KR20180092765A - Apparatus and method for identifying lanes by using top view image - Google Patents

Abstract

Disclosed are a method and an apparatus for recognizing lanes, which obtain images from a plurality of cameras provided in a vehicle, generate a top view image by matching the images, detect a driving lane when the vehicle is running, detect a parking line when the vehicle is being parked, notify a driver that the lane of the vehicle cannot be changed if the driving lane is detected as a solid line, and notify the driver that the lane of the vehicle can be changed if the driving lane is detected as a dotted line. Accordingly, the present invention can improve driving performance by accurately recognizing the lane.

Description

[0001] APPARATUS AND METHOD FOR IDENTIFYING LANES BY USING TOP VIEW IMAGE [0002]

The present invention relates to a system and method for recognizing a lane using a top view image, and more particularly, to a device and a method for recognizing a driving lane for driving a vehicle and a parking line for parking by analyzing a top view image .

With the development of vehicle technology, there is a growing interest in autonomous vehicles (intelligent vehicles or smart cars) that travel on their own without the assistance of drivers. The autonomous vehicle recognizes the surrounding obstacles while driving and adjusts the speed and brakes, and confirms the current position of the autonomous vehicle and determines the route to the target position by itself.

As interest in these autonomous vehicles has increased, both domestic and international automakers are developing their own self-propelled vehicles, and they are also investing and researching in universities and research institutes.

On the other hand, one of the most important requirements in an autonomous vehicle is to acquire accurate peripheral information, and if the technology is supported, the vehicle can be driven with stability. There are various technologies for recognizing the driving lane and the parking line of the vehicle. However, the technology for obtaining the accurate information by reducing the influence from the surrounding environment does not yet appear, so that the needs of the market are not sufficiently satisfied.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to propose a method for accurately recognizing a driving lane at the time of driving a vehicle.

Another object of the present invention is to propose a method for accurately recognizing a parking line when parking a vehicle.

Yet another object of the present invention is to improve the performance of an autonomous vehicle at low cost by realizing lane recognition without additional sensors.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are not intended to limit the invention to the particular form disclosed. ≪ / RTI >

According to another aspect of the present invention, there is provided a lane recognition method comprising: acquiring images photographed in different directions from a plurality of cameras provided in a vehicle equipped with a lane recognition device; A step of generating a top view image of the vehicle, a step of detecting a driving lane using a top view image when it is determined that the vehicle is running, and a step of detecting a parking line using the top view image And the driving lane is a solid line or a dotted line. When the driving lane is detected as a solid line, it is informed that the vehicle can not be changed, and when the driving lane is detected as a dotted line,

The lane recognizing method may further include the step of detecting a candidate area in which the vehicle is parkable when it is determined that the vehicle is parked but the parking line is not detected.

The step of detecting the driving lane may detect the driving lane by extracting the area of interest from the top view image and determining the candidate points through boundary detection in the area of interest.

The step of detecting the driving lane may determine the driving lane as a solid line when the detected driving lane continues to the vanishing point continuously, and may determine the driving lane as a dotted line when the detected driving lane does not exceed the predetermined pixel.

The step of detecting the parking line may detect the vertices of the parking space in the top view image and determine the border of the space defined by the vertexes as a parking line.

The candidate area can be detected by comparing the area detected by the stereo algorithm with the size of the space required for parking the vehicle.

According to another aspect of the present invention, there is provided a lane recognition apparatus comprising: a plurality of image capturing units provided in a vehicle equipped with a lane recognizing device and capturing images in different directions; A traveling lane recognizing section for determining whether a vehicle is running or parked; a driving lane recognizing section for detecting a driving lane using a top view image when the vehicle is judged to be traveling; And a parking line recognizing unit for detecting a parking line by using the top view image when it is judged that the vehicle is parked. The driving lane is a solid line or a dotted line, and when the driving lane is detected as a solid line, When the lane is detected as a dotted line, it is informed that the vehicle is permitted to change.

According to the embodiments of the present invention, the following effects can be expected.

