JP4941843B2 - Road marking image processing apparatus, road marking image processing method, and road marking image processing program - Google Patents

Description

  The present invention relates to an apparatus for recognizing road markings drawn on a road surface using a camera installed in a vehicle.

  As a method for recognizing an achromatic object from an image, such as a white road marking painted on asphalt or concrete, Patent Document 1 discloses a technique for detecting a white line using a luminance image. Yes. In this way, when an achromatic object is recognized from an image, a luminance image is generally used. When the image input device is a color camera or the like, a luminance image is generated using the following equation for each pixel. Can do. The calculation of the luminance Y is also disclosed in Patent Document 2.

  It is also generally known to use an image having a green component value that contributes most to luminance as an image value (hereinafter referred to as a green component image) without performing luminance conversion.

  However, in the luminance image and the green component image, it may be difficult to separate the road marking from the pavement surface. A specific example is shown using FIG. FIG. 2 is a diagram illustrating a situation in which the shade 203 is superimposed on a part of the road marking. At this time, in the luminance image or the green component image, the distribution of pixel values corresponding to the paved road surface 201 existing in the sun and the road marking 202 existing in the shade often overlaps as shown in FIG. . In such a case, it is possible to correctly recognize the road marking by a method of performing binarization processing and recognizing that the pixel value corresponding to the road marking is larger than the pixel value corresponding to the paved road surface. Have difficulty.

  Therefore, as shown in FIG. 2, as a conventional method for obtaining a relatively good binary image even in a situation where the intensity of illumination light varies depending on the part, a local binary as disclosed in Patent Document 3 is used. A method using a chemical method is known. In the local binarization method, a threshold value for binarization is obtained for each partial region, and after smoothing the value, binarization is performed with a different threshold value for each partial region. Since it is based on the relative magnitude relationship between the luminance values in the local area, it is easy to extract the road markings without being influenced by the difference in luminance between the shade and the sun that exist at distant locations.

  Patent Document 4 discloses a method for recognizing a yellow line and a white line by scattered light.

JP 2000-331148 A, FIG. Japanese Patent Laid-Open No. 2001-218070 Japanese Patent Laid-Open No. 7-093689, FIG. JP 2006-338556 A

  However, since both the shade and the sun are included in the local area at the boundary between the shade and the sun, there is a problem that it is easy to extract the sun area as an area corresponding to the road marking. In particular, when the sun and shade are mixed with a fine period like a tree shade, it becomes easy to determine the sun area as a road marking, and it becomes difficult to correctly separate the paved road surface and the road marking.

  Therefore, in order to solve the above-described problem, the object of the present invention is to provide a road marking that exists in the shade even when the shade and the sun are superimposed on the road marking, compared to the case where the luminance image and the green component image are used. An object of the present invention is to provide a road marking recognition device that converts a pixel value of a corresponding pixel into a value larger than a pixel value of a pixel corresponding to a paved road surface existing in the sun.

  In order to solve the above problems, a road marking image processing apparatus according to the present invention is a color image acquisition function block that captures a color image of a road surface in a road marking recognition apparatus installed in the host vehicle so as to capture the road surface. A luminance image generation functional block that generates a luminance image based on the color image acquired by the color image acquisition functional block, and a blue component by calculating the intensity of the blue component included in the color image for each pixel. By adding the blue component intensity image generation function block for generating the intensity image, and the luminance image and the blue component intensity image, the difference between the pixel values corresponding to the sunny and shaded areas is smaller than that of the luminance image. A shade correction image generation function block that generates a shade correction image, and a road marking recognition function block that recognizes a road marking from the shade correction image, That.

