JP6230482B2 - Lane mark recognition device, vehicle, and lane mark recognition method - Google Patents

Description

The present invention relates to a lane mark recognition apparatus , a vehicle, and a lane mark recognition method for recognizing a lane mark provided on a road based on a captured image of a camera mounted on the vehicle .

  2. Description of the Related Art Conventionally, a lane mark recognition device that recognizes a lane mark by identifying the color of a lane mark provided on the road from an image of a road ahead of the vehicle captured by an in-vehicle camera is known (for example, a patent) Reference 1).

  The lane mark recognition device described in Patent Document 1 uses a chromatic color light reflected on a lane mark by using a difference between an RGB value and a brightness value as a correction value for a road portion in a captured image. Has reduced the impact.

  More specifically, the lane mark recognition device described in Patent Document 1 averages the average values Rmean, Gmean, and Bmean of the RGB signals of the road area AR1 input from the imaging device, The brightness value Gray of the road area is calculated.

Gray = (Rmean + Gmean + Bmean) / 3
Then, the lane mark recognition device is a color component (Roff, Goff, Boff) of the road area AR1.
(Roff, Goff, Boff) = (Gray-Rmean, Gray-Gmean, Gray-Bmean)
Is calculated.

  This value is a correction value for reducing the influence of chromatic light emitted from a road lamp or the like from a captured image when the road area AR1 is assumed to be achromatic (Gray = Rmean = Gmean = Bmean). is there.

  The lane mark recognition apparatus adds the color components (Roff, Goff, Boff) of the road area AR1 obtained by the above formula to the signal values (Rsource, Gsource, Bsource) of each pixel of the input image signal, thereby The image signal values (Rmodify, Gmodify, Bmodify) corrected as follows are obtained.

(Rmodify, Gmodify, Bmodify) = (Rsource + Roff, Gsource + Goff, Bsource + Boff)
This value is an RGB value indicating a color such as a road, a white line, or a yellow line when chromatic light derived from a road lamp or the like is not irradiated. The lane mark recognition apparatus is configured to recognize a lane mark by detecting a white line and a yellow line based on the RGB values.

Japanese Patent No. 4526963

  However, the technique of Patent Document 1 reduces the influence caused by irradiating achromatic roads with chromatic light derived from road lights or the like. For this reason, when there is an influence other than the irradiation of chromatic light, for example, the influence of shadow, the correction value is calculated from the difference between the RGB signal value of each color and the brightness for the road portion of the technique of Patent Document 1. It is difficult to do. As a result, with the technique of Patent Document 1, it is difficult to accurately recognize the lane mark when there is an influence of a shadow.

The present invention has been made in view of such a background, and provides a lane mark recognition device , a vehicle, and a lane mark recognition method capable of improving the detection accuracy of a lane mark even when there is an influence of a shadow. Objective.

  The present invention has been made to achieve the above object, and recognizes a lane mark provided on a road based on an image around the vehicle imaged by an RGB color camera mounted on the vehicle. The present invention relates to a mark recognition device, a vehicle, and a lane mark recognition method.

And the lane mark recognition device of the present invention,
A shadow determination unit that determines whether the region is a shadow region based on the RGB values of the captured image of the color camera;
For the previous SL IMAGING image, of a region of interest in the road region is estimated to be the image portion of the road, the target shadow region is the target region is determined to be a region of the shadow by the shadow determining unit The value of B of the comparison shadow region that is the comparison region that is determined to be a shadow region by the shadow determination unit among the comparison regions in the road region that includes the image portion of the road other than the lane mark. The requirement is that the value obtained by subtracting the value of B from the R value of the target shadow region is greater than the value obtained by subtracting the value of B from the R value of the comparative shadow region. the color of the target shadow region, characterized in that it comprises a color determining unit configured to regard yellow.

  According to the study of the inventor of the present invention, the yellow lane mark in the road region has a B value smaller than the B value in the region including the image portion of the road other than the yellow lane mark or the R value. In many cases, the value obtained by subtracting the B value from the value is larger than the value obtained by subtracting the B value from the R value of the road image portion other than the yellow lane mark.

