KR20230070080A - Lane Departure Prevention Apparatus - Google Patents

Abstract

The present invention provides a lane departure prevention device which comprises: an object detection unit generating information about an object outside a vehicle; a sensing unit detecting position and direction changes of the vehicle; a position generation unit generating position data of the vehicle; an autonomous driving unit collecting object information generated through the object detection unit, sensing information detected by the sensing unit, and the position data generated by the position generation unit to determine whether the vehicle departures a lane; and a driving control unit controlling a vehicle driving device based on a signal received from the autonomous driving unit.

Description

Lane Departure Prevention Apparatus}

The present invention relates to a lane departure prevention device for an autonomous vehicle.

As navigation, audio and air conditioning functions are operated through the screen of a display device mounted on a vehicle, the rate of traffic accidents caused by diverting the driver's gaze to the display device while driving is greatly increasing.

As described above, when forward gaze is not properly performed while driving due to drowsy driving or manipulation of a display device, a dangerous situation in which the vehicle leaves the driving lane may occur, which is highly likely to lead to a traffic accident.

Accordingly, there is a demand for a technology for predicting departure of the vehicle from the driving lane and controlling a driving-related device of the vehicle so that the vehicle does not depart from the driving lane when departure from the driving lane is predicted.

An object of the present invention is to provide a lane departure prevention device for an autonomous vehicle capable of reliably preventing lane departure by obtaining various types of information during vehicle driving.

The solution means of the present invention includes an object detection unit generating information on an object outside the vehicle; a sensing unit that detects changes in the position and direction of the vehicle; a location generating unit generating location data of the vehicle; an autonomous driving unit that determines whether the vehicle departs from a lane by collecting object information generated through the object detection unit, sensing information detected by the sensing unit, and location data generated by the location generating unit; A lane departure prevention device for an autonomous vehicle may include a driving control unit that controls a vehicle driving device based on a signal received from the autonomous driving unit.

As described above, the present invention provides information to the driving control unit to enable autonomous driving by collecting various sensing information including an object detection unit for generating information on an object outside the vehicle, a driving control unit for controlling a driving device of the vehicle, and the like. Autonomous driving unit, location generation unit that generates vehicle location data, sensing unit that generates information from various sensors such as vehicle posture, lane curvature calculation unit that automatically calculates the curvature of the lane while driving the vehicle, and steering angle sensor constituting the sensing unit Lane departure during vehicle driving can be reliably prevented by a driving direction prediction unit that calculates the driving curvature of the vehicle using steering angle information and predicts the driving direction of the vehicle based on the calculated driving curvature.

1 is an overall configuration diagram of the present invention.
2 is an explanatory diagram for explaining a process of detecting a lane, which is an object, by an object detection unit of the present invention.
3 is a schematic diagram illustrating a process of detecting a forward object in a state in which a camera, which is an object detection unit of the present invention, is installed inside a vehicle.
4 is a diagram for explaining a process of recognizing a lane by a camera of the present invention and processing a standard lane detection unit and a lane detection unit in a control unit.
5 is a diagram for explaining a correction edge operation of a curved lane according to the present invention.
6 is a diagram for explaining a process of setting reinforcement lane data through a reinforcement edge operation by dividing a rough recognized curve into a division curve according to the present invention.

Referring to FIG. 1 , a lane departure prevention device 100 according to the present invention includes an object detection unit 110, a drive control unit 120, an autonomous driving control unit 130, a sensing unit 140, and a location generator 150. can do.

The object detector 110 may generate information about an object outside the vehicle. The information about the object may include at least one of information about the existence of the object, location information of the object, distance information between the vehicle and the object, and relative speed information between the vehicle and the object.

The object detector 110 may include at least one of a camera, radar, lidar, ultrasonic sensor, and infrared sensor. The camera may generate information about an object outside the vehicle using an image. A camera may include at least one lens, at least one image sensor, and at least one image signal processor. The image signal processor may be electrically connected to the image sensor, process a received signal, and generate data about an object based on the processed signal.

The camera may obtain location information of an object, distance information from the object, or relative speed information from the object using various image processing algorithms.

The camera may be mounted in a position where a field of view (FOV) can be secured in the vehicle in order to photograph the outside of the vehicle. That is, for example, the camera may be disposed close to the front windshield inside the vehicle to obtain an image of the front of the vehicle.

The radar may generate information about an object outside the vehicle using radio waves.

LIDAR may generate information about an object outside the vehicle by using laser light.

The driving control unit 120 may control the vehicle driving device based on a signal received from the autonomous driving control unit 130 . For example, the driving controller 120 may control a power train, a steering device, and a brake device based on a signal received from the autonomous driving controller 130 .