  In order to achieve the above object, a road marking recognition method according to the present invention includes a color image acquisition function in a road marking recognition method using a road marking recognition device installed in the host vehicle so as to photograph a road surface. The block is a step of photographing a color image of a road surface, the luminance image generation function block is a step of generating a luminance image based on the color image acquired by the color image acquisition unit, and the blue component intensity image generation function block is The step of generating a blue component intensity image by calculating the intensity of the blue component included in the color image for each pixel, and the shade correction image generation function block, the luminance image and the blue component intensity image By adding, a step of generating a shade-corrected image in which a difference between pixel values corresponding to the sunny and shaded areas is smaller than that of the luminance image, and a road marking recognition function block, Characterized in that the serial shade corrected image and a step of recognizing the road marking.

  Furthermore, in order to achieve the above object, a road marking recognition program according to the present invention is a color marking acquisition function in a road marking recognition program that uses a road marking recognition device installed in a host vehicle so as to photograph a road surface. The block is a step of photographing a color image of a road surface, the luminance image generation function block is a step of generating a luminance image based on the color image acquired by the color image acquisition unit, and the blue component intensity image generation function block is The step of generating a blue component intensity image by calculating the intensity of the blue component included in the color image for each pixel, and the shade correction image generation function block, the luminance image and the blue component intensity image A step of generating a shade-corrected image with a smaller difference in pixel values corresponding to the sunlit and shaded regions than the luminance image by adding, and a road marking recognizer Block, characterized in that it comprises a step of recognizing the road marking from the shade corrected image.

  This object can also be achieved by a road marking recognition method corresponding to the road marking recognition apparatus having the above configuration.

  The object is also achieved by a road marking recognition method having the above-described configurations and a computer program for realizing the corresponding device by a computer, and a computer-readable storage medium storing the computer program. Achieved.

  According to the present invention, a larger value is added to the luminance image, the green component image, or the blue component image as the pixel value of the blue component of the input color image is larger than the pixel value of the red component or the green component. Generate images and recognize road markings. The shade is a region where light from the sun, which is the light source, is not directly applied, but it is usually bright because scattered light and reflected light are incident. There are many features. Therefore, assuming that there are only achromatic regions in the image, even in two regions where the luminance and green component values are almost equal, the region where the ratio of the blue component to the red component and the green component is large is It exists in the area | region where incident light quantity is comparatively small, and it can be estimated that the reflective characteristic of the surface is conversely larger than the other area | region. Therefore, by adding a value corresponding to the intensity of the blue component to the luminance image, green component image, etc., it becomes possible to generate an image closer to the reflection characteristics of the object surface, and the shade and the sun are superimposed on the road marking. Even in this case, as shown in FIG. 4 or FIG. 8, the pixel value of the road marking and the paved road surface can be easily separated.

  Hereinafter, embodiments of the present invention will be described using a plurality of examples.

  In addition, this invention is not limited to said embodiment, Embodiment can be changed suitably within the range of the technical idea of this invention.

  Embodiments of the present invention will be described in detail with reference to the drawings.

Example 1
Referring to FIG. 1, an embodiment of the present invention includes a color image acquisition module 101 that acquires a color road surface image, a luminance image generation module 102 that generates a luminance image from the color image acquired by the color image acquisition module 101, and A blue component intensity image generation module 103 that generates a blue component intensity image representing the intensity of the blue component from the color image acquired by the color image acquisition module 101, a luminance image output from the luminance image generation module 102, and a blue component intensity image The image forming module 103 includes a shade correction image generation module 104 that reduces the influence of the shade from the image indicating the blue component intensity output from the generation module 103, and a road marking recognition module 105 that recognizes the road marking from the shade correction image generation module 104. FIG. 9 is a diagram illustrating an example of a hardware configuration that implements the present embodiment. The CPU 901 controls the overall operation of the road marking recognition apparatus by executing various software programs (computer programs). More specifically, in the present embodiment, the color camera 902 acquires a color image, and the CPU 901 refers to a storage medium 903 such as a memory as appropriate, while referring to a luminance image generation module 102, a blue component intensity image generation module 103, and a shade. Software programs such as the corrected image generation module 104 and the road marking recognition module 105 are executed.

  The color image acquisition module 101 is composed of a color camera or the like installed so as to photograph a road surface on the own vehicle. Image data is output every moment at a predetermined frame rate.