First, the driving lane can be accurately recognized while the autonomous driving vehicle is running, and the driving performance can be improved.

Secondly, the parking line can be accurately recognized at the time of parking the autonomous vehicle, and the parking performance can be improved.

Third, the increase in cost required for performance improvement can be minimized, and a high-efficiency performance improvement can be achieved even at a low cost.

The effects obtainable in the embodiments of the present invention are not limited to the effects mentioned above, and other effects not mentioned can be obtained from the description of the embodiments of the present invention described below by those skilled in the art Can be clearly understood and understood. In other words, undesirable effects of implementing the present invention can also be obtained by those skilled in the art from the embodiments of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention. It is to be understood, however, that the technical features of the present invention are not limited to the specific drawings, and the features disclosed in the drawings may be combined with each other to constitute a new embodiment.
1 is a view for explaining a process of acquiring a top view image related to the proposed embodiment.
2 is a block diagram showing a configuration of a lane recognizing apparatus according to the proposed embodiment.
FIG. 3 is a view for explaining an image capturing process according to the embodiment of the present invention.
4 is a view for explaining a driving lane recognition process according to the embodiment of the present invention.
5 to 7 are diagrams for explaining a parking line recognition process according to the embodiment of the present invention.
8 is a flowchart showing a lane recognition method according to the proposed embodiment.
9 is a diagram for explaining a simulation process for the proposed embodiment.
10 is a view for explaining an actual road test procedure for the proposed embodiment.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings, which will be readily apparent to those skilled in the art. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In order to clearly illustrate the present invention, parts not related to the description are omitted, and similar parts are denoted by like reference characters throughout the specification.

The terms used in this specification will be briefly described and the present invention will be described in detail.

While the present invention has been described in connection with what is presently considered to be the most practical and preferred embodiment, it is to be understood that the invention is not limited to the disclosed embodiments. Also, in certain cases, there may be a term selected arbitrarily by the applicant, in which case the meaning thereof will be described in detail in the description of the corresponding invention. Therefore, the term used in the present invention should be defined based on the meaning of the term, not on the name of a simple term, but on the entire contents of the present invention.

When an element is referred to as "including" an element throughout the specification, it is to be understood that the element may include other elements as well, without departing from the spirit or scope of the present invention. Also, the terms "part," " module, "and the like described in the specification mean units for processing at least one function or operation, which may be implemented in hardware or software or a combination of hardware and software . In addition, when a part is referred to as being "connected" to another part throughout the specification, it includes not only "directly connected" but also "connected with other part in between".

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

1 is a view for explaining a process of acquiring a top view image related to the proposed embodiment.

Recently, research on panorama technology for synthesizing a plurality of images into a single image has been actively conducted. Techniques for applying the panoramic technology to a vehicle and shooting the surroundings of the vehicle without a square are also being discussed. The method of shooting the surroundings of the vehicle by installing cameras on the front, rear, left, and right sides of the vehicle is called "top view ) Or an 'around view'.

It is necessary to perform a calibrating process to reflect the characteristics of each camera before matching the captured images using a plurality of cameras disposed in the vehicle. This correction process can be performed according to the following equation (1).

[Equation 1]

In Equation (1), fc1 and fc2 represent different focal lengths, and cc1 and cc2 represent different principal points.

When a CCD (Charge-Coupled Device) sensor is used, fc1 and fc2 generally have similar values due to the existence of an error. However, when the CCD pixel is rectangular, fc2 / fc1 has a value other than 1. Since a wider angle of view of the camera results in severe distortion, a camera having an angle of view of about 180 'is typically used to obtain a forward image of the vehicle.

For correction of distortion, a radial distortion model and a tangent distortion model can be used. The radial distortion compensates the distance away from the radius r by using three parameters kc1, kc2, and kc3, while the tangent distortion compensates the amount of movement by using two parameters kc4 and kc5. The process of calculating the parameters for correcting the distortion of the camera can be performed according to Equation (2) below, and the coordinates of the image can be corrected by applying the calculated distortion correction parameters to each pixel.

