JP2007264712A - Lane detector - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a lane detector for appropriately excluding wide signs of a pedestrian crossing etc. that are unnecessary for detecting lane positions, and for accurately detecting the lane positions by detecting necessary wide lanes and signs. <P>SOLUTION: A lane detector 1 comprises a photographic means 2 for photographing a front of an own vehicle to output a pair of images having a brightness value p1ij for each pixel; an image processing means 6 for calculating distance Lij in an actual space for each pixel of one image T; and a detection means 9 for detecting pixels corresponding to one edge part of a lane as lane candidate points, on the basis of the brightness value and the distance at the image T so as to detect lane positions LR, LL, on the basis of the detected candidate points. The detection means 9 calculates the actual distances between the candidate points and the edge parts on the opposite sides of the lanes, and uses the candidate points to detect the lane positions, only when the distances are smaller than the detected lane wide threshold Wthd. The threshold Wthd is variable, depending on the traveling condition of own vehicle or the road conditions. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a lane detection device, and more particularly to a lane detection device that detects a lane from a captured image in front of a vehicle.

  In recent years, roads that have been premised on improving safety and automatic control by applying image processing to images taken with in-vehicle stereo cameras and video cameras, etc., in order to improve driving safety of automobiles and automatic control of vehicles, etc. Development of a road recognition device for recognizing a shape is in progress. In order to accurately recognize the road shape accurately, it is necessary to reliably detect the positions of the lanes marked on the left and right of the host vehicle from an image obtained by capturing a landscape including the road ahead of the host vehicle. (For example, see Patent Documents 1 and 2).

  In these devices, in order to detect the left and right lane positions of the host vehicle, attention is paid to the luminance value and the luminance differential value of each pixel of the captured image, and the edge portion of the lane is detected from the change, and based on the edge portion. Often configured to detect lane position.

In the present invention, a continuous line or a broken line marked on a road surface such as a carriageway center line such as an overtaking prohibition line, a vehicle lane boundary line, a lane line dividing a roadside belt and a roadway is referred to as a lane, The width of itself is called the lane width.
JP 2001-92970 A JP 7-28975 A

  However, if the lane position is detected based only on the luminance value of each pixel of the captured image, as shown in FIG. 26, the zebra pattern paint portion of the pedestrian crossing is defined as the lane especially in the image in which the pedestrian crossing is imaged. Incorrect detection may result in erroneous detection of the left and right lane positions LR and LL of the host vehicle.

  In order to avoid such a situation, for example, a normal lane with a lane width standard of 10 to 15 cm and a zebra pattern paint on a pedestrian crossing with a width standard of 45 to 50 cm A method of distinguishing a wide sign such as a part by providing a threshold value can be considered. Thus, if the threshold value about the lane width is provided, the edge portion of the wide lane mark whose lane width corresponding to the detected edge portion is larger than the threshold value can be excluded from the detection target.

  However, when the threshold value is set so as to exclude the wide sign in this way, this time, as shown in FIG. 27, for example, the lane width in which three lanes of yellow and white are marked in parallel is about 45 cm wide. The lane TL representing the roadway center line cannot be detected, and the right lane position cannot be specified in the case of FIG. Further, it is impossible to detect a thick broken-line marking BL having the same width as that shown at a junction of an expressway as shown in FIG.

  Thus, simply setting a threshold value for the lane width makes it impossible to detect a wide lane or a thick sign to be detected, and the left and right lane positions of the host vehicle cannot be specified.

  The present invention has been made in view of such circumstances, and appropriately eliminates wide signs unnecessary for lane position detection such as pedestrian crossings, and detects necessary wide lanes and signs accurately. An object of the present invention is to provide a lane detection device capable of detecting a lane position.

In order to solve the above problem, the first invention provides:
In the lane detector,
Imaging means for imaging a landscape including a road ahead of the host vehicle and outputting a pair of images having a luminance value for each pixel;
Image processing means for calculating a distance in real space for each pixel of at least one image based on the pair of captured images;
Detecting means for detecting a pixel corresponding to one edge portion of the lane as a lane candidate point based on the luminance value and the distance for the one image, and detecting a lane position based on the detected lane candidate point; ,
The detection means calculates a distance in real space between the lane candidate point and an edge portion on the opposite side of the lane corresponding to the lane candidate point, and the detected lane candidate point is detected only when the distance is equal to or smaller than a detected lane width threshold. Used to detect the lane position,
In addition, the detected lane width threshold value is variable depending on a traveling environment or a road environment of the host vehicle.

  According to a second aspect of the present invention, in the lane detection device according to the first aspect of the present invention, when the detection means detects a sign wider than the standard of one lane width, the detected lane width threshold is set to the wide sign. Is expanded to a detectable value.

  According to a third aspect of the present invention, in the lane detection device according to the second aspect of the invention, the detection means is a predetermined distance ahead of the host vehicle or closer to the host vehicle than the predetermined distance in the one image for searching for lane candidate points. In the search area, the detected lane width threshold is set to a value corresponding to the standard of one lane width, and in the search area far from the predetermined distance ahead of the host vehicle, the detected lane width threshold is set to an enlarged value. It is characterized by setting.

  According to a fourth aspect of the present invention, in the lane detection device according to the third aspect, the detection unit shortens the predetermined distance according to a travel distance of the host vehicle.

A fifth aspect of the present invention is the lane detection device according to any one of the first to fourth aspects of the invention,
Highway running detection means for detecting whether or not the host vehicle is running on a highway,
The detecting means can detect a sign that is wider than the standard of one lane width for the detected lane width threshold when the highway running detecting means detects that the host vehicle is running on the highway. It is characterized by enlarging to various values.

A sixth invention is the lane detection device of the fifth invention,
Pedestrian crossing detection means for detecting whether there is a pedestrian crossing in front of the vehicle,
The detection means does not detect that the host vehicle is traveling on an expressway by the expressway running detection means, and if the pedestrian crossing is detected by the pedestrian crossing detection means, the detection lane width threshold is set. It is set to a value corresponding to the standard of the one lane width.

  According to a seventh aspect of the present invention, in the lane detection device according to the sixth aspect of the present invention, the detection means is not detected by the highway travel detection means that the host vehicle is traveling on a highway, but is detected by the pedestrian crossing detection means. When a pedestrian crossing is not detected, if it is highly possible that the host vehicle is traveling on an expressway based on the speed of the host vehicle or the distance between the left and right lane positions of the host vehicle, the detected lane width The threshold value is expanded to a value that allows detection of a wider sign than the specification of one lane width.

  According to an eighth aspect of the present invention, in the lane detection device according to any one of the first to seventh aspects, the detection means detects a sign that is wider than a standard of one lane width continuously for a certain distance or more. The detection lane width threshold value is expanded to a value capable of detecting the wide sign.

  According to a ninth aspect of the present invention, in the lane detection device according to the eighth aspect of the invention, the fixed distance is a distance that is equal to or greater than the length of the pedestrian crossing in the traveling direction of the host vehicle.

  According to the first invention, the traveling environment of the host vehicle such as the host vehicle traveling on the highway or having a high possibility of traveling, or the road environment such as a pedestrian crossing marked on the road By appropriately setting the detection lane width threshold to be enlarged or reduced according to the situation, unnecessary signs can be properly excluded, and necessary signs can be accurately detected to detect lane candidate points. It becomes.

  Therefore, it is possible to accurately detect the left and right lane positions of the host vehicle. In a normal standard driving state, a lane candidate point is set by narrowing the detection lane width threshold so as to detect a normal lane having a lane width standard of 10 to 15 cm. This makes it possible to prevent false detection.

  According to the second aspect of the present invention, when a lane or sign wider than the one lane width standard is detected, the detected lane width threshold value is expanded to appropriately detect such a wide lane or the like. And the effects of the present invention are more appropriately exhibited.

  According to the third aspect of the present invention, the detection lane width threshold is set narrow in the search area of the image within a predetermined distance when viewed from the host vehicle, and is set wide in the search area far from the predetermined distance. The detection lane width threshold value can be narrowed in a region where a lane exists, and the detection lane width threshold value can be increased in a region where a wide lane or the like exists. Therefore, it is possible to prevent erroneous detection of lane candidate points in an area where one normal lane exists, and to appropriately detect such a wide lane in an area where a wide lane exists, etc. The effect of each invention is exhibited more effectively.