  The luminance image generation module 102 converts the color image output from the color image acquisition module 101 into a luminance image. The luminance image is created by combining the red, green, and blue components constituting the color image for each pixel by the following calculation.

や

などで表現する。ここでαは正の定数を表す。また、上式におけるＲはＧに置き換えてもよい。日陰は大気中で散乱されやすい青系の光を比較的多く含むため、青成分強度画像は舗装道路面や路面標示などの無彩色領域に対して、日陰ほど赤成分、緑成分に比して大きな値を持つ特徴を有する。定数αの値が大きさほど、日陰の路面標示に対する画素値も大きくできるが、αの値が大きすぎると後述する日陰補正画像において、日陰の舗装道路面に対応する画素値と日向の路面標示に対応する画素値とが類似しやすくなり認識が困難となるため、太陽光の分光特性および舗装道路面と路面標示の反射特性を考慮して、日陰補正画像において日陰の舗装道路面に対応する画素値が日向の路面標示に対応する画素値よりも常に小さくなるような定数αを設定する。 The blue component intensity image generation module 103 calculates the amount of the blue component with respect to the red component or the green component for each pixel, and outputs an image representing the intensity of the blue component. For example, the intensity C of the blue component is

And

Express with Here, α represents a positive constant. R in the above formula may be replaced with G. Because the shade contains a relatively large amount of blue light that is easily scattered in the atmosphere, the blue component intensity image is more shaded compared to the red and green components for shaded areas such as paved road surfaces and road markings. It has features with large values. The larger the value of the constant α, the larger the pixel value for the shaded road marking, but if the value of α is too large, the pixel value corresponding to the shaded paved road surface and the sunny road surface marking will be displayed in the shade correction image described later. Pixels corresponding to shaded paved road surfaces in shade-corrected images, taking into account the spectral characteristics of sunlight and the reflection characteristics of paved road surfaces and road markings, because the corresponding pixel values are likely to be similar and difficult to recognize. A constant α is set such that the value is always smaller than the pixel value corresponding to the sunny road marking.

  Further, the value of the positive constant α may be set to a different value determined in advance based on the time, date, and weather when the image is acquired. As a method for determining the value of the constant α based on the time, date and weather, for example, under these conditions, the average spectral characteristics of the sun and the average spectral characteristics of the shade are examined in advance. The value of α is calculated such that α times the difference in values (BR) in the shade or α times the difference in values (B / R) is exactly the difference in the green component value G, and this is used as a lookup table. You can memorize and refer to it. This greatly reduces the difference between the green component value between the sun and the shade, and the green component value on the road marking easily exceeds the green component value on the paved road surface, regardless of the sun or shade, and recognizes the road marking. An easy-to-use image can be generated.

  The shade correction image generation module 104 adds the luminance image output from the luminance image generation module 102 and the image indicating the number of blue components output from the blue component intensity calculation module 103 for each pixel, and outputs a shade correction image. . As shown in FIG. 3, the density histogram of the entire shade-corrected image when the pixels corresponding to the shaded road marking on the luminance image and the pixels corresponding to the sunny paved road surface have similar luminance values is shown. It is shown in FIG. The shade-corrected image as a whole corresponds to the paved road surface 803 in the shade, the paved road surface 801 in the sun, the road marking 802 in the shade, and the road marking 804 in the sun in descending order of density value. Therefore, if a binarization process or the like is performed with the threshold value 805, it is possible to easily extract only pixels corresponding to the road marking from the image.

  Here, the shade correction image has a wider range of pixel values than the luminance image and requires more storage capacity, so if you want to keep the storage capacity to the same extent as the luminance image, clip the maximum and minimum values. Alternatively, the minimum value may be limited to 0 and the maximum value may be limited to 255, for example, by performing conversion using a monotonically increasing function.

