JPH05347000A - Travelling lane detecting device - Google Patents

Abstract

PURPOSE:To detect a travel lane fast at a high speed and at a high detection ratio by dynamically varying and setting the position and range of a linear approximation calculation processing area. CONSTITUTION:A video camera 1 picks up an image of the road in front of a vehicle to obtain a road image and a threshold value setting means 2 sets a threshold value used for white line extraction. A white line extracting means 3 processes the road image obtained from the video camera 1 by using the threshold value set by the threshold setting means 2 to extract a white line drawn on the road surface. A straight line approximating means 4 approximates the white line extracted by the white line extracting means 3 with a straight line and a straight line approximation calculation processing area setting means 5 sets the calculation processing area for the straight line approximation by using a history of past travel lane detection results. A travel lane detecting means 6 detects the traveling lane by using two white lines approximated by the straight line approximating means 4 with straight lines.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a traveling lane detecting device used for detecting a traveling lane in which a vehicle is traveling.

[0002]

2. Description of the Related Art Conventionally, as a traveling lane detecting device of this type, as described in Japanese Patent Laid-Open No. 63-142478, an edge of an image captured from an image pickup means is detected and surrounded by the edge. The lane is detected by detecting a white line candidate area drawn on the road surface based on the size of the area and approximating the white line to a straight line using a straight line detecting means such as HOUGH conversion on the skeleton of the white line candidate area. It was

[0003]

In such a conventional traveling lane detection apparatus, since the entire image is processed, not only a sufficient processing speed cannot be obtained, but also the possibility of erroneous detection due to the influence of background noise. It also had the problem of being expensive.

The present invention solves the above problems and reduces and detects the amount of calculation by dynamically changing and setting the position and range of the calculation processing area of the HOUGH conversion which is a straight line detection algorithm. A first object of the present invention is to provide a traveling lane detection device which realizes reduction of influence from pixels other than the target white line and has a high speed and a high detection rate.

Also, regarding the area for white line extraction,
A second object of the present invention is to provide a traveling lane detection device having a more excellent white line extraction capability by dynamically changing and setting the position and range as in the calculation processing region of HOUGH conversion.

A third object of the present invention is to provide a traveling lane detection device capable of more stable white line extraction by dynamically setting the processing area for calculating the threshold used for white line extraction. ..

Further, the calculation of the threshold value used for white line extraction is
A fourth object of the present invention is to provide a traveling lane detection device capable of surely extracting a white line even when the contrast of an image changes by performing a combination of the maximum value and the average value of the brightness in the threshold setting processing area. ..

Further, there is provided a traveling lane detection device which gives a warning to the driver about the lane departure by calculating the lane departure degree of the vehicle using the linear approximation result of the detected traveling lane to prevent an accident. This is the fifth purpose.

[0009]

In order to achieve the first object of the present invention, a video camera mounted on a vehicle for picking up an image of a road ahead of the vehicle, and a road image taken from the video camera, is displayed on a road surface. Threshold setting means for setting a threshold used for extracting the drawn white line, and extracting the white line drawn on the road surface from the road image captured from the video camera using the threshold set by the threshold setting means. White line extracting means, straight line approximating means for approximating the white line extracted by the white line extracting means with a straight line, and straight line approximating calculation for setting a calculation processing area in the straight line approximating means by using a traveling lane detection result up to one screen before. The processing area setting means and the traveling lane detecting means for detecting the traveling lane by the white line linearly approximated by the linear approximating means are provided. And a first means for solving problems that dynamically changing the set position and extent of the processing region.

In order to achieve the second object, in the traveling lane detecting device of the first problem solving means, further, the position and range of the processing area of the white line extracting means are detected up to one screen before the traveling lane. The second problem solving means is to dynamically change and set the white line extraction area setting means using the result.

Further, in order to achieve the third object, in the first problem solving means or the second problem solving means, a processing area for calculating a threshold value used for extracting a white line is displayed one screen before. The third problem solving means is to set the threshold calculation area setting means using the traveling lane detection result and dynamically change the position and range thereof.