  According to the lane mark recognizing device of the present invention constructed focusing on this point, the value of B in the target area is smaller than the value of B in the comparison area, and the value of B is subtracted from the value of R in the target area. Since the color of the target area is regarded as yellow on the condition that at least one of the obtained value is larger than the value obtained by subtracting the B value from the R value in the comparison area, a yellow lane mark can be detected accurately. it can.

  As a result, according to the lane mark recognition apparatus of the present invention, the detection accuracy of the yellow lane mark can be improved even when there is an influence of a shadow.

The lane mark recognition apparatus of the present invention is
B of the comparison area in the road area where the value of B of the target area in the road area estimated to be the image part of the road includes the image part of the road other than the lane mark with respect to the captured image of the color camera And the target area is required to be at least one of a value obtained by subtracting the B value from the R value in the target area and a value obtained by subtracting the B value from the R value in the comparison area. A color determination unit configured to regard the color of the region as yellow;
A shadow determination unit that determines whether the region is a shadow region based on the RGB values of the captured image ;
With
In the color determination unit, the value of B of the target shadow region that is the target region that is determined to be a shadow region by the shadow determination unit in the target region is determined by the shadow determination unit in the comparison region. The value obtained by subtracting the B value from the R value of the target shadow area is smaller than the B value of the comparative shadow area that is the comparison area determined to be the R area, and the R value of the comparative shadow area It is characterized in that the color of the target shadow area is regarded as yellow, with at least one of being larger than the value obtained by subtracting the value of B from.

  According to the lane mark recognition apparatus having the configuration, the target shadow area that is the target area determined as the shadow area is compared with the comparative shadow area that is the comparison area determined as the shadow area. The color of is determined. As a result, the yellow lane mark is detected with high accuracy even in the shadow area.

In the lane mark recognition apparatus of the present invention,
It is preferable to include a candidate image extraction unit that recognizes a candidate image of a lane mark based on an image portion determined to be yellow by the color determination unit.

  As described above, the color of the image portion of the lane mark in the non-shadow region and the color of the image portion of the lane mark in the shadow region are often different. Even if the color of the image portion of the lane mark in the area is different from the assumed color (yellow), if the color of the image portion is determined to be yellow by the color determination unit, the color is determined based on the image portion. The candidate image portion of the lane mark is recognized. As a result, the lane mark detection accuracy can be improved.

The vehicle of the present invention
A shadow determination unit that determines whether the region is a shadow region based on the RGB values of the captured image of the color camera;
For the previous SL IMAGING image, of a region of interest in the road region is estimated to be the image portion of the road, the target shadow region is the target region is determined to be a region of the shadow by the shadow determining unit The value of B of the comparison shadow region that is the comparison region that is determined to be a shadow region by the shadow determination unit among the comparison regions in the road region that includes the image portion of the road other than the lane mark. as at least one requirement of the value obtained by subtracting the value from the value of R B of less than the value and the target shadow region is greater than the value obtained by subtracting the value from the value of R B of the comparison shadow region, the the color of the target shadow region, characterized in that it comprises a color determining unit configured to regard yellow.

  According to the vehicle of the present invention, the same operational effects as those of the lane mark recognition apparatus of the present invention described above can be obtained.

In the vehicle having the configuration described above, a lane mark recognition unit configured to recognize a yellow lane mark based on an image portion determined to be yellow by the color determination unit;
It is preferable to include a steering control unit that performs steering control of the vehicle so as not to exceed the yellow lane mark recognized by the lane mark recognition unit.

  According to the vehicle of the configuration, when the yellow lane mark instructs to prohibit the lane change, it is possible to perform the steering control according to this instruction and to keep the vehicle in the lane partitioned by the lane mark. Become.

The lane mark recognition method of the present invention is
A shadow determination step for determining whether the region is a shadow region based on the RGB values of the image captured by the color camera;
For the previous SL IMAGING image, of a region of interest in the road region is estimated to be the image portion of the road, the target shadow region is the target region is determined to be a region of the shadow by the shadow determining step The value of B of the comparison shadow region that is the comparison region that has been determined to be a shadow region by the shadow determination step among the comparison regions in the road region including the image portion of the road other than the lane mark. The requirement is that the value is smaller than the value and the value obtained by subtracting the B value from the R value in the target shadow region is greater than the value obtained by subtracting the B value from the R value in the comparative shadow region , the color of the target shadow region, characterized in that it comprises a color judgment step that is configured regarded as yellow.

