KR102192844B1 - Color image enhancement device for lane detection and method thereof - Google Patents

Abstract

An apparatus and method for improving a color image for lane detection are provided. According to an embodiment of the present invention, the apparatus for improving the color image for lane detection comprises: a sensing unit which acquires an image of the road on which a vehicle is driving; and a control unit communicatively connected to the sensing unit. Here, the control unit is configured to convert an acquired RGB image to a LAB space, improve the contrast for white color, yellow color, and blue color using a lookup table including an exponential coefficient for improving contrast in the LAB space, create a gray image from the improved image, amplify the contrast for the blue color in the gray image, and sharpen the amplified gray image.

Description

Color image enhancement device for lane detection and method thereof

The present invention relates to an apparatus and method for improving color images for lane detection.

ADAS (Advanced Driver Assistance Systems) is integrated into almost all vehicles to assist the driver while driving. ADAS adapts, automates and improves vehicle systems, enabling a smooth and safe driving experience.

On the other hand, LD (Lane Detection) guarantees the position of the vehicle with respect to the lane marking, and to safely switch lanes (assistant lane change) or to accidentally manipulate lane invasion (lane departure warning or lane keeping system). This is the important part.

However, detection of lane markings is a difficult task considering constraints such as other lane colors, lane patterns, road textures, road signs, lighting conditions with tunnels, or some shade.

KR 2012-0089528 A

In order to solve the problems of the prior art as described above, an object of the present invention is to provide a color image improving apparatus and method for detecting lanes capable of improving color lane display from other background objects. .

However, the problem to be solved by the present invention is not limited to the problems mentioned above, and other problems that are not mentioned will be clearly understood by those skilled in the art from the following description.

According to an aspect of the present invention for solving the above problem, a sensing unit for obtaining an image of a road on which the vehicle is driving; And a control unit that is communicatively connected to the sensing unit. Here, the control unit converts the obtained RGB image into a LAB space, and in the LAB space, improves the contrast for white, yellow and blue by using a lookup table including an index coefficient for improving the contrast, and the improvement And generating a gray image from the generated image, amplifying the contrast for blue in the gray image, and sharpening the amplified gray image.

In one embodiment, the exponent coefficient may be preset to suppress dark pixels and emphasize bright pixels.

In one embodiment, the control unit separates the LAB spatially transformed image into each channel image, and generates an inverse B channel image by inverting the bit of the B channel image, and the L channel image and the yellow color for the white color. The B channel image for the blue color and the inverse B channel image for the blue color may be configured to improve contrast by using the lookup table.

In an embodiment, the controller may be configured to sharpen the improved B channel image and the improved inverse B channel image using contrast limit adaptive histogram equalization (CLAHE).

In one embodiment, the controller may be configured to generate a gray image from the improved L channel image, the sharpened B channel image, and the sharpened inverse B channel image.

In one embodiment, the control unit separates the image for each channel from the RGB image, inverts the bits of the R channel image to generate an inverse R channel image, and uses the R channel image and the inverse R channel image to generate the It can be configured to amplify the blue contrast in the gray image.

According to another aspect of the present invention, the method comprising: obtaining an RGB image of a road on which a vehicle is running; Converting the obtained RGB image into LAB space; Improving contrast for white, yellow, and blue in the LAB space by using a lookup table including an index coefficient for improving contrast; Generating a gray image from the improved image; Amplifying the contrast for blue in the gray image; And sharpening the amplified gray image. A method for improving a color image for detecting lanes is provided.

In one embodiment, the exponent coefficient may be preset to suppress dark pixels and emphasize bright pixels.

In one embodiment, the step of improving comprises: separating the LAB spatially transformed image into each channel image; Inverting the bits of the B channel image to generate an inverse B channel image; And improving the contrast of the L channel image for white, the B channel image for yellow, and the inverse B channel image for blue, respectively, using the lookup table.

In an embodiment, the step of improving may further include sharpening the improved B channel image and the improved inverse B channel image using contrast limit adaptive histogram equalization (CLAHE).