The autonomous driving control unit 130 may implement at least one Advanced Drive Assistance System (ADAS) function. ADAS can be implemented through at least one of a smart cruise control system and a lane keeping assist system (LKAS).

The sensing unit 140 may detect the state of the vehicle and may include an inertial measurement unit (IMU) sensor 141 . An inertial measurement unit (IMU) sensor may include one or more of an acceleration sensor, a gyro sensor, and a magnetic sensor.

The IMU sensor 141 may detect changes in the position and direction of the autonomous vehicle based on the inertial acceleration force.

The sensing unit 140 may generate vehicle state data based on a signal generated by at least one sensor. The vehicle state data may be information generated based on data sensed by various sensors provided inside the vehicle. The sensing unit 140 includes vehicle posture data, vehicle motion data, vehicle yaw data, vehicle roll data, vehicle pitch data, vehicle collision data, vehicle direction data, vehicle angle data, and vehicle speed. data, vehicle acceleration data, vehicle inclination data, vehicle forward/reverse data, vehicle weight data, battery data, fuel data, tire pressure data, vehicle internal temperature data, vehicle internal humidity data, steering wheel rotation angle data, external illuminance of the vehicle Data, pressure data applied to the accelerator pedal, pressure data applied to the brake pedal, etc. can be generated.

The sensing unit 140 may include a steering sensor 142 that detects a rotational angle of a steering wheel of a vehicle and a wheel sensor 143 that detects a wheel speed of a vehicle.

The sensing unit 140 may be referred to as a sensor unit for autonomous driving of a vehicle. Accordingly, the sensing unit 140 may generate reference values for vehicle speed, steering wheel angle, and vehicle location information.

The location generator 150 may generate vehicle location data, and may include at least one of a Global Positioning System (GPS) and a Differential Global Positioning System (DGPS).

The location generating unit 150 may generate vehicle location data based on a signal generated by at least one of GPS and DGPS. The location generating unit 150 may correct location data based on at least one of an Inertial Measurement Unit (IMU) sensor of the sensing unit 140 and a camera constituting the object detection unit 110 .

In addition, the present invention may include a driving direction predictor 170 capable of predicting the driving direction of the autonomous vehicle using steering angle information, which is information sensed by the sensing unit 140 during driving of the autonomous vehicle.

In other words, the driving direction predictor 170 calculates the driving curvature of the vehicle in the lane curvature calculator 160 using steering angle information from the steering sensor 142 constituting the sensing unit 140, and the calculated driving Based on the curvature value, it is possible to predict the traveling direction of the vehicle.

The lane curvature information calculated by the lane curvature calculation unit 160 may be sent to the autonomous driving control unit 130 together with the lane information acquired by the object detector 110 .

In addition, information on the driving direction of the vehicle predicted by the driving direction predictor 170 may be sent to the autonomous driving control unit 130 .

The autonomous driving control unit 130 collects lane information (straight line, curved line, solid line, dotted line), lane curvature information, and information on the driving direction of the vehicle, and the driving control unit 120 controls the vehicle driving device with data determined through calculation. can be reflected in

If the vehicle is driven at night or on a rainy day, the lane may not be accurately recognized because the lane is not properly seen.

Since the possibility of leaving the lane may increase when the lane is not accurately recognized, the degree of lane recognition may vary depending on the driving environment. Therefore, clearly recognizing these lanes can be of paramount importance.

In the present invention, in order to acquire information through more accurate lane recognition by reflecting this situation, the object detection unit 110 sets detailed areas for each lane and sends the recognized lane information to the autonomous driving control unit 130. can

3 is a schematic diagram showing a process of detecting a lane, which is a forward object, in a state in which a camera, which is an object detection unit of the present invention, is installed inside a vehicle. Referring to FIG. 3, the camera 112 constituting the object detection unit 110 may be placed on the inside of the windshield of the vehicle 10 to keep an eye on the front.

The camera 112 of the present invention may be an interlocking camera that follows an object that is not in a fixed state. That is, the camera 112 can automatically move with a tracking function to change the camera gaze angle up, down, left, and right.

In addition, the camera 112 may be capable of close-up or long-distance photography so as to perform a zoom function. Accordingly, the camera 112 may include an image sensor and a zoom sensor, and may include a tracking unit that looks at an object and a driving unit that moves the camera 112 based on information provided by the tracking unit.

According to the present invention, when driving a vehicle, the camera 112 constituting the object detection unit 110 provided inside the vehicle can drive while looking forward, and at this time, the camera 112 can obtain image information of the driving lane. there is.