  The road marking recognition module 105 processes the shade correction image to recognize the road marking. Although various methods can be considered as the recognition method, for example, after performing binarization or multi-leveling, template matching or a statistical pattern recognition method may be applied to the extracted connected region. Alternatively, a feature point such as a corner point or a linear edge point may be extracted and collated based on the positional relationship of the feature points.

  Next, the operation of the present embodiment will be described with reference to the drawings. FIG. 5 shows a flowchart of the present embodiment. This figure represents a processing procedure executed by the road marking recognition apparatus shown in FIG. 9 in the present embodiment.

  As shown in FIG. 9, the road marking recognition apparatus according to the present invention includes a CPU 901 that is a central processing element, a color camera 902 that is connected to the CPU 901 via an interface, and a memory 903 that is connected to the CPU 901. The In particular, in the memory 903, a color image acquisition module program 904 that is a program for operating the color image acquisition module 101, a luminance image generation module program 906 that is a program for operating the luminance image generation module 102, and a blue component intensity A blue component intensity image generation module program 908 which is a program for operating the image generation module 103, a shade correction image generation module program 910 which is a program for operating the shade correction image generation module 104, and a program for operating the road marking recognition module 105 And a road marking recognition module program 912.

  First, the color image acquisition module program 904 captures a color image capturing a road surface with the color camera 902 and outputs it to the memory 903 (step S501). Next, the luminance image generation module program 906 generates a luminance image from the color image stored in the memory 903 and stores it in the memory 903 (step S502), and the blue component intensity image generation module program 908 A blue component intensity image is generated from the color image and output to the memory 903 (step S503). Further, the shade correction image generation module program 910 reads the luminance image and the blue component intensity image from the memory 903 and adds them to generate a shade correction image and outputs it to the memory 903 (step S504). Finally, the road marking recognition module program 912 recognizes the road marking from the shade correction image and outputs the recognition result to the memory 903 (step S505).

  Since the blue component intensity image tends to have a large value such as shade, the pixel value of the luminance image is added as the shade increases. As a result, even if the pixel corresponding to the sunny paved road surface and the pixel corresponding to the shaded road marking have similar pixel values on the luminance image, the pixel value of the road marking is large on the shade corrected image. Since it becomes easy to take a value, it becomes easy to distinguish between a road pavement surface and a road marking, and it becomes possible to make road marking recognition highly accurate.

  Although the case where the blue component intensity image is added to the luminance image has been described in the above embodiment, the same effect can be obtained by using the green component image that contributes most to the luminance instead of the luminance image. . When a green component image is used, it is not necessary to generate a luminance image, so that the processing speed can be increased. Furthermore, since the recognition target is an achromatic color, it is assumed that the values of the green and red components are almost equal, and if the red component image is used instead of the green component image, there is no need to use the green component. The number of times can be reduced and the processing can be further accelerated.

  Also, instead of the blue component intensity image, a blue component weak image having a larger pixel value as the blue component is smaller is generated, and the shade correction image generation module subtracts the blue component weak image from the luminance image or the green component image. It is clear that it is good.

  The present invention described by taking each of the above embodiments as an example, after supplying a computer program capable of realizing the function of the flowchart (FIG. 5) referred to in the description to the road marking recognition apparatus described above, This is accomplished by reading the computer program to the CPU 901 of the apparatus and executing it. Further, instead of the CPU 901, a dedicated arithmetic unit such as an image processor may be used.

  The computer program supplied to the apparatus may be stored in a readable / writable memory or a storage device (storage medium) such as a hard disk device. In such a case, the present invention is constituted by the code of the computer program or a storage medium.

(Example 2)
With reference to FIG. 6, another embodiment of the present invention will be described. Referring to FIG. 6, this embodiment replaces the luminance conversion image generation module, the blue component image generation module, and the shade correction image generation module in the embodiment shown in FIG. Have FIG. 9 is a diagram illustrating an example of a hardware configuration that implements the present embodiment. The CPU 901 controls the overall operation of the road marking recognition apparatus by executing various software programs (computer programs). More specifically, in the present embodiment, the color camera 902 acquires a color image, and the CPU 901 appropriately refers to a storage medium 903 such as a memory, and the shaded area luminance enhancement image generation module 601 and the road marking recognition module 105. Etc. are executed.