  By applying the lane mark recognition method of the present invention to a vehicle and implementing it, the same operational effects as those of the lane mark recognition device of the present invention described above can be obtained.

The block diagram of a lane mark recognition apparatus. The flowchart of a lane mark recognition process. It is a figure explaining a captured image and an edge image, (a) is a figure explaining a captured image, (b) is a figure explaining an edge image. The figure explaining the relationship between recognition of a lane mark and an obstruction, and steering control. The block diagram of the lane mark recognition apparatus of other embodiment. The flowchart of the lane mark recognition process of other embodiment.

  An embodiment of a lane mark recognition apparatus of the present invention will be described with reference to FIGS.

(Configuration of lane mark recognition device)
Referring to FIG. 1, a lane mark recognizing device 10 is applied to a vehicle 1 (corresponding to a vehicle according to the present invention) provided with a camera 2 (RGB color camera), a speaker 5, a display 6, and a steering mechanism 7. It is installed.

  The lane mark recognition device 10 is an electronic unit that includes a CPU, a memory, various interface circuits, and the like (not shown), and a lane mark recognition and steering control program held in the memory is executed by the CPU, thereby taking a captured image. It functions as an acquisition unit 11, an edge image generation unit 12, a color determination unit 14, a candidate image part extraction unit 15, a lane mark recognition unit 16 and a steering control unit 17. Further, the lane mark recognition apparatus 10 implements the lane mark recognition method of the present invention.

(Lane mark recognition process)
Hereinafter, processing for recognizing a lane mark provided on a road by the lane mark recognition apparatus 10 will be described with reference to the flowchart shown in FIG. The lane mark recognition device 10 executes processing according to the flowchart shown in FIG. 2 for each predetermined control period, and recognizes the lane mark (continuous line lane mark and broken line lane mark) of the road on which the vehicle 1 is traveling. To do.

  FIG. 2 / STEP 1 is processing by the captured image acquisition unit 11. The captured image acquisition unit 11 inputs a video signal output from the camera 2 in front of the vehicle 1 (corresponding to the periphery of the vehicle of the present invention), and color components (R value, G value, B value) of the video signal. ) To obtain a color captured image 21 having R value, G value, and B value as data of each pixel. The acquired data of the captured image 21 is held in the image memory 20.

  Here, Im1 shown in FIG. 3A is an example of an image in which the captured image 21 is shown in black and white, and the image portion of the left solid line lane mark that divides the lane in which the vehicle 1 (own vehicle) is traveling. 51a (yellow line), right broken line lane mark image portions 61a to 64a (white line), and solid line lane mark image portion 71a (white line) are included. Im1 also includes a shadow area 81a extending to the near side area of the left solid line lane mark image portion 51a (area close to the vehicle 1 (camera 2) in real space).

  In the captured image Im1, among the image portion 51a of the yellow lane mark, the RGB values of the pixel P1 not included in the shadow area 81a are R = 200, G = 150, and B = 30, and the color of the pixel P1 is yellow. It is. In the image portion 51a of the yellow lane mark, the RGB values of the pixel P2 included in the shadow area 81a are R = 60, G = 75, and B = 60, and the color of the pixel P2 is different from yellow. In addition, among the portions included in the shadow area 81a, the RGB values of the road pixel P3 are R = 45, G = 68, and B = 70, and the color of the pixel P3 is different from yellow.

  The subsequent FIG. 2 / STEP 3 to FIG. 2 / STEP 6 are processes by the color determination unit 14. In addition, the process process by the color determination part 14 is corresponded to the color determination process of the lane mark recognition method of this invention.

  The color determination unit 14 selects one pixel of the road area (corresponding to the “target area” of the present invention) among the pixels included in the captured image 21 (FIG. 2 / STEP 3). The road region is a region on a captured image corresponding to a road in real space, and is defined as a region having a predetermined lateral width with the center in the left-right direction of the vehicle 1 being the center in the lateral direction of the road region. The predetermined lateral width is a width that includes a road and a lane mark determined from the lateral width of the vehicle 1 and the legal road width.

