JP2014123301A - White line detection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a white line detection device capable of more effectively detecting a white line.SOLUTION: A white line detection device 1 comprises: division means 3a that divides an image captured from a vehicle into a plurality of areas; determination means 3b that determines a degree of brightness in each area; and determining means 3c that determines an edge threshold for each area so that the edge threshold is made greater as the degree of brightness is higher.

Description

  The present invention relates to a white line detection device that is suitable for steering assistance and driving assistance when applied to vehicles such as passenger cars, trucks, and buses.

  As an apparatus for detecting a white line on a road from an image obtained by imaging the front of a vehicle, for example, there is one described in Patent Document 1. In Patent Document 1, it is proposed that the threshold value is set based on the luminance of the road surface area excluding the white line in the comparison between the binarized data acquired from the image and the threshold value, thereby improving the detection accuracy of the white line. Yes.

JP 2005-157670 A

  However, in such a device, the brightness difference between the vicinity where the headlight hits at night and the distance where the headlight does not hit and the brightness difference between the vicinity inside the tunnel and the distance outside the tunnel are considered. A difference occurs in the luminance difference between the edges of the white lines that are greatly different. For this reason, when a luminance difference is taken into consideration, there arises a problem that a white line cannot always be detected effectively.

  An object of this invention is to provide the white line detection apparatus which can detect a white line more effectively in view of the said problem.

  In order to solve the above problem, a white line detection device according to the present invention includes a dividing unit that divides an image captured from a vehicle into a plurality of regions, a determination unit that determines the brightness in each of the regions, And determining means for determining an edge threshold value for each of the regions so that the edge threshold value is increased as the brightness becomes brighter. Here, the dividing unit may divide the image at least in the front-rear direction of the vehicle, and the determination unit may determine the brightness using an average luminance in the region. The dividing unit may divide the image into left and right with reference to the vanishing point of the white line in the image, and divide the image in the front-rear direction so that the distance between the vanishing point and the lower edge of the image is equal.

  According to the present invention, it is possible to detect a white line more efficiently regardless of the difference in brightness for each region.

It is a schematic block diagram which shows one Embodiment of the white line detection apparatus 1 of Example 1 which concerns on this invention. It is a schematic diagram which shows the division | segmentation aspect of the image in one Embodiment of the white line detection apparatus 1 of Example 1. FIG. It is a flowchart which shows the control content of the white line detection apparatus 1 of Example 1 including a search and white line parameter calculation. FIG. 6 is a schematic diagram illustrating a horizontal distribution pattern of luminance of a road surface and a white line in an image that is a basis for determining an edge threshold value in the white line detection device 1 of the first embodiment. It is a schematic diagram which shows the aspect of the straight line approximation in a white line detection apparatus 1 of Example 1, a distant search, and a neighborhood search. It is a schematic diagram which shows the range expansion of the edge detection in the white line detection apparatus 1 of Example 1 based on a comparison with a prior art. It is a schematic diagram which shows the division | segmentation aspect of the image in one Embodiment of the white line detection apparatus 1 of Example 2. FIG. It is a flowchart which shows the control content of the white line detection apparatus 1 of Example 2. FIG.

  Hereinafter, embodiments for carrying out the present invention will be described with reference to the accompanying drawings.

  As shown in FIG. 1, a white line detection device 1 according to the first embodiment includes a front camera 2 that is provided in a vehicle and images a road surface in front of the vehicle, and an image ECU 3 (Electronic that processes an image captured by the front camera 2). Control Unit).

  The image ECU 3 is composed of, for example, a CPU, a ROM, a RAM, a data bus that connects them to each other, and an input / output interface. According to a program stored in the ROM, the image ECU 3 performs each control described below, thereby dividing the unit 3a, It functions as determination means 3b, determination means 3c, edge detection means 3d, integration means 3e, search means 3f, and calculation means 3g.

  The dividing unit 3a divides an image captured from the vehicle by the front camera 2 into a plurality of regions. More specifically, as shown in FIG. 2, the dividing means 3a divides the white line vanishing point LP in the image into the left and right, the left side is the region Left, the right side is the region Right, and the white line vanishing point LP and the lower edge of the image Are divided in the front-rear direction so that the distances are equal, and the far side is the region Far and the near side is the region Near.

  In the first embodiment, the white line vanishing point LP is fixed without being updated every processing cycle of the image ECU 3. Accordingly, the actual distance between the area Near and the area Far is also fixed. For example, when the actual distance of the white line vanishing point is 100 m, the area Near is 0 to 50 m, and the area Far is 50 to 100 m. A general projective transformation is used for the conversion from the actual distance to the position on the image. The horizontal position of the white line vanishing point LP is always fixed so as to be the center (center) of the image width. As a result, the width of the region Left and the region Right on the image is always equal. It should be noted that correction based on the presence or absence of a road gradient or a curve is performed as appropriate.

  The determination unit 3b determines the brightness in each of the front, rear, left, and right regions in the image using the average brightness in the region, and the determination unit 3c detects a white line as the average brightness is higher (the brightness is brighter). The edge threshold value for each region is determined so as to increase the edge threshold value. As shown in FIG. 3, the luminance distribution in the X direction in the image is in the form of a rectangular wave in which the white line is offset relative to the road surface and has one or two pairs of rising and falling edges (two pairs in FIG. 3). The entire rectangular wave moves up and down in the Y-axis direction according to the above, but the determining means 3c determines the edge threshold within a range in which the white line and the road surface can be distinguished according to the vertical movement.