  The color image acquisition module 101 outputs a color image as in the above embodiment.

を算出し、日陰に対応する領域の画素の輝度を強調した画像を出力する。また、前記実施の形態と同様に、輝度画像にかえて緑成分画像を用いて、

とする画像を出力してもよい。また、緑成分画像にかえて赤成分画像を用いて

とする画像を出力してもよい。 The shaded area luminance-enhanced image generation module 601 multiplies the color image acquired by the color image acquisition module by a predetermined constant to a value obtained by subtracting the blue component value from the red component value for each pixel. An image added to the constituent green component value or luminance value is output. That is, the shaded area luminance enhancement image generation module

And an image in which the luminance of the pixels in the area corresponding to the shade is emphasized is output. Similarly to the above embodiment, using a green component image instead of a luminance image,

May be output. In addition, using a red component image instead of a green component image

May be output.

  The road marking recognition module 105 recognizes the road marking from the image output by the shade area luminance enhancement image generation module.

  Next, the operation of the present embodiment will be described with reference to the drawings. FIG. 7 shows a flowchart of the present embodiment. This figure shows the processing procedure which the road marking recognition apparatus in this embodiment performs.

  First, the color image acquisition module program 904 captures and outputs a color image capturing a road surface with the color camera 902 (step S701). Next, the shade correction image generation module program 910 multiplies the value obtained by subtracting the blue component value from the red component value for each pixel from the color image, and multiplies a predetermined constant to determine the green component value or red component constituting the color image. The image added to the component value or luminance value is output to the memory 903 (step S702). Finally, the road marking recognition module program 912 performs road marking recognition on the image stored in the memory in step S702, and outputs the recognition result to the memory 903 (step S703).

  According to this embodiment, an image used for recognition of road markings can be generated with a single module, and the amount of calculation can be reduced.

  The present invention described by taking each of the above embodiments as an example, after supplying a computer program capable of realizing the functions of the flowchart (FIG. 7) referred to in the description to the above-described road marking recognition device, This is accomplished by reading the computer program to the CPU 901 of the apparatus and executing it. Further, instead of the CPU 901, a dedicated arithmetic unit such as an image processor may be used.

  The computer program supplied to the apparatus may be stored in a readable / writable memory or a storage device (storage medium) such as a hard disk device. In such a case, the present invention is constituted by the code of the computer program or a storage medium.

It is a block diagram in 1st embodiment which concerns on the road marking recognition apparatus of this invention. It is a road surface figure which shows the condition where the shadow is superimposed on a part of conventional road marking. It is the brightness | luminance histogram of the pixel corresponding to the paved road surface which exists in the conventional sun, and the road marking which exists in shade. It is the brightness | luminance histogram of the pixel corresponding to the paved road surface which exists in the sun, and the road marking which exists in the shade which the conventional shade correction | amendment image generation module outputs. It is a flowchart in 1st embodiment which concerns on the road marking recognition method of this invention. It is a block diagram in 2nd embodiment which concerns on the road marking recognition apparatus of this invention. It is a flowchart in 2nd embodiment which concerns on the road marking recognition method of this invention. It is the density histogram of the pixel corresponding to the pavement road surface which exists in the shade, the pavement road surface which exists in the shade, the road surface indication which exists in the shade, and the road surface indication which exists in the shade in the shade correction image. It is a block diagram in 1st embodiment which concerns on the road marking recognition apparatus of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 Color image acquisition module 102 Luminance image generation module 103 Blue component intensity | strength image generation module 104 Shade correction | amendment image generation module 105 Road marking recognition module 201 Pavement road surface 202 which exists in the sun Road marking 203 which exists in the shade 203 Shade 301 It exists in the sun Luminance histogram 302 of pixels corresponding to a paved road surface Luminance histogram 401, 801 of pixels corresponding to a road marking existing in the shade Luminance histograms 402, 802 of pixels corresponding to a paved road surface existing in the sun after blue component enhancement Luminance histogram 601 of pixels corresponding to road markings existing in the shade after blue component enhancement Color image acquisition module 602 Shade area luminance enhancement image generation module 603 Road marking recognition module 803 Pavement existing in the shade after blue component enhancement surface Luminance histogram 804 of pixels corresponding to, luminance histogram 805 of pixels corresponding to road markings existing in the sun after blue component enhancement Threshold threshold 901 for CPU
902 Color camera 903 Memory 906 Luminance image generation module program 908 Blue component intensity image generation module program 910 Shade correction image generation module program 912 Road marking recognition module program