  According to the fourth aspect of the invention, by reducing the initially set predetermined distance in accordance with the travel distance of the host vehicle, different detection lane width thresholds are set in the two search areas of the image as in the third aspect of the invention. It is possible to set appropriately according to the traveling of the host vehicle, and it is possible to exert the effects of the respective inventions more accurately.

  According to the fifth invention, since there is no pedestrian crossing on the highway, when the highway travel detection means detects that the vehicle is traveling on the highway, the detected lane width threshold is increased to increase the speed It is possible to accurately detect a thick broken-line marking that is marked at a junction or the like on the road. Therefore, it is possible to more accurately detect the left and right lane positions of the host vehicle, and the effects of the above-described inventions are more accurately exhibited.

  According to the sixth aspect of the present invention, when a pedestrian crossing is detected by the pedestrian crossing detection means while traveling on a general road where the vehicle is not detected to be traveling on the highway by the highway travel detection means, the detected lane is detected. By setting the width threshold value narrow, it is possible to prevent the pedestrian crossing from being erroneously detected as a lane and erroneously detecting the left and right lane positions of the host vehicle as in the conventional example. Therefore, the effects of the above inventions can be more accurately exhibited.

  According to the seventh aspect of the invention, the highway running detection means does not detect that the host vehicle is traveling on the highway, but the pedestrian crossing detection means does not detect the pedestrian crossing, and the own vehicle is, for example, at high speed. If you are traveling, you may be traveling on a highway. In such a case, by enlarging the detection lane width threshold, it becomes possible to accurately detect a thick broken-line marking that is marked at a junction of an expressway, etc., and the effects of each of the above-described inventions can be exhibited more accurately. Is done.

  According to the eighth invention, when wide lanes or markings are continuously detected for a certain distance or more, the detected lane width threshold value is expanded so that three lanes such as yellow and white are marked in parallel. Thus, the roadway center line can be reliably detected, and the effects of the above-described inventions can be exhibited more accurately.

  According to the ninth aspect, by making the constant distance in the eighth aspect more than the general length of the pedestrian crossing, it is possible to reliably detect the zebra pattern paint portion of the pedestrian crossing. It becomes possible to prevent, and the effect of said 8th invention is exhibited more effectively.

  Embodiments of a lane detection device according to the present invention will be described below with reference to the drawings.

  As shown in FIG. 1, the lane position detection apparatus 1 according to the present embodiment mainly includes an imaging unit 2, a conversion unit 3, an image processing unit 6, and a detection unit 9.

  The imaging means 2 is for imaging the periphery of the vehicle, and is configured to capture a landscape including a road ahead of the vehicle at a predetermined sampling period and output a pair of images. In the present embodiment, a stereo camera including a pair of main camera 2a and sub-camera 2b each incorporating a synchronized image sensor such as a CCD or CMOS sensor is used. In the present embodiment, CCD cameras are used for the main camera 2a and the sub camera 2b.

  The main camera 2a and the sub camera 2b are attached, for example, in the vicinity of a rearview mirror with a predetermined interval in the vehicle width direction. Of the pair of stereo cameras, a camera closer to the driver, as will be described later, a main camera 2a that captures an image serving as a basis for detecting distances and detecting lanes for each pixel, and a camera farther from the driver Is a sub-camera 2b that captures an image to be compared in order to obtain the distance or the like.

  A / D converters 3a and 3b as conversion means 3 are connected to the main camera 2a and the sub camera 2b, respectively. In the A / D converters 3a and 3b, each of the pair of analog images output from the main camera 2a and the sub camera 2b has a luminance value of a predetermined luminance gradation such as a gray scale of 256 gradations for each pixel. It is configured to be converted into an image. The digital image converted from the image on which the distance is calculated for each pixel and the lane is detected is used as the reference image, and the digital image converted from the image to be compared for obtaining the distance is compared. It is output as an image.

  An image correction unit 4 is connected to the A / D converters 3a and 3b. In the image correction unit 4, the luminance includes a deviation caused by an error in the attachment positions of the main camera 2 a and the sub camera 2 b and noise removal with respect to the reference image and the comparison image output from the A / D converters 3 a and 3 b. Image correction such as value correction is performed using affine transformation or the like.

  For example, as shown in FIG. 2, the reference image T is image data composed of luminance values of 512 pixels in the horizontal direction and 200 pixels in the vertical direction. The image correction unit 4 is configured to output image data composed of luminance values for 200 pixels in the vertical direction. Each image data is stored in an image data memory 5 connected to the image correction unit 4 and is simultaneously transmitted to the detection means 9.

  As described above, in this embodiment, a CCD camera is used for the main camera 2a and the like. The main camera 2a and the sub camera 2b scan and capture the landscape in front of the host vehicle in a line in the horizontal direction sequentially from the lower side, and sequentially transmit the data captured for each horizontal line to the conversion means 3. Then, after the imaging data is converted into a luminance value of a predetermined luminance gradation for each pixel of the horizontal line transmitted by the conversion means 3 and image correction is performed by the image correction unit 4, Image data is stored in the image data memory 5 and simultaneously transmitted to the detection means 9 at the same time.

  An image processing unit 6 is connected to the image correction unit 4, and the image processing unit 6 mainly includes an image processor 7 and a distance data memory 8.

  In the image processor 7, each setting area composed of each pixel or a block composed of a plurality of pixels of the reference image T based on the digital data of the reference image T and the comparison image output from the image correction unit 4 by the stereo matching process and the filtering process. The parallax dp for calculating the distance in the real space is calculated. The calculation of the parallax dp is described in detail in Japanese Patent Application Laid-Open No. 5-1114099 previously filed by the applicant of the present application.

  The image processor 7 calculates one parallax dp for each pixel block of 4 × 4 pixels for the reference image T having 512 × 200 pixels. As described above, the luminance value p1ij of 0 to 255 is assigned to each of the 16 pixels constituting one pixel block, and the luminance value p1ij of the 16 pixels forms a luminance value characteristic unique to the pixel block. Yes.

  The subscripts i and j of the luminance value p1ij are the i coordinates of the pixel at the lower left corner of the pixel block when the lower left corner of the image plane of the reference image T is the origin, the horizontal direction is the i coordinate axis, and the vertical direction is the j coordinate axis. And the j coordinate. For the comparison image, the i-coordinate and the j-coordinate are similarly taken with the pixel previously associated with the origin of the reference image T as the origin.

In the stereo matching process in the image processor 7, the reference image T is divided into a maximum of 128 × 50 pixel blocks every 4 × 4 pixels as described above, and the comparison image is a horizontal line having a width of 4 pixels extending in the horizontal direction. Divide into Then, one pixel block of the reference image T is taken out and the city block distance CB obtained by the following equation (1) is minimized while shifting the horizontal line of the corresponding comparison image one pixel at a time in the horizontal direction, that is, in the i direction. The pixel block on the horizontal line to be, that is, the pixel block on the comparative image having the luminance value characteristic similar to the pixel block of the reference image T is searched.
CB = Σ | p1ij−p2ij | (1)

  P2ij represents the luminance value of the pixel at the coordinates (i, j) on the comparison image. In the present embodiment, as described above, the image data for each horizontal line is output from the image correction unit 4. Therefore, the image processor 7 receives the 4 × horizontal image data every time the image data for four horizontal lines is input. Stereo matching processing is performed on a pixel block of 4 pixels.

  The image processor 7 calculates the amount of deviation between the pixel block on the comparison image specified in this way and the pixel block on the reference image T, and uses the amount of deviation as the parallax dp to the pixel block on the reference image T. It is designed to be assigned.

  The parallax dp is a relative displacement amount in the horizontal direction with respect to the mapping position of the same object in the reference image T and the comparison image derived from the main camera 2a and the sub camera 2b separated by a certain distance. The distance from the center position of the camera 2b to the object and the parallax dp can be associated based on the principle of triangulation.

