized. The allowable vehicle speed stored in advance in the memory is obtained from the radius.

The signal processing means compares the obtained allowable vehicle speed with the current vehicle speed detected by the vehicle speed detecting means, and when the current vehicle speed is higher than the allowable vehicle speed, activates the alarm device for a certain period of time to give a signal. The processing means can issue an alarm to the driver.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows an embodiment of a curved road warning device according to the present invention. In the drawing, 1 is a photograph of a curve curvature radius display sign attached to a vehicle and installed on a curved road of a highway. In order to do so, a two-dimensional camera 2, a vehicle speed sensor as a vehicle speed detecting means, an alarm device 3, and a control unit 4 as a signal processing means.

The control unit 4 also includes an A / D converter 41 for converting the output image of the two-dimensional camera 1 into a digital signal, and a frame for storing the digital signal from the A / D converter 41. The memory 42 and the input circuit 4 for taking in the vehicle speed signal from the vehicle speed sensor 2.
3 and a CP for extracting the luminance edge portion in the captured image by binarizing the data stored in the frame memory 42.
U44, a frame memory 45 that stores data binarized by the CPU 44, a memory 46 that stores a character pattern (template) in advance, and a CPU
The ROM 47 stores programs and data to be executed by the CPU 44 in advance, and an output circuit 48 for outputting the processing result calculated by the CPU 44 to the alarm device 3.

FIG. 2 shows a flowchart of a processing program executed by the CPU 44 shown in FIG.
The operation of the embodiment of the present invention shown in FIG. 1 will be described below with reference to this flowchart.

First, initialization processing is performed on the CPU 44 (step S1).

Next, the timer counter incorporated in the CPU 44 is reset to "0", and the output circuit 48 is initialized to OFF (step S2).

Further, a vehicle speed signal is read from the vehicle speed sensor 2 via the input circuit 43, and the result is stored as the own vehicle speed V (step S3).

Further, the digital signal of the two-dimensional camera 1 output from the A / D converter 41 is stored in the frame memory 42 (step S4).

The two-dimensional image data once stored in the frame memory 42 is read out, the binarization processing is performed as described above, the edge processing for extracting the luminance edge portion in the captured image is performed, and the processing result data is obtained. It is stored in the frame memory 45 (step S5).

Therefore, by the binarization processing and the edge processing, the contours of character portions such as objects and signs in the image captured by the camera 1 are extracted and stored in the frame memory 45 as character pattern data. Become.

Next, the CPU 44 uses the frame memory 45.
The character pattern data stored in (4) is compared with the character pattern preset in the memory 46.

[0028] As shown in Table 1 below as the registration pattern, the capital letters "R", the number "0" to "9", lowercase "m" is the template data _{t R, t 0 ~t 9,} t
Each template is registered and set as _m , and template matching processing is executed to determine whether or not the registered pattern exists in the frame memory 45 (step S
6).

[0029]

[Table 1]

Then, in step S6, "R ○○○
When a pattern corresponding to “m” (the circle is a number) is detected, the three-digit “XX” section is used to calculate the curve curvature radius value (m
It is recognized as a unit (step S7).

Description of Template Matching Here, the process of recognizing the curve curvature radius value in step S7 will be described in detail below.

FIG. 3 shows a state in which the vehicle 10 is entering the clockwise curved road 20. At the entrance of the curved road 20, a curve curvature display sign 30 is erected as shown.

An example of an image taken by the two-dimensional camera 1 mounted on the vehicle 10 is shown in FIG.

Now, the two-dimensional image on the screen shown in FIG. 4 is a two-variable real function z = f (representing the entire density value f (x, y) of the point coordinates (x, y) on the screen. It is represented by x, y).

When the density value g (x, y) of the processed image is obtained from the density value f (x, y) of the original image by a certain process (for example, the binarization process described above), if the linear process is performed, The formula holds.

[0036]

[Equation 1]

Here, α (x, y) is a parameter defined by the processing content and is called a weighting function.

That is, the above equation (1) is applied to the point (u, v) of the input image at the density value g (x, y) of the processed image at the point (x, y) on the screen imaged by the camera 1. ), The concentration value f (u, v) contributes with the weight α (x−u, y−v),
It indicates that this is determined by the difference (x−u, y−v) in the coordinate values between the point (x, y) and the point (u, v).

In the template matching, the similarity, the identity and the difference between the standard image (template image) t (x, y) prepared in advance and the original image are evaluated and recognized using the following scales. -Perform a process to make a distinction.

Typical template manching is (1)
In the equation, the weighting function α (x, y) is set to the inverted value t (−x, −y) of the template (α (x, y) = t (−
x, -y)). This is expressed by the following equation using a normalization factor.

