JP2003044863A - Device for recognizing partition line - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a partition line recognizing device capable of appropriately setting a partition line recognized state and a partition line unrecognized state even when stains are attached to a camera lens and a windshield. SOLUTION: This partition line recognizing device for photographing a white line on a vehicle traveling road with a camera 1 mounted on a vehicle, which judges that a partition line is recognized when the recognition rate of the photographed white line is larger than a preset threshold, and judges that the partition line is not recognized when the recognition rate is smaller than the threshold, has a stain detecting part 5b for detecting stains attached to the camera or the windshield of the vehicle, and changes thresholds when the stain detecting part 5b detects the attached stains.

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention is mounted on a vehicle,
The present invention relates to a partition line recognition device that recognizes a partition line such as a white line on a vehicle traveling path.

[0002]

2. Description of the Related Art A white line recognition device is used, for example, in a lane departure warning device, a drowsiness driving warning device, or the like to determine the lane in which the vehicle is traveling.

For example, in a white line recognition device used in a dozing driving warning device disclosed in Japanese Patent Laid-Open No. 9-390601, a camera is attached in front of a vehicle, and a luminance is differentiated from an image taken by the camera. A white line in front of the vehicle is recognized by applying a filter.

At this time, even if the white line cannot be partially recognized due to reflection of illumination at night or water pool on the road surface,
If the white line is recognizable in a certain threshold value, for example, 50% or more, it is determined that the white line is recognizable,
The dozing driving alarm control is continuously performed. On the other hand, when the white line area is below the threshold value, the white line is not recognized and the dozing driving warning control is not performed.

[0005]

In the conventional white line recognition apparatus described above, for example, when the white line cannot be recognized in a part of the image, no matter what the cause of the recognition is, it is uniform. White line recognition is judged based on the threshold value. Therefore, for example, when dirt is attached to the lens portion of the camera, the dirty area is treated as a white line detection failure area even though the road surface is not photographed in the first place. .

Therefore, in reality, although the white line is recognized at a high rate, the dirty area is treated as the white line non-detection portion, and the white line non-recognition state is highly likely. There was a problem that it would become.

The present invention has been made to solve such conventional problems, and the purpose thereof is to:
If dirt on the camera is detected, and if dirt is found, the threshold used for judging the recognition of the partition line (white line) will be changed according to the size of the dirt, to indicate that the partition line is not recognized. An object of the present invention is to provide a partition line recognition device that can reduce the possibility that

[0008]

In order to achieve the above-mentioned object, the invention according to claim 1 of the present application was photographed by photographing a partition line of the vehicle traveling path by a photographing means mounted on the vehicle. In the partition line recognition device, when the recognition rate of the partition line is larger than a preset threshold value, the partition line recognition state is determined, and when it is smaller than the threshold value, the partition line non-recognition state is determined. It is characterized in that it has means or dirt detecting means for detecting dirt on the windshield of the vehicle, and the threshold value is changed when the dirt is detected by the dirt detecting means.

According to a second aspect of the present invention, the threshold value is set according to a ratio of an area of dirt detected by the dirt detecting means to a shooting area of the shooting means. .

According to a third aspect of the present invention, when the ratio of the area of the dirt detected by the dirt detecting means to the shooting area by the shooting means is equal to or larger than a first predetermined value, The threshold value is a fixed value, and when it is less than the first predetermined value, the threshold value is changed according to the ratio of the area of the stain.

According to a fourth aspect of the present invention, the ratio of the area of the dirt detected by the dirt detecting means to the shooting area by the shooting means is a second predetermined value larger than the first predetermined value. If the above is the case, it is characterized in that it is determined that the partition line is not recognized.

[0012] The invention according to claim 5 is characterized by comprising notifying means for notifying an occupant of a vehicle when dirt is detected by the dirt detecting means.

[0013]

According to the first aspect of the present invention, the threshold value used for judging the partition line recognition is detected by detecting the dirt on the photographing means or the windshield of the vehicle and by the presence or absence of the dirt. Since the value is changed, it is possible to reduce the ratio of the partition line non-recognition state when dirt is present.

According to the second aspect of the present invention, since the threshold value is variable according to the ratio of the area of dirt to the photographing area, an appropriate threshold value can be set according to the area of dirt and the partition line. The rate of non-recognition can be reduced.