Claims (18)

In the road marking recognition device installed in the vehicle so as to photograph the road surface,
A color image acquisition function block that captures a color image of a road surface;
A luminance image generation functional block that generates a luminance image based on the color image acquired by the color image acquisition functional block;
A blue component intensity image generation functional block for generating a blue component intensity image by calculating the intensity of the blue component included in the color image for each pixel;
A shade-corrected image generation function block that generates a shade-corrected image in which the difference between the pixel values corresponding to the sun and shade areas is smaller than that of the brightness image by adding the brightness image and the blue component intensity image;
A road marking recognition apparatus comprising: a road marking recognition function block that recognizes a road marking from the shade correction image. In the road marking recognition device installed in the vehicle so as to photograph the road surface,
A color image acquisition function block that captures a color image of a road surface;
A green component image generation functional block that generates a green component image obtained by extracting only the green component value from the color image acquired by the color image acquisition functional block;
A blue component intensity image generation functional block for generating a blue component intensity image by calculating the intensity of the blue component included in the color image for each pixel;
A shade-corrected image generation function block that generates a shade-corrected image having a smaller difference in pixel values corresponding to the areas of the sun and shade than the green component image by adding the green component image and the blue component intensity image. When,
A road marking recognition apparatus comprising: a road marking recognition function block that recognizes a road marking from the shade correction image.   2. The blue component intensity image generation functional block generates a blue component intensity image by multiplying a positive constant by a value obtained by subtracting a red component value from a blue component value for each pixel. The road marking recognition apparatus according to 2.   3. The blue component intensity image generation functional block generates a blue component intensity image by multiplying a positive constant by a value obtained by dividing a blue component by a red component for each pixel. Road marking recognition device. In the road marking recognition device installed in the vehicle so as to photograph the road surface,
A color image acquisition function block that captures a color image of a road surface;
Based on the color image acquired by the color image acquisition functional block, a green component value or a red component value constituting the color image is obtained by multiplying a value obtained by subtracting a blue component value from a red component value for each pixel by a positive constant. Alternatively, a shaded area brightness-enhanced image generation function block that outputs an image added to the brightness value;
A road marking recognition device comprising: a road marking recognition functional block for recognizing road markings from an image output by the shade area luminance enhancement image generation functional block.   6. The road surface according to claim 3, wherein the blue component intensity image generation functional block determines the value of the positive constant based on any one or more of a date, time, and weather. Sign recognition device. In the road marking recognition method using the road marking recognition device installed in the host vehicle so as to photograph the road surface,
The color image acquisition function block includes a step of taking a color image of the road surface,
The luminance image generation functional block generates a luminance image based on the color image acquired by the color image acquisition unit;
The blue component intensity image generation functional block generates a blue component intensity image by calculating the intensity of the blue component included in the color image for each pixel;
The shade-corrected image generation functional block generates a shade-corrected image in which the difference between the pixel values corresponding to the shaded and shaded areas is smaller than that of the brightness image by adding the brightness image and the blue component intensity image. Process,
The road marking recognition function block includes a step of recognizing a road marking from the shade correction image. In the road marking recognition method using the road marking recognition device installed in the host vehicle so as to photograph the road surface,
The color image acquisition function block includes a step of taking a color image of the road surface,
The green component image generation functional block generates a green component image obtained by extracting only the green component value from the color image acquired by the color image acquisition functional block;
The blue component intensity image generation functional block generates a blue component intensity image by calculating the intensity of the blue component included in the color image for each pixel;
The shade-corrected image generation functional block adds the green component image and the blue component intensity image to obtain a shade-corrected image having a smaller difference in pixel values corresponding to the shaded and shaded areas than the green component image. Generating step;