&Quot; (2) "

In Equation (2), xd is the x-coordinate where the distortion is corrected, r is the distance to the center, x is the x-coordinate of the distorted image, and y is the y-coordinate of the distorted image.

In addition, internal and external parameters must be calculated to use the camera. At this time, the rotation and movement between the coordinate system of the camera and the actual ground plane can be performed using the internal parameters and the external parameters calculated above, and the images can be accurately matched only if all the cameras are aligned at the same point.

In order to proceed with the error correction, the vehicle is placed on a predetermined size plate as shown in FIG. The reason for using such a chess board type plate is that the matching point for matching the image can be accurately found and accurate homography value can be obtained through the homography matrix used in the matching process. (3) is a homography transformation matrix, and h11, h12, h13 and h14 represent rotation, scaling, shearing and reflection, h13 and h23 represent parallel movement, h31 and h32 represent a perspective change . x and y represent the respective pixels.

&Quot; (3) "

In Equation (3), the image is corrected and matched using the homography parameters calculated for each image, and the images obtained from the different cameras are converted, and the images are matched based on the converted image. When the matching is completed, a top view image is generated to confirm the vicinity of the vehicle.

2 is a block diagram showing a configuration of a lane recognizing apparatus according to the proposed embodiment. The lane recognizing device of FIG. 2 shows specific configurations of a device for recognizing a lane using the top view image described above with reference to FIG. The lane recognizing apparatus 1000 according to an exemplary embodiment includes an image photographing unit 110, an image processing unit 120, an image matching unit 130, a travel determining unit 140, a driving lane recognizing unit 150, (160) and a controller (170). However, the present invention is not limited to such a configuration, and may further include other general configurations.

Hereinafter, the term " lane " is used to mean both a driving lane existing around the vehicle when driving the vehicle and a parking line existing around the vehicle when the vehicle is parked.

The lane recognizing apparatus 1000 is a device mounted on an autonomous driving vehicle (intelligent vehicle, smart car) to acquire and process a top view image to assist driving and parking. Hereinafter, the specific operation of the lane recognizing apparatus 1000 will be described taking the components of the lane recognizing apparatus 1000 shown in FIG. 1 as an example.

First, the image capturing unit 110 refers to an image capturing unit (e.g., a camera) installed in a vehicle and capturing an image. The lane recognizing apparatus 1000 may include a plurality of image capturing units 110. Each of the image capturing units 110 is mounted in four directions of front, rear, left, and right of the vehicle, and photographs the surroundings of the vehicle in different directions. The image capturing unit 110 can generate image data by photographing the surroundings of the vehicle.

The image processing unit 120 processes or corrects the image data so that the images captured by the image capturing unit 110 can be registered. The image processor 120 may correct image distortion as described above with reference to FIG. 1, and may perform a preprocessing process to improve image quality or remove noise. Also, the image processing unit 120 may process the image data through a method of adjusting the size of the image or adjusting the tilt.

The image matching unit 130 matches the image data captured by the plurality of image capturing units 110 to generate a top view image. The top view image can be generated in such a way that the images taken by the image capturing units 110 arranged on the front, rear, left and right sides of the vehicle are connected in a panoramic form. The generated top view image may be stored in the form of data, and the lane recognition device 1000 is a reference for recognizing the driving lane and the parking line.

The travel determining unit 140 determines whether the vehicle on which the lane recognizing apparatus 1000 is mounted is traveling or parked. Whether or not the vehicle is traveling and whether or not the vehicle is parked can be determined from various information such as the moving speed of the vehicle, the gear state of the vehicle, and the driving pattern of the driver. The travel determining unit 140 collects data from various sensors mounted on the vehicle Determine whether the vehicle is traveling or not.