  Also, for example, the color determination unit 14 extracts pixels having different RGB values from the surroundings from the captured image before the processing of FIG. 2 / STEP 3, and the extracted pixels constitute a part of the image portion of the lane mark. (Whether or not the extracted pixel constitutes a part of an image portion of a linear or curved object) may be determined. In this case, the color determination unit 14 selects one of the extracted pixels in FIG. 2 / STEP3 only when the extracted pixel constitutes the image portion of the lane mark. 2 / STEP5B processing is executed.

The color determination unit 14 determines that the value T _R−B obtained by subtracting the B value from the R value of the selected pixel (hereinafter referred to as “target pixel”) is the R value of the comparison area including the road image portion other than the lane mark. It is determined whether or not it is larger than a value CR _−B obtained by subtracting the B value from FIG. 2 (STEP 4A).

  Here, the comparison region including the image portion of the road other than the lane mark includes a pixel having a high probability of becoming the image portion of the road other than the lane mark when the target pixel is the pixel of the image portion of the lane mark. An area.

  More specifically, an area including a point that is separated from the target pixel by a predetermined distance (for example, 1 m) in the real space in the horizontal direction from the target pixel toward the center of the vehicle 1 can be adopted as the comparison area. For example, as shown in FIG. 3A, when the target pixel is P2, an area including a pixel P3 that exists at a predetermined distance on the center side of the vehicle 1 (camera 2) in the horizontal direction of the target pixel P2. Can be adopted as a comparison region.

In this case, for the target pixel P2, the color determination unit 14 determines that the value TR _−B obtained by subtracting the B value from the R value of the target pixel P2 is 0, whereas the value T _R−B is 0 from the R value of the pixel P3 as the comparison region. Since the value obtained by subtracting the value CR _−B is −25, the determination result is a positive result.

  Further, as a comparison region, a region having a predetermined range centered on the target pixel (for example, a size range corresponding to a circle having a radius of 50 cm in real space) can be adopted.

A plurality of pixels are selected as the comparison region, and the minimum value, median value, average value, etc., of the values obtained by subtracting the B value from the R value among these pixels may be used as the comparison threshold value CR _{-B in} FIG. Good.

  When the determination result of FIG. 2 / STEP 4A is affirmative (FIG. 2 / STEP 4A... YES, for example, when the determination result for the pixel P2 is affirmative), the color determination unit 14 determines that the color of the target pixel is yellow ( FIG. 2 / STEP5A). On the other hand, when the determination result of FIG. 2 / STEP4A is negative (FIG. 2 / STEP4A... NO, for example, when the determination result for the pixel P3 is negative), the color determination unit 14 sets the color of the target pixel to a color other than yellow. The color is determined (FIG. 2 / STEP5B). As the color other than yellow, for example, the color of a road portion that is not a shadow area is determined as the color of the target pixel.

  Further, the color determination unit 14 determines whether or not the RGB value of the target pixel is a yellow RGB value before the processing of FIG. 2 / STEP 4A, and when the RGB value of the target pixel is not a yellow RGB value. As long as the processing shown in FIG. 2 / STEP4A is performed. Thereby, the color determination part 14 can extract a yellow image part more reliably. Note that whether or not the RGB value of the target pixel is a yellow RGB value is determined based on a predetermined range of R, B, and G values indicating that the RGB value of the target pixel is yellow and its similar colors (for example, R It is determined by whether or not the value is included in a range of 150 to 256, a G value of 150 to 256, and a B value of 0 to 50.

  In addition, when the color determination unit 14 determines that the RGB value of the target pixel is not a yellow RGB value in the above process, the color determination unit 14 refers to the RGB values of the pixels around the target pixel before the process of FIG. Then, only when the value obtained by subtracting the B value from the R value of the surrounding pixels is lower than the predetermined value, the process of FIG. 2 / STEP 4A may be performed. Here, the RGB values of the pixels around the target pixel are the average value, median value, and maximum value of the RGB values of one pixel around the target pixel or the RGB values of a plurality of pixels around the target pixel. Or it means the lowest value.