[0041]

[Equation 2]

By substituting g '(x, y) obtained by the equation (2) into g (x, y) of the equation (1), | g'
(x, y) | can be used as a measure of similarity evaluation between the template and the original image.

In the case of the above embodiment, the original image or the template is enlarged or reduced using the area within the outline of the character candidate as a parameter so that the size of the character pattern in the original image and the size of the template substantially match. Later, the above | g ′ (x, y) | is obtained, and if the value is equal to or larger than the set value, it can be recognized that the template character exists in the image. (If the template and the original image completely match, | g '(x, y) | = 1.)

In this way, it is determined in step S8 whether or not the curve curvature radius value is recognized. If it is determined that the curve curvature radius value is recognized, the recognition value is set to R (step S9).

Then, based on this recognized R value, RO
The allowable vehicle speed on the curved road, which is set in advance in M47 by an experiment or a calculation, is obtained from the mapping data shown in the figure, and the allowable vehicle speed is stored as V _R (step S1).
0, S11). The allowable vehicle speed with respect to the radius of curvature of the curve is publicly known as disclosed in Japanese Patent Laid-Open No. 54-55923.

Then, a set value is set in a built-in timer counter for generating an overspeed warning for a predetermined time (step S12).

[0047] Thereafter, the comparison determination and the allowable vehicle speed V _R stored in the host vehicle speed V and step S11 read at step S3 (step S13), and when it was found to be V> V _R, the current timer counter Is "1" (step S14), when it is not "1", the value of the timer counter is decremented by "1" (step S15), and when it is "1", the value of the timer counter is set to "1". It is reset to 0 "(step S16), and an alarm signal is given to the alarm device 3 via the output circuit 48 (step S17).

[0048] When it is found to be V ≦ V _R in step S13, it resets the timer counter "0" (step S18), and the output circuit 48 so as not to give an alarm signal to the alarm unit 3 (step S19).

Further, since the above-mentioned recognition of the curvature radius value of the curve (step S7) is usually performed in an instant, the CPU4
Since the recognition value does not exist the next time step S8 is passed due to the processing cycle of 4, the step S20
Check the value of the timer counter and proceed to steps S13 to S17 if the timer counter value is not "0". If the timer counter value is "0", jump to step S19 and generate an alarm output. Try not to.

With the above operation, the alarm output operates as follows.

Each time the curve curvature radius display sign on the expressway is recognized, the own vehicle speed is compared with the previously stored allowable vehicle speed. If the own vehicle speed ≦ the allowable vehicle speed, no alarm is issued and the own vehicle speed> allowed. In the case of vehicle speed, thereafter, an alarm is issued until the vehicle speed becomes equal to or less than the allowable vehicle speed or a predetermined time elapses after the alarm is issued.

[0052]

As described above, according to the curved road warning device of the present invention, the image signal obtained from the two-dimensional camera for picking up the curve curvature display sign installed on the curved road is binarized. The radius of curvature of the curved road is detected by collating the character pattern detected by performing the edge processing with the character pattern of the curve curvature stored in advance, and the allowable vehicle speed stored in advance is calculated from this radius of curvature. When the detected vehicle speed is higher than the vehicle speed, it is configured as an overspeed warning for a certain period of time, so an accurate curve curvature can be obtained before entering a curved road without requiring complicated correction calculation processing. It is possible to issue an alarm for the speed over a curved road at an accurate and appropriate timing, which contributes to safe operation of the vehicle.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of a curved road warning device according to the present invention.

FIG. 2 is a flowchart showing a processing procedure executed by a control unit used in the curved road warning device according to the present invention.

FIG. 3 is a diagram showing an installation example of a display sign when a vehicle enters a curved road.

FIG. 4 is a diagram showing an example of an image when a curvature display sign of a curved road is photographed by a camera.

[Explanation of symbols]

 1 two-dimensional camera 2 vehicle speed sensor 3 alarm device 4 control unit 42, 45 frame memory 44 CPU 46 memory In the drawings, the same reference numerals indicate the same or corresponding parts.

Claims (2)

[Claims] 1. A two-dimensional camera mounted on a vehicle for photographing a curve curvature radius display sign of a road, and a vehicle speed detecting means,
The curvature of the curved road is determined by collating the alarm device with a character pattern determined by performing binarization processing and edge processing on the image signal obtained from the camera and a character pattern having a curve curvature radius stored in advance. Signal processing means for detecting a radius and for activating the alarm device for a certain period of time when a previously stored allowable vehicle speed is obtained from the curve curvature radius and the detected vehicle speed is higher than the allowable vehicle speed. Curve road warning device. 2. The signal processing means according to claim 1, further comprising a memory storing a character pattern of a curve radius of curvature and a memory storing an allowable vehicle speed corresponding to the curve radius of curvature. Curve road warning device.