Further, in order to achieve the fourth object, calculation of a threshold value used for extracting a white line in the first problem solving means, the second problem solving means, or the third problem solving means. Fourth combination of the maximum brightness value in the threshold value calculation area detected by the maximum brightness value detection means and the average brightness value in the threshold value calculation area calculated by the average brightness value calculation means is performed. As a means.

Furthermore, in order to achieve the fifth object, the traveling lane of the first problem solving means, the second problem solving means, the third problem solving means, or the fourth problem solving means. Lane departure degree calculating means for calculating the running lane departure degree of the vehicle from the linear approximation result of the running lane detected by the detection device, and giving a warning to the driver according to the lane departure degree calculated by the lane departure degree calculating means A fifth problem solving means is to provide a warning means and dynamically monitor the running state.

[0014]

According to the first means for solving the above problems, the present invention provides:
By dynamically changing and setting the position and range of the calculation processing area of the HOUGH conversion which is the straight line detection processing, it is possible to detect a traveling lane with high speed and a high detection rate.

Further, by the second problem solving means, it is possible to detect the traveling lane excellent in the white line extraction ability by dynamically setting the white line extraction area.

According to the third problem solving means, stable white line extraction can be performed by dynamically setting the threshold value calculation area used for white line extraction, and the detection rate of the traveling lane can be improved.

Further, by the fourth problem solving means, the threshold value used for white line extraction is calculated by combining the maximum value and the average value of the brightness in the threshold setting processing area, so that even when the contrast of the image changes, it is possible to be sure. It is possible to accurately detect the driving lane by extracting the white line in the line.

Still further, according to the fifth means for solving the problems,
By giving a warning to the driver according to the degree of deviation from the driving lane calculated from the straight line approximation result of the white line, the risk of accident can be significantly reduced.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the first problem solving means of the present invention will be described below with reference to FIG.

As shown in FIG. 1, the video camera 1 captures a road ahead of the vehicle to obtain a road image, and the threshold setting means 2 sets a threshold used for white line extraction. The white line extraction means 3 performs image processing on the road image obtained from the video camera 1 using the threshold value set by the threshold value setting means 2, and extracts the white line drawn on the road surface. The straight line approximation means 4 approximates the white line extracted by the white line extraction means 3 with a straight line, and the straight line approximation calculation processing area setting means 5 sets a calculation processing area for the straight line approximation using the history of past traveling lane detection results. To do. The traveling lane detecting means 6 detects the traveling lane by using the two white lines linearly approximated by the linear approximating means 4.

The operation of the above configuration will be described with reference to FIG. 2. First, when the apparatus is started, step 2
In 1, a road image ahead of the vehicle is captured from the video camera 1. FIG. 3 shows a road image captured from the video camera 1.

Next, in step 22, the threshold setting means 2
Threshold setting processing is performed. The threshold value can be set by detecting the maximum value of the brightness from the threshold value setting area set in the image and considering the brightness distribution of the white line on the basis of the detected maximum value.

Next, in step 23, the white line extracting means 3
White line extraction processing is performed. The white line is extracted by searching its contour. As shown in FIG. 4, pixels that exceed the threshold set in step 22 are searched from the center of the image toward the outside of the left and right, and the pixels that exceed the threshold first are extracted as white line outlines. White line contours are extracted independently. Note that the white line extraction is SOB
It is also possible to extract an edge from the image using an edge extraction filter represented by an EL filter.

Next, the process proceeds to step 24, where the linear approximation calculation processing region setting means 5 sets the linear approximation calculation processing region. An algorithm called HOUGH conversion is used for linear approximation, and this HOUGH conversion will be briefly described with reference to FIG. What is HOUGH conversion?
ρ = x · cos θ + y · sin θ (1) XY
It transforms a point on the plane into a curve on the ρθ plane,
As shown in FIG. 5, the curves obtained by converting the points on the same straight line intersect at one point on the ρθ plane.
Therefore, a straight line on the XY plane can be detected by converting the point sequence on the XY plane and then extracting the point on the ρθ plane where the most curves intersect.