  After any one of the processes of FIG. 2 / STEP 5A to FIG. 2 / STEP 5B, the color determination unit 14 determines whether or not the processes of FIG. 2 / STEP 3 to FIG. 2 / STEP6).

  If the determination result is affirmative (FIG. 2 / STEP6... YES), the edge image generation unit 12 performs the processing of STEP7.

  When the determination result is negative (FIG. 2 / STEP 6... NO), the color determination unit 14 executes the process from STEP 3 for the next target pixel.

  The edge image generation unit 12 generates a yellow line feature image obtained by extracting only the pixels determined to be yellow from the captured image 21, and the color of each pixel of the yellow line feature image and the pixels that are not determined to be yellow. A process of converting the component into luminance is performed to generate a grayscale image (multi-valued image). Then, the edge image generation unit 12 is a pixel in which the luminance difference (luminance change amount) from the edge point (the surrounding pixel (image portion)) is a predetermined value or more from the grayscale image. The luminance changes from dark to bright. The edge image 22 (see FIG. 1) is generated by extracting positive edge points and negative edge points whose luminance changes from light to dark (FIG. 2 / STEP 7).

  Im2 shown in FIG. 3B is an edge image generated by performing edge extraction processing on the grayscale image. In the edge image Im2, the left solid line lane mark image portion 51c has an edge point between the shadow region 81a and the left solid line lane mark portion not included in the shadow region 81a and the shadow region 81a. Extraction is performed so that no edge point exists between the left solid line lane mark portion. The edge image Im2 includes many image points 51c of the left solid line lane mark, image parts 61c to 64c of the right broken line lane mark, and an image part 71c of the solid line lane mark, and many edge points that are linearly continuous. It becomes the image part. Further, the shadow area 81c is an area surrounded by edge points.

  The subsequent FIG. 2 / STEP 8 to FIG. 2 / STEP 10 are processes performed by the candidate image portion extraction unit 15. The candidate image portion extraction unit 15 labels the edge image Im2 with a label on a group of edge points (hereinafter referred to as edge portions) and the adjacent edge portions as image portions of the same object. Perform clustering to associate as there is.

  Further, the candidate image portion extraction unit 15 in FIG. 2 / STEP9, among the image portions extracted by clustering, has an image portion having a line characteristic (for example, the ratio of linearly continuous edge points is a predetermined level or more). A certain image portion) is extracted as a candidate image portion that is a candidate for the image portion of the lane mark.

  The candidate image portion extraction unit 15 performs reverse projection conversion from the camera coordinates to the real space coordinates for each candidate image portion extracted in FIG. 2 / STEP 9, in the real space corresponding to each candidate image portion. The position of the object is calculated (FIG. 2 / STEP 10). Note that an image portion whose size is equal to or smaller than a predetermined size (a size that is not assumed to be an image portion of a lane mark) may be excluded from the candidate image portions.

  Next, FIG. 2 / STEP 11 to FIG. 2 / STEP 12 are processes by the lane mark recognition unit 16. The processing step by the lane mark recognition unit 16 corresponds to the lane mark recognition step of the lane mark recognition method of the present invention.

  The lane mark recognition unit 16 has a length in the corresponding real space obtained in FIG. 2 / STEP 10 within a predetermined range (set based on the length (specified length) of the broken lane mark defined by law). In this case, a candidate for a broken line lane mark is detected based on candidate image portions that are continuous in a certain direction.

  The lane mark recognizing unit 16 uses a known method described in, for example, Japanese Patent Laid-Open No. 11-219435 when the lane is divided by continuous line lane marks (white line, yellow line, etc.). Detect candidates for.

  The lane mark recognition unit 16 recognizes the lane mark based on these lane mark candidates (FIG. 2 / STEP 11).

  For example, as shown in FIG. 4, the left lane mark 51b is recognized on the left side of the lane in which the vehicle 1 is traveling, and the right lane marks 61b to 64b, 71b are recognized on the right side.

  In subsequent FIG. 2 / STEP 12, the lane mark recognition unit 16 recognizes the center position of the lane based on the positions of the left and right lane marks. In the example of FIG. 4, the center position Xc between the inner position XL1 of the left lane mark 51b on the left side and the inner position XR1 of the right lane marks 61b to 64b on the right side is recognized.

