KR101584907B1 - Method and Apparatus for recognizing lane using region of interest - Google Patents

Abstract

Disclosed is a method for recognizing a lane using a region of interest. According to an embodiment of the present invention, the method for recognizing a lane using a region of interest comprises the steps of: acquiring image information including the two lanes which are the left lane and the right lane; setting multiple regions of interest in the acquired image information to detect the edge of the two lanes; applying the Hough transformation to the edge of the two lanes detected for each of the multiple regions of interest to detect multiple straight lines corresponding to the edge; and using the multiple straight lines detected for each of the multiple regions of interest to recognize the border of the two lanes.

Description

[0001] The present invention relates to a method and apparatus for recognizing a lane using a region of interest,

One embodiment of the present invention relates to lane recognition, and more particularly, to a lane recognition method and apparatus using a region of interest.

In recent years, there has been developed an intelligent vehicle in which a driving operation such as running, stopping, turning, decelerating, or decelerating is automatically performed by a computer or the like without directly operating the vehicle, even if there is no driver or there is a driver.

The main tasks of these intelligent vehicles are to maintain driving lanes, to establish safety distances to adjacent vehicles, to detect and avoid collision obstacles, to control collision avoidance, and to control vehicle speed in response to traffic conditions and road conditions.

Further, according to the progress of information and communication technology, a safety driving assist system such as a lane departure warning system (LDWS) or a lane keeping system, and a vehicle automatic control system have been developed, and the practical use of the intelligent vehicle is progressing more rapidly . Especially, the detection of the driving lane is one of the core technologies for solving the main problems in the intelligent vehicle, and many researches are actively conducted.

Since the detection of such a driving lane has a great influence on the safe driving, an accurate driving lane is detected by using various sensors in order to estimate and determine the position of the lane. For example, various sensors such as an image sensor, a RADAR or a LIDAR sensor have been used to implement an intelligent vehicle control system in a stand-alone or fused form for lane detection or object recognition in front of the vehicle.

Especially, image based system using image sensor has been widely used because it can extract a lot of information at low cost and can utilize various image processing algorithms.

Such an image-based lane detection system can extract feature information from an input image, apply a parametric model for lane detection, update algorithms such as matching, Kalman filter or particle filtering, A lane is detected using a method such as an approximation method or a non-parametric model matching by conversion such as a Hough Transform (HT).

Such conventional lane detection algorithms have a problem of recognizing linear noises other than the lane as lanes and have a problem that the amount of computation is large by performing image processing for all areas of the input image and the lane recognition rate is decreased in the rapid curve section.

An object of an embodiment of the present invention is to provide a lane recognition method and apparatus which can reduce a calculation amount for lane recognition and reduce a lane recognition accuracy.

According to an aspect of the present invention, there is provided a lane recognition method including: obtaining image information including two lanes of left and right lanes; Detecting edges of the two lanes by setting a plurality of regions of interest in the obtained image information; Applying a Hough transform to edges of the two lanes detected for each of the plurality of ROIs to detect a plurality of straight lines corresponding to the edges; And recognizing the boundaries of the two lanes using the plurality of straight lines detected for each of the plurality of ROIs.

Preferably, the step of detecting a plurality of straight lines corresponding to the edge by applying the Hough transform may include detecting a plurality of straight lines represented by a set of parameters of angles and distances according to the coordinates of pixels included in the edge, Only straight lines within the angular range can be detected.

Preferably, the predetermined angle range is from an LLA (Left Low Angle) to an LH (Left High Angle), which is an angle to an uppermost point in an edge of the left lane, A second angular range from an RLA (Right Low Angle), which is an angle with respect to the lowermost point on the edge of the right lane to an RHA (Left High Angle), which is an angle with respect to the uppermost point on the edge of the right lane, . ≪ / RTI >

Advantageously, the step of detecting edges of the two lanes comprises the steps of converting the plurality of ROIs in the acquired image information into monochrome image information; Removing noise by applying a Gaussian blur filter to the monochrome image information; And detecting a contour of the two lanes by applying a sobel mask to the noise-removed monochrome image information.

Preferably, recognizing the boundaries of the two lanes using the plurality of straight lines includes calculating an average angle that is an average value of a plurality of angles corresponding to the plurality of straight lines with respect to the edges of the two lanes; And recognizing one lane boundary for each of the two lane edges using the average angle.

Preferably, the plurality of regions of interest may be individually or grouped into a predetermined group and subjected to parallel processing.