The driving lane recognizing unit 150 recognizes the driving lane by analyzing the top view image generated by the image matching unit 130 when the vehicle is running. A driving lane means a line which is located on both sides of the vehicle when the vehicle is running and separates the lane, and may be a solid line or a dotted line. The driving lane recognizing unit 150 can distinguish between a dotted line and a solid line, and in the case of a dotted line, the vehicle can change lanes to a side lane unlike a solid line.

The parking line recognizing unit 160 recognizes the parking line by analyzing the top view image generated by the image matching unit 130 when the vehicle is parked. The parking line means a line for identifying a parking space when the vehicle is parked. On the other hand, a vehicle may be parked in a space without a parking line, not in a parking space divided by a parking line. The parking line recognizing unit 160 recognizes the parking space even when there is no parking line but the parking lot is not parked when it is judged that the vehicle is parked. Details will be described later.

The control unit 170 controls overall operation of the lane recognition apparatus 1000 between the components so that the lane recognizing apparatus 1000 operates according to its function and purpose. For example, when the control unit 170 controls the image matching unit 130 to match the image data photographed by the image capturing unit 110, or when the driving determination unit 140 determines that the vehicle is running, The control unit 150 can be controlled to recognize the lane around the vehicle.

FIG. 3 is a view for explaining an image capturing process according to the embodiment of the present invention. As shown in FIG. 3, when four image radiographing units are arranged along the front, back, right, and left four directions of the vehicle 300, each image radiographing unit captures images in different directions to generate image data. The images 310, 320, 330 and 340 photographed in different directions are corrected and processed by the image processor, and then registered. The lane recognition device acquires the top view image through matching.

Hereinafter, a detailed process of recognizing the driving lane and the parking line by the lane recognizing device through the above-described components will be described. First, a process of recognizing a driving lane will be described with reference to FIG. 4, and a process of recognizing a parking line will be described with reference to FIGS. 5 to 7. FIG.

4 is a view for explaining a driving lane recognition process according to the embodiment of the present invention. In the case of ordinary roads, the lanes are primary colors such as white or yellow, and there are two or more lanes on the road. If two lanes on either side of the vehicle are straight and parallel to each other, the two lanes meet at a vanishing point located in front of the vehicle.

According to the proposed embodiment, the lane recognizing device recognizes the driving lane using the feature of the lane. First, the lane recognition apparatus detects a boundary in a top view image, designates a candidate area having a lane as an area of interest, and extracts a lane candidate point. The region of interest (ROI) may include an area ranging from a portion where a lane appears to a vanishing point in a top view image, and the lane recognizing device may convert an image corresponding to a region of interest into a gray scale image Convolution of the Gaussian kernel for back-noise cancellation is performed.

The lane recognizing device detects a boundary in order to extract a lane from the ROI image. The lane recognizing device detects two points (for example, a (ax, ay), b (bx, ay)) to see if the distance between two points satisfies the width of the lane (w). The distance between two points can be expressed as wab = bx-ax. If two points with an error between the width of the lane, which is a preset value, are less than the threshold value, the lane recognizing device compares the strength and direction of the boundary where the two points are located and recognizes the lane when the strengths of the two boundaries are similar, And the midpoint of the two selected points is determined as a candidate point of the lane.

&Quot; (4) "

Equation (4) means coordinates of the above-mentioned candidate point. On the other hand, the lane recognition apparatus performs Probabilistic Hough Transform (PHT) using candidate points selected to be located in the lane, and then extracts only the lines toward the vanishing point by applying the RANSAC algorithm. Through this process, lanes can be extracted from the top view image.

On the other hand, if the lane is a solid line, it continues to the vanishing point, but if it is a dotted line, the lane is broken at regular intervals. According to the probabilistic Hough transform, only a line segment having a predetermined length or more is detected, and the lane recognizing device determines a solid line when the extracted line segment exceeds a predetermined pixel (for example, 30 pixels) after performing the probabilistic Hough transform. If not, it is decided by dotted line. At this time, since it may be more important to correct the external error than the length of the dotted lane itself, the lane recognizing device can recognize the dotted line from a line having a size smaller than the actually detected dotted line.