  Next, the entire control content of the white line detection device 1 of the first embodiment will be described with reference to the flowchart shown in FIG. As shown in step S1, the image ECU 3 acquires an image captured by the front camera 2, and the dividing unit 3a divides the image into four parts on the lower edge side from the white line vanishing point LP by the method described above in step S2. In step S3, the determination unit 3b determines the brightness for each region using the average brightness, and in step S4, the edge threshold is increased as the brightness, that is, the average brightness for each region determined in step S3 is brighter. The edge threshold value is determined for each region so that the darker the edge threshold value is, the darker the edge threshold value is.

  In step S5 to step S11, using the edge threshold value determined in step S4, the edge detection unit 3d scans the image in the horizontal direction for each divided region, and detects rising and falling edges exceeding the edge threshold value. To detect. In step S6, the integration unit 3e performs a process of integrating these and projecting them onto the road surface plane based on the three-dimensional coordinates of the edges for each region detected in step S5.

  In step S7, the search means 3f extracts edge line segments from the edge group in the range set as the linear approximation region on the road surface plane projected in step S6, and then the white line position of the previous frame of the previous process or the like. The white line edge line segment is selected from the width of the rising and falling edge line segments.

  In step S8, the search means 3f extracts the white line edge of the neighboring area LN before the straight line approximation area L in FIG. 5 existing within a certain range from the white line position of the previous frame of the previous process.

  In step S9, the search means 3f applies the clothoid model separately for the left and right white lines to the white line edge group detected in steps S7 and S8. In step S10, the search means 3f detects a white line edge in the far region LF on the far side of the linear approximation region L in FIG. 5 obtained by extending the clothoid estimated in step S9 far.

  In step S11, the calculation unit 3g applies a clothoid model to the white line edge group detected in steps S7, S8, and S10 separately to extract a white line.

  According to the white line detection device 1 of the first embodiment executed by the above control, the following operational effects can be obtained. That is, as shown in FIG. 6 (a), it is difficult to detect the edge indicated by the broken line for the white line with low contrast in the distance shown in the ellipse LC in FIG. 6 (a). As shown in FIG. 6B, detection can be performed in the first embodiment.

  Further, in the first embodiment, since the image is divided based on the white line vanishing point LP fixed in advance, it is not necessary to detect the white line in advance for the division. For this reason, it is possible to reduce the processing load in the division, and to make it easier to divide the distant and neighboring areas with low contrast.

  In the above-described first embodiment, the form in which the white line vanishing point LP is fixed is shown. However, the white line vanishing point LP can be updated every processing cycle of the image ECU 3. The second embodiment will be described below.

  The white line detection device 1 of the second embodiment is the same as that shown in FIG. 1 of the first embodiment described above, and the processing content is the same as that shown in FIG. 4 of the first embodiment up to step S11. The difference will be mainly described.

  As shown at the end of the flowchart shown in FIG. 7, in the white line detection device 1 of the second embodiment, the image ECU 3 calculates the white line vanishing point LP from the result of the previous process. When white lines are detected on both the left and right sides in the image, the intersection of the white line parameters on both sides is set as the white line vanishing point LP.

  As shown in FIG. 8, the left side is the region Left and the right side is the region Right based on the obtained white line vanishing point LP, and the region in front of the white line vanishing point LP is divided into two so that the actual distance is equal, A region Near is set, and a far side is set as a region Far. As a result, the width of the region Left and the region Right on the image is variable.

  Also by the white line detection device 1 of the second embodiment that is executed by the above control, the following operational effects can be obtained as in the first embodiment. That is, as shown in FIG. 6 (a), the edge of the white line shown in the ellipse LC shown in FIG. As shown in FIG. 4, it can be detected in the second embodiment.

  According to the white line detection device 1 of the second embodiment, since the edge of the white line in the region with low contrast in the distance can be detected more easily, the white line is particularly applied to a vehicle with automatic driving in view. The white line detection range required for the detection apparatus 1 is required to include a farther region, and this request can be reliably met.

  In the second embodiment, since the image is divided based on the white line vanishing point LP calculated every calculation cycle, the accuracy of the division can be further improved. For this reason, the accuracy of edge detection, straight line approximation area search, neighborhood area search, white line model fitting, far area search, white line parameter calculation after division is improved, especially in the far area search, accuracy, fitting performance, and robustness are improved. Can do.

  Although the preferred embodiments of the present invention have been described in detail above, the present invention is not limited to the above-described embodiments, and various modifications and substitutions are made to the above-described embodiments without departing from the scope of the present invention. be able to.

  For example, in the above-described embodiment, the number of divisions in the front-rear direction is divided into two, but the number of divisions can be increased as necessary. Similarly, the number of divisions in the left-right direction is not limited to two divisions, and can be divided into multiple divisions with a larger number.

  In the embodiment described above, the average luminance is used as a parameter for determining the brightness (exposure) in the image, but other parameters can be used as a matter of course.

  The white line detection device of the present invention can detect a white line more effectively, and in particular, can detect a white line with a low contrast in the distance, so that automatic driving, more predictive steering assistance (LKA, LDW) and collision prevention (PCS) can be performed. It is useful when applied to various vehicles such as passenger cars, trucks, and buses.

1 White line detection device 2 Front camera 3 Image ECU
3a Division means 3b Determination means 3c Determination means 3d Edge detection means 3e Integration means 3f Search means 3g Calculation means

Claims (3)

  A dividing unit that divides an image captured from a vehicle into a plurality of regions, a determination unit that determines the brightness in each of the regions, and an edge for each region so that the edge threshold is increased as the brightness is brighter A white line detecting apparatus comprising a determining means for determining a threshold value.   The white line detection device according to claim 1, wherein the dividing unit divides the image in at least a front-rear direction of the vehicle.   The white line detection apparatus according to claim 2, wherein the determination unit determines the brightness using an average luminance in the region.

Images