This HOU is used to set the linear approximation calculation processing area.
This is performed on the ρθ plane of GH conversion, and first, a feature point on the ρθ plane that represents the straight line extracted on the previous screen is set as the center point of the region. Then take the history of the variation of the past several screens of ρ and theta, taking out the amount of movement of the greatest fluctuating ρ direction therein mρ _max, the amount of movement theta direction m.theta _max, using these maximum travel The size of the area to be set. Here, if three times the maximum movement amount of each is set as the area size, the calculation processing area can be set as shown in FIG. The linear approximation calculation processing areas are set independently for the left and right white lines.

Next, in step 25, the linear approximation means 4
The linear approximation processing is performed. This process is H
OUGH conversion is performed independently for the left and right white line contours obtained in step 23, and calculation is performed only in the calculation processing area for each of the left and right white lines set in step 24. As a result, white lines existing on the left and right of the vehicle can be obtained as straight lines. FIG. 7 shows a white line linearly approximated.

Next, in step 26, the traveling lane detection processing by the traveling lane detection means 6 is performed. The running lane is detected by forming a triangle by two straight lines approximating the white line and the bottom of the image in step 25, and this triangle represents an area indicating the lane currently running.
FIG. 8 shows the driving lane detection results expressed by triangles.

Next, the routine proceeds to step 27, where the history of data relating to the current traveling lane detection result is updated.

As described above, according to the present embodiment, since the position and range of the area of the straight-line approximation calculation processing in the driving lane detection are dynamically set, the calculation amount is reduced and the influence other than the white line to be detected is influenced. Can be removed at the same time, and the performance of detecting the driving lane can be improved.

Next, an embodiment of the second means for solving the problems of the present invention, that is, an embodiment of a traveling lane detecting device for dynamically setting the position and range of the white line extraction area will be described with reference to FIG. To do.

As shown in FIG. 9, a white lane extraction area setting means 7 for dynamically setting the white lane extraction area using the history of the lane detection results of the previous screen is added to the traveling lane detection device described in the above embodiment. In addition, by improving the white line extraction rate in the white line extraction processing, the detection rate of the traveling lane is also improved.

The operation of the above structure will be described with reference to FIG. 10. First, when the apparatus is started, the road image in front of the vehicle is captured from the video camera 1 in step 21 as in the above embodiment, and step 22 is executed. The threshold value setting means sets the threshold value used for white line contour extraction.

Next, in step 101, the white line extraction area setting means 7 sets the white line extraction area. First, regarding the position, two straight lines, which are the white line approximation results on the previous screen, are set as the center lines of the left and right regions. Next, the maximum horizontal movement amount ml _max of the white line approximation straight line is detected for each of the left and right from the history of straight line approximation results of the past several screens, and the width of each region is determined using this. Here, if a constant value is used for the height of the area and the width of the area is set to 3 times each ml _max , the white line extraction area becomes as shown in FIG.

Next, in step 23, the white line extraction processing is performed by the white line extraction means as in the above embodiment. Here, what is different from the above-described embodiment is that the point where the search for the outline of the white line is started is not the center of the image but the boundary inside the image of each white line extraction area as shown in FIG.

After the white line is extracted, the processing from step 24 to step 27 is sequentially performed to detect the traveling lane, as in the above embodiment.

As described above, according to the present embodiment, the position and the range of the white line extraction area are dynamically changed and set, so that it is possible to detect the running lane excellent in the white line extraction ability.

Next, an embodiment of a third means for solving the problems of the present invention, that is, a running lane detection apparatus for dynamically changing the position and range of a region for calculating a threshold value used for white line extraction, will be described. An example will be described with reference to FIG.

As shown in the figure, in the driving lane detection apparatus described in the above embodiment, the threshold calculation area for dynamically setting the calculation area of the threshold used for white line extraction using the history of the travel lane detection results on the previous screen. Setting means 8 is added to more appropriately set the threshold value used in the white line extraction process, and the white line extraction rate is improved to improve the detection rate of the traveling lane.