Referring to FIG. 2 , the autonomous driving control controller 130 may recognize and process lane information acquired by the camera 112 .

The autonomous driving control unit 130 may receive data (through CAN communication) from the sensing unit 140 to enable autonomous driving of the autonomous vehicle.

The sensing unit 140 may include an IMU sensor 141 , a steering sensor 142 , and a wheel sensor 143 .

The IMU sensor 141 can detect changes in the position and direction of the self-driving vehicle based on the inertial acceleration force, the steering sensor 142 detects the steering angle caused by steering wheel manipulation, and the wheel sensor 143 detects the speed of the wheel. can detect

Furthermore, the autonomous driving control unit 130 may receive an image obtained from the camera 112 and, based on this, recognize various types of information about a lane on which the vehicle is currently located.

Referring to FIG. 2 , the present invention may include a lane processing unit 180 capable of more accurately recognizing a lane by detecting and correcting a lane (straight or curved) during driving of an autonomous vehicle, and the lane processing unit 180 Overall operation may be controlled by the autonomous driving controller 130 .

That is, the lane processing unit 180 includes a lane detecting unit 181 that detects a lane by finding an edge of image information obtained from the camera 112, and a reference lane data generation unit for comparing with lane data detected by the lane detecting unit 181. (182), and a lane correction unit 183 that compares lane data from the lane detection unit 181 and standard lane data from the reference lane data generation unit 182 and corrects them.

The lane detection unit 181 may detect a lane based on current lane image information acquired by the camera 112 .

When detecting a lane, the lane detection unit 181 may determine whether the lane is a straight line or a curve, and find and detect the edge of the lane through the first pixel, the second pixel to the Nth pixel from the obtained lane data.

The reference lane data generation unit 182 may provide reference lane data for comparison with the lane data detected by the lane detection unit 181 . The reference lane data may be a state in which reference lane data for a reference pattern of straight or curved lane data is stored in advance.

The reference lane data may be data generated using lane information detected through the lane detection unit 181, and lane information received for a certain period of time is analyzed and, accordingly, when a similar pattern is continuously input, it is converted into the reference lane data. can be stored as data.

The lane correcting unit 183 may extract characteristics of reference lane data and current lane data provided at the present time.

The information characteristics of the lane data may include a degree of change of a lane angle and a degree of change of a lane width over time.

The lane correction unit 183 may generate and store lane angles corresponding to reference lane data and current lane data, and may generate and store lane angles corresponding to the stored current lane data and change values of lane angles corresponding to the reference lane data. You can check.

In addition, the lane correction unit 183 extracts characteristics of the standard lane data and the current lane data provided at the current point in time, and then detects whether the current lane data is incorrectly detected based on the extracted characteristics.

In other words, if the lane data is normally detected, the signal for the lane data detected through the lane detection unit 181 is transmitted to the autonomous driving control unit 130 as it is, and if the lane data is incorrectly detected, the generated standard lane data past lane data may be transmitted to the autonomous driving control unit 130 .

When erroneous lane data is detected, the lane angle change value corresponding to the degree of lane angle change may be a difference between the lane angle of the current lane data (left and right lanes) and the lane angle of the reference lane data.

By comparing the lane angle difference value with a threshold value, it may be determined whether the currently received lane data is erroneously detected.

As a result of determining whether the currently received lane data is erroneously detected, the reference lane data may be transmitted to the autonomous driving control unit 130, and if it is determined that the currently received lane data is normally detected, the lane correction unit 183 ), the currently received lane data may be transmitted to the autonomous driving control unit 130 as it is.

On the other hand, if there is an error in the standard lane data, lane erroneous detection may be continuously detected. Therefore, in the reference lane data generator 182, if the lane correction unit 183 repeatedly detects erroneous lane detection a certain number of times or more, Reference lane data may be deleted, and reference lane data may be regenerated using lane data input at the present time.

Therefore, the present invention utilizes past lane data to detect the instantaneous error detection status of lane data and corrects the incorrectly detected lane data accordingly, so that actual lanes are worn out or complex noise objects in the lanes or around the lanes are not generated. If included, it is possible to solve an instantaneously increasing false detection situation, and thereby improve reliability for achieving the lane departure prevention function.

For example, referring to FIG. 4 , the lane correcting unit 183 compares the standard lane data L2 from the lane reference data generation unit 182 and the lane data L1 detected by the lane detection unit 181. If a difference occurs through , it can be determined as an erroneous detection.

The lane correction unit 183 may first compare current lane data L1 and past reference lane data L2 on a straight line. As a result of the comparison, if a difference exceeds a predetermined threshold value, it may be recognized as an erroneous detection and the reference lane data L2 may be transmitted to the autonomous driving control unit 130 to be corrected.

