KR20150117555A - Lane Detecting Method and Apparatus for Estimating Front and Rear Lane Based on Camera Image and Vehicle Information - Google Patents

Abstract

The present invention relates to a method and an apparatus for detecting a lane, further accurately estimating a lane and recognizing a front/rear lane in comparison with a conventional technique that depends on a method that combines vehicle information such as camera image information, deviation speed/heading information or the like in case of a sudden change of characteristics of a road such as a junction or an interchange, thereby more accurately performing the process of determining the degree of danger to achieve a lane maintaining autonomous drive, a lane changing autonomous drive or the like in a lane-based autonomous drive system. The apparatus for detecting a lane comprises: a front lane detection part; a movement distance estimation part; a rear lane estimation part; and a lane information providing part.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a lane detecting method and apparatus for detecting a front lane by combining camera image information and vehicle information,

The present invention relates to a lane detecting method and apparatus, and more particularly, to a lane detecting method and apparatus capable of recognizing front and rear lanes of a vehicle traveling road by fusing vehicle information such as camera image information and vehicle speed / heading information.

The conventional lane recognition method is a method of estimating a lane by extracting a straight line component of a lane of a driving lane using only the front image of the vehicle, and the lane detection performance may be degraded in a rapidly varying section due to characteristics of the road environment.

For example, when a road is changed from a straight line / curve to a curve / straight line, curvature or slope of a road may suddenly change at a branch point or a confluence point, so estimating a lane with a partial image may cause distortion There is a problem.

Domestic Registration No. 10-1339255 (issued on December 09, 2013)

SUMMARY OF THE INVENTION It is therefore an object of the present invention to improve lane recognition performance by using vehicle information such as camera image information, The present invention provides a lane detection method and apparatus capable of providing accurate lane information reflecting characteristics of a lane.

According to another aspect of the present invention, there is provided a lane detecting apparatus for detecting a lane on a driving road, the apparatus comprising: A front lane detecting unit detecting the front lane information by relative position and updating the lane information by a predetermined period; A moving distance estimating unit that receives the vehicle speed and heading information and estimates a moving distance of the vehicle in the corresponding path at every predetermined period; A backward lane estimator for estimating the rear lane information for each moving distance of the vehicle using information corresponding to a rear position according to a moving position of the vehicle in the front lane information; And a lane information providing unit for performing forward road modeling using the front lane information and backward road modeling using the rear lane information for the present vehicle position for each moving distance of the vehicle.

The rear lane estimator may estimate the rear lane information by correcting the information corresponding to the backward position in the front lane information to lane information for the relative position to the rear in the current child position.

The moving distance estimating unit may vary the estimation period of the moving distance in accordance with the lane curvature or slope estimated from the camera image information, in addition to the vehicle speed and heading information.

The lane information providing unit may perform the modeling for each section having the same lane curvature or slope.

According to another aspect of the present invention, there is provided a lane detection method for detecting a lane on a driving road in a vehicle, the method comprising: detecting a front lane information by a relative position with respect to a vehicle from camera image information about a front lane of the vehicle; ; Estimating a moving distance of the vehicle on the corresponding path at predetermined intervals by receiving the vehicle speed and the heading information; Estimating the rear lane information for each moving distance of the vehicle using the information corresponding to the rear position according to the moving position of the vehicle in the front lane information; And performing forward road modeling using the front lane information and backward road modeling using the rear lane information for a current child position for each moving distance of the vehicle.

According to the lane detecting method and apparatus of the present invention, it is possible to estimate the lane more accurately than in the past, for example, when the road suddenly changes in characteristics such as a confluence point or a turning point depending on the vehicle information fusion method such as the camera image information and the vehicle speed / have.

Further, according to the lane detecting method and apparatus of the present invention, it is possible to more precisely recognize the front / rear lane, so that it is more accurate in determining the risk for lane keeping autonomous driving, Thereby enabling processing.

In addition, the lane departure warning system and the lane keeping system can be utilized in various systems requiring lane information to provide accurate lane information.