The operation of the above structure will be described with reference to FIG. 14. First, when the apparatus is started, a road image ahead of the vehicle is captured from the video camera 1 in step 21 as in the above embodiment.

Next, in step 141, the threshold value calculation area setting means 8 sets the threshold value calculation area. First, as to the position, as in the case of setting the white line extraction region in the above embodiment, two straight lines, which are the white line approximation results on the previous screen, are set as the center lines of the left and right regions. As for the width of the area, similarly to the above embodiment, the maximum horizontal movement amount ml _max of the straight line approximation line is detected for each of the left and right from the history of the linear approximation results of the past several screens, and this is used for each area. Determine the width. However, since continuity is not particularly required for the threshold value calculation area, it is possible to select a specific scanning line and set a range satisfying the above condition on each selected scanning line as the threshold value calculation area.
FIG. 15 shows a diagram in which the threshold value calculation area is set with the width of the area on nine scanning lines being three times the ml _max . The number of scanning lines to be selected and the number of scanning lines to be selected may be fixed, but this can also be dynamically changed.

Next, in step 22, the threshold setting means 2
Although the threshold value setting process is performed according to the above step, it is different from the above embodiment in that the threshold value is calculated using only the data in the area set in step 141.

After the threshold value is set, the processing from step 23 to step 27 is sequentially performed to detect the traveling lane, as in the above embodiment.

As described above, according to the present embodiment, the setting is performed while dynamically changing the position and range of the threshold calculation area used for white line extraction, so that stable white line extraction can be performed and the detection rate of the traveling lane can be achieved. Can be improved.

Next, an embodiment of a fourth problem solving means of the present invention, that is, a threshold value used for extracting a white line is calculated by combining the maximum value and the average value of the brightness in the threshold value calculation area. An example of the lane detection device will be described with reference to FIG.

As shown in the figure, the driving lane detecting apparatus described in the above embodiment calculates the average of the maximum luminance value detecting means 9 for detecting the maximum luminance value in the threshold value calculating area and the average luminance value in the threshold value calculating area. The brightness average value calculation means 10 is added, and the threshold value setting means 2 sets the threshold value for white line extraction using these two values, and the white line extraction can be reliably performed even when the contrast of the image changes, and the running It reduces the decrease in lane detection rate.

The operation of the above structure will be described with reference to FIG. 17. First, when the apparatus is started, the road image in front of the vehicle is captured from the video camera 1 in step 21 as in the above embodiment, and then, In step 141, the threshold calculation area is set.

Next, in step 171, the maximum brightness value detection means 9 detects the maximum brightness value in the threshold value calculation area, and in step 172 the brightness average value calculation means 10 is detected.
The average value of the brightness in the threshold value calculation area is calculated. Since these processes do not interfere with each other,
Processing can proceed in parallel.

Next, in step 22, the threshold value setting process is performed, which is different from the above embodiment in steps 171 and 17.
The difference is that the threshold value is calculated using the two values of the maximum brightness value and the average brightness value obtained in 2. The threshold value is calculated by threshold value = A, maximum brightness value + B, average brightness value (2)
The brightness maximum value and the brightness average value are added at a fixed ratio, and the constants A and B are determined depending on the brightness of the camera used in the system. FIG. 18 shows how the threshold value is set when the threshold value is calculated with A = 0.75 and B = 0.25. As is clear from the figure, when the contrast is good, the threshold is set low, and when the contrast is poor, the threshold is set high, and a threshold suitable for distinguishing the white line from other pixels is obtained. I understand.

After the threshold value is set, the processing from step 23 to step 27 is sequentially performed in the same manner as in the above embodiment to detect the traveling lane.

As described above, according to the present embodiment, the threshold value used for white line extraction is calculated from the maximum luminance value and the average luminance value. Therefore, even when the contrast of the image changes, the white line can be appropriately extracted and the driving lane The detection rate of can be improved.