In an embodiment, the generating may generate a gray image from the improved L channel image, the sharpened B channel image, and the sharpened inverse B channel image.

In one embodiment, the amplifying step comprises: separating an image for each channel from the RGB image and inverting bits of the R channel image to generate an inverse R channel image; And amplifying a blue contrast in the gray image by using the R channel image and the inverse R channel image.

According to an embodiment of the present invention, an apparatus and method for improving color images for lane detection according to an embodiment of the present invention can improve the contrast of white, blue, and yellow by converting an RGB image into a LAB space, thereby improving color lane display from other background objects. Therefore, it can be supported to easily detect lanes having colors in various environments.

In addition, according to the present invention, by improving the contrast for each channel using the lookup table, the amount of computation can be reduced and the computation time can be reduced, thereby improving the efficiency of the improvement process.

1 is an exemplary view showing an environment in which a vehicle having a color image improving device for detecting a lane according to an embodiment of the present invention travels on a road,
2 is a block diagram of an apparatus for improving a color image for detecting a lane according to an embodiment of the present invention.
3 is a flow chart of a color image improvement method for detecting a lane according to an embodiment of the present invention.
4 is a photograph showing a captured image and a result image by a color image improvement method for lane detection according to an embodiment of the present invention, wherein a) is a blue lane with a shade, b) a blue lane on a road with a different texture, c) is a blue lane with a different pattern,
5 is a photograph showing a captured image and a result image by a color image improvement method for lane detection according to an embodiment of the present invention, wherein a) is a faint blue lane, b) a faint white lane, c) a white lane and blue Lane,
6 is a photograph showing a captured image and a result image by a color image improvement method for lane detection according to an embodiment of the present invention, wherein a) is a white lane and a yellow lane, b) a white lane and a yellow lane inside a tunnel , c) shaded white and yellow lanes.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art can easily implement the present invention. The present invention may be implemented in various different forms, and is not limited to the embodiments described herein. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and the same reference numerals are assigned to the same or similar components throughout the specification.

The embodiments of the present invention are provided to more completely describe the present invention to those of ordinary skill in the art, and the embodiments described below may be modified in various different forms, and The scope is not limited to the following examples. Rather, these examples are provided to make the present invention more faithful and complete, and to fully convey the spirit of the present invention to those skilled in the art.

The terms used in this specification are used to describe specific embodiments, and are not intended to limit the present invention. As used herein, the singular form may include a plural form unless the context clearly indicates another case. Further, as used herein, "comprise" and/or "comprising" specifies the presence of the mentioned shapes, numbers, steps, actions, members, elements and/or groups thereof. And does not exclude the presence or addition of one or more other shapes, numbers, actions, members, elements, and/or groups. As used herein, the term "and/or" includes any and all combinations of one or more of the corresponding listed items.

In the present specification, terms such as first and second are used to describe various members, regions and/or parts, but it is obvious that these members, parts, regions, layers and/or parts are not limited by these terms. . These terms do not imply any particular order, top or bottom, or superiority, and are only used to distinguish one member, region, or region from another member, region, or region. Therefore, the first member, region, or region to be described below may refer to the second member, region or region without departing from the teachings of the present invention.

In the present specification, terms such as “or” and “at least one” may represent one of words listed together, or a combination of two or more. For example, "A or B" and "at least one of A and B" may include only one of A or B, and may include both A and B.

Hereinafter, embodiments of the present invention will be described with reference to the drawings schematically showing embodiments of the present invention. In the drawings, for example, depending on manufacturing techniques and/or tolerances, variations of the illustrated shape can be expected. Therefore, the embodiments of the present invention should not be construed as being limited to the specific shape of the region shown in the present specification, but should include, for example, a change in shape caused by manufacturing.

1 is an exemplary view showing an environment in which a vehicle including a color image improving apparatus for detecting a lane according to an embodiment of the present invention travels on a road.