According to another aspect of the present invention, there is provided a lane recognition apparatus comprising: an image acquisition unit for acquiring image information including two lanes of left and right lanes; An edge detector configured to detect edges of the two lanes by setting a plurality of ROIs in the obtained image information; A Hough transform unit for applying a Hough transform to edges of the two lanes detected for each of the plurality of ROIs to detect a plurality of straight lines corresponding to the edges; And a lane boundary recognition unit for recognizing the boundaries of the two lanes using the plurality of straight lines detected for each of the plurality of ROIs.

Preferably, the Hough transform unit detects a plurality of straight lines represented by a set of parameters of angles and distances according to the coordinates of the pixels included in the edge, and detects the straight lines within a predetermined angle range, Can be detected.

Preferably, the predetermined angle range is from an LLA (Left Low Angle) to an LH (Left High Angle), which is an angle to an uppermost point in an edge of the left lane, A second angular range from an RLA (Right Low Angle), which is an angle with respect to the lowermost point on the edge of the right lane to an RHA (Left High Angle), which is an angle with respect to the uppermost point on the edge of the right lane, . ≪ / RTI >

Preferably, the edge detecting unit comprises: a monochrome image converting unit for converting the plurality of interest regions in the acquired image information into monochrome image information; A noise removing unit for removing noise by applying a Gaussian blur filter to the monochrome image information; And an edge processing unit for detecting a contour of the two lanes by applying a sobell mask to the noise-removed monochrome image information.

Preferably, the lane boundary recognizing unit includes: an angle calculating unit for calculating an average angle that is an average value of a plurality of angles corresponding to the plurality of straight lines with respect to edges of the two lanes; And a lane processing unit for recognizing one lane boundary for each of the edges of the two lanes using the average angle. .

Preferably, each of the plurality of ROIs may be grouped individually or in a predetermined group so as to be subject to parallel processing.

According to the embodiment of the present invention, there is an effect that the amount of calculation for lane recognition can be reduced and the lane recognition rate can be lowered.

According to another embodiment of the present invention, there is an effect that the calculation time for lane recognition can be shortened through parallel processing for a plurality of ROIs.

Further, according to another embodiment of the present invention, the lane can be accurately recognized even in the rapid curve section.

1 is a flowchart illustrating a lane recognition method according to an embodiment of the present invention.
2 is a diagram illustrating a plurality of regions of interest according to an embodiment of the present invention.
FIG. 3 is a flowchart illustrating an edge detection method in a plurality of ROIs according to an exemplary embodiment of the present invention. Referring to FIG.
4 is a diagram illustrating an edge detection method based on a Sobel mask according to an embodiment of the present invention.
5 is a diagram for explaining a Hough transform method according to an embodiment of the present invention on a Huff plane.
6 is a diagram for explaining a Hough transform method according to an embodiment of the present invention on a coordinate plane.
FIG. 7 is a diagram illustrating a Hough transform method according to an embodiment of the present invention in video information.
FIG. 8 illustrates an example of recognizing a lane on image information using a lane recognition method according to an embodiment of the present invention.
9 is a view for explaining a lane recognizing apparatus according to an embodiment of the present invention.
10 is a diagram illustrating an edge detector according to an embodiment of the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. Like reference numerals are used for like elements in describing each drawing.

The terms first, second, A, B, etc. may be used to describe various elements, but the elements should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component. And / or < / RTI > includes any combination of a plurality of related listed items or any of a plurality of related listed items.

It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, . On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

The terminology used in this application is used only to describe a specific embodiment and is not intended to limit the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the terms "comprises" or "having" and the like are used to specify that there is a feature, a number, a step, an operation, an element, a component or a combination thereof described in the specification, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries are to be interpreted as having a meaning consistent with the contextual meaning of the related art and are to be interpreted as either ideal or overly formal in the sense of the present application Do not.

Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

1 is a flowchart illustrating a lane recognition method according to an embodiment of the present invention.

In step 110, the lane recognition device obtains image information including two lanes of the left and right lanes.

At this time, the lane recognizing device may be a black box mounted on a car or a portable terminal of a user installed in a car.

In step 120, the lane recognizing device detects a two-lane edge by setting a plurality of ROIs in the acquired image information.

Preferably, the plurality of ROIs are configured to include two lanes in the image information, and may be positioned vertically or horizontally adjacent to each other. This will be described in more detail in Fig.

A specific method of detecting the edges of the two lanes will be described later with reference to Figs. 3 and 4. Fig.