  Next, FIG. 2 / STEP 13 is processing by the steering control unit 17. The steering controller 17 assists the driving operation of the driver of the vehicle 1 by operating the steering mechanism 7 so that the vehicle 1 travels near the center position Xc in the lane. In addition, the steering control unit 17 alerts the driver of the vehicle 1 with at least one of the speaker 5 and the display 6 as necessary.

  Here, the steering controller 17 increases the deviation between the vehicle 1 and the center position Xc, and the interval WR between the vehicle 1 and the right lane marks 61b to 64b or the interval WL between the vehicle 1 and the left lane mark 51b is increased. When the value is equal to or less than a predetermined determination value, the steering mechanism 7 is operated to control the vehicle 1 closer to the center position Xc.

  In addition, the steering control unit 17 uses different determination values for the colors of the left and right lane marks for the interval WR between the vehicle 1 and the right lane marks 61b to 64b or the interval WL between the vehicle 1 and the left lane mark 51b. Also good. For example, when the left lane mark 51b is a yellow line and the right lane marks 61b to 64b are white lines, the determination value of the interval WL between the vehicle 1 and the left lane mark 51b is set as the vehicle 1 and the right lane marks 61b to 64b. The interval WL may be set smaller than the determination value. By doing so, it is possible to perform the steering control so that the vehicle 1 does not deviate from the left lane mark 51b compared to the right lane mark.

(Operation and effect of this embodiment)
According to the inventor's study, the yellow lane mark in the road area is obtained by subtracting the B value from the R value of the image portion of the road other than the lane mark. Often greater than

  According to the lane mark recognizing device of the present invention configured by paying attention to this point, the value (0) obtained by subtracting the value of B from the value of R of the target region (pixel P2) is the value of the comparison region (pixel P3). Since the color of the target area (pixel P2) is regarded as yellow on the condition that it is at least one greater than the value obtained by subtracting the value of B from the value of R (−25), it is possible to detect a yellow lane mark. it can.

  As described above, the color of the image portion (pixel P1) of the lane mark in the non-shadow region is often different from the color of the image portion (pixel P2) of the lane mark in the shadow region 81a. According to the mark recognition device 10, even if the color of the image portion (pixel P2) of the lane mark in the shadow area 81a is different from the assumed color (yellow), the color of the image portion (pixel P2) is the same. When the color determination unit determines that the color is yellow, the lane mark candidate image portion 51c is recognized based on the image portion (pixel P2). As a result, the lane mark detection accuracy can be improved.

  According to the lane mark recognizing device 10 having the configuration, the steering of the vehicle is controlled in a different manner between the yellow lane mark 51b and the white lane marks 61b to 64b. As a result, it is possible to perform steering control according to the significance of each lane mark.

(Modification)
As illustrated in FIG. 5, the lane mark recognition apparatus 10 may include a shadow determination unit 13. In the following, description of the same configuration and processing as those of the above-described embodiment will be omitted, and configuration or processing different from the above-described embodiment will be described.

  In this case, before FIG. 6 / STEP3, the shadow determination unit 13 determines whether each image part of the captured image 21 is an image part included in the shadow region by determining the brightness level of the other image parts. It is determined whether or not (FIG. 6 / STEP2). An image portion whose luminance is lower than a value obtained by subtracting a predetermined luminance from the average luminance of other image portions is determined as an image portion included in the shadow area.

  Note that the luminance L (P) of each part is obtained by the following equation, for example.

L (P) = 0.3 * R + 0.6 * G + 0.1 * B
For example, in FIG. 3A, if the value obtained by subtracting a predetermined luminance from the average luminance of other image portions is 100, the pixel P1 has an average luminance (L (P1) = 153) of other image portions. Since it is higher than the value obtained by subtracting the predetermined luminance from the luminance (100), it is not determined as a shadow region.

  On the other hand, since the brightness of pixels P2 and P3 (L (P2) = 69, L (P3) = about 61) is lower than the value (100) obtained by subtracting the predetermined brightness from the average brightness of other image portions, It is determined as an area.