Referring to FIG. 1, a color of a lane may exist on a road on which a vehicle 10 including a color image improving apparatus 100 for detecting a lane according to an exemplary embodiment of the present invention travels. In addition, shades S1 and S2 may be cast on the road to form part of the lane.

At this time, objects with shades of white, blue and yellow are more prominent in a single gray image. Therefore, when at least one of the white lane W, the blue lane B, and the yellow lane Y exists on the road, a lot of labor is required to detect the lane.

Accordingly, the present invention is for image enhancement to easily detect objects having white, blue, and yellow colors regardless of scenarios such as sunny weather, tunnel interior, and shade conditions. In particular, the present invention is to assist in easily detecting a white lane (W), a blue lane (B) and a yellow lane (Y).

In other words, the color image improving apparatus 100 for detecting lanes according to an embodiment of the present invention includes white, blue, and yellow lanes in order to easily detect a white lane W, a blue lane B, and a yellow lane Y. Image enhancement can be performed on.

To this end, the color image improvement apparatus 100 for detecting lanes may perform color image improvement by converting the image of white, blue, and yellow into a LAB space to improve contrast for individual channels.

Here, the L value in the LAB space represents the brightness. When L = 0, it is black, and when L = 100, it is white. a indicates whether the color is skewed toward red or green. If a is negative, the color is skewed toward green, and if a is positive, it is skewed toward red/purple. b represents yellow and blue. If b is negative, it is blue, and if b is positive, it is yellow.

2 is a block diagram of an apparatus for improving a color image for lane detection according to an embodiment of the present invention.

The color image improvement apparatus 100 for detecting lanes includes a detection unit 110, a control unit 120, and a storage unit 130.

The sensing unit 110 may detect surrounding information of the vehicle 10. As an example, the sensing unit 110 may be a camera disposed at various positions in the front and rear sides and sides. In this case, the sensing unit 110 may acquire an image having an RGB color of a road on which the vehicle 10 is driving.

The control unit 120 may be communicatively connected to the sensing unit 110 and the storage unit 130. Here, the controller 120 may be an ECU of the vehicle 10. Optionally, the controller 120 may be a controller separately provided for image improvement.

The controller 120 may be configured to perform image improvement for white, blue, and yellow in order to easily detect the white lane W, the blue lane B, and the yellow lane Y.

Here, the controller 120 may be configured to convert the RGB image obtained from the sensing unit 110 into a LAB space. In this case, the control unit 120 may first reduce the size of the RGB image to QVGA (Quarter Video Graphics Array) in order to reduce the processing time.

In addition, the controller 120 may be configured to improve contrast for white, yellow, and blue using a lookup table including exponential coefficients for improving contrast in the LAB space. Here, the exponent coefficient may be preset to suppress dark pixels and emphasize bright pixels.

As a result, there is no limit on the threshold for color image improvement, and thus, the computational time can be reduced by reducing the computational amount, thereby improving the efficiency of the improvement process.

More specifically, the controller 120 may be configured to separate the LAB spatially converted image into each channel image. In this case, the LAB spatially converted image may be divided into an L channel image, an A channel image, and a B channel image.

The controller 120 may be configured to generate an inverse B channel image B_rev for the B channel image. In this case, the controller 120 may be configured to invert the bit of the B channel image.

The controller 120 may be configured to improve the contrast of the divided images for each channel. That is, the controller 120 may be configured to improve an L channel image for white, a B channel image for yellow, and an inverse B channel image B_rev for blue. At this time, the control unit 120 may use a lookup table. Here, an L channel enhancement image Enh_L, a B channel enhancement image Enh_B, and an inverse B channel enhancement image Enh_B_rev may be generated.

The controller 120 may be configured to perform additional contrast improvement for blue and yellow. That is, the controller 120 may be configured to sharpen the improved B channel image Enh_B and the improved inverse B channel image Enh_B_rev, respectively. In this case, the control unit 120 may be configured to configure sharpening using Contrast-Limited Adaptive Histogram Equalization (CLAHE). Here, a B-channel sharpening image (B_sharpened) and an inverse B-channel sharpening image (B_rev_sharpened) may be generated.