On the other hand, when the detected lane is determined to be a solid line, the lane recognizing device can transmit a signal such that the mounted vehicle does not change the lane. On the other hand, when the lane recognizing device determines that the lane recognizing device determines that the lane is recognized, It is possible.

5 to 7 are views for explaining a parking line recognizing process according to the embodiment of the present invention. 5 shows a top view image for detecting an obstacle behind the vehicle. Generally, parking lots are mostly primary colors, such as white or yellow, similar to road lanes. Further, the parking space has a shape of a rectangle or a parallelogram, and exists continuously at regular intervals.

Fig. 6 shows various types of parking spaces. In Fig. 6, white indicates parking lines and black indicates parking spaces.

In the case of a rectangular parking space as shown in Figs. 6 (a) and 6 (b), the lane recognizing device detects a candidate area in which the angles of the rectangle vertices constituted by the parking lines have 90 '. In the case of the parking space of the parallelogram shape as shown in FIG. 6 (c), the lane recognizing device detects a candidate area in which the angles of the vertexes of the parallelogram formed by the parking lines are within a predetermined predetermined angle range. Then, the lane recognizing device recognizes an area that is close to the current position of the vehicle and has no obstacle in the candidate area as a parking space.

The process of detecting the vertices described above can be performed by the Harris Corner algorithm. This process can be understood as a method in which the image change amount E becomes locally maximized when moving by 1 pixel on the top view image, and can be performed according to the following equation (5).

&Quot; (5) "

In Equation (5), I (x, y) denotes coordinates of x and y in the image, and W denotes a local window.

On the other hand, assuming that the amount of movement is very small in Equation (5), if I is differentiated and linearly approximated, the following Equation (6) is derived.

&Quot; (6) "

From Equation (5) and Equation (6), the following Equation (7) is derived,

&Quot; (7) "

From Equation (7), the following Equation (8) is derived.

&Quot; (8) "

When the eigenvalues of the derived 2x2 matrix of Equation (8) are denoted by [lambda] 1 and [lambda] 2, the amount of image change becomes maximum when moving in the direction of the eigenvector of [lambda] 1 and minimized when moving in the direction of the eigenvector of [lambda] 2. When two eigenvalues are obtained, the vertex is determined when the two values are both large.

If any one vertex is determined, other vertexes present in the horizontal and vertical directions of the detected vertex can be easily detected. The lane recognizing device calculates a pixel value of one of the vertexes and forms a straight line connecting the distant vertexes of two adjacent vertexes in the horizontal direction to form a straight line connecting the vertexes of two adjacent vertexes in the vertical direction, A possible candidate region can be detected as a rectangle.

FIG. 7 shows a process of recognizing a parking space even if there is no parking line, unlike FIG. The lane recognizing device can determine that the vehicle is parked by the travel judging unit even in an area where there is no parking line, thereby recognizing the surrounding parking space. To recognize the parking space in the absence of a parking line, a stereo algorithm is used to measure the distance between the obstacle and the vehicle.

In the process of matching the images to obtain the top view image, the lane recognition device partially overlaps with the images captured by the other two cameras captured by the specific camera. Applying a stereo algorithm based on the triangulation method on the overlapping part, the coordinate system of the overlapping part is converted from two dimensions to three dimensions, so that it is possible to check how far the vehicle or obstacle around it is separated.

Generally, since the information acquired most by the camera attached to the vehicle is the color information of the road surface, the RGB value of the pixel acquired from the captured image becomes the color of the road, have. Considering that there is a high likelihood of parallel parking in the absence of a parking line, the lane recognition device may calculate the minimum space required for the backward parallel parking and compare it with the detected candidate area using the stereo algorithm, have.

On the other hand, the stereo algorithm shown in FIG. 7 will be described in detail. The lane recognition device calculates the distance between the camera and the object according to Equation (9) below.

&Quot; (9) "

Equation 9 is derived by a trigonometric function, and each parameter means the length or angle shown in Fig. According to the theorem of Equation (9) and Pythagoras,? 1 can be summarized as the following Equation (10), and a similar method is applied to? 2.