According to the third aspect of the present invention, when the ratio of the area of dirt to the photographing area is equal to or larger than the first predetermined value, the threshold value is fixed, so that the area of dirt becomes large. In this case, it is possible to prevent the threshold value from falling too low, and it is possible to improve the partition line recognition accuracy.

According to another aspect of the present invention, when the ratio of the area of dirt to the photographing area is equal to or larger than the second predetermined value,
Since the partition line is not recognized, the area of dirt becomes very large, and even if the partition line can be recognized in the area where there is no dirt, it cannot be practically used as a partition line. Such a situation can be reliably detected.

In the fifth aspect of the present invention, when dirt is detected by the dirt detecting means, the passenger of the vehicle is immediately notified, so that the passenger takes appropriate measures such as wiping off the dirt. be able to.

[0018]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration when a partition line recognition device according to an embodiment of the present invention is applied to a lane departure warning device. In the present embodiment, the partition line drawn on the vehicle traveling path is described as a “white line”, but the present invention is not limited to this and is also applied to a yellow line or a partition line of another color. be able to.

The lane departure warning device recognizes the white line in front of the own vehicle, judges whether the own vehicle deviates from the driving lane based on this data, and when it is detected that the vehicle departs from the lane. , An alarm signal is output to notify the passengers of the vehicle.

As shown in FIG. 1, the lane departure warning system 10 detects a camera (image taking means) 1 for taking an image of a road surface outside the vehicle, an arithmetic circuit section 5, and a departure of the vehicle from the driving lane. Camera 1, and as described below
When the lens or the windshield is dirty, a display (notifying means) 6 for notifying it by screen display and a speaker (notifying means) 7 for notifying by voice
And a partition line recognition device 11 composed of

Further, a steering angle sensor 2, a left wheel speed sensor 3 and a right wheel speed sensor 4 are provided, and each detection signal detected by each of these sensors 2, 3 and 4 is sent to an arithmetic circuit section 5. To be given.

The arithmetic circuit section 5 has a dirt detecting section (dirt detecting means) 5a for detecting the presence or absence of dirt adhering to the lens of the camera 1 or the windshield portion in front of the camera 1 and the size of the dirt. is doing.

FIG. 2 is a flowchart of the processing procedure of the vehicle deviation warning device 10 described above, and the operation of this embodiment will be described below with reference to the flowchart.

The flowchart shown in FIG. 2 is executed in the arithmetic circuit section 5 shown in FIG. 1 at a predetermined control cycle (for example, 10 msec). First, step S1
Then, the camera 1 captures a road surface in front of the vehicle (for example, 5 to 30 meters), and the captured black and white image information is loaded into the arithmetic circuit unit 5.

Then, in step S2, the dirt detector 5
“A” detects whether the lens of the camera 1 or the windshield of the vehicle is dirty.

In this processing, when the movement of the vehicle is detected by the movement detecting means (the steering angle sensor 2, the left wheel speed sensor 3, the right wheel speed sensor 4) for detecting the movement of the vehicle, a certain image area The presence of dirt is determined by whether or not a fixed pattern that does not move is detected. That is, if the fixed pattern exists even though the vehicle is moving, it is determined that the fixed pattern is due to dirt. On the other hand, when the fixed pattern does not exist, it is determined that the stain does not exist.

When it is determined that stains exist, the size of the stains is measured from the area of the fixed pattern.

Here, as the method of detecting the above-mentioned movement of the vehicle, for example, a method of judging that the vehicle is moving when the steering angle data obtained from the steering angle sensor 2 is changed, or the wheel speeds of the left and right wheels are determined. A method of detecting whether or not the vehicle is moving based on a change in the difference between the left and right vehicle speed data obtained from the sensors 3 and 4 can be adopted.

In step S3, it is judged in the process of step S2 whether or not the lens of the camera 1 or the windshield has stains.
If YES in step S3, and in the process of step S4,
The presence of dirt is notified to the vehicle occupant by the display 6 or the speaker 7.

Next, in step S5, when it is determined whether the white line is recognized (whether the partition line is recognized or the partition line is not recognized) according to the size of the detected stain, The threshold value D is calculated. Details of this processing will be described later.