Specifically, in real space, a point on the road surface directly below the center of the main camera 2a and the sub camera 2b is set as the origin, and the vehicle width direction of the own vehicle, that is, the X axis in the left-right direction, the Y axis in the vehicle height direction, When the Z axis is taken in the vehicle length direction, that is, the distance direction, coordinate conversion from the point (i, j, dp) on the distance image to the point (X, Y, Z) on the real space is as follows (2) to ( 4) Performed based on the equation.
X = CD / 2 + Z * PW * (i-IV) (2)
Y = CH + Z × PW × (j−JV) (3)
Z = CD / (PW × (dp−DP)) (4)

  That is, the center position of the main camera 2a and the sub camera 2b, more precisely, the distance L from the point on the road surface just below the center to the object and the parallax dp are obtained by setting Z in the above equation (4) as the distance L. Uniquely associated. Here, CD is the distance between the main camera 2a and the sub camera 2b, PW is the viewing angle per pixel, CH is the mounting height of the main camera 2a and the sub camera 2b, and IV and JV are infinity points in front of the host vehicle. I-coordinate and j-coordinate, and DP on the distance image represent vanishing point parallax.

  Further, for the purpose of improving the reliability of the parallax dp, the image processor 7 performs a filtering process on the parallax dp thus obtained, and outputs only the valid parallax dp. That is, for example, even if a 4 × 4 pixel block consisting of only a road image is scanned on a horizontal line having a width of 4 pixels of the comparison image, all the portions of the comparison image where the road is imaged are correlated. Even if the corresponding pixel block is specified and the parallax dp is calculated, the reliability of the parallax dp is low. Therefore, such a parallax dp is invalidated in the filtering process, and 0 is output as the value of the parallax dp.

  Accordingly, the distance L between the pixels of the reference image T output from the image processor 7, that is, the parallax dp for calculating the distance in the real space for each pixel block of the reference image T is usually in the left-right direction of the reference image T. The data has an effective value only for a so-called edge portion in which the difference in luminance value p1ij is large between adjacent pixels.

  The parallax dp of each pixel block of the reference image T calculated by the image processor 7 is converted into Z, that is, a distance L based on the above equation (4), and stored in the distance data memory 8 of the image processing means 6. It has become. In the processing in the detection means 9, one pixel block of the reference image T is treated as 4 × 4 pixels, and the 16 pixels belonging to one pixel block are independent pixels (i, j). Hereinafter, the distance L of the pixel (i, j) is expressed as Lij.

  The detection means 9 is composed of a microcomputer in which a CPU, ROM, RAM, input / output interface and the like (not shown) are connected to a bus. The detection means 9 is connected to a navigation device B equipped with a vehicle speed sensor A, a GPS (Global Positioning System) receiver, and the like.

  In this embodiment, the detection means 9 includes an expressway traveling detection unit 91, a pedestrian crossing detection unit 92, and a lane detection unit 93.

  The highway travel detection unit 91 is configured to determine whether or not the host vehicle is traveling on the highway. In the present embodiment, the highway travel detection unit 91 makes the determination based on information from the navigation device B.

  Moreover, the highway driving | running | working detection part 91 is based on the information from the vehicle speed sensor A, for example, when the state whose vehicle speed of the own vehicle is 80 km / h or more continues for 1 minute or more, the own vehicle drive | works the highway. It is determined that there is a high possibility that the vehicle has started driving on the highway. Note that the interval between the left and right lane positions detected by the lane detection unit 93 to be described later is used to determine whether or not the host vehicle is traveling on an expressway or is likely to be traveling on an expressway. It is also possible to measure the width of the traveling lane based on the above and make a positive determination if the width is equal to or greater than a predetermined distance.

  The pedestrian crossing detection unit 92 scans the horizontal line j having a width of one pixel on the reference image T corresponding to, for example, 50 m ahead of the host vehicle from the left end to the right, reads the luminance value p1ij of each pixel, and If there are, for example, three or more high-luminance pixel portions and low-luminance pixel portions that are approximately 45 to 50 cm continuous as a distance in the real space, such as three or more pixels, the pedestrian crossing is detected. It has become. It is also possible to configure to detect a pedestrian crossing by other methods.

  The lane detector 93 detects left and right lane positions of the host vehicle on the reference image T based on the luminance value p1ij and the distance Lij of each pixel (i, j) of the reference image T.

  Specifically, as described above, the lane detecting unit 93 performs image correction for each horizontal line corresponding to one pixel width of the reference image T so that the data of the luminance value p1ij of each pixel (i, j) of the reference image T is displayed. It is input from the part 4. Then, the luminance value p1ij of each pixel is input while the horizontal line is sequentially offset from the lower side by one pixel width, and finally the data of the luminance value p1ij of all the pixels (i, j) of the reference image T is the lane. The signal is input to the detection unit 93.

  Since the j coordinate of each pixel (i, j) on the horizontal line is the same, a horizontal line in which the j coordinate of each pixel on the line is j is represented as a horizontal line j.

  The lane detection unit 93 searches the horizontal line j input from the image correction unit 4 according to the basic flow shown in FIG. It is detected as a lane candidate point, and the left and right lane positions of the host vehicle are detected based on the detected lane candidate point.

  The search for lane candidate points can be performed on the entire image of the reference image T. That is, the entire area of the reference image T can be set as a search area.

  However, in the present embodiment, from the viewpoint of suppressing erroneous detection of lane candidate points and improving the processing speed, as shown in FIG. 4, the right lane position LRlast and the left lane on the left and right of the host vehicle are detected in the previous sampling period as shown in FIG. When the position LLlast is detected, in this search, only the surroundings of the previous right lane position LRlast and left lane position LLlast on the reference image T are set as the search areas Sr and Sl, and only within those areas. Search for lane candidate points.

  In the present embodiment, the search areas Sr and Sl are set to positions separated from the previous right lane position LRlast and left lane position LLlast by a certain distance in the left-right direction in real space. Specifically, as shown in FIG. 5, the prescribed values of the search areas on the side close to the host vehicle with respect to the right lane position LRlast and the left lane position LLlast, that is, the left side of the right lane position LRlast and the right side of the left lane position LLlast. The inner specified value Wth_in is set as the specified value for the search area on the far side from the host vehicle with respect to the right lane position LRlast and the left lane position LLlast, that is, the right lane position LRlast and the left lane position LLlast. The outer specified value Wth_out is set. The inner specified value Wth_in and the outer specified value Wth_out are appropriately set to values of about 40 cm.

  Further, the search for the lane candidate point is performed by using the horizontal line j to be searched, the left edge of the right lane side search area Sr, and the right edge of the left lane side search area S1 as intersection pixels as search start points is of the respective areas, and the right lane. On the side, the search is performed by offsetting one pixel at a time on the horizontal line j to the right, and on the left lane side to the left on the horizontal line j.

  Further, the search is performed up to each search end point ie using the pixel at the intersection of the horizontal line j and the right end of the right lane side search region Sr and the left end of the left lane side search region S1 as the search end point ie of each region. It has become. As will be described later, when the start point Ps is found and the first lane width threshold value Wth1 is set, the first lane width threshold value Wth1 is exceeded until the first lane width threshold value Wth1 is reached. The search continues.

  The lane detection unit 93 detects lane candidate points according to the flowchart shown in FIG. 6 in the lane candidate point detection conversion process (step S10) which is the first process of the basic flow of FIG. Since detection of lane candidate points for the left and right search areas Sr and S1 is performed in the same manner, a case where the right lane side search area Sr is searched will be described below as a representative. The search for the left lane side search area S1 is configured to be performed simultaneously with the search for the right lane side search area Sr for the same horizontal line j.

  When the luminance value p1ij of each pixel (i, j) on the horizontal line j of the reference image T is transmitted from the image correction unit 4 to the lane detection unit 93, the lane detection unit 93 performs the search for the search region Sr in the manner shown in FIG. A search start point is and a search end point ie are set (step S101). Subsequently, the lane detector 93 determines whether or not the search pixel satisfies the following first start point condition while offsetting the pixel to be searched to the right (step S102) (step S103).

[First start point condition]
Condition 1: The luminance value p1ij of the search pixel is greater than the road surface luminance value proad by the first start point luminance threshold pth1 and the edge intensity Eij represented by the luminance differential value is greater than or equal to the first start point edge intensity threshold Eth1. There is.
Condition 2: A point on the real space corresponding to the search pixel is on the road surface.
Here, the road surface brightness value proad is the horizontal lines j-1, j-2, j-3, j-4 for four rows that have already been searched immediately below the horizontal line j currently being searched. It is calculated for each horizontal line j as a luminance value that maximizes the frequency of appearance of the luminance value histogram of the above pixel.