Next, an embodiment of the fifth means for solving the problems of the present invention, that is, by calculating the running lane departure degree of the vehicle from the linear approximation result of the detected running lane, the driver is given a warning regarding the lane departure. An embodiment of a traveling lane detection device for preventing an accident will be described with reference to FIG.

As shown in the figure, the lane departure degree calculating means 11 for calculating the running lane departure degree of the vehicle from the straight line approximation result of the white line in the running lane detection device described in the above embodiment.
And warning means 1 for giving a warning to the driver according to the lane departure degree calculated by the lane departure degree calculating means 11.
2 is added to dynamically monitor the running state.

The operation of the above structure will be described with reference to FIG. 20. First, when the apparatus is started, the processing from step 21 to step 25 is performed and the white line on the road is straight as in the above embodiment. Is approximated by.

Next, in step 201, the lane departure degree calculation means 11 calculates the lane departure degree. The lane departure degree was calculated using the HOUGH used for the straight line approximation of the white line.
The conversion can be performed using the value of θ of the feature point on the ρθ plane. In the straight line approximating process of the white line, two feature points on the left and right are extracted, and one having a value of θ close to 90 degrees is adopted. This takes advantage of the fact that the straight line approximation result has a white line that is closer to vertical in the image at a position closer to the vehicle, and depending on which of the left and right feature points was selected, the white line You can judge whether you are approaching at the same time. Here, it is possible to regard the value of θ as the lane departure degree as it is, but in the present embodiment, a region regarding θ is set as shown in FIG. 21, and it is determined to which region the θ belongs. And

Next, in step 202, step 2
It is judged whether the value of θ of the feature point selected in 01 is in the warning region, and if a warning is necessary, the process proceeds to step 203, where the warning means 12 gives a warning to the driver, and the warning is not necessary. Returns to the first image capturing process. Here, the alarm is set for each of a plurality of areas as shown in FIG. 21, and information is provided to the driver by voice or display.

As described above, according to the present embodiment, the degree of deviation of the traveling lane of the vehicle is calculated from the straight line approximation result of the white line and the information is provided to the driver. Can be prevented.

[0057]

As is apparent from the above-described embodiments, according to the present invention, a video camera mounted on a vehicle to image a road ahead of the vehicle and a road image captured from the video camera are drawn on the road surface. Threshold setting means for setting a threshold used for extracting a white line, and white line extracting means for extracting a white line drawn on a road surface from a road image captured from the video camera by using the threshold set by the threshold setting means. A linear approximation means for approximating a white line extracted by the white line extraction means with a straight line; a linear approximation calculation processing area setting means for setting a calculation processing area in the linear approximation means; and a linear approximation by the linear approximation means. With a driving lane detecting means for detecting a driving lane by a white line, in order to dynamically change the set position and range of the linear approximation calculation processing area. Strategy by reducing influence from the white line other pixels is a calculation amount reduction detection target, it is possible to perform a high running lane detecting the speed and detection rate.

Further, by dynamically changing and setting the position and range of the processing area of the white line extracting means by the white line extracting area setting means, it is possible to detect the traveling lane excellent in the white line extracting ability.

Further, the processing area for calculating the threshold value used for extracting the white line is set by the threshold value calculation area setting means, and its position and range are dynamically changed, so that the white line extraction is stabilized by the appropriate threshold value setting. It is possible to realize the traveling lane detection which can be performed by doing so.

Further, by calculating the threshold value used for extracting the white line by combining the maximum value and the average value of the brightness in the threshold value calculation area, the white line can be surely extracted even when the contrast of the image changes. It is possible to realize possible lane detection.

Furthermore, a lane departure degree calculating means for calculating the running lane departure degree of the vehicle from the straight line approximation result of the white line, and an alarm for giving a warning to the driver according to the lane departure degree calculated by the lane departure degree calculating means. It is possible to realize a traveling lane detection device which is provided with means and dynamically monitors the traveling state to give an alarm to the driver about the lane departure and prevent an accident.