&Quot; (10) "

The two angles &thetas; 1 and &thetas; 2 are summarized as the following equation (11).

&Quot; (11) "

Based on Equation (11), the distance d between the camera and the object is expressed by Equation (12) below.

&Quot; (12) "

Therefore, the parking space can be constantly searched by applying the stereo algorithm to the top view image, with the space required for parking the vehicle in advance according to the size of the vehicle.

8 is a flowchart showing a lane recognition method according to the proposed embodiment. FIG. 8 shows a process of the lane recognizing device operating according to the above-described embodiments according to a time series flow. Therefore, although the detailed description is omitted in FIG. 8, it is needless to say that the description of FIGS. 2 to 7 may be applied to the same points.

When the vehicle starts running (S810), the lane recognizing device determines whether the vehicle is traveling at a constant speed or more (S820). Driving above a certain speed may mean traveling on the road, while driving below a certain speed may mean parking. As described above, it is needless to say that the criteria for distinguishing between driving and parking can be further considered in addition to the speed, the gear state, and the driving pattern of the driver.

If the lane recognizing device determines that the vehicle is traveling on the road at a predetermined speed or more, the lane recognizing device analyzes the top view image to recognize the driving lane (S830). The top view image can be generated by matching images obtained from four or more cameras disposed in the vehicle, and the lane recognition device can recognize the driving lane by extracting the area of interest from the top view image. On the other hand, the lane recognizing device can distinguish the recognized lane by the solid line and the dotted line (S840). The lane recognizing device can visually or audibly inform the driver of the vehicle or the vehicle equipped with the lane recognizing device whether the lane recognizing device can be changed You can tell.

The lane recognizing device continuously recognizes whether the vehicle is traveling at a constant speed or more after recognizing the lane (S850), and the process is performed in the same manner as the process of S820. On the other hand, if it is determined that the vehicle travels less than the predetermined speed and is parked, the lane recognizing device analyzes the top view image to recognize the parking line (S860).

As described above, the process of recognizing the parking line can be performed by extracting the vertices of the candidate area that can be parked, and the lane recognition device allows the vehicle to move to a parking space where there is no obstacle among the candidate areas identified through the parking line (S870). When parking is completed, the running of the vehicle is terminated (S880). In the meantime, the lane recognizing device detects a candidate area in which parking is possible even if there is no parking space in steps S860 and S870, and provides a notification to the driver or the vehicle can do.

Meanwhile, the lane recognition method described in Fig. 8 can be implemented in a general-purpose digital computer that can be created as a program that can be executed in a computer and operates the program using a computer-readable medium. Further, the structure of the data used in the above-described method can be recorded on a computer-readable medium through various means. Recording media that record executable computer programs or code for carrying out the various methods of the present invention should not be understood to include transient objects such as carrier waves or signals. The computer-readable medium may comprise a storage medium such as a magnetic storage medium (e.g., ROM, floppy disk, hard disk, etc.), optical readable medium (e.g., CD ROM, DVD, etc.).

9 is a diagram for explaining a simulation process for the proposed embodiment. 9 shows a simulation result of the operation of the lane recognizing apparatus described above. In the simulation process of FIG. 9, a camera having a parameter of 1 / 3.6 'CMOS, an angle of view 185', a resolution of 640 * 480, Of the vehicle. The lane was simulated with a width of 15cm on the basis of the Road Traffic Act and a length of 5m and a length of 2.3m according to the building method.

9A and 9B, the binocular image is binarized by the above-described lane recognizing apparatus and method, and the area disappearing from the vanishing point and recognized as a lane can be confirmed by a red line. In addition to the straight lane in FIG. 9 (a), corner lanes in FIG. 9 (b) are also sensed, and in the case of a corner lane, the lane disappears through two vanishing points.