  As shown in FIGS. 7A and 7B, the condition 1 is that the luminance value p1ij increases from the road surface luminance value proad to the threshold value pth1 or more and the edge intensity Eij as the luminance differential value is the threshold value Eth1 or more. This is a condition for finding a search pixel corresponding to one edge portion as a start point that can be a lane candidate point. Condition 2 is a condition for excluding search pixels which are body parts such as pillars and bumpers of the preceding vehicle above the road surface and edge parts such as guardrails and telephone poles from the start point.

  When a pixel satisfying the condition 1 appears, the lane detection unit 93 reads the distance Lij for the pixel from the distance data memory 8 and determines whether or not a point in the real space corresponding to the pixel is on the road surface. To do.

Here, a method for determining whether or not the condition 2 is satisfied will be briefly described. As shown in FIG. 8, first, assuming that a vertical line drawn from the vanishing point Vp to the horizontal line j to which the search pixel M on the reference image T belongs is a pixel m, the point m in the real space corresponding to the pixel m is Located in front of the host vehicle. If the point m is on the road surface, the distance from the main camera 2a mounted on the vehicle to the point m is L _m , the focal length of the main camera 2a is f, and the mounting height of the main camera 2a is h. The deviation y between the image of the point m and the vanishing point Vp at the imaging position of the main camera 2a is
y = h × f / L _m (5)
It is represented by

Assuming that the pixel interval between the pixel m and the vanishing point Vp on the reference image T is ypixel, the relationship ypixel = y / p is established when the pixel length is p. Therefore, the pixel interval ypixel between the pixel m and the vanishing point Vp. And the distance L _m from the main camera 2a to the point m is
ypixel = y / p = h × f / (L _m × p) (6)
That is,
L _m = h × f / (p × ypixel) (7)
The relationship holds. Since the j coordinate of the vanishing point Vp is known in advance, the distance L _m from the host vehicle to the point m in the real space is calculated from the j coordinate of the horizontal line j based on the equation (7).

Similarly, the distance L _mM between the point m on the real space and the point M on the road surface corresponding to the search pixel M can be calculated from the pixel interval xpixel between the pixel m of the reference image T and the search pixel M. From this distance L _mM and the distance L _m , a distance L _{M in} real space between the vehicle and a point M on the road surface is calculated.

Then, the distance L _M between the vehicle in the real space and the point M on the road surface is compared with the distance Lij read from the distance data memory 8, and if they match within a certain error range, the search pixel is obtained. M points on the real space corresponding to M can be determined that are on the road surface, if the distance Lij is smaller than the distance L _M, M points in the real space corresponding to the search pixel M in the above road surface position It can be determined that the point M is a pillar or the like of an existing preceding vehicle and is not on the road surface.

  If it is determined that the search pixel satisfies the first start point condition (step S103 in FIG. 6: YES), the lane detection unit 93 sets the search pixel as the start point Ps on the RAM, and the first lane width threshold Wth1. Is set (step S104). The first lane width threshold value Wth1 is a threshold value for securing a search range when the start point Ps is detected, and is set to 1 m as a distance in the real space, for example. It is also possible to make the first lane width threshold Wth1 variable depending on j of the horizontal line j.

  When the lane detection unit 93 sets the start point Ps and the first lane width threshold Wth1, the lane detection unit 93 continues to search rightward on the horizontal line j, and determines whether or not the search pixel satisfies the following second start point condition. Judgment is made (step S105). This is the pixel Psold corresponding to the end of the old lane Lold where the pixel set as the start point Ps has actually disappeared as shown in FIG. 9, and is newly overlapped with the old lane Lold. When the repainted lane Lnew is drawn, the original start point Ps and the like are canceled, and the pixel Psnew corresponding to the end of the repainted lane Lnew is newly set as the start point Ps. It is a judgment standard.

[Second starting point condition]
Condition 3: The luminance value p1ij of the search pixel is larger than the road surface luminance value proad by the second starting point luminance value threshold value pth2, and the edge strength Eij represented by the luminance differential value is equal to or larger than the second starting point edge strength threshold value Eth2. Be. However, pth2> pth1.
Condition 4: The difference between the average luminance value from the start point Ps to the pixel adjacent to the left of the search pixel and the road surface luminance value proad is equal to or less than the first lane average luminance value threshold Ath1.
Condition 5: A point on the real space corresponding to the search pixel is on the road surface.

  When the search pixel satisfies the second start point condition (step S105: YES), the lane detection unit 93 cancels the set of the original start point Ps and the first lane width threshold Wth1, and performs the current search. The pixel is reset as a new start point Ps, and the first lane width threshold value Wth1 is also reset (step S106).

  FIG. 10A is a diagram for explaining the second start point luminance threshold value pth2 of the second start point condition, and FIG. 10B is a diagram for explaining the second start point edge intensity threshold Eth2. If the difference between the average luminance value of the region K from the original start point to the pixel adjacent to the left of the current search pixel and the road surface luminance value proad is larger than the threshold value Ath1 in the condition 4, the region K is corresponded. Since it is considered that the lane is currently marked as a living lane, there is no need to reset the starting point Ps.

  Further, in the normal lane position detection, the normal lane is often detected from the beginning instead of the lane Lnew repainted on the old lane Lold. Satisfactory search pixels simultaneously satisfy the second start point condition. Therefore, when the first start point condition is satisfied, it is set as the start point Ps, the first lane width threshold Wth1 is set, it is immediately determined that the second start point condition is satisfied, and the start point Ps and the like are reset. The

The lane detector 93 continues to search rightward on the horizontal line j, and determines whether the search pixel satisfies the following end point condition (step S107 in FIG. 6).
[End condition]
Condition 6: The edge intensity Eij represented by the luminance differential value of the search pixel is equal to or less than the end point edge intensity threshold −Eth2, or the luminance value of the search pixel is smaller than the luminance value at the start point Ps.

  Although not shown, the end point Pe is a point corresponding to the edge portion on the opposite side of the lane, from the high luminance pixel corresponding to the lane to the low luminance value pixel corresponding to the road surface. Represents. In the present embodiment, the absolute value of the end point edge intensity threshold −Eth2 is set to be the same as the absolute value of the second start point edge intensity threshold Eth2.

  If a point that satisfies the end point condition, that is, the end point Pe is not detected (step S107: NO), the lane detection unit 93 searches until the first lane width threshold Wth1 is reached, and reaches the first lane width threshold Wth1. However, if the end point Pe is not detected (step S108: YES), the setting of the start point Ps and the first lane width threshold Wth1 is canceled.

  If the search end point ie has not been reached (step S102: NO), the search on the horizontal line j is continued and the determination from the determination whether the search pixel satisfies the first start point condition (step S103). Repeat the above process. As can be seen from this flowchart, even if the search continues to the right from the start point Ps and exceeds the search end point ie, the horizontal line is not reached until the first lane width threshold Wth1 is reached. The search on j continues.

Further, when the lane detection unit 93 detects the end point Pe (step S107: YES), does the average luminance value from the start point Ps to the pixel adjacent to the left of the end point Pe satisfy the following first average luminance value condition? It is determined whether or not (step S109).
[First average luminance value condition]
Condition 7: The difference between the average luminance value from the start point Ps to the pixel adjacent to the left of the end point Pe and the road surface luminance value proad is equal to or less than the first lane average luminance value threshold Ath1.

  In the present embodiment, the first lane average luminance value threshold Ath1 in the condition 7 is the same value as the first lane average luminance value threshold Ath1 in the condition 4 of the second start point condition. As described above, the first lane average luminance value threshold value Ath1 is a threshold value that divides the average luminance value corresponding to the old lane Lold that has disappeared and the average luminance value corresponding to the newly repainted lane Lnew.

  The lane detector 93 corresponds to a lane in which the average luminance value from the start point Ps to the pixel adjacent to the left of the end point Pe satisfies the first average luminance value condition (step S109: YES), that is, the average luminance value is low. If there is a possibility that there is a newly painted lane near this lane, the lane candidate point is not already stored in the storage means (not shown) (step S110: YES). The coordinates (i, j) of the start point Ps are temporarily stored as lane candidate points (step S111), and the search on the horizontal line j is continued.

  On the other hand, the lane detecting unit 93 does not satisfy the first average luminance value condition from the average luminance value from the start point Ps to the pixel adjacent to the left of the end point Pe (step S109: NO), that is, the average luminance value is high. If it is determined that the average luminance value corresponds to the lane, if there is a lane candidate point corresponding to a lane having a low average luminance value already saved in the search on the same horizontal line j (step S112: YES), that The lane candidate point is deleted (step S113), the start point Ps having the higher average luminance value is stored as the lane candidate point (step S114), and the search on the same horizontal line j is terminated.