  In this case, after the color determination unit 14 selects one pixel (FIG. 6 / STEP 3), the color determination unit 14 determines the pixel (referred to as “ It corresponds to the “target shadow region”) (FIG. 6 / STEP 4B). For example, the determination result is negative for the pixel P1 not included in the shadow area 81a, while the determination result is positive for the pixel P2 and the pixel P3 included in the shadow area 81a.

  When the determination result in FIG. 6 / STEP4B is negative (FIG. 6 / STEP4B... NO, for example, when the determination result for the target pixel P1 is negative), the color determination unit 14 determines the color corresponding to the RGB value of the target pixel. Is determined as the color of the pixel (FIG. 6 / STEP5C). For example, the color determination unit 14 determines that the color of the target pixel P1 not included in the shadow area 81a is yellow based on the RGB value.

When the determination result of FIG. 6 / STEP4B is affirmative (FIG. 6 / STEP4B... YES, for example, when the determination result for the target pixels P2 and P3 is affirmative), the color determination unit 14 determines the target pixel (target region). Whether the value T _{R−B obtained} by subtracting the B value from the R value is greater than the value C _R−B obtained by subtracting the B value from the R value of the comparative shadow region including the image portion of the road other than the lane mark, or yellow It is determined whether or not (FIG. 6 / STEP4C).

  Here, the comparative shadow region including the image portion of the road other than the lane mark is the image portion of the road other than the lane mark when the target pixel is a pixel of the image portion of the lane mark in the shadow region. An area including a pixel having a high probability.

  More specifically, as the comparative shadow area, an area including a point away from the target pixel in the horizontal direction toward the center of the vehicle 1 by a predetermined distance (for example, 1 m) in the real space can be adopted as the comparative shadow area.

  In addition, as a comparative shadow region, a region within a predetermined range (for example, a range corresponding to a circle with a radius of 50 cm in real space) out of the shadow region can be employed.

When a plurality of pixels are designated as the comparative shadow region, the minimum value, median value, average value, etc. of the values obtained by subtracting the B value from the R value of those pixels are the comparison threshold value T in FIG. 6 / STEP4C. You may employ _| adopt as _RB .

  According to the lane mark recognition apparatus 10 having the configuration, the target shadow area (pixel P2) that is the target area determined to be the shadow area 81a is the comparative shadow area (the comparison shadow area that is determined to be the shadow area 81a). The color of the target shadow area (pixel P2) is determined by comparison with the pixel P3). As a result, the yellow lane mark is detected with high accuracy even in the shadow area.

Further, instead of or in addition to the determination of the value TR _−B obtained by subtracting the B value from the R value in FIG. 2 / STEP 4A or FIG. 6 / STEP 4C, the value of B of the target pixel is the value of B in the comparison region or the comparison shadow region. If the value is smaller than the value, the target pixel may be determined to be yellow.

  In this case, when a plurality of pixels are designated as the comparison region or the comparison shadow region, the highest value, median value, average value, etc., of the B values among these pixels may be used as the comparison threshold value for the determination.

  In addition, although one pixel is set as a target pixel (target region), a plurality of pixels within a predetermined range may be adopted as the target region instead. In this case, the lowest value, median value, average value, etc. of the values obtained by subtracting the B value from the R value, or the highest value, median value, average value, etc. of the B value are compared with the comparison value in the above determination. May be.

  Further, in FIG. 2 / STEP 7 (FIG. 6 / STEP 7), instead of creating a yellow line feature image, in FIG. 2 / STEP 9 (FIG. 6 / STEP 9), the candidate image partial extraction unit 15 includes a plurality of candidate images. Even if the RGB values are different, the color determination unit 14 recognizes a plurality of adjacent candidate images as candidate images of one lane mark on the condition that the candidate images are determined to be yellow. May be configured.

  DESCRIPTION OF SYMBOLS 1 ... Vehicle (own vehicle), 2 ... Camera, 7 ... Steering mechanism, 10 ... Lane mark recognition apparatus, 11 ... Captured image acquisition part, 12 ... Edge image generation part, 13 ... Shadow determination part, 14 ... Color determination part, DESCRIPTION OF SYMBOLS 15 ... Candidate image part extraction part, 16 ... Lane mark recognition part, 17 ... Steering control part, 20 ... Image memory, 21 ... Captured image, 22 ... Edge image.

