JP2000285379A - Image type vehicle sensor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a vehicle sensor capable of surely detecting a vehicle stopping at a stop line without being influenced by a road surface condition, the color of the vehicle, etc. SOLUTION: A 1st camera 21 picks up the image of a passing vehicle measurement line 14 and outputs a 1st luminance pattern signal S1. A 2nd camera 22 picks up the image of a measurement designated area 12 and outputs a 2nd luminance pattern signal S2. A signal processor 3 stores the signal S1, also compares the signal S1 with the signal S2 and detects the existence/absence of a stopping vehicle within the area 12.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] The present invention relates to an image type vehicle sensor.

[0002]

2. Description of the Related Art In general, an image-type vehicle sensor captures an image of a passing vehicle measurement line set in advance by a camera. The vehicle is sensed from the difference from the vehicle brightness pattern when the vehicle is running. At night, the vehicle is sensed by detecting the peak brightness value of the headlights of the vehicle. In the case of twilight, vehicle detection is performed when a vehicle is detected in either of the above modes.

[0003] When vehicle detection is performed in the daytime by the above-described vehicle detection method, there is no problem because the difference between the vehicle luminance pattern and the road luminance pattern is large for white vehicles. The difference between the vehicle luminance pattern and the road luminance pattern cannot be sufficiently obtained. Therefore, the detection of a black vehicle is not a black portion of the vehicle body, for example,
Vehicle detection is performed by detecting a luminance pattern corresponding to a white portion such as a headlight portion.

One of the problems of the above-described vehicle sensing method is to focus on a whole or partial white portion of the vehicle and detect the vehicle based on the difference between the vehicle luminance pattern and the road luminance pattern. For this reason, on a whitish road or a road that has become white due to snowfall or snowfall, the vehicle may not be able to be identified and detected from the road.

[0005] In particular, at an intersection where a vehicle that stops and starts must be detected in a certain area, if the road becomes a white road as described above, the presence or absence of a white portion of the vehicle or the vehicle body It is impossible to detect a vehicle only by the ratio of the white portion.

[0006] Another problem is that since various marks such as traffic signs are provided on the road surface, there is a risk of confusion between these marks and vehicle brightness patterns.

[0007]

SUMMARY OF THE INVENTION An object of the present invention is to provide an image processing apparatus that is not affected by road surface conditions or vehicle color.
An object of the present invention is to provide a vehicle sensor capable of reliably detecting a vehicle stopped at a stop line.

[0008]

To solve the above-mentioned problems, a vehicle detector according to the present invention comprises a first camera and a second camera.
And a signal processing device.

The first camera captures an image of a passing vehicle measurement line and outputs a first luminance pattern signal. The passing vehicle measurement line is set on the road and before the vehicle stop line when viewed in the traveling direction of the vehicle.

[0010] The second camera captures an image of the designated measurement area and outputs a second luminance pattern signal. The measurement designated area is set on a road in front of the passing vehicle measurement line as viewed in the traveling direction of the vehicle and immediately before the vehicle stop line.

The signal processing device stores a first luminance pattern signal supplied from the first camera, and stores the first luminance pattern signal and a second luminance pattern signal supplied from the second camera. In contrast to the luminance pattern signal,
The presence or absence of a stopped vehicle in the measurement designated area is detected.

In the vehicle sensor described above, the first camera captures an image of the passing vehicle measurement line and outputs a first luminance pattern signal. The first luminance pattern signal obtained by imaging the passing vehicle measurement line includes different luminance patterns depending on whether there is a vehicle passing through the passing vehicle measurement line or not. Therefore, the first camera outputs a first luminance pattern signal having a different luminance pattern depending on whether or not the vehicle is on the passing vehicle measurement line. The passing vehicle measurement line is set before the vehicle stop line when viewed in the traveling direction of the vehicle. What is the vehicle stop line?
This is a white stop line set immediately before entering the intersection. Because the passing vehicle measurement line is set before the vehicle stop line, the vehicle passing through the passing vehicle measurement line before the vehicle reaches the vehicle stop line set immediately before the intersection,
Detected by the first camera.