Then, in step S6, a process of detecting a white line from the image taken by the camera 1 is performed. Although various methods are known as white line detection methods, in the present embodiment, the brightness value of the image is differentiated along the lateral direction of the image, and the brightness value changes abruptly based on the differentiated value. Find a point to do and detect this as a white line.

That is, as shown in FIG. 4A, when an image including the white line 8 is photographed, the differential value of the luminance is obtained for the data on the scanning line 9 in the horizontal direction. Then, as shown in FIG. 6B, the peaks of the differential values appear in the order of minus and plus from the left side, and the intervals between the peaks are considered to be appropriate as the white line 8 (for example, 30 dots). If it is within (inside), it is determined that this is a white line, and the intermediate points are saved as M1 and M2.

In step S7, white line recognition is determined based on the processing result of step S6. That is, since there is one lane on the left and one on the right with respect to the vehicle, when scanning for white line detection is performed from the left side to the right side of the screen, two white lines are detected in one scan. Also,
If the number of screen pixels is, for example, 300 in the vertical direction,
If white lines exist from the top to the bottom of the screen and white lines can be detected for all the scanning lines (300), a total of 600 white lines will be recognized.

However, all white lines may not be recognized due to snow, fog, heavy rain, or reflection of illumination at night. Here, the number of recognized white lines is calculated in step S5.
A process of comparing whether or not it is larger than the threshold value D obtained in the process of is performed. In this case, it is determined that the partition line is not recognized.
For example, when the threshold value D is set to 50%, it is determined that the partition line is recognized when 300 or more of the 600 white lines in total are recognized.

When the partition line is recognized,
If YES in step S7 and the lane departure warning function is turned on in step S8, an alarm is output when the vehicle deviates from the white line included in the image captured by the camera 1. Set.

On the other hand, when the partition line is not recognized,
If NO in step S7, the lane departure warning function is turned off in step S9. In this case, since the white line cannot be normally recognized from the image captured by the camera 1, it is not determined whether the host vehicle deviates from between the two white lines.

Next, the specific processing of the threshold value calculation processing shown in step S5 will be described with reference to the flowchart shown in FIG.

When dirt is detected in step S2 shown in FIG. 2, it is determined in step S51 of FIG. 3 whether the size of the dirt is larger than zero, and it is larger than zero. If not, step S51 results in NO, and the threshold value D is set to B (for example, B = 0.5) in the process of step S56.

If the size of dirt is larger than zero, YES is obtained in step S51 and step S5 is performed.
The size Y of the stain is 25% of the entire screen in the process 2
It is determined whether (first predetermined value) is exceeded.
If the stain size Y is less than 25% of the entire screen, YES is obtained in step S52, and the threshold value D is set to the following equation (1).

D = {(A−Y) / A} × B (1) where A is the number of pixels and B is the threshold value when there is no stain.

For example, the pixel A of the photographed image is 300 × 400 =
When the amount of dirt is 120000 (dot) and the size Y of dirt is 24000 (dot), the ratio of dirt is 20%. Further, in this case, when the threshold value D is set to B = 0.5 and is applied to the above equation (1), D = 0.4 (40%).

On the other hand, the stain size Y is 25 on the entire screen.
If it is more than%, NO is obtained in step S52,
In the process of step S54, it is determined whether or not the stain size Y is less than 50% (second predetermined value) of the entire screen. If it is less than 50%, step S5
If YES in step 4, the size of dirt Y
Is set to 25%, the threshold value (= C) obtained by the above equation (1) is set.

That is, the value C of the threshold value D is 0.38 (fixed value) according to the following equation (2).

{(120000-30000) / 120000} × 0.5≈0.38 (2) That is, the threshold value D varies within the range of 0.38 to 0.5.

Further, the dirt size Y is 50 on the entire screen.
If it exceeds%, NO is determined in the step S54, and FIG.
The process proceeds to step S9 shown in. That is, when more than half of the screen is covered with dirt, it is determined that normal white line recognition cannot be performed, and the vehicle deviation warning function is turned off.