FIG. 9 (c) shows a simulation result for a parking line, and simulation is performed by considering that a parking line constitutes a rectangular or parallelogram. According to the proposed lane recognition device and method, it is confirmed that the parking line is correctly recognized and outputted in red.

10 is a view for explaining an actual road test procedure for the proposed embodiment. In FIG. 10, there is shown a result of performing an actual experiment using a camera with the same parameters as the simulation of FIG. 9 and an RC car (L490XW380XH280, motor BLDC 540 class).

In the actual experiment, the reaction speed of the algorithm was recognized without any problem even in a vehicle traveling at about 60 km / h, and the result is shown in FIG. 10 (a). As can be seen from Fig. 10 (a), it can be seen that both the yellow solid line and the white dotted lane are normally recognized.

10 (b) and 10 (c) show the result of recognizing the parking line and the result of recognizing the parking space when there is no parking line. In the case of parking, the experiment was conducted based on the vehicle running at a speed of 10 km / h. The area where the parking is not possible is red. When the candidate area where the parking is possible is detected, the green line is displayed on the screen, Respectively.

It will be understood by those skilled in the art that the foregoing description of the present invention is for illustrative purposes only and that those of ordinary skill in the art can readily understand that various changes and modifications may be made without departing from the spirit or essential characteristics of the present invention. will be. It is therefore to be understood that the above-described embodiments are illustrative in all aspects and not restrictive. For example, each component described as a single entity may be distributed and implemented, and components described as being distributed may also be implemented in a combined form.

The scope of the present invention is defined by the appended claims rather than the detailed description and all changes or modifications derived from the meaning and scope of the claims and their equivalents are to be construed as being included within the scope of the present invention do.

110: image capturing unit 120:
130: image matching unit 140:
150: driving lane recognition unit 160: parking line recognition unit

Claims (7)

A method for recognizing a lane by using a top view image of a lane recognition device,
Acquiring images photographed in different directions from a plurality of cameras provided in a vehicle equipped with the lane recognition device;
Generating a top view image of the vehicle by matching images photographed by the plurality of cameras;
Detecting a driving lane using the top view image when it is determined that the vehicle is running; And
And detecting the parking line using the top view image when it is determined that the vehicle is parked,
Wherein the driving lane is a solid line or a dotted line, and when the driving lane is detected as a solid line, it is informed that it is impossible to change to the vehicle, and when the driving lane is detected as a dotted line, Recognition method. The method according to claim 1,
In the lane recognition method,
Further comprising the step of detecting the candidate area in which the vehicle is parkable if it is determined that the vehicle is parked but the parking line is not detected. The method according to claim 1,
Wherein the detecting of the driving lane comprises extracting a region of interest from the top view image and determining candidate points through boundary detection within the region of interest to detect the driving lane. The method of claim 3,
Wherein the step of detecting the driving lane comprises the step of determining the driving lane as a solid line when the detected driving lane continues to the vanishing point continuously and determining the driving lane as a dotted line when the detected driving lane does not exceed the predetermined pixel The lane recognition method. The method according to claim 1,
Wherein the step of detecting the parking line detects the vertices of the parking space in the top view image and determines the border of the space defined by the vertices as the parking line. 3. The method of claim 2,
Wherein the candidate area is detected by comparing the area detected by the stereo algorithm with the size of the space required for parking the vehicle. A lane recognition apparatus for recognizing a lane using a top view image,
A plurality of image capturing units provided in a vehicle on which the lane recognizing apparatus is mounted and each capturing an image in different directions;
An image matching unit for matching the images photographed by the plurality of image capturing units and generating a top view image of the vehicle;
A traveling determining unit determining whether the vehicle is traveling or being parked;
A driving lane recognition unit for detecting a driving lane using the top view image when it is determined that the vehicle is running; And
And a parking line recognition unit for detecting a parking line using the top view image when it is determined that the vehicle is parked,
Wherein the driving lane is a solid line or a dotted line, and when the driving lane is detected as a solid line, it is informed that it is impossible to change to the vehicle, and when the driving lane is detected as a dotted line, Recognition device.

Images