  In the present embodiment, as described above, when the search on the horizontal line j ends, when a lane with a high average luminance value is found and the search is terminated, only the lane with a low average luminance value is found and the search is completed. When the end point ie is reached, there are three cases where the lane is not found and the search end point ie is reached. Then, if a live lane is found with a high average luminance value, priority is given to the lane, and among the pixels satisfying the first start point condition or the second start point condition, the pixel closest to the host vehicle is set as the lane candidate point. To detect.

  When the search on the horizontal line j ends, the lane detector 93 determines whether or not lane candidate points are stored (step S115). If the lane candidate point is not stored (step S115: NO), it is determined whether or not the horizontal line j has reached the 200th line (step S116). If the lane candidate point has not reached the 200th line (step S116: NO) Until the 200th line is reached, the above processing procedure is repeated for the horizontal line j + 1 that is one pixel above and transmitted from the image correction unit 4.

  If the lane candidate points are stored (step S115: YES), the lane detector 93 subsequently performs a thick marking number counting process (step S117).

  Thick sign count processing is performed on white, yellow, and other lanes that are marked wider than the lanes marked with normal continuous lines or broken lines, such as zebra patterns on pedestrian crossings or multiple lines in the center of the road. Alternatively, it is a process of counting the number of lane candidate points corresponding to a thick broken-line marking marked at a junction of an expressway or the like.

  In the present embodiment, the number of lane candidate points corresponding to thick markings in the range of about 10 m to 20 m ahead of the host vehicle is counted, and the determination materials for detection mode switching (step S40 in FIG. 3) to be described later are tabulated. It has become.

  Specifically, in the thick marking number counting process (step S117), the lane detecting unit 93, as shown in FIG. 11, first, the horizontal line j currently being searched is about 10 m ahead of the host vehicle as described above. It is determined whether or not a region within a range of 20 m from the search is being searched (step S118). If not within the range (step S118: NO), the detected lane width threshold value changing process without performing this process (step S121) Migrate to

  If it is determined that the horizontal line j is within the above range (step S118: YES), the lane detecting unit 93 subsequently calculates a real space distance between the lane candidate point and the corresponding end point Pe. It is supposed to be. The distances in the real space can be obtained from the distance Lij from the own vehicle to the lane candidate point and the number of pixels between the lane candidate point and the end point Pe in the manner described above. Note that the distance between the lane candidate point and the end point Pe corresponds to the length or width in the left-right direction of the lane and other signs.

  The lane detector 93 determines whether or not the calculated distance between the lane candidate point and the end point Pe is within the second lane width threshold Wth2 (step S119), and the distance, that is, the width of the indication is the second lane width threshold. If it is larger than Wth2 (step S119: NO), the thick marking number count is incremented by 1 (step S120), and the number of lane candidate points corresponding to the thick marking is counted. The second lane width threshold value Wth2 is a threshold value for determining whether the sign is a normal lane or the thick display, and is set to 30 cm, for example.

  Subsequently, the lane detection unit 93 performs a detection lane width threshold value changing process (step S121 in FIG. 6).

  The detected lane width threshold Wthd is the Hough transform (step S132) described later, leaving the lane candidate point detected according to the distance between the lane candidate point and the corresponding end point Pe, that is, the lane width or the width of the marking. This is a threshold value for determining whether to be a target or to be deleted.

  As described above, the first lane width threshold value Wth1 is a threshold value for securing a search range when the start point Ps is detected, and the second lane width threshold value Wth2 is an indication that the sign is a normal lane or It is a threshold value for determining whether the display is thick, and both are fixed to a set numerical value.

  On the other hand, the detected lane width threshold value Wthd is variable according to the conditions of the traveling environment of the host vehicle and the road environment as described below, and is automatically changed. Note that the detected lane width threshold Wthd is a value that can detect a lane width standard, that is, a lane with a width of 10 to 15 cm and does not detect a zebra pattern of a pedestrian crossing with a standard of 45 to 50 cm in the standard detection mode. That is, for example, it is set to 30 cm.

  In the present embodiment, two types of lane position detection modes can be set: a standard detection mode and a thick lane detection mode. As described above, the standard detection mode is a detection mode in which a lane having a normal lane width is detected and a lane wider than that is not detected. The thick lane detection mode is a detection mode in which a wide lane can be detected.

  In the detected lane width threshold value changing process (step S121), the lane detection unit 93 determines that the host vehicle is traveling on the highway by the highway travel detection unit 91 as shown in FIG. If this is the case (step S122: YES), the detected lane width threshold Wthd is increased to, for example, 55 cm (step S123).

  If the host vehicle is not traveling on the expressway (step S122: NO) and the pedestrian crossing is detected in front of the host vehicle by the pedestrian crossing detection unit 92 (step S124: YES), the pedestrian crossing from the host vehicle is performed. The distance R1 is set to 50 which means 50m ahead (step S125), and the detected lane width threshold Wthd is set to 30 cm, for example (step S126). R1 is set to 0 as an initial value.

  Even after the pedestrian crossing is no longer detected at a point 50 m ahead of the host vehicle, the distance R1 from the host vehicle to the pedestrian crossing is updated in the detection mode switching process (step S40) described later. 2, until the distance R1 between the pedestrian crossing and the host vehicle is in the real space corresponding to the lowest horizontal line j of the reference image T. The vehicle is tracked until the distance between the vehicle and the vehicle is less than or equal to Runseen.

  Therefore, even if a pedestrian crossing is not detected in front of the host vehicle by the pedestrian crossing detection unit 92 (step S124: NO), if the distance R1 between the pedestrian crossing and the host vehicle is larger than the distance Runseen (step S127: YES) ) Since the pedestrian crossing is still imaged on the reference image T, the detected lane width threshold Wthd is set to be narrow, for example, 30 cm (step S126). The distance Runseen is determined in advance according to the installation state of the main camera 2a.

  On the other hand, if no pedestrian crossing is detected or the distance R1 is equal to or less than the distance Runseen (step S127: NO), the lane detection unit 93 continues with the highway traveling detection unit 91. If it is determined that the vehicle is traveling on the highway or that it is highly likely that the vehicle has started traveling on the highway (step S128: YES), the detected lane width threshold Wthd is increased to, for example, 55 cm (step S123).

  Further, the lane detection unit 93 is unlikely that the host vehicle is traveling on the highway or has started to travel on the highway (step S128: NO), and the current detection mode is not a thick lane detection mode ( Step S129: NO), the detected lane width threshold Wthd is set to 30 cm, for example (step S126).

  Further, if the current detection mode is a thick lane detection mode (step S129: YES), the lane detection unit 93 determines that the distance of the line in the real space corresponding to the horizontal line j currently being searched for to the own vehicle. It is determined whether the distance is within a predetermined distance R2, which will be described later (step S130). If the distance is within the predetermined distance R2 (step S130: YES), the detected lane width threshold Wthd is set to, for example, 30 cm (step S126). If the distance is greater than the predetermined distance R2 (step S130: NO), the detected lane width threshold Wthd. Is enlarged to 55 cm, for example (step S123).

  In the present embodiment, the case where only two types of detection lane width threshold values Wthd are provided, for example, 30 cm or 55 cm will be described. However, it is also possible to configure so that the set value is appropriately changed according to the various conditions. It is.

  When the lane detection unit 93 sets the detected lane width threshold value Wthd as described above, the lane candidate point calculated in the thick count number count (step S117 in FIG. 6) and the end point Pe corresponding to the lane candidate point are calculated. It is determined whether or not the distance in the real space is within the set detection lane width threshold Wthd (step S131).

  If the distance between the lane candidate point and the end point Pe is larger than the detected lane width threshold Wthd (step S131: NO), the lane candidate point is deleted (step S132).

  If the distance between the lane candidate point and the end point Pe is within the detected lane width threshold Wthd (step S131: YES), the lane detection unit 93 continues to the pixel on the left side of the end point Pe from the lane candidate point. It is determined whether or not the average luminance value up to the following second average luminance value condition is satisfied (step S133).

[Second average luminance value condition]
Condition 8: The difference between the average luminance value from the lane candidate point to the left adjacent pixel of the corresponding end point Pe and the road surface luminance value proad is equal to or greater than the second lane average luminance value threshold Ath2.