In step 130, the lane recognition device applies a Hough transform to the edges of two lanes detected for each of a plurality of ROIs to detect a plurality of straight lines corresponding to the edges.

At this time, a method of detecting a plurality of straight lines by applying the Hough transform will be described later with reference to FIG. 5 to FIG.

In step 140, the lane recognition device recognizes the boundaries of two lanes using a plurality of straight lines detected for each of a plurality of ROIs.

As described above, according to the embodiment of the present invention, the image processing is not performed on all the areas of the inputted image information, but the lane is recognized by performing image processing only on a plurality of ROIs.

In addition, according to the embodiment of the present invention, it is possible to accurately recognize the lane in the rapid curve section as compared with the conventional art in which lanes are detected only in one region of interest by appropriately setting the size and position of a plurality of ROIs .

2 is a diagram illustrating a plurality of regions of interest according to an embodiment of the present invention.

In FIG. 2, four areas of interest 212, 214, 222 and 224 are shown, with the two left area of interest 212 and 214 being the area of interest including the left yellow centerline, , 224) is the area of interest that contains the white lane to the right.

The four regions of interest 212, 214, 222, and 224 shown in FIG. 2 are set within an image frame in the input image. As shown in FIG. 2, the regions of interest may be set adjacent to each other, The position may be set such that each of the regions of interest is separated by a predetermined distance.

At this time, the ROI may be set by the user, but may be automatically set by the lane recognition device so that a plurality of regions estimated to include lanes are included as ROIs.

However, in another embodiment, the ROIs may be set over a plurality of image frames rather than one image frame. If the ROI is set across a plurality of image frames, even if the car changes lanes, .

On the other hand, the ROIs may be individually grouped into a predetermined group, and may be subject to parallel processing. For example, in FIG. 2, interest areas 212 and 214 for the left lane are grouped into a first group, interest areas 222 and 224 for the right lane are grouped into a second group, Parallel operations may be performed on each of the second groups. In addition, an operation for lane recognition may be performed in parallel for each of the four ROIs 212, 214, 222, and 224 in Fig.

 Through this parallel processing, the present invention has a shorter computation time than the prior art in which lane recognition is performed by performing image processing on all areas of image information without setting a region of interest.

FIG. 3 is a flowchart illustrating an edge detection method in a plurality of ROIs according to an exemplary embodiment of the present invention. Referring to FIG.

In step 310, the lane recognition device converts a plurality of ROIs in the image information into monochrome image information.

The reason for converting the image information into the monochrome image information is to reduce the amount of calculation for lane recognition.

In step 320, the lane recognizing device applies a Gaussian blur filter to the monochrome image information to remove noise.

The Gaussian blur filter can be defined as Equation (1).

는 가우스 분포의 표준 편차를 나타낸다. Where x is the x coordinate value of the pixel on the coordinate plane, y is the y coordinate value of the pixel on the coordinate plane,

Represents the standard deviation of the Gaussian distribution.

As described above, according to the embodiment of the present invention, the Gaussian blur filter is applied to the black and white image information to remove the noise, and the image processing for lane recognition is performed. Therefore, the problem of the prior art that the straight line noise other than the lane is recognized as the lane Can be solved.

In step 330, the lane recognition device detects a contour of two lanes by applying a sobel mask to the black and white image information from which noise has been removed.

The outline detection method using the Sobel mask will be described as follows.

First, a Sobel mask for detecting a contour is determined. In the case of a 3x3 Sobel mask, a horizontal mask Gx and a vertical mask Gy are defined as in Equation (2).

Next, the magnitude of the inclination of the Sobel mask is calculated according to Equation (3), and the direction of the inclination is calculated according to Equation (4).

Next, the angle of the gradient value of the image information to which the Sobel mask is applied is calculated by Equation (5).

Finally, the contour is detected by comparing the angle of the gradient value with the high threshold and the low threshold. More specifically, when the angle d of the gradient value has a value higher than the high threshold value, it is displayed as a strong edge pixel, and when the angle d of the gradient value has a value smaller than the low threshold value, display is suppressed as a weak edge pixel The contour is detected.

4 is a diagram illustrating an edge detection method based on a Sobel mask according to an embodiment of the present invention.

4, the input image information is converted into monochrome image information, and then a Gaussian blur filter is applied. Finally, a contour line of the objects in the image information is detected by the regions of interest 212, 214, 222, and 224 by applying a Sobel mask Fig.

In Fig. 4, the outline of the vehicle and the edge of the lane are displayed. After applying the Gaussian Blur filter, the noise near the lane is removed by applying the Sobel mask.