FIG. 1 is a view for explaining a lane detecting apparatus for recognizing front and rear lanes according to an embodiment of the present invention.
FIG. 2 is a flowchart for explaining the operation of the lane detecting apparatus of FIG. 1. FIG.
3 is a diagram for explaining front lane information detection in FIG.
FIG. 4 is a diagram for explaining rear lane information detection in FIG. 2. FIG.
Fig. 5 is a diagram for explaining road modeling in Fig. 2. Fig.
FIG. 6 is a diagram for explaining the accuracy of extraction of the lane of the present invention from the conventional technique.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings. In the drawings, the same components are denoted by the same reference symbols as possible. In addition, detailed descriptions of known functions and / or configurations are omitted. The following description will focus on the parts necessary for understanding the operation according to various embodiments, and a description of elements that may obscure the gist of the description will be omitted.

Also, some of the elements of the drawings may be exaggerated, omitted, or schematically illustrated. The size of each component does not entirely reflect the actual size, and therefore the contents described herein are not limited by the relative sizes or spacings of the components drawn in the respective drawings.

FIG. 1 is a view for explaining a lane sensing apparatus 100 recognizing front and rear lanes according to an embodiment of the present invention.

1, the lane detecting apparatus 100 according to an embodiment of the present invention is applied to a lane-based autonomous driving system, a lane departure warning system, a lane keeping system, and the like, A forward lane detecting unit 110, a moving distance estimating unit 120, a backward lane estimating unit 130, and a lane information providing unit 140. In addition, the front lane detecting unit 110 may further include a camera for providing image information on the front of the vehicle. However, a camera that can be installed at an appropriate position such as on the vehicle roof or on the front glass so as to face the front of the vehicle is manufactured separately from the lane detecting device 100 and operates as an external device to detect the front lane of the lane detecting device 100, And may provide the image information of the vehicle ahead to the vehicle 110.

The components of the lane detecting apparatus 100 according to an embodiment of the present invention may be hardware, software, or a combination thereof. For example, a predetermined application program in the form of software may be stored in a predetermined memory and executed to interoperate with each component of the lane detecting apparatus 100 to perform or assist the corresponding function.

The front lane detecting unit 110 detects information of a front lane (e.g., lane driving lane) by position relative to the vehicle from the camera image information about the front lane of the vehicle (see FIG. 3) Can provide star front lane information.

The travel distance estimating unit 120 receives the vehicle information including the vehicle speed and the heading information, and estimates the travel distance of the vehicle in the corresponding route at predetermined intervals. Information on the speed of the running vehicle can be extracted from the CAN (Controller Area Network) signal of the vehicle, and in some cases, it is possible to directly receive the information from the vehicle speed or RPM sensor mounted on the vehicle instead of the CAN signal It is also possible. The heading information is information such as an angle indicating the traveling direction of the vehicle in the clockwise direction with respect to the north, and may be a gyro sensor, an inertial sensor, a geomagnetic sensor, or a GPS (Global Positioning System) Or may be obtained by operating one or more positioning systems in combination, or may be obtained by combining two or more such sensors. In some cases, heading information may also be included in the CAN signal and extracted from it.

The backward lane estimating unit 130 estimates the backward lane information for each moving distance of the vehicle using the information corresponding to the rear position according to the moving position of the vehicle among the front lane information detected by the front lane detecting unit 110 4).

The lane information providing unit 140 performs a lane information modeling based on the front lane information detected by the front lane detecting unit 110 and the back lane information estimated by the rear lane estimating unit 130, To perform a predetermined rearward road modeling. The lane information providing unit 140 can provide the lane information in the form of a predetermined function for the front road and the rear road through road modeling (refer to FIG. 5, described below).

Hereinafter, the operation of the lane detecting apparatus 100 according to an embodiment of the present invention will be described in detail with reference to the flowchart of FIG.

Lane ahead detection unit 110 the camera image information detected by the relative position information of the front lane deviation, and L _0, the front-specific relative position and updated every predetermined period lane information (for deviation from the front L _N .., .. L _{N + M} ) may be detected and stored in a memory (not shown) and provided (S110).