  In the present embodiment, as described above, the first lane average luminance value threshold Ath1 defines the average luminance value corresponding to the old lane Lold that has disappeared and the average luminance value corresponding to the newly repainted lane Lnew. The second lane average luminance value threshold value Ath2 in condition 8 defines the difference between the minimum required average luminance value as the lane and the road surface luminance value proad, and is the first lane average luminance. A value smaller than the value threshold Ath1 is set as appropriate.

If the lane detection unit 93 determines that the lane candidate point does not satisfy the second average luminance value condition (step S133: NO), the lane detection point 93 deletes the lane candidate point (step S132). On the other hand, if it is determined that the lane candidate point satisfies the second average luminance value condition (step S133: YES), the lane candidate point is left without being deleted, and finally on the horizontal line j as shown in FIG. A lane candidate point (I _j , J _j ) is detected at.

  Subsequently, the lane detector 93 performs Hough transform on the lane candidate points detected on the horizontal line j as described above (step S134). The Hough transform is intended to detect a straight line that serves as a reference for lane position detection based on a plurality of lane candidate points on the reference image T.

In the present embodiment, a known method is used for the Hough transform. Specifically, for example, the detected lane candidate point is a straight line on the reference image T i = aj + b (8)
Assuming that it exists above, I _j and J _j are
I _j = aJ _j + b (9)
Meet.

The equation (9) is
b = −J _j × a + I _j (10)
And can be transformed. As can be seen from the equation (10), when the lane candidate point (I _j , J _j ) is detected on the horizontal line j in the lane candidate point detection process, −J _j , I _j are shown as slopes and b intercepts. As shown in FIG. 14, one straight line can be drawn on the ab plane which is a Hough plane.

  The ab plane is divided into cells having a predetermined size as shown in FIG. 15, and when a straight line represented by the equation (10) is drawn, a straight line like a hatched cell in FIG. The count value of the mesh that passes is increased by one. It should be noted that since the predetermined a and b values correspond to each cell on the ab plane, selecting the cell on the ab plane means selecting the corresponding a and b values, This is synonymous with selecting a straight line having an inclination a and an i-intercept b on the reference image T as expressed by equation (8).

  The lane detector 93 searches the reference image T while shifting the horizontal line upward by one pixel at a time, and performs the Hough transform each time a lane candidate point is detected on each horizontal line to obtain the ab plane. The count value of the square is added. When the search is finally completed up to the top 200 horizontal line of the reference image T, a plurality of lane candidate points are obtained as shown in region A of FIG. Are in a state in which the respective count values are added.

In FIG. 16 and subsequent figures, lane candidate points are shown to be sparsely detected on the reference image T or in real space, but in reality, a large number are detected very finely. In the above description, the case of searching the right lane side search area Sr has been described. However, in the case of searching the left lane side search area S1 as well, the Hough transform is performed on the detected lane candidate point (I _j , J _j ). The ab plane is created separately from the case of the right lane side search area Sr, and a count value is added to each cell of the ab plane.

  After completing the Hough transform for all lane candidate points detected in the rightward and leftward searches on the horizontal line j and adding the count value to each square of the two ab planes, the horizontal line j of the 200th row When the search ends (step S116 in FIG. 6: YES), the lane candidate point detection conversion process (step S10), which is the first process of the basic flow shown in FIG. 3, ends.

  Next, the lane detector 93 proceeds to a lane line detection process (step S20), which is the second process of the basic flow.

  The lane detecting unit 93 extracts one or a plurality of cells having a large count value from the ab planes of the right lane side search area Sr and the left lane side search area S1 obtained by the Hough transform, respectively. Peak straight lines r1, r2, l1, and l2 as shown in FIG. 18 are extracted.

  The lane detection unit 93 rejects a plurality of extracted peak straight lines that meet the following selection conditions, and performs selection for each of the right search and the left search until the peak straight line is narrowed down to one. It has become.

[Selection condition]
Condition 9: The position of the peak straight line at a predetermined forward distance exists inside the position of the vehicle width converted from the center of the host vehicle. However, when the lane position is continuously detected, the condition 9 is not applied if the change amount from the previously detected lane position is within a predetermined threshold.
Condition 10: Parallelism with the estimated trajectory Lest of the host vehicle is larger than a certain threshold value.
Condition 11: When there is a peak straight line in which the left-right difference distance from the center of the vehicle at a predetermined forward distance represented by the one-dot broken line Ld in FIGS. 17 and 18 is more than a specified value far_th and a peak straight line less than a specified value near_th The peak straight line is more than the specified value far_th. However, far_th> near_th.
Condition 12: A peak straight line closest to the lane estimation center position estimated from the previously detected lane position and width is left.

  The lane detector 93 shows left and right peak straight lines r1 and l1 representing straight lines suitable as lane positions from the peak straight lines extracted in the right lane side search area Sr and the left lane side search area S1 in this way, respectively. The lane straight lines r1 and l1 are selected one by one, and the lane straight line detection process (step S20) in FIG. 3 is terminated.

  Subsequently, the lane detection unit 93 proceeds to lane position detection processing (step S30), which is the third process of the basic flow. In the lane position detection process, linear or curved lanes existing on the left and right of the host vehicle are detected with reference to the lane straight lines r1 and l1 obtained in the lane straight line detection process.

  Specifically, the lane detection unit 93 performs processing while scanning rightward and leftward on a horizontal line j having a single pixel width with a constant j coordinate in the reference image T, similarly to the lane candidate point detection conversion processing. . First, the right lane straight line r1 or the left lane straight line l1 is used as a reference, and lane candidate points whose difference in i direction from each straight line is within a threshold are recorded as lane positions until a predetermined number is reached.

  When the lane detection unit 93 records the predetermined number of lane positions on or near the lane straight lines r1 and l1, the lane position detected last in the horizontal line j above the lane straight lines r1 and l1 regardless of the lane straight lines r1 and l1. Based on the above, the lane is tracked in the following manner, and even when the lane is curved, the lane position is detected following the curve.

  That is, when a predetermined number of lane positions are recorded, the lane detection unit 93 performs the i direction between the lane candidate point detected next on the horizontal line j above and the predetermined number of lane positions detected at the end, It is determined whether the displacement in the j direction is within a specified value. When it is determined that the value is within the specified value, the lane candidate point is recorded as a lane position.

  Thereafter, similarly, as shown in FIG. 19A, when a lane candidate point is detected on the horizontal line j, whether or not the displacement in the i direction and the j direction with respect to the previously detected lane position a is within a specified value. If it is determined that it is within the specified value, the lane candidate point detected this time is recorded as the lane position b.

  In addition, when a lane candidate point is detected on the horizontal line j and the displacement in the i direction and the j direction with respect to the previously detected lane position is not within the specified value, a straight line obtained by extending the previously detected lane position, That is, in FIG. 19B, if the displacement in the i direction from the straight line connecting the lane position b and the lane position a is within a specified value, the detected lane candidate point is recorded as the lane position c, d,. .

  The lane detector 93 sequentially detects the lane position while shifting the horizontal line j above the reference image T while confirming the space in the real space between the left and right lanes, that is, the road width, and finally, as shown in FIG. In addition, the lane positions LR and LL are detected on the left and right of the host vehicle. In this manner, the lane position detection process (step S30) in FIG. 3 ends.

  Subsequently, the lane detector 93 proceeds to a detection mode switching process (step S40), which is the fourth process of the basic flow. This detection mode switching is a pre-process for the next detection process.

  In the detection mode switching process (step S40), the lane detection unit 93 first determines whether or not the current detection mode is the thick lane detection mode (step S401), and the thick lane detection mode, as shown in FIG. If so (step S401: YES), the lane candidate points corresponding to the thick markings in the range of about 10m to 20m ahead of the host vehicle counted in the thick marking number counting process (step S117) of the lane candidate point detection conversion process. It is determined whether or not the number is equal to or greater than a threshold value (step S402). The threshold is appropriately set to 6, for example.

  If the number of lane candidate points corresponding to the thick marking is equal to or greater than the threshold (step S402: YES), the distance r by which the host vehicle advances between the current detection process and the next detection process is calculated (step S402). S403). The distance r is calculated as the product of the vehicle speed V and the processing cycle Δt during the current processing.