The second camera captures an image of the designated measurement area and outputs a second luminance pattern signal. The measurement designated area is set in front of the passing vehicle measurement line and immediately before the vehicle stop line when viewed in the traveling direction of the vehicle. Since the measurement designated area imaged by the second camera includes the vehicle passing through the passing vehicle measurement line, the second luminance pattern signal output from the second camera is output from the first camera. Vehicle brightness pattern that matches the vehicle brightness pattern included in the first brightness pattern signal.

The signal processing device stores the first luminance pattern signal supplied from the first camera. Then, the stored first luminance pattern signal is compared with the second luminance pattern signal supplied from the second camera. When there is a stopped vehicle, a luminance pattern that matches the first luminance pattern signal always exists in the second luminance pattern.
Therefore, the vehicle stopped in the designated measurement area can be reliably detected by distinguishing it from various marks such as traffic signs on the road surface.

In the case where a black vehicle is detected in the passing vehicle measurement line and the designated measurement area, if the difference between the vehicle luminance pattern and the road luminance pattern cannot be sufficiently obtained, the detection of the black vehicle is performed in the black portion of the vehicle body. Instead, for example, vehicle detection is performed by detecting a luminance pattern corresponding to a white portion such as a headlight portion. Thereby, the luminance pattern of the black vehicle can be reliably obtained in both the passing vehicle measurement line and the measurement designated area.

When detecting a black vehicle, if the road is whitish or the road is white due to snowfall or snowfall and the difference between the vehicle luminance pattern and the road luminance pattern can be sufficiently obtained. The vehicle is detected by detecting a luminance pattern corresponding to a black portion of the vehicle body. Thereby, the luminance pattern of the black vehicle can be reliably obtained in both the passing vehicle measurement line and the measurement designated area.

In the case of detecting a white vehicle, when the road is whitish or the road is white due to snowfall or snowfall, and the difference between the vehicle luminance pattern and the road luminance pattern cannot be sufficiently obtained. In the detection of a white vehicle, vehicle detection is performed by detecting a luminance pattern corresponding to a black portion included in the vehicle, not a white portion of the vehicle body. Thereby, the luminance pattern of the white vehicle can be reliably obtained in both the passing vehicle measurement line and the measurement designated area.

When detecting a white vehicle, if the road is black and there is no snowfall or snowfall and the difference between the vehicle luminance pattern and the road luminance pattern can be sufficiently obtained, the luminance corresponding to the white portion of the vehicle body is detected. Vehicle detection is performed by detecting a pattern. Thereby, the luminance pattern of the white vehicle can be reliably obtained in both the passing vehicle measurement line and the measurement designated area.

Therefore, the vehicle stopped at the stop line can be reliably detected without being affected by the surface condition of the road or the color of the vehicle.

The other features of the present invention and the operation and effects thereby will be described in more detail with reference to the accompanying drawings.

[0021]

FIG. 1 is a plan view showing the relationship between the configuration of a vehicle sensor according to the present invention and a passing vehicle measurement line and a designated measurement area. In FIG. 1, a road 10 on one side immediately before entering an intersection CR is shown. Road 10
Is separated from other lanes by a center line 13. On the road 10, at the position just before entering the intersection CR,
A vehicle stop line 11 is provided. In the embodiment of FIG.
In the case where the black vehicle 4 is detected, the case where the difference between the vehicle luminance pattern and the road luminance pattern cannot be sufficiently obtained is shown.

The vehicle detector comprises a first camera 21 and a second camera 21.
Camera 22 and the signal processing device 3. The first camera 21 captures an image of the passing vehicle measurement line 14 and outputs a first luminance pattern signal S1.
Is set before the vehicle stop line 11 when viewed in the traveling direction a of the vehicle 4. In the drawing, only one first camera 21 is provided, but a plurality of first cameras 21 may be provided.

The second camera 22 captures an image of the measurement designated area 12 and outputs a second luminance pattern signal S2. The measured area 12 is viewed from the passing vehicle measurement line 14 when viewed in the traveling direction a of the vehicle 4. It is set ahead and immediately before the vehicle stop line 11. In the figure, only one second camera 22 is provided, but a plurality of second cameras 22 may be provided.