5 is a diagram for explaining a Hough transform method according to an embodiment of the present invention on a Huff plane.

Fig. 5 (a) shows four pixels on a coordinate plane included in an edge for two lanes, which are converted and displayed on a Hough plane according to a conventional Hough transform scheme. The coordinates (x, y) The pixel can be Huff transformed by the equation (6) and displayed on the Huff plane.

는 원점에서 (x,y)의 좌표를 가지는 픽셀까지의 직선 거리를 나타내고,

는

와 y축이 이루는 각도를 나타낸다. At this time,

Represents the straight line distance from the origin to a pixel having coordinates of (x, y)

The

And the y-axis.

을 지나는 직선이 검출되고, 제3 픽셀 그래프와 제4 픽셀 그래프의 교점인

를 지나는 직선이 검출될 수 있다. 이와 같은 방식으로 2개 차선에 대한 에지에 포함된 좌표 평면상에 포함된 모든 픽셀에 대해 허프 변환을 적용하면 에지에 포함되는 픽셀의 좌표별로 각도 및 거리의 매개변수 집합으로 표시되는 복수의 직선을 검출할 수 있다. 하지만, 이와 같은 방식은 픽셀별로 모든 각도에 대하여 직선을 검출하기 때문에 연산량이 지나치게 많다는 문제점이 있다. In FIG. 5 (a), four pixels on the coordinate plane are Huff transformed according to Equation (6) and are represented by four curves on the Huff plane. In order to detect a straight line on such a Huff plane, On the other hand. The graph of FIG. 5A is referred to as a first pixel graph, a second pixel graph, a third pixel graph, and a fourth pixel graph sequentially from the top according to the starting position at the origin, The intersection of a pixel graph

Is detected, and the intersection of the third pixel graph and the fourth pixel graph

Can be detected. In this way, applying a Hough transform to all the pixels contained on the coordinate plane included in the edge for the two lanes will cause a plurality of straight lines to be displayed as a set of parameters of angles and distances in the coordinates of the pixels contained in the edges Can be detected. However, such a method has a problem that a calculation amount is excessively large because a straight line is detected for every angle for each pixel.

In an embodiment of the present invention, when detecting a straight line corresponding to an edge through Hough transform, as shown in FIG. 5 (b), only a straight line within a predetermined angle range is detected to reduce the amount of calculation.

More specifically, a first angular range from LLA (Left Low Angle), which is an angle to the lowermost point on the edge of the left lane to LHA (Left High Angle), which is an angle to the uppermost point on the edge of the left lane, The straight line is detected only in the second angle range from the right low angle (RLA) to the lowermost point in the lane edge to the left high angle (RHA), which is the angle with respect to the uppermost point in the right lane edge. In Fig. 5 (b), it is not necessary to perform the straight line detection calculation through the Hough transform in the angular ranges other than the first angular range and the second angular range, so that the calculation amount is much reduced as compared with Fig. 5 (a) .

A detailed description of the first angular range and the second angular range applied when detecting a straight line corresponding to an edge will be described later with reference to Fig.

6 is a diagram for explaining a Hough transform method according to an embodiment of the present invention on a coordinate plane.

의 좌표를 가지는 픽셀을 지나는 빨간색의 왼쪽 차선에 대한 직선(L1)을 검출하기 위한 제1 각도 범위를 설명하기 위한 도면이다. Fig. 6 is a view showing a state in which two lanes included in the image information

And the first angular range for detecting the straight line L1 to the left lane of red passing through the pixel having the coordinates

를 지나는 직선을 나타내고, 2번으로 표시되는 직선은

의 좌표의 픽셀과 왼쪽 차선의 에지에서의 최하단 지점에 위치하는 픽셀을 연결하는 직선을 나타내고, 3번으로 표시되는 직선은

의 좌표의 픽셀과 왼쪽 차선의 에지에서의 최상단 지점에 위치하는 픽셀을 연결하는 직선을 나타낸다. The straight line indicated by numeral 1 in Fig.

And the straight line indicated by 2 indicates the straight line passing through

And a straight line connecting the pixel at the lowest point in the left lane edge and the pixel at the coordinates of

And the pixel located at the uppermost point in the edge of the left lane.