[Brief description of drawings]

FIG. 1 is a block diagram showing a basic configuration of a traveling lane detection device according to an embodiment of the present invention.

FIG. 2 is an operation flowchart of the traveling lane detection device.

FIG. 3 is a diagram showing a road image captured from a video camera of the traveling lane detection device.

FIG. 4 is a diagram illustrating white line contour extraction processing of the same-behind lane detection device.

FIG. 5 is a diagram illustrating a linear approximation process of the traveling lane detection device.

FIG. 6 is a diagram showing a region setting on a ρθ plane in a linear approximation process of the traveling lane detection device.

FIG. 7 is a diagram showing a linear approximation result of the traveling lane detection device.

FIG. 8 is a diagram showing a traveling lane detection result of the traveling lane detection device.

FIG. 9 is a block diagram showing a basic configuration of a traveling lane detection device according to an embodiment of the present invention.

FIG. 10 is an operation flowchart of the traveling lane detection device.

FIG. 11 is a diagram showing a white line extraction area set by the traveling lane detection device.

FIG. 12 is a diagram illustrating a white line extraction process of the traveling lane detection device.

FIG. 13 is a block diagram showing a basic configuration of a traveling lane detection device according to an embodiment of the present invention.

FIG. 14 is an operation flowchart of the destination lane detection device.

FIG. 15 is a diagram showing a threshold value setting processing region of the preceding lane detection device.

FIG. 16 is a block diagram showing a basic configuration of a traveling lane detection device according to an embodiment of the present invention.

FIG. 17 is an operation flowchart of the destination lane detection device.

FIG. 18 is a diagram for explaining the effect of threshold setting of the preceding lane detection device.

FIG. 19 is a block diagram showing a basic configuration of a traveling lane detection device according to an embodiment of the present invention.

FIG. 20 is an operation flowchart of the destination lane detection device.

FIG. 21 is a diagram showing setting of a warning judgment area of the preceding lane detection device.

[Explanation of symbols]

 1 Video Camera 2 Threshold Setting Means 3 White Line Extracting Means 4 Linear Approximation Means 5 Linear Approximation Calculation Processing Area Setting Means 6 Driving Lane Detection Means

Claims (5)

[Claims] 1. A video camera mounted on a vehicle for imaging a road ahead of the vehicle, and threshold setting means for setting a threshold used for extracting a white line drawn on a road surface from a road image captured by the video camera. A white line extracting means for extracting a white line drawn on a road surface from a road image captured from the video camera by using a threshold value set by the threshold setting means, and a white line extracted by the white line extracting means by a straight line Straight line approximation means, a straight line approximation calculation processing area setting means for setting the calculation processing area in the straight line approximation means by using the traveling lane detection results up to the previous screen, and a white line linearly approximated by the straight line approximation means. It consists of a driving lane detection means for detecting lanes, and is characterized by dynamically changing the set position and range of the linear approximation calculation processing area. Traveling lane detection apparatus according to. 2. The traveling lane detection device according to claim 1, further comprising: dynamically determining the position and range of the processing area of the white line extraction means by the white line extraction area setting means using the traveling lane detection result up to one screen before. A traveling lane detection device characterized by being changed and set. 3. The traveling lane detection device according to claim 1 or 2, wherein a processing area for calculating a threshold used for extracting a white line is set as a threshold calculation area by using a traveling lane detection result up to one screen before. A driving lane detection device characterized by being set by means and dynamically changing its position and range. 4. The driving lane detection device according to claim 1, wherein the threshold value used for extracting the white line is calculated by a maximum brightness value in a threshold value calculation area detected by the maximum brightness value detection means. And a brightness average value in the threshold value calculation area calculated by the brightness average value calculation means are combined to perform the driving lane detection device. 5. A lane departure degree calculating means for calculating a lane departure degree of a vehicle from a straight line approximation result of a white line in the running lane detection device according to claim 1.
A traveling lane detection device comprising: warning means for giving a warning to a driver in accordance with the lane departure degree calculated by the lane departure degree calculating means, and dynamically monitoring a running state.