In the embodiment of FIG. 1, when the black vehicle 4 is detected, it is assumed that the difference between the vehicle luminance pattern and the road luminance pattern cannot be sufficiently obtained. The vehicle is detected by detecting a luminance pattern corresponding to, for example, a white portion such as a headlight portion instead of the black portion.

The signal processing device 3 stores the first luminance pattern signal S 1 supplied from the first camera 21, and stores the stored first luminance pattern signal S 1 and the first luminance pattern signal S 1 supplied from the second camera 22. The presence or absence of the stopped vehicle 4 that is stopped in the measurement designated area 12 is detected by comparing with the second luminance pattern signal S2 thus obtained.

In the vehicle sensor described above, the first camera 21 images the passing vehicle measurement line 14 and outputs a first luminance pattern signal S1. Passing vehicle measurement line 1
4, the first luminance pattern signal S1 obtained by imaging
Different luminance patterns are included depending on whether there is a vehicle 4 passing through the passing vehicle measurement line 14 or not.

In the illustrated embodiment, it is assumed that the black vehicle 4 is detected and that the difference between the vehicle luminance pattern and the road luminance pattern cannot be sufficiently obtained. In this case, when detecting the vehicle 4, a luminance pattern corresponding to, for example, a white portion such as a headlight portion is detected instead of a black portion of the vehicle body.

FIG. 2 shows the passing vehicle measurement line 14
A vehicle luminance pattern A11 and a road luminance pattern B1 included in a first luminance pattern signal S1 obtained by the first camera 21 are shown. In FIG. 2, the passing vehicle measurement line pixel M (M = m1 to mp) is plotted on the horizontal axis, and the luminance is plotted on the vertical axis. The passing vehicle measurement line pixel M is
In the case of the present embodiment, the case where the passing vehicle measurement line 14 is divided into p pixels m1 to mp is shown. The set vehicle brightness pattern A11 is a pattern having two brightness peaks corresponding to a white portion such as a headlight portion. Although the road luminance pattern B11 includes a ripple component,
The pattern has a substantially constant average value. The two luminance peak values included in the vehicle luminance pattern A11 indicate higher luminance than the road luminance pattern B11.

As shown in FIG. 2, the first camera 21 outputs the vehicle luminance pattern A11 when the vehicle 4 is on the passing vehicle measurement line 14, and outputs the vehicle luminance pattern A11 when the vehicle 4 is not present. Is a different road luminance pattern B11
Will be output. Since the passing vehicle measurement line 14 is set before the vehicle stop line 11 when viewed in the traveling direction a of the vehicle 4, the vehicle 4
Before the vehicle reaches the vehicle stop line 11 set immediately before the intersection CR, the vehicle 4 passing through the passing vehicle measurement line 14
Is detected by the first camera 21.

The second camera 22 captures an image of the designated measurement area 12 and outputs a second luminance pattern signal S2. The measurement designated area 12 is set in front of the passing vehicle measurement line 14 and immediately before the vehicle stop line 11 when viewed in the traveling direction a of the vehicle 4.

FIG. 3 is a diagram for explaining a state in which the vehicle 4 enters the designated measurement area 12 and is stopped immediately before the vehicle stop line 11, and an image of the designated measurement area 12 by the second camera 22 at that time. It is. The measurement designated area 12 has its bad area determined in consideration of the area occupied by the stopped vehicle.
Then, the designated measurement area 12 is divided on the second camera 22 into pixels (n1 to nr) and (m1 to Mp) in r rows and p columns.

Since the measurement designated area 12 imaged by the second camera 22 includes the vehicle 4 that has passed the passing vehicle measurement line 14, the second luminance pattern signal S2 output from the second camera 22 Is the first luminance pattern signal S
A luminance pattern that matches the luminance pattern 1 is included.

FIG. 4 shows a vehicle brightness pattern A11 included in the first brightness pattern signal S1 obtained by the first camera 21 on the passing vehicle measurement line 14, and obtained by the second camera 22 in the measurement designated area 14. FIG. 9 is a diagram showing a relationship with a vehicle luminance pattern A12 included in a second luminance pattern signal S2. FIG. 4 also shows a road luminance pattern B1. In FIG. 4, the vehicle luminance pattern A12 included in the second luminance pattern signal S2 is
It is shown as obtained at pixel N = n2.