가 제1 각도 범위가 된다. 본 발명의 일 실시예에서는 제1 각도 범위 내에서만 왼쪽 차선의 에지에 대응되는 직선들을 검출하므로, 모든 각도 범위 내에서 직선들을 검출하는 종래 기술에 비해 연산량이 줄어들게 된다. In this case, the angle of the second straight line in FIG. 6 is LLA (Left Low Angle), the angle of the third straight line is LHA (Left High Angle), and the angle range from LLA to LHA

Becomes the first angle range. In one embodiment of the present invention, since the straight lines corresponding to the edges of the left lane are detected only within the first angle range, the amount of computation is reduced as compared with the prior art in which straight lines are detected within all angular ranges.

On the other hand, the calculation of the second angular range with respect to the right lane is performed based on a process substantially the same as the first angular range with respect to the left lane in Fig. 6, and thus detailed description will be omitted.

FIG. 7 is a diagram illustrating a Hough transform method according to an embodiment of the present invention in video information.

FIG. 7 is a diagram showing a plurality of straight lines indicated by purple corresponding to the edges of the detected lanes in FIG. 4, by regions of interest 712, 714, 722, and 724. At this time, a plurality of straight lines corresponding to the edges shown in Fig. 7 are straight lines detected within the first angular range and the second angular range.

In FIG. 7, the straight lines detected in correspondence with the edge of the lane overlap with the actual lane area, but occupy a relatively large area, making it difficult to determine an accurate boundary. Therefore, it is necessary to clearly display the boundary of the lane by one straight line, which will be described later with reference to Fig.

FIG. 8 illustrates an example of recognizing a lane on image information using a lane recognition method according to an embodiment of the present invention.

In FIG. 8, the width of the detected straight line corresponding to the edge of the lane is narrowly indicated as green compared to FIG. 7. The reason why the width of the green straight line in FIG. 8 is narrow is that, This is because only one straight line among the detected plurality of straight lines is displayed in Fig. 8, only one straight line is displayed because a plurality of straight lines are not required to display the boundary of the lane, and the boundary of the lane can be displayed with only one straight line corresponding to each of the left lane and the right lane .

In FIG. 7, an average angle corresponding to a plurality of straight lines is used to detect one straight line among a plurality of straight lines detected to display a lane, as shown in FIG.

The average angle of the plurality of straight lines corresponding to the left lane is calculated by Equation (7), and the average angle of the plurality of straight lines corresponding to the right lane can be calculated by Equation (8).

를 계산하고, 수학식 8에서는 m개의 직선들의 각도의 평균 값인 평균 각도

를 계산한다. In Equation (7), an average value of angles of n straight lines

In Equation (8), an average value of the angles of m straight lines

.

Next, when the average angles of the plurality of straight lines in the left lane and the right lane are calculated by Equations (7) and (8), one straight line is calculated for each of the left lane and the right lane by Equations (9) and .

과 각도

를 갖는 한개의 직선이 검출되고, 수학식 10에 의해서 복수의 직선들 중에서 거리

와 각도

를 갖는 한개의 직선이 검출된다. That is, according to Equation (9), the distance

And angle

One of the straight lines is detected, and the distance < RTI ID = 0.0 >

And angle

Is detected.

9 is a view for explaining a lane recognizing apparatus according to an embodiment of the present invention.

9, the lane recognizing apparatus according to an embodiment of the present invention includes an image obtaining unit 910, an edge detecting unit 920, a Hough transform unit 930, and a lane boundary recognizing unit 940. [

The image obtaining unit 910 obtains image information including two lanes of the left and right lanes.

The edge detection unit 920 detects edges of two lanes by setting a plurality of ROIs in the image information acquired through the image acquisition unit 910. [

The detailed configuration of the edge detecting unit 920 will be described later with reference to FIG.

The Hough transform unit 930 applies a Hough transform to edges of two lanes detected for each of a plurality of ROIs to detect a plurality of straight lines corresponding to the edges.

The lane boundary recognition unit 940 recognizes the boundaries of two lanes using a plurality of straight lines detected for each of a plurality of ROIs.

Preferably, the lane boundary recognizing unit 940 includes an angle calculating unit (not shown) for calculating an average angle that is an average value of a plurality of angles corresponding to a plurality of straight lines with respect to edges of two lanes, and an angle calculating unit And a lane processing unit (not shown) for recognizing one lane boundary for each of the edges of the lane.

10 is a diagram illustrating an edge detector according to an embodiment of the present invention.

10, an edge detector 920 according to an exemplary embodiment of the present invention includes a monochrome image converter 922, a noise eliminator 924, and an edge processor 926. Referring to FIG.

The monochrome image converting unit 922 converts a plurality of regions of interest in the image information into monochrome image information.