  Further, in the detected lane width threshold value changing process (step S121), the predetermined distance R2 serving as a reference for switching the detected lane width threshold value Wthd when the current detection mode is the thick lane detection mode is determined from the current value. Calculation is performed by subtracting the distance r by which the host vehicle moves forward (step S404). That is, the predetermined distance R2 that is a reference for switching the detected lane width threshold Wthd is shortened by the amount of advance of the host vehicle.

  If the current detection mode is a thick lane detection mode, but the number of lane candidate points corresponding to the thick marking is less than the threshold (step S402: NO), the detection mode is switched to the standard detection mode and set (step S402). S405).

  On the other hand, when the current detection mode is not the thick lane detection mode but the standard detection mode (step S401: NO), if the number of lane candidate points corresponding to the thick marking is less than the threshold value (step S406: NO), it remains as it is. Set to standard detection mode.

Further, when the current detection mode is the standard detection mode and the number of lane candidate points corresponding to the thick marking is equal to or greater than the threshold (step S406: YES), the detection mode is switched to the thick lane detection mode and set. (Step S407), the predetermined distance R2 is set to the reference distance _R0 (Step S408). The reference distance _R0 is appropriately set to 20 m, 18 m, 16 m, etc., for example.

  If the distance R1 between the pedestrian crossing and the host vehicle is larger than the distance Runseen (step S409: YES), the lane detection unit 93 determines the distance r by which the host vehicle moves forward between the current detection process and the next detection process. If the distance r is calculated (step S410) or if the distance r has been calculated in step S403, the value is used, and the distance R1 is shortened from the current value by the distance r that the host vehicle has advanced (step S411). To come to track.

  In this way, the detection mode switching process (step S40) in FIG. 3 ends.

  Next, the operation of the lane detector 1 according to this embodiment will be described.

  In the initial state of the detection means 9, the detection mode is the standard detection mode, and the detection lane width threshold Wthd is set to 30 cm in this embodiment. Therefore, in this standard detection mode, a normal lane having a standard of 10 to 15 cm is detected, but a zebra pattern paint portion of a pedestrian crossing repeated at intervals of 45 to 50 cm is not detected.

  However, if the detection lane width threshold value Wthd is set to be narrow as described above, for example, a thick broken-line sign BL indicated at a junction of an expressway as shown in FIG. 22 cannot be detected.

  Therefore, in the present embodiment, when the detected lane width threshold value changing process (step S121) of the lane candidate point detection and conversion process (step S10) is detected (step S122: YES) or crossing is detected. When there is a high possibility that the vehicle is traveling on a highway without detecting a sidewalk (step S128: YES), the detected lane width threshold Wthd is enlarged to 55 cm, for example. Therefore, a lane candidate point corresponding to such a thick broken-line marking BL can be reliably detected.

  In this case, when the sign BL is recognized as a thick sign in the thick sign count process (step S117) in the lane candidate point detection conversion process (step S10), the detection mode is changed in the detection mode switching process (step S40). The standard detection mode is switched to the thick lane detection mode. However, in the detected lane width threshold value changing process (step S121), the detected lane width threshold value Wthd is increased if the vehicle is traveling on the highway or its possibility is high regardless of the detection mode switching. If the vehicle is in the middle or the possibility is high, a lane candidate point and a lane position corresponding to the marking BL are detected in each detection process.

  In addition, since there is no pedestrian crossing on the highway, if it is detected that the vehicle is traveling on the highway, the detection lane width threshold Wthd may be unconditionally expanded without waiting for the detection result of the pedestrian crossing (step S122). , S123).

  However, even if there is a high possibility that the vehicle is traveling on a highway due to the vehicle speed V or the like of the host vehicle, it is possible to travel on a general road at high speed as long as it is not reliably detected that the vehicle is traveling on the highway. Sex remains. Therefore, in this embodiment, it is determined whether or not there is a high possibility that the vehicle is traveling on a highway only when no pedestrian crossing is detected (steps S124, S127, and S128).

  On the other hand, in the standard detection mode in which the detection lane width threshold value Wthd is set narrow, for example, it is not possible to detect the lane TL representing the roadway center line in which about three lanes as shown in FIG.

  Therefore, in this embodiment, first, in the thick sign number count process (step S117) of the lane candidate point detection conversion process (step S10), the standard detection mode is performed in the range of about 10 m to about 20 m ahead of the host vehicle MC. In this detection process, the number of lane candidate points corresponding to thick markings that are deleted as lane candidate points is counted to detect thick markings.

If the number of lane candidate points corresponding to the thick marking is equal to or greater than the threshold, the detection mode is switched to the thick lane detection mode (step S407) in the detection mode switching process (step S40), and the detection lane width threshold Wthd is switched. The reference predetermined distance R2 is set to the reference distance _R0 (step S408).

In the next detection process, since the detected lane width threshold value changing process (step S121) of the lane candidate point detection conversion process (step S10) is the thick lane detection mode (step S129: YES), the lane candidate point is searched. If the distance from the host vehicle MC in the real space of the horizontal line j that is performing is a predetermined distance R2, that is, within the reference distance _{R0 in} this case (step S130: YES), the detected lane width threshold Wthd is set to 30 cm, for example It is set narrower (step S126).

Since the lane width is narrow as shown in FIG. 23 when the distance from the host vehicle MC in the real space of the horizontal line j is within the reference distance R ₀ , the horizontal line j is on the horizontal line j even if the detected lane width threshold Wthd is narrow. The left and right lane positions can be detected by searching.

Further, in the range where the distance between the horizontal line j and the host vehicle MC is greater than the reference distance _R0 (step S130: NO), the lane width is wide, so that the detected lane width threshold Wthd is increased (step S123), it becomes possible to detect the position of the lane that has become wide in the search on the horizontal line j. In other words, the reference distance _R0 is set so that a wide lane can be accurately detected.

  In addition, as the host vehicle MC travels, the host vehicle MC moves forward in the next detection process. If the processing cycle of this apparatus is Δt, the host vehicle MC moves forward by V · Δt obtained by multiplying the current vehicle speed V by Δt. When this is viewed in the stationary system of the host vehicle MC, as shown in FIG. 24, the thick lane TL approaches the host vehicle MC by V · Δt.

Therefore, in the present embodiment, at the time when the previous lane position detection is completed, the detection mode switching process (step S40) performed as a pre-process for the next detection process is performed from the vehicle speed V of the host vehicle and the processing cycle Δt. The distance r = V · Δt traveled by the host vehicle MC is calculated (step S403), and the distance r is subtracted from the currently set predetermined distance R2 (step S404). In the example of FIG. 24, for the next detection process, the predetermined distance R2 is set by subtracting r = V · Δt from the reference distance _R0 currently set as the predetermined distance R2.

  In this way, by measuring the vehicle speed V for each detection process and subtracting r = V · Δt from the predetermined distance R2, the position for switching the detected lane width threshold Wthd is appropriately shifted to the own vehicle side. It becomes possible to accurately detect the lane position.

  If a pedestrian crossing is detected by the pedestrian crossing detection unit 92, the distance R1 from the host vehicle to the pedestrian crossing is 50 in the detected lane width threshold value changing process (step S121) of the lane candidate point detection conversion process (step S10). (Step S125), and the detected lane width threshold Wthd is set to be narrow, for example, 30 cm (step S126).

  Even after the pedestrian crossing detection unit 92 no longer detects the pedestrian crossing, the distance R1 between the pedestrian crossing and the host vehicle is larger than the distance Runseen in the detection mode switching process (step S40) (step S409: YES). The distance R1 is shortened by the distance r that the host vehicle has advanced (step S411), and the pedestrian crossing is traced until the distance R1 between the pedestrian crossing and the host vehicle is equal to or less than the distance Runseen (step S409: NO). . During that time (step S127: YES), the detected lane width threshold Wthd is set to be narrow, for example, 30 cm (step S126).

  Therefore, the zebra pattern paint portion of the pedestrian crossing is not detected until the pedestrian crossing is not visible from the reference image T.

  As described above, according to the lane detection device 1 according to the present embodiment, the traveling environment of the host vehicle such as the host vehicle traveling on the highway or having a high possibility of traveling, or crossing on the road. By appropriately setting the detection lane width threshold Wthd by enlarging or reducing it according to the road environment where the sidewalk is marked, unnecessary signs are properly excluded, and necessary signs are detected accurately. Thus, lane candidate points can be detected.

  Therefore, it is possible to accurately detect the left and right lane positions of the host vehicle.