As shown in FIG. 4, the passing vehicle measurement line 1
In 4, the vehicle luminance pattern A11 obtained by the first camera 21 and the vehicle luminance pattern A12 obtained by the second camera 22 in the measurement designated area 14 are almost similar. From this relationship, it is assumed that the vehicle has the vehicle luminance pattern A11 and the passing vehicle measurement line 14
The signal processing device 3, which can determine that the vehicle 4 detected in step S 1 is stopped in the measurement designated area 12, outputs the first luminance pattern signal S 1 supplied from the first camera 21.
Is stored. Then, the stored first luminance pattern signal S1 and the second luminance pattern signal S2 supplied from the second camera 22 are compared, and the measurement designated area 12 is compared.
The presence or absence of the vehicle 4 stopped inside is detected. Stopped vehicle 4
If there is, as shown in FIG. 4, the second luminance pattern signal S2 always includes a vehicle luminance pattern A12 that matches the vehicle luminance pattern A11 included in the first luminance pattern signal S1. Therefore, it is possible to reliably detect the vehicle 4 stopped in the measurement designated area 12. Stopped vehicle 4 even if a traffic sign is attached on the road surface
Can be reliably detected separately from the traffic sign.

Although illustration is omitted, when the black vehicle 4 is detected, the road 10 may be whitish, or the road 10 may be white due to snowfall or snowfall.
When the difference between the vehicle luminance pattern and the road luminance pattern is sufficiently obtained, vehicle detection is performed by detecting a luminance pattern corresponding to a black portion of the vehicle body. Thereby, the luminance pattern of the black vehicle 4 can be reliably obtained in both the passing vehicle measurement line 14 and the measurement designated area 12.

FIG. 5 shows a case where the white vehicle 4 is detected, and the road 10 is white due to whitening, snowfall or snowfall, and the difference between the vehicle luminance pattern and the road luminance pattern is sufficiently large. It is a figure which shows vehicle detection when it cannot be taken. In the figure, the same reference numerals as those in FIG. 1 indicate the same components.

In this case, the white vehicle 4 senses by detecting a luminance pattern corresponding to a black portion included in the vehicle 4 instead of the white portion of the vehicle body. Even in the white vehicle 4, not all of the vehicle body is white, and, for example, tires and the like are black. Or, a part of the vehicle body such as a bumper usually has a black portion.

The first camera 21 captures an image of the passing vehicle measurement line 14 and outputs a first luminance pattern signal S1. The first luminance pattern signal S1 obtained by imaging the passing vehicle measurement line 14 includes different luminance patterns depending on whether there is a vehicle 4 passing through the passing vehicle measurement line 14 or not.

FIG. 6 shows the passing vehicle measurement line 14
The vehicle luminance pattern A21 and the road luminance pattern B11 included in the first luminance pattern signal S1 obtained by the first camera 21 are shown. In FIG. 2, the passing vehicle measurement line pixels M (M = m1 to m11) are plotted on the horizontal axis, and the luminance is plotted on the vertical axis. The passing vehicle measurement line pixel M is
This is a pixel viewed on the first camera 21 or the display, and in the case of the embodiment, shows a case where the passing vehicle measurement line 14 is divided into p pixels m1 to mp. The set vehicle brightness pattern A21 is a pattern having two brightness minimum values corresponding to the black portion. The road luminance pattern B11 includes a ripple, but has a substantially constant average value. The two luminance minimum values included in the vehicle luminance pattern A21 are more than the road luminance pattern B11.
Shows fairly low brightness.

As shown in FIG. 6, the first camera 21 outputs a vehicle luminance pattern A21 when the vehicle 4 is on the passing vehicle measurement line 14, and outputs the vehicle luminance pattern A21 when the vehicle 4 is not present. Is a different road luminance pattern B11
Will be output. Since the passing vehicle measurement line 14 is set before the vehicle stop line 11 when viewed in the traveling direction a of the vehicle 4, the vehicle 4
Before the vehicle reaches the vehicle stop line 11 set immediately before the intersection CR, the vehicle 4 passing through the passing vehicle measurement line 14
Is detected by the first camera 21.