The noise removing unit 924 removes noise by applying a Gaussian blur filter to the monochrome image information.

The edge processing unit 926 detects a two-lane edge by applying a Sobel mask to the monochrome image information from which noise has been removed.

The above-described embodiments of the present invention can be embodied in a general-purpose digital computer that can be embodied as a program that can be executed by a computer and operates the program using a computer-readable recording medium.

The computer readable recording medium includes a magnetic storage medium (e.g., ROM, floppy disk, hard disk, etc.), optical reading medium (e.g., CD ROM, DVD, etc.).

The present invention has been described with reference to the preferred embodiments. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims. Therefore, the disclosed embodiments should be considered in an illustrative rather than a restrictive sense. The scope of the present invention is defined by the appended claims rather than by the foregoing description, and all differences within the scope of equivalents thereof should be construed as being included in the present invention.

Claims (12)

Obtaining image information including two lanes of left and right lanes;
Detecting edges of the two lanes by setting a plurality of regions of interest in the obtained image information;
Applying a Hough transform to edges of the two lanes detected for each of the plurality of ROIs to detect a plurality of straight lines within a predetermined angle range corresponding to the edges; And
Recognizing a boundary of the two lanes using the plurality of straight lines detected for each of the plurality of ROIs,
The predetermined angular range
A first angular range from an LLA (Left Low Angle), which is an angle with respect to a lowermost point on the edge of the left lane to an LHA (Left High Angle), which is an angle with respect to an uppermost point on the edge of the left lane, And a second angle range from RLA (Right Low Angle), which is an angle to the lowermost point on the edge, to RHA (Left High Angle), which is an angle with respect to the uppermost point on the edge of the right lane, Way. The method according to claim 1,
The step of detecting the plurality of straight lines corresponding to the edge by applying the Hough transform
Wherein a plurality of straight lines represented by a set of parameters of angles and distances are detected for each of the coordinates of pixels included in the edge. delete The method according to claim 1,
The step of detecting the edges of the two lanes
Converting the plurality of ROIs in the acquired image information into monochrome image information;
Removing noise by applying a Gaussian blur filter to the monochrome image information; And
And detecting a contour of the two lanes by applying a sobel mask to the noise-removed monochrome image information. The method according to claim 1,
The step of recognizing the boundaries of the two lanes using the plurality of straight lines
Calculating an average angle that is an average value of a plurality of angles corresponding to the plurality of straight lines with respect to edges of the two lanes; And
And recognizing one lane boundary for each of the two lane edges using the average angle. The method according to claim 1,
The plurality of regions of interest
Are grouped individually or in a predetermined group so as to be subject to parallel processing. An image obtaining unit for obtaining image information including two lanes of left and right lanes;
An edge detector configured to detect edges of the two lanes by setting a plurality of ROIs in the obtained image information;
A Hough transform unit for applying a Hough transform to edges of the two lanes detected for each of the plurality of ROIs to detect a plurality of straight lines within a predetermined angle range corresponding to the edges; And
And a lane boundary recognizing unit for recognizing a boundary of the two lanes using the plurality of straight lines detected for each of the plurality of ROIs,
The predetermined angular range
A first angular range from an LLA (Left Low Angle), which is an angle with respect to a lowermost point on the edge of the left lane to an LHA (Left High Angle), which is an angle with respect to an uppermost point on the edge of the left lane, And a second angle range from RLA (Right Low Angle), which is an angle to the lowermost point on the edge, to RHA (Left High Angle), which is an angle with respect to the uppermost point on the edge of the right lane, Device. 8. The method of claim 7,
The Hough transform unit
Wherein the plurality of straight lines are detected by detecting a plurality of straight lines represented by a set of parameters of angles and distances according to the coordinates of the pixels included in the edge. delete 8. The method of claim 7,
The edge detector
A monochrome image conversion unit for converting the plurality of ROIs in the acquired image information into monochrome image information;
A noise removing unit for removing noise by applying a Gaussian blur filter to the monochrome image information; And
And detecting an outline of the two lanes by applying a Sobel mask to the noise-removed monochrome image information. 8. The method of claim 7,
The lane boundary recognizing unit
An angle calculating unit for calculating an average angle that is an average value of a plurality of angles corresponding to the plurality of straight lines with respect to the edges of the two lanes; And
And a lane processing unit for recognizing one lane boundary for each of the edges of the two lanes using the average angle. 8. The method of claim 7,
Each of the plurality of regions of interest
And are grouped individually or in a predetermined group so as to be subject to parallel processing.

Images