For example, the front lane detecting unit 110 may detect the lane by analyzing the brightness, angle, and the like of the camera image information. The lane detecting unit 110 may detect a lane in front of the lane of the lane ahead of the lane by a predetermined relative distance ₀ to d ₁ , d _N to d _{N +1} , d _{N + M} to d _{N + M + 1} . As shown in FIG. 3, the relative lane-by-position lane information L _{t , 0} (or 0, 1, 2, ...) at each time index L ₀ , L _{t , N} (or L _N ), L _{t , N + M} (or L _{N + M} )

Distance estimation section 120 receives the vehicle information including a deviation speed and heading information to a predetermined update period (for example, the number to several tens msec) for each deviation movement distance Δd _t of in the _path to estimate the _first memory (Not shown) (S120). The moving distance Δd _{t - 1} of the vehicle can be calculated as a distance value on the corresponding direction path according to the heading information by multiplying the vehicle speed by the corresponding time interval. Here, the period for estimating the moving distance? D _{t- 1} of the vehicle is described as an example of the period in which the front lane detecting unit 110 updates the front lane information for each position, but the present invention is not limited thereto, There is no need and can be set differently. Further, in addition to the moving distance estimation unit 120 is the above deviation speed and heading information, in accordance with the lane curvature and the slope estimated from the camera image information, it is also possible to change the moving distance of the estimated period Δd _{t -1.} For example, when the variation of the lane curvature or the slope is large, the estimation period can be reduced for more accurate (or fine) estimation, and if not, the estimation period can be increased.

The backward lane estimating unit 130 compares the previous position and the current position and calculates a backward lane based on the moving position of the lane of the front lane information L ₀ , .., L _N , .., L _{N + M} detected by the front lane detecting unit 110 The rear lane information is estimated for each moving distance DELTA dt _-1 of the vehicle using the information corresponding to the rear position (S130).

First, each time index (t = 0,1,2, ..) the relative distance of the periodic movement with respect to the distance d _{t, i t} Δd of ₁ predetermined function for determining the position of the vehicle using the f (d _{t , i} ,? d _{t -1} ) is applied to calculate the position of the vehicle at a certain point relative to the reference point. For example, in each time index (t = 0, 1, 2, ..) from the reference time t = 0, vehicle information V _t such as the vehicle speed and heading information and the vehicle position D _{t , 0} to D _{t , 1} is changed. At this time, for example, as for the change of the vehicle position D _{t , 0} to D _{t , 1} , in each time index (t = 0, 1, 2, The position D _{t , 0} to D _{t , 1} of each current vehicle can be estimated by accumulating Δd _{t -1} .

[Equation 1]

When the position D _{t , 0} to D _{t , 1} of the current vehicle is estimated at each time index (t = 0, 1, 2, ..), the front lane detecting section 110, which is stored in the memory, lane information (L _0, .. L _N, L .. _{N + M)} of the deviation in the movement of the position of the corresponding predetermined rear position according to the position (e.g., Fig. 4, at time index t = 2 L _{0, 2)} but replacing the information (for example, L _{0, 2} ..) each corresponding to a rear lane information, which corrects the relative position information of a specific lane rearward of the deviation from the current position because it was about the front. Such a method is the moving distance Δd _t-1 each front lane information L _{t, i,} the function _{f (Δd t -1, L t} , i) for correcting the relative position information of a specific lane rearward by replacing the rear lane information For example, to a predetermined plurality of positions that are located in a rearward position.

Lane information providing unit 140 is connected to a distance Δd _t is detected ahead lane information for the front and rear road modeling lane estimation unit 130 using the front lane detector 110 to the current deviation for each position _-1 of the estimated deviation The backward lane information is used to perform a predetermined rear road modeling (S140). As shown in FIG. 5, the lane information providing unit 140 may provide lane information in the form of a predetermined function for the front road and the rear road through road modeling.

For example, according to the manner described above, the front or back lane information, Equation (2) and the like, the function f (d _{t, i, t -1} Δd of the relative position by (i) the time index (t) , to the location information that is determined by the L _{t, i)} using the lane information i _{t, i} is, and, given a polynomial function of the thus calculated lane information i _{t, i} f (i _{t, i)} relatively By summing a plurality of positions with respect to the time index (t) by position (i), the front or rear lane information I can be calculated for a certain distance by a curve fitting method or the like.