The second camera 22 captures an image of the designated measurement area 12 and outputs a second luminance pattern signal S2. The measurement designated area 12 is set in front of the passing vehicle measurement line 14 and immediately before the vehicle stop line 11 when viewed in the traveling direction a of the vehicle 4.

FIG. 7 is a view for explaining a state in which the vehicle 4 enters the measurement designated area 12 and is stopped immediately before the vehicle stop line 11, and an image of the measurement designated area 12 by the second camera 22 at that time. It is.

Since the measurement designated area 12 imaged by the second camera 22 includes the vehicle 4 passing through the passing vehicle measurement line 14, the second luminance pattern signal S2 output from the second camera 22 Is the first luminance pattern signal S
A luminance pattern that matches the luminance pattern 1 is included.

FIG. 8 shows a vehicle brightness pattern A21 included in the first brightness pattern signal S1 obtained by the first camera 21 on the passing vehicle measurement line 14, and obtained by the second camera 22 in the measurement designated area 14. FIG. 10 is a diagram showing a relationship with a vehicle luminance pattern A22 included in a second luminance pattern signal S2. FIG. 8 also shows a road luminance pattern B2. In FIG. 8, the vehicle luminance pattern A12 included in the second luminance pattern signal S2 is
It is shown as obtained at pixel N = n2.

As shown in FIG. 8, passing vehicle measurement line 1
In 4, the vehicle luminance pattern A21 obtained by the first camera 21 and the vehicle luminance pattern A22 obtained by the second camera 22 in the measurement designated area 14 are almost similar. From this relationship, it can be determined that the vehicle 4 detected on the passing vehicle measurement line 14 is stopped in the measurement designated area 12 as the vehicle luminance pattern A21. The first luminance pattern signal S1 supplied from is stored. Then, the stored first luminance pattern signal S1 is compared with the second luminance pattern signal S2 supplied from the second camera 22, and the vehicle 4 stopped in the measurement designated area 12 is compared. Detect presence / absence. When there is a stopped vehicle 4, as shown in FIG. 8, the second luminance pattern signal S2 includes a vehicle luminance pattern A21 included in the first luminance pattern signal S1.
Is always present. Therefore, it is possible to reliably detect the vehicle 4 stopped in the measurement designated area 12.

Although not shown, when the white vehicle 4 is detected, when the road 10 is black, there is no snowfall or snowfall, and the difference between the vehicle luminance pattern and the road luminance pattern can be sufficiently obtained, Vehicle detection is performed by detecting a luminance pattern corresponding to a white portion of the vehicle body. Thereby, the passing vehicle measurement line 14 and the measurement designated area 1
In both cases, the luminance pattern of the white vehicle 4 can be reliably obtained.

Therefore, the surface condition of the road 10 or the vehicle 4
The vehicle 4 stopped at the stop line can be reliably detected without being affected by the color of the vehicle.

The signal processing device 3 outputs the luminance pattern signal S
1, S2 may be Fourier-transformed and a luminance pattern may be detected from the spectrum pattern.
The Fourier transform of the luminance pattern signals S1 and S2 and the extraction of the spectrum pattern can be easily realized by existing techniques. Fourier transform is fast Fourier transform (FFT transform)
It is desirable to do this by means. FFT transforms are well known in the field of signal processing.

FIG. 9 is a flowchart for explaining the stopped vehicle detecting operation of the vehicle detector shown in FIGS. As described above, when the first camera 21 detects the vehicle 4 passing through the passing vehicle measurement line 14, the vehicle luminance pattern of the passing vehicle 4 is stored in the signal processing device 3. The signal processing device 3 compares the stored vehicle luminance pattern with the vehicle luminance pattern obtained by the second camera 22. If the two vehicle luminance patterns can be considered to be similar or overlap, the measurement designated area 12 It is determined that there is a stopped vehicle. Vehicle 4 passing through passing vehicle measurement line 14
Is not detected, and when the vehicle luminance pattern obtained by the first camera 21 does not correspond to the vehicle luminance pattern obtained by the second camera 22, the process returns to the start point of the stopped vehicle detection operation. .