In this way, in the lane departure warning apparatus 10 according to the present embodiment, the size of the dirt existing on the lens of the camera 1 or the windshield, that is, the size of the screen area covered by the dirt is determined. Therefore, since the threshold value for judging the partition line recognition state and the partition line non-recognition state is appropriately changed, even when the lens of the camera 1 or the windshield is contaminated, The partition line recognition state or the partition line non-recognition state can be accurately determined. As a result, the lane departure warning function can be switched on and off with higher accuracy.

When it is detected that the lens of the camera 1 or the windshield is dirty,
Since it is possible to inform the passengers of the vehicle as a display on the display 6 or as an audio signal from the speaker 7,
The occupant of the vehicle can quickly notice this and can take appropriate measures such as wiping off dirt.

In this embodiment, the partition line recognition device 1
Although the example in which 1 is mounted in the lane departure warning device 10 has been described, the present invention is not limited to this and can be applied to other devices that require white line recognition processing.

In this embodiment, the example in which the camera is installed inside the vehicle has been described. However, when the camera is installed outside the vehicle such as the front grill and side mirrors, the lens of the camera or the cover for protecting the lens is not stained. It is also possible to detect and perform the same processing.

The partition line recognition device and the lane departure warning device of the present invention have been described above based on the illustrated embodiment.
The present invention is not limited to this, and the configuration of each unit can be replaced with any configuration having the same function.

For example, in the above embodiment, the value B of the threshold value D when there is no stain is 0.5 (that is, 5).
However, the present invention is not limited to this, and it is possible to set other values.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a lane departure warning device including a partition line recognition device according to an embodiment of the present invention.

FIG. 2 is a flowchart showing an operation of a lane departure warning device including a partition line recognition device according to an embodiment of the present invention.

FIG. 3 is a flowchart showing a detailed processing procedure of threshold value calculation processing shown in FIG.

FIG. 4A is an explanatory diagram showing an example of an image taken by a camera, and FIG. 4B is a characteristic diagram showing a result when a luminance value is differentiated along a horizontal scanning line.

[Explanation of symbols]

1 camera (shooting means) 2 Rudder angle sensor 3 Left wheel speed sensor 4 Right wheel speed sensor 5 Arithmetic circuit 5a Dirt detection unit (dirt detection means) 6 display (notification means) 7 speaker (notification means) 8 white line (partition line) 9 scan lines 10 lane departure warning device 11 Partition line recognition device

─────────────────────────────────────────────────── ─── Continuation of front page (51) Int.Cl. ⁷ Identification code FI theme code (reference) G06T 7/00 G06T 7/00 T // G08G 1/16 G08G 1/16 FF term (reference) 5B057 AA16 BA11 CH01 DA02 DA07 DA15 DA16 5H180 AA01 CC04 CC24 LL01 LL02 LL07 LL08 LL20 5L096 BA04 CA02 DA03 EA45 FA31 GA51 JA25

Claims (5)

[Claims] 1. A partition line recognition is carried out by photographing a partition line of the vehicle traveling path by a photographing means mounted on a vehicle, and when the recognition rate of the photographed partition line is larger than a preset threshold value. A partition line recognizing device that determines that the partition line is not recognized when the state is smaller than a threshold value, and has a dirt detecting unit that detects dirt on the image capturing unit or the windshield of the vehicle. A partition line recognizing device, wherein the threshold value is changed when dirt is detected by the detecting means. 2. The photographing area by the photographing means,
The partition line recognition device according to claim 1, wherein the threshold value is set according to a ratio of an area of dirt detected by the dirt detection unit. 3. A photographing area by the photographing means,
The ratio of the area of dirt detected by the dirt detection unit is
If it is equal to or more than the first predetermined value, the threshold value is fixed, and if it is less than the first predetermined value, the threshold value is changed according to the ratio of the area of the dirt. The partition line recognition device according to claim 2, wherein 4. The photographing area by the photographing means,
The ratio of the area of dirt detected by the dirt detection unit is
The partition line recognition apparatus according to claim 3, wherein when the partition line recognition state is equal to or larger than a second predetermined value which is larger than the first predetermined value, it is determined that the partition line is not recognized. 5. The notification means for notifying an occupant of a vehicle when dirt is detected by the dirt detection means, according to any one of claims 1 to 4. The partition line recognition device described in.