  Specifically, according to the lane detection device 1 according to the present embodiment, in a normal standard traveling state, a normal single continuous line or broken line having a lane width standard of 10 to 15 cm. The detected lane width threshold Wthd is set to 30 cm, for example, so as to detect a lane-like lane. Thus, by setting it comparatively narrow, it becomes possible to prevent the erroneous detection of a lane candidate point.

  On the other hand, as shown in FIGS. 23 and 24, when a wide lane or sign is detected in the range of about 10 m to about 20 m ahead of the host vehicle MC, the detected lane width threshold Wthd is increased to, for example, 55 cm. Thus, such a wide lane or the like can be appropriately detected.

At this time, the detected lane width threshold value Wthd is set to be narrow when searching on the horizontal line j in the region of the reference image T within the reference distance R _{0 when} viewed from the host vehicle, and is farther than the reference distance R _0. When searching on the horizontal line j in a region, the detection lane width threshold Wthd is narrow in the region where one normal lane exists, and the detection lane width threshold in the region where a wide lane exists. Wthd can be widened.

  Therefore, it is possible to prevent erroneous detection of lane candidate points in an area where one normal lane exists, and to appropriately detect such a wide lane in an area where a wide lane exists.

Furthermore, by reducing the predetermined distance R initially set to the reference distance R ₀ according to the travel distance r of the host vehicle, the detected lane width threshold Wthd that differs between the two regions of the reference image T as described above can be set. It is possible to set appropriately according to the running of the vehicle, and it is possible to achieve the effect more accurately.

  Further, since there is no pedestrian crossing on the expressway, for example, by setting the detection lane width threshold Wthd widely over the entire area of the reference image T, a thick broken line shown at the junction of the expressway as shown in FIG. It is possible to accurately detect the sign BL of the shape, and it is possible to accurately detect the left and right lane positions LR and LL of the host vehicle.

  Further, when a pedestrian crossing is detected on a general road, by setting the detection lane width threshold Wthd to be narrow, the pedestrian crossing is erroneously detected as a lane and the left and right lane positions of the host vehicle are detected as in the conventional example. It is possible to prevent erroneous detection of LR and LL.

  It should be noted that, in the reference image T, it is possible to configure so that the lane candidate point is not detected for the region where the pedestrian crossing exists, and the lane position detection according to the present embodiment is performed for the other region.

  In addition, as shown in FIG. 25, a thick sign for counting the number of corresponding lane candidate points in the range of about 10 m to 20 m ahead of the host vehicle is, for example, the general length of the pedestrian crossing in the traveling direction of the host vehicle. When detected continuously, the thick marking is not the zebra pattern paint portion of the pedestrian crossing, but can be determined to be the roadway center line in which three lanes such as yellow and white are marked in parallel.

  For this reason, it is possible to increase the detection lane width threshold Wthd in such a case. If comprised in this way, it will become possible to detect reliably the wide sign which is not the zebra pattern paint part of a pedestrian crossing, and it becomes possible to detect a wide roadway center line exactly.

It is a block diagram which shows the structure of the lane position detection apparatus which concerns on this embodiment. It is a figure which shows an example of a reference | standard image. It is a flowchart which shows the basic flow of the process performed by a detection means. It is a figure explaining the search area | region limited to the circumference | surroundings of the lane detected last time. It is a figure explaining the inner side regulation value and outer side regulation value which define a search area | region. It is a flowchart which shows the procedure of a lane candidate point detection conversion process. It is a figure explaining (A) 1st starting point brightness | luminance threshold value pth1 and (B) 1st starting point edge intensity | strength threshold value Eth1. It is a figure explaining the calculation method of the distance to the point on the road surface corresponding to the pixel on a reference | standard image. It is a figure explaining the new lane partially overlapped with the old lane. It is a figure explaining (A) 2nd starting point brightness | luminance threshold value pth2 and (B) 2nd starting point edge intensity | strength threshold value Eth2. It is a flowchart which shows the procedure of a thick marking number count process. It is a flowchart which shows the procedure of a detection lane width threshold value change process. It is a figure explaining the lane candidate point detected on the horizontal line j. It is a figure explaining the straight line drawn by the ab plane which is a Hough plane. It is a figure explaining the cell of the ab plane through which a straight line passes. It is a figure which shows the lane candidate point detected by the search on a horizontal line. It is the figure which represented the peak straight line on the reference | standard image. It is the figure which represented the peak straight line of FIG. 17 on real space. It is a figure explaining the conditions which record a lane candidate point as a lane position. It is a figure which shows the right lane position and the left lane position which were finally detected. It is a flowchart which shows the procedure of a detection mode switching process. It is a figure showing the thick marking marked at the junction of the expressway. It is a figure showing the roadway center line in which the number of lane candidate points corresponding to a thick mark in the range ahead of the own vehicle is counted. It is a figure showing the roadway center line which the own vehicle advanced and relatively approached the own vehicle. It is a figure showing the roadway center line detected continuously more than the length of a pedestrian crossing. It is a figure which shows the example of the right-and-left lane position misdetected by the influence of the pedestrian crossing. It is a figure showing the roadway centerline imaged on the image.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Lane position detection apparatus 2 Imaging means 6 Image processing means 9 Detection means 91 Expressway driving | running | working detection part 92 Crosswalk detection part p1ij Luminance value Lij Distance T Reference | standard image LR, LL Lane position Wthd Detection lane width threshold value MC Own vehicle BL Wide Sign TL Roadway center line R2 Predetermined distances Sr and Sl in front of own vehicle Search area r Travel distance of own vehicle V Vehicle speed

Claims (9)

Imaging means for imaging a landscape including a road ahead of the host vehicle and outputting a pair of images having a luminance value for each pixel;
Image processing means for calculating a distance in real space for each pixel of at least one image based on the pair of captured images;
Detecting means for detecting a pixel corresponding to one edge portion of the lane as a lane candidate point based on the luminance value and the distance for the one image, and detecting a lane position based on the detected lane candidate point; ,
The detection means calculates a distance in real space between the lane candidate point and an edge portion on the opposite side of the lane corresponding to the lane candidate point, and the detected lane candidate point is detected only when the distance is equal to or smaller than a detected lane width threshold. Used to detect the lane position,
And the said detection lane width threshold value is made variable with the driving | running | working environment or road environment of the own vehicle, The lane detection apparatus characterized by the above-mentioned.   2. The detection means, when detecting a sign wider than one lane width standard, expands the detected lane width threshold to a value capable of detecting the wide sign. The lane detection device described in 1.   The detection means sets the detected lane width threshold to the one lane width in a predetermined distance in front of the host vehicle or in a search region closer to the host vehicle than the predetermined distance in the one image for searching for lane candidate points. The lane detection device according to claim 2, wherein the detection lane width threshold is set to an enlarged value in a search area farther than the predetermined distance ahead of the host vehicle. .   The lane detection device according to claim 3, wherein the detection unit shortens the predetermined distance according to a travel distance of the host vehicle. Highway running detection means for detecting whether or not the host vehicle is running on a highway,
The detecting means can detect a sign that is wider than the standard of one lane width for the detected lane width threshold when the highway running detecting means detects that the host vehicle is running on the highway. The lane detection device according to any one of claims 1 to 4, wherein the lane detection device expands to a large value. Pedestrian crossing detection means for detecting whether there is a pedestrian crossing in front of the vehicle,
The detection means does not detect that the host vehicle is traveling on an expressway by the expressway running detection means, and if the pedestrian crossing is detected by the pedestrian crossing detection means, the detection lane width threshold is set. 6. The lane detection device according to claim 5, wherein the lane detection device is set to a value corresponding to the standard of the one lane width.   The detecting means detects that the own vehicle is not traveling on the highway by the highway travel detecting means, and if the pedestrian crossing is not detected by the pedestrian crossing detecting means, When there is a high possibility that the host vehicle is driving on an expressway based on the distance between the left and right lane positions of the vehicle, the detection lane width threshold can be detected to be wider than one lane width standard. The lane detection device according to claim 6, wherein the lane detection device expands to a large value.   When the detection means detects a sign wider than a standard of one lane width continuously for a certain distance or more, the detection means expands the detection lane width threshold to a value capable of detecting the wide sign. The lane detection device according to claim 1, wherein the lane detection device is a lane detection device.   The lane detection device according to claim 8, wherein the certain distance is a distance greater than or equal to a length of a pedestrian crossing in the traveling direction of the host vehicle.

Images