The road marking recognition function block includes a step of recognizing a road marking from the shade correction image.   The blue component intensity image generation functional block generates, for each pixel, a blue component intensity image by multiplying a value obtained by subtracting a red component value from a blue component value by a positive constant. The road marking recognition method according to 8.   9. The blue component intensity image generation functional block generates a blue component intensity image by multiplying a positive constant by a value obtained by dividing a blue component by a red component for each pixel. Road marking recognition method. In the road marking recognition method using the road marking recognition device installed in the host vehicle so as to photograph the road surface,
The color image acquisition function block includes a step of taking a color image of the road surface,
The shaded area luminance-enhanced image generation function block is configured to multiply the value obtained by subtracting the blue component value from the red component value for each pixel by a positive constant based on the color image acquired by the color image acquisition function block. Outputting an image added to the green component value or red component value or luminance value constituting
The road marking recognition function block includes a step of recognizing the road marking from the image output by the shade area luminance enhancement image generation function block.   12. The road surface according to claim 7, wherein the blue component intensity image generation function block determines the value of the positive constant based on any one or more of a date, time, and weather. Sign recognition method. In the road marking recognition program using the road marking recognition device installed in the host vehicle so as to photograph the road surface,
The color image acquisition function block includes a step of taking a color image of the road surface,
The luminance image generation functional block generates a luminance image based on the color image acquired by the color image acquisition unit;
The blue component intensity image generation functional block generates a blue component intensity image by calculating the intensity of the blue component included in the color image for each pixel;
The shade-corrected image generation functional block generates a shade-corrected image in which the difference between the pixel values corresponding to the shaded and shaded areas is smaller than that of the brightness image by adding the brightness image and the blue component intensity image. Process,
The road marking recognition function block includes a step of recognizing a road marking from the shade correction image. In the road marking recognition method using the road marking recognition device installed in the host vehicle so as to photograph the road surface,
The color image acquisition function block includes a step of taking a color image of the road surface,
The green component image generation functional block generates a green component image obtained by extracting only the green component value from the color image acquired by the color image acquisition functional block;
The blue component intensity image generation functional block generates a blue component intensity image by calculating the intensity of the blue component included in the color image for each pixel;
The shade-corrected image generation functional block adds the green component image and the blue component intensity image to obtain a shade-corrected image having a smaller difference in pixel values corresponding to the shaded and shaded areas than the green component image. Generating step;
The road marking recognition function block includes a step of recognizing a road marking from the shade correction image.   The blue component intensity image generation functional block generates a blue component intensity image by multiplying a positive constant by a value obtained by subtracting a red component value from a blue component value for each pixel. 14. A road marking recognition program according to 14.   15. The blue component intensity image generation functional block generates a blue component intensity image by multiplying a positive constant by a value obtained by dividing a blue component by a red component for each pixel. Road marking recognition method. In the road marking recognition program using the road marking recognition device installed in the host vehicle so as to photograph the road surface,
The color image acquisition function block includes a step of taking a color image of the road surface,
The shaded area luminance-enhanced image generation function block is configured to multiply the value obtained by subtracting the blue component value from the red component value for each pixel by a positive constant based on the color image acquired by the color image acquisition function block. Outputting an image added to the green component value or red component value or luminance value constituting
The road marking recognition function block includes a step of recognizing road marking from an image output by the shade area luminance enhancement image generation function block.   The road surface according to any one of claims 13 to 17, wherein the blue component intensity image generation function block determines the value of the positive constant based on any one or more of a date, time, and weather. Sign recognition program.

Images