 &Quot; (2) "

In accordance with the road modeling method as described above, the lane information providing unit 140 generates the lane information corresponding to the vehicle information (e.g., V ₀ , V ₁ , ... V _N ) such as the vehicle speed and heading information about the front _0, L _1, .. front using the L _N) in a predetermined polynomial function type that indicates a lane position value corresponding to the relative position in the forward direction (or the distance) _(F x) from the deviation, as shown in FIG. 5 with respect to the road ahead Lane information y _F can be provided. Similarly, the lane information providing unit 140 the vehicle information such as speed and heading deviation information on the rear (e.g., V _N, V _{N +1, N} .. V _{+ M),} the rear lane information (for example, L corresponding to _N , L _{N + 1} , L _{N + M} ), a relative distance from the vehicle to the rear as shown in FIG. 5 for the rear road, or a predetermined polynomial function representing the lane position value along the position (x _R ) And can provide the rear lane information y _R. At this time, the lane information providing unit 140 may set a lane curvature or a slope of the lane according to the lane curvature or slope estimated from the camera image information, in addition to the vehicle information (V ₀ , V ₁ , ..) It is possible to provide the front lane information y _F and the rear lane information y _R in the form of a predetermined polynomial function as the curve fitting method of the section function.

As shown in FIG. 6, in the conventional technique of the rear lane information estimated for the rear lane and the comparison result of the present invention, there may be some similar results for the straight road, but in the curved road, It was confirmed that the result can provide a more accurate result than the result according to the prior art.

As described above, in the lane detecting apparatus 100 according to the present invention, when the characteristics of the road suddenly change, such as a confluence point or a bifurcation point, according to the vehicle information fusion method such as the camera image information and the vehicle speed / It is possible to estimate the lane more precisely and to recognize the forward / rear lane more accurately. Therefore, it is possible to more precisely handle the lane keeping autonomous driving and the lane changing autonomous driving in the subsequent lane based autonomous driving system do. In addition, it can be applied to various systems requiring lane information such as lane departure warning system, lane keeping system, and accurate lane information can be provided.

As described above, the present invention has been described with reference to particular embodiments, such as specific elements, and specific embodiments and drawings. However, it should be understood that the present invention is not limited to the above- Those skilled in the art will appreciate that various modifications, additions and substitutions are possible, without departing from the essential characteristics of the invention. Therefore, the spirit of the present invention should not be construed as being limited to the embodiments described, and all technical ideas which are equivalent to or equivalent to the claims of the present invention are included in the scope of the present invention .

The front lane detecting unit 110 detects the front lane,
The movement distance estimating unit 120
The backward-
The lane information providing unit 140

Claims (5)

A lane detecting device for detecting a lane on a traveling road during a vehicle operation,
A front lane detecting unit detecting the front lane information by relative position with respect to the vehicle from the camera image information about the front lane of the vehicle and updating the lane lane information at every predetermined period;
A moving distance estimating unit that receives the vehicle speed and heading information and estimates a moving distance of the vehicle in the corresponding path at every predetermined period;
A backward lane estimator for estimating the rear lane information for each moving distance of the vehicle using information corresponding to a rear position according to a moving position of the vehicle in the front lane information; And
A lane information providing unit for performing forward road modeling using the forward lane information and backward road modeling using the rear lane information for a current vehicle position for each moving distance of the vehicle,
Wherein the lane detection device comprises: The method according to claim 1,
Wherein the rear lane estimator comprises:
And estimates the rear lane information by correcting the information corresponding to the rear position of the front lane information to the lane information of the rear position relative to the current lane position. The method according to claim 1,
The movement distance estimating unit estimates,
And changes the estimation period of the travel distance according to the lane curvature or slope estimated from the camera image information, in addition to the vehicle speed and heading information. The method of claim 3,
The lane information providing unit,
Wherein the modeling is performed for each section having the same lane curvature or slope. A lane detection method for detecting a lane on a driving road in a vehicle,
Detecting the front lane information by position relative to the vehicle from the camera image information about the front of the vehicle, and updating the detected lane information at every predetermined period;
Estimating a moving distance of the vehicle on the corresponding path at predetermined intervals by receiving the vehicle speed and the heading information;
Estimating the rear lane information for each moving distance of the vehicle using the information corresponding to the rear position according to the moving position of the vehicle in the front lane information; And
Performing forward road modeling using the forward lane information and backward road modeling using the rear lane information for a current child position for each moving distance of the vehicle;
And detecting the lane of the lane.

Images