The controller 120 may be configured to generate a gray image with an image with improved contrast. In this case, the controller 120 may be configured to generate a gray image LAB_result from the improved L channel image Enh_L, the sharpened B channel image B_sharpened, and the sharpened inverse B channel image B_rev_sharpened.

The controller 120 may be configured to amplify the contrast for blue in the gray image LAB_result. That is, the contrast for an element corresponding to blue in the gray image LAB_result may be adjusted.

In this case, the controller 120 may be configured to generate an inverse R channel image R_rev by separating an image for each channel from an RGB image whose size has been reduced to QVGA. In this case, the control unit 120 may be configured to invert the bits of the R channel image.

The controller 120 may be configured to amplify the blue contrast in the gray image LAB_result by using the R channel image and the inverse R channel image R_rev. That is, the elements B_rev_sharpened corresponding to blue in the gray image LAB_result may be further improved by using the components corresponding to blue in the original RGB color image.

The controller 120 may be configured to sharpen the amplified gray image LAB_result. At this time, the edge can be detected according to sharpening. Therefore, it is possible to easily detect lane detection from the improved image (LAB_result) in the future.

The storage unit 130 may store a lookup table for improving the contrast of each channel in the LAB space.

With such a configuration, the color image improving apparatus for detecting lanes according to an embodiment of the present invention can improve color lane display from other background objects, and thus support to easily detect lanes having colors in various environments. , It is possible to reduce the amount of computation and reduce the computation time, thereby improving the efficiency of the improvement process.

Hereinafter, a color image improvement method for detecting a lane according to the present invention will be described with reference to FIGS. 3 to 6.

3 is a flowchart of a color image improvement method for detecting a lane according to an embodiment of the present invention.

The color image improvement method 200 for lane detection includes: obtaining a road image (S201), converting it to a LAB space (S202), improving the contrast for a specific color (S203 to S206), and converting a gray image. It includes generating (S207), amplifying blue contrast (S208 and S209), and sharpening the image (S210).

In more detail, as shown in FIG. 3, first, the color image improving apparatus 100 for detecting a lane acquires an RGB image of a road on which the vehicle 10 is driving (step S201). In this case, the apparatus 100 for improving a color image for lane detection may reduce the size of the RGB image to QVGA in order to reduce processing time.

Next, the color image improvement apparatus 100 for lane detection converts the obtained RGB image into a LAB space (step S202).

Next, the color image improving apparatus 100 for detecting lanes performs contrast improvement for a specific color in the LAB space from steps S203 to S206. In this case, the color image improving apparatus 100 for lane detection may improve the contrast for white, yellow, and blue using a lookup table including an index coefficient for improving contrast in the LAB space.

More specifically, the color image improvement apparatus 100 for lane detection divides each channel (step S203). In this case, the color image improvement apparatus 100 for detecting lanes may separate the LAB spatially converted image into an L channel image, an A channel image, and a B channel image.

Next, the color image enhancement apparatus 100 for lane detection generates an inverse B channel image B_rev for the B channel image (step S204). In this case, the color image improving apparatus 100 for detecting lanes may invert a bit of the B channel image.

Next, the color image improvement apparatus 100 for lane detection improves the LAB spatial division image for each channel (step S206). In this case, the apparatus 100 for improving a color image for lane detection may improve contrast of an L channel image for white, a B channel image for yellow, and an inverse B channel image B_rev for blue. Here, the color image enhancement apparatus 100 for lane detection may generate an L channel enhancement image Enh_L, a B channel enhancement image Enh_B, and an inverse B channel enhancement image Enh_B_rev using the lookup table.