FIG. 10 is a block diagram of a vehicle sensor according to the present invention. The video amplifier 311 is connected to the first camera 21
Alternatively, the luminance pattern signals S1 and S2 supplied from the second camera 22 are amplified to an appropriate level. Image A / D
The conversion circuit 315 converts the luminance pattern signals S1 and S2 into numerical values and converts the data into dual image memories 321 and 322.
To supply.

The synchronization signal extracting circuit 312 extracts a horizontal synchronization signal and a vertical synchronization signal based on the luminance pattern signals S1 and S2. The horizontal / vertical address counter 313 digitizes the current scanning position on the scanning line based on the horizontal synchronizing signal and the vertical synchronizing signal, and obtains the corresponding horizontal address and vertical address. Then, the horizontal address and the vertical address are supplied to the image memories 321 and 322.

The ROM 332 stores a program and is connected to the CPU 34. RAM 331 is
It constitutes an external storage area of the CPU 34 and is connected to the CPU 34 via a data bus.

The CPU 34 detects the presence or absence of the vehicle 4 based on the waveform correlation between the normalized data when the vehicle 4 does not exist in the passing vehicle measurement line 14 and the designated measurement area 12 and the normalized data obtained by the measurement. The CPU 34
When the presence of the vehicle 4 is detected, the output circuit 35
, The vehicle detection signal S3 is output.

[0054]

As described above, according to the present invention, a vehicle sensor capable of reliably detecting a vehicle stopped at a stop line without being affected by the surface condition of the road or the color of the vehicle. Can be provided.

[Brief description of the drawings]

FIG. 1 is a plan view showing a relationship between a configuration of a vehicle sensor according to the present invention and a passing vehicle measurement line and a designated measurement area.

FIG. 2 is a diagram showing a vehicle luminance pattern and a road luminance pattern obtained by a first camera on a passing vehicle measurement line.

FIG. 3 shows a state in which a vehicle enters a measurement designated area and is stopped immediately before a vehicle stop line, and a second camera 22 at that time.
FIG. 9 is a diagram for describing imaging of a measurement designated area by the following.

FIG. 4 is a diagram illustrating a relationship between a vehicle luminance pattern obtained by a first camera on a passing vehicle measurement line and a vehicle luminance pattern obtained by a second camera in a measurement designated area.

FIG. 5 is a diagram showing a case where a road is whitish, or white due to snowfall or snow cover, and a difference between a vehicle luminance pattern and a road luminance pattern cannot be sufficiently obtained when a white vehicle is detected. It is a figure showing vehicle detection.

FIG. 6 is a diagram showing a vehicle luminance pattern and a road luminance pattern obtained by a first camera on a passing vehicle measurement line.

FIG. 7 is a diagram illustrating a state in which a vehicle enters a designated measurement area and stops immediately before a vehicle stop line, and imaging of the designated measurement area by a second camera at that time.

FIG. 8 is a diagram showing a relationship between a vehicle luminance pattern obtained by a first camera on a passing vehicle measurement line and a vehicle luminance pattern obtained by a second camera in a measurement designated area.

FIG. 9 is a flowchart illustrating signal processing in the signal processing device.

FIG. 10 is a block diagram of a vehicle sensor according to the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 Vehicle stop line 12 Measurement designated area 14 Passing vehicle measurement line 21, 22 Camera 3 Signal processing device S1, S2 Luminance pattern signal S3 Vehicle detection signal

Claims (1)

[Claims] 1. A vehicle sensor including a first camera, a second camera, and a signal processing device, wherein the first camera captures an image of a passing vehicle measurement line to generate a first luminance pattern. A signal is output, and the passing vehicle measurement line is on the road, and is set before the vehicle stop line when viewed in the traveling direction of the vehicle, and the second camera captures an image of the measurement designated area. A second luminance pattern signal is output, and the measurement designated area is set on a road ahead of the passing vehicle measurement line and just before the vehicle stop line, as viewed in the traveling direction of the vehicle, The signal processing device stores a first luminance pattern signal supplied from the first camera, and stores the first luminance pattern signal and a second luminance pattern signal supplied from the second camera. And the measurement designated area A vehicle detector that detects the presence or absence of a stopped vehicle in the building.