Claims (6)

A lane mark recognition device for recognizing a lane mark provided on a road based on an image around the vehicle imaged by an RGB color camera mounted on the vehicle,
A shadow determination unit that determines whether the region is a shadow region based on the RGB values of the captured image of the color camera;
For the previous SL IMAGING image, of a region of interest in the road region is estimated to be the image portion of the road, the target shadow region is the target region is determined to be a region of the shadow by the shadow determining unit The value of B of the comparison shadow region that is the comparison region that is determined to be a shadow region by the shadow determination unit among the comparison regions in the road region that includes the image portion of the road other than the lane mark. The requirement is that the value obtained by subtracting the value of B from the R value of the target shadow region is greater than the value obtained by subtracting the value of B from the R value of the comparative shadow region. lane mark recognition apparatus characterized by comprising a color determination unit that the color of the target shadow region are configured to regarded as yellow. A lane mark recognition device for recognizing a lane mark provided on a road based on an image around the vehicle imaged by an RGB color camera mounted on the vehicle,
B of the comparison area in the road area where the value of B of the target area in the road area estimated to be the image part of the road includes the image part of the road other than the lane mark with respect to the captured image of the color camera And the target area is required to be at least one of a value obtained by subtracting the B value from the R value in the target area and a value obtained by subtracting the B value from the R value in the comparison area. A color determination unit configured to regard the color of the region as yellow;
A shadow determination unit that determines whether the region is a shadow region based on the RGB values of the captured image;
With
In the color determination unit, the value of B of the target shadow region that is the target region that is determined to be a shadow region by the shadow determination unit in the target region is determined by the shadow determination unit in the comparison region. The value obtained by subtracting the B value from the R value of the target shadow area is smaller than the B value of the comparative shadow area that is the comparison area determined to be the R area, and the R value of the comparative shadow area A lane mark recognizing device configured to regard the color of the target shadow region as yellow, with at least one of being larger than a value obtained by subtracting the value of B from A. In the lane mark recognition apparatus according to claim 1 or 2,
A lane mark recognition apparatus comprising: a candidate image extraction unit that recognizes a candidate image of a lane mark based on an image portion determined to be yellow by the color determination unit. A vehicle that recognizes a lane mark provided on a road based on an image around the vehicle imaged by an RGB color camera mounted on the vehicle,
A shadow determination unit that determines whether the region is a shadow region based on the RGB values of the captured image of the color camera;
For the previous SL IMAGING image, of a region of interest in the road region is estimated to be the image portion of the road, the target shadow region is the target region is determined to be a region of the shadow by the shadow determining unit The value of B of the comparison shadow region that is the comparison region that is determined to be a shadow region by the shadow determination unit among the comparison regions in the road region that includes the image portion of the road other than the lane mark. as at least one requirement of the value obtained by subtracting the value from the value of R B of less than the value and the target shadow region is greater than the value obtained by subtracting the value from the value of R B of the comparison shadow region, the vehicle; and a color determination unit that the color of the target shadow region are configured to regarded as yellow. The vehicle according to claim 4, wherein
A lane mark recognizing unit configured to recognize a yellow lane mark based on an image portion determined to be yellow by the color determining unit;
A vehicle comprising: a steering control unit that performs steering control of the vehicle so as not to exceed a yellow lane mark recognized by the lane mark recognition unit. A lane mark recognition method for recognizing a lane mark provided on a road based on an image around the vehicle captured by an RGB color camera mounted on the vehicle,
A shadow determination step for determining whether the region is a shadow region based on the RGB values of the image captured by the color camera;
For the previous SL IMAGING image, of a region of interest in the road region is estimated to be the image portion of the road, the target shadow region is said target area which is determined to be shadow areas in the shadow determination step The value of B of the comparison shadow region that is the comparison region determined to be a shadow region in the shadow determination step among the comparison regions in the road region including the road image portion other than the lane mark is The requirement is that the value is smaller than the value and the value obtained by subtracting the B value from the R value in the target shadow region is greater than the value obtained by subtracting the B value from the R value in the comparative shadow region , lane mark recognition method which comprises a color determination step of the color of the target shadow region are configured to regarded as yellow.