Next, the color image improvement apparatus 100 for lane detection performs additional contrast improvement of the LAB spatial division image (step S206). In this case, the color image improving apparatus 100 for detecting lanes may sharpen the improved B-channel image Enh_B and the improved inverse B-channel image Enh_B_rev, respectively. Here, the color image improvement apparatus 100 for detecting lanes can generate a B-channel sharpening image (B_sharpened) and an inverse B-channel sharpening image (B_rev_sharpened) by sharpening using contrast-limited adaptive histogram equalization (CLAHE). have.

Next, the color image improving apparatus 100 for lane detection generates a gray image from the improved image (step S207). At this time, the color image improvement apparatus 100 for detecting lanes generates a gray image (LAB_result) from the improved L channel image (Enh_L), the sharpened B channel image (B_sharpened), and the sharpened inverse B channel image (B_rev_sharpened). Can be generated.

Next, the color image improvement apparatus 100 for lane detection amplifies the contrast for blue in the gray image LAB_result through steps S208 and S209. That is, the color image improving apparatus 100 for detecting lanes may adjust the contrast of an element corresponding to blue in the gray image LAB_result.

More specifically, the color image enhancement apparatus 100 for lane detection generates an inverse R channel image R_rev (step S208). In this case, the color image improvement apparatus 100 for lane detection may separate an image for each channel from an RGB image whose size is reduced by QVGA, and may invert a bit of the separated R channel image.

Next, the color image improving apparatus 100 for lane detection amplifies the blue contrast (step S209). In this case, the color image improvement apparatus 100 for detecting lanes may amplify the blue contrast in the gray image LAB_result by using the R channel image and the inverse R channel image R_rev. That is, the elements B_rev_sharpened corresponding to blue in the gray image LAB_result may be further improved by using the components corresponding to blue in the original RGB color image.

Next, the color image improving apparatus 100 for detecting lanes sharpens the amplified gray image LAB_result (step S210). In this case, the color image improving apparatus 100 for detecting lanes may detect an edge by sharpening. Therefore, it is possible to easily detect lane detection from the improved image (LAB_result) in the future.

4 is a photograph showing a captured image and a result image by a color image improvement method for lane detection according to an embodiment of the present invention, wherein a) is a blue lane with a shade, b) a blue lane on a road with a different texture, c) Blue lanes with different patterns.

Referring to FIG. 4, when a part of the blue lane B has a shade as shown in a), the entire blue lane B may be improved so that the entire blue lane B is easily identified. b) When the texture of the road is partially different, it can be improved so that the entire blue lane B is easily identified regardless of the texture of the road. c) When the blue lane B has a different pattern, it may be improved so that the entire blue lane B is identified regardless of the pattern.

In this way, the contrast of the blue lane B may be improved under various conditions. Therefore, detection of the blue lane B can be easily performed later.

5 is a photograph showing a captured image and a result image by a color image improvement method for lane detection according to an embodiment of the present invention, wherein a) is a faint blue lane, b) a faint white lane, c) a white lane and blue It's the second best.

Referring to FIG. 5, when a part of the blue lane B is blurred as shown in a), it may be improved so that the entire blue lane B is easily identified. As shown in b), when a part of the white lane W is dim, it may be improved so that the entire white lane W is easily identified. If the white lane W and the blue lane B coexist as shown in c), it may be improved so that both the white lane W and the blue lane B are easily identified.

As such, the contrast of the blue lane B and/or the white lane W may be improved under various conditions. Accordingly, detection of the blue lane B and/or the white lane W may be easily performed later.

6 is a photograph showing a captured image and a result image by a color image improvement method for lane detection according to an embodiment of the present invention, wherein a) is a white lane and a yellow lane, b) a white lane and a yellow lane inside a tunnel , c) shaded white and yellow lanes.

Referring to FIG. 6, when the white lane W and the yellow lane Y coexist as shown in a), it may be improved so that both the white lane W and the yellow lane Y are easily identified. In the case of the inside of the tunnel as shown in b), it can be improved so that both the white lane W and the yellow lane Y are easily identified as a whole. If there is a shade in a part of the lane as shown in c), it can be improved so that both the white lane W and the yellow lane Y are easily identified as a whole.

In this way, the contrast of the white lane W and the yellow lane Y may be improved under various conditions. Accordingly, detection of the white lane W and/or the yellow lane Y may be easily performed later.

By such a method, the present invention can improve color lane display from other background objects, so that it can support to easily detect lanes having colors in various environments, can reduce the amount of computation, and can reduce computation time. The efficiency of the improvement treatment can be improved.

The above-described methods may be implemented by the color image enhancement apparatus 100 for lane detection as shown in FIG. 1, and in particular, may be implemented as a software program that performs these steps, and in this case, such a program They may be stored on a computer-readable recording medium or transmitted by a computer data signal combined with a carrier wave in a transmission medium or a communication network. In this case, the computer-readable recording medium may include all types of recording devices in which data readable by a computer system is stored.

Although an embodiment of the present invention has been described above, the spirit of the present invention is not limited to the embodiment presented in the present specification, and those skilled in the art who understand the spirit of the present invention can add components within the scope of the same idea. Other embodiments may be easily proposed by changes, deletions, additions, etc., but it will also be said to fall within the scope of the present invention.

100: color image enhancement device for lane detection
110: sensing unit 120: control unit
130: storage unit 10: vehicle

Claims (12)

A sensing unit that acquires an image of a road on which the vehicle is driving; And
And a control unit communicatively connected to the sensing unit,
The control unit,
Convert the obtained RGB image to LAB space,
In the LAB space, contrast for white, yellow and blue is improved by using a lookup table including an index coefficient for contrast improvement,
Generating a gray image from the improved image,
Amplifies the contrast for blue in the gray image, separates the image for each channel from the RGB image, and generates an inverse R channel image by inverting the bits of the R channel image, and the R channel image and the inverse R channel image To amplify the blue contrast in the gray image, and
A color image enhancement device for detecting lanes, configured to sharpen the amplified gray image. The method of claim 1,
The color image enhancement apparatus for detecting lanes, wherein the exponential coefficient is preset to suppress dark pixels and emphasize bright pixels. The method of claim 1,
The control unit,
Separating the LAB spatially converted image into each channel image,
Inverts the bits of the B channel image to create an inverse B channel image, and
The color image improvement apparatus for detecting lanes, configured to improve the contrast of the L channel image for white, the B channel image for yellow, and the inverse B channel image for blue, respectively, using the lookup table. The method of claim 3,
The control unit is configured to sharpen the improved B-channel image and the improved inverse B-channel image using contrast limit adaptive histogram equalization (CLAHE). The method of claim 1,
The control unit is configured to generate a gray image from the improved L-channel image, the sharpened B-channel image, and the sharpened inverse B-channel image. delete Obtaining an RGB image of a road on which the vehicle is driving;
Converting the obtained RGB image into LAB space;
Improving contrast for white, yellow, and blue in the LAB space by using a lookup table including an index coefficient for improving contrast;
Generating a gray image from the improved image;
Amplifying the contrast for blue in the gray image; And
Sharpening the amplified gray image;
Including,
The amplifying step,
Separating the image for each channel from the RGB image and inverting the bits of the R channel image to generate an inverse R channel image; And
And amplifying a blue contrast in the gray image using the R channel image and the inverse R channel image. The method of claim 7,
The exponential coefficient is preset to suppress dark pixels and emphasize bright pixels. The method of claim 7,
The improving step,
Separating the LAB spatially converted image into each channel image;
Inverting the bits of the B channel image to generate an inverse B channel image; And
A color image improvement method for lane detection comprising the step of improving the contrast of the L channel image for white, the B channel image for the yellow, and the inverse B channel image for the blue, respectively, using the lookup table. . The method of claim 9,
The improving step further comprises sharpening the improved B-channel image and the improved inverse B-channel image using contrast-limited adaptive histogram equalization (CLAHE). The method of claim 7,
The generating step comprises generating a gray image from the improved L-channel image, the sharpened B-channel image, and the sharpened inverse B-channel image. delete

Images