On the other hand, even in the case of a curved lane, similar to a straight lane, if the difference value is compared with the reference lane data and the difference exceeds the threshold value, it may be determined as an erroneous detection, and the reference lane data autonomous driving control unit 130 for the curve ) and can be corrected.

Here, since the curved lane data detected by the lane detection unit 181 is changed with a predetermined curvature compared to the straight lane data, lane data may be input significantly inaccurately.

Therefore, the lane correcting unit 183 does not directly compare with the reference lane data for the curve, but determines the correction edge of the curved lane through a correction operation, sets the corrected lane data, and then compares it with the curved lane reference data to determine the difference part can be corrected.

Referring to FIG. 5 , when working on a correction edge of a curved lane, two first and second curves S1 and S2 corresponding to the currently obtained lane data are applied to a first area A1 and a second area, respectively. The first reference area A3 and the second reference area for the first reference curve S3 and the second reference curve S4, which are two curves corresponding to the current lane data subdivided in (A2), corresponding to the reference lane data. Compared to (A4), it can be reinforced so that the difference does not exceed the threshold value.

Since the two first curves (S1) and the second curve (S2) corresponding to the currently obtained lane data are not recognized as one curve and may be discontinuously recognized in a truncated state, such inaccurate current lane If the data is directly compared with the reference lane data, the difference value is inevitably large. Therefore, in the state where the lane data for the current curved lane is reinforced, if it is compared with the reference lane data for the past curve and does not differ by more than the threshold value, it is recognized as normal detection and the lane data for these curves is recognized by the autonomous driving control unit. (130).

Moreover, the differential coefficient, which is the slope of the tangent between the first reference curve S3 and the second reference curve S2 existing in the first reference area A3 and the second reference area A4, is set to a first threshold. ) (T1) and may be set to a range smaller than the value of the second threshold (T2). If the differential coefficient, which is the slope of the tangent between the first reference curve S3 and the second reference curve S2, is smaller than the first threshold T1 value or larger than the second threshold T2 value, It may not be used as lane data provided to the autonomous driving control unit 130.

In addition, in the present invention, reinforcement lane data may be set through a reinforcement edge operation for a curve recognized as rough as follows.

That is, referring to FIG. 6, the first division curve S11, the second division curve S12, and the third division curve S13 of the first curve S10 recognized as being roughly divided into a plurality of , By setting areas for the first splitting curve S21, the second splitting curve S22, and the third splitting curve S23 of the second curve S20, the first splitting area A11 to the sixth splitting area (A16).

The first split area A11, the second split area A12, and the third split area A13 are areas for the first curve S10, and the fourth split area A14 and the fifth split area A15 , The sixth divided area A16 may be an area for the second curve S20.

At this time, the first division curve S11 and the second division curve S12 of the first curve S10 that are roughly divided and recognized, and the first region A11 and the second region of the third division curve S13 (A12) and the third area A13 detect only the connected areas, and similarly, only the connected areas of the fourth area A14, the fifth area A15, and the sixth area A16 can be detected.

In addition, the first split curve S11 to the third split curve S13 may not be detected as they are, but the average value of the first split curve S11 to third split curve S13 may be calculated and adopted.

The first split curve S21 to the third split curve S23 of the second curve S20 may not be detected as they are, but the average value of the first split curve S21 to third split curve S23 may be calculated and adopted. there is.

Compensation curve lane data corresponding to this average value is extracted, and the reinforcement curve lane data is compared with the reference lane data of the curve, and the lane correction unit 183 determines that the difference is greater than a threshold value as an erroneous detection, and the curve standard Correction may be performed by transmitting lane data to the autonomous driving control unit 130 .

As such, the present invention obtains lane information by the object detection unit 110, grasps the vehicle attitude by the IMU sensor of the sensing unit 140, and vehicle location data by the location generator 150 made of GPS. is generated, the lane curvature is calculated every moment by the lane curvature calculation unit 160, and the driving direction prediction unit 170 predicts the vehicle's traveling direction in advance to prevent the vehicle from departing from the lane. The autonomous driving control unit 130 may apply a signal to the driving control unit 120 so as to prevent the vehicle from departing from its lane.

Accordingly, the present invention quickly and accurately obtains and reflects various lane information while driving the vehicle, and compares and determines the obtained information to control the vehicle so as to more quickly and reliably prevent lane departure. .

10... vehicle
100... lane departure prevention device
110.. object detection unit 111.. detection area
111a... first area 111b... second area
112 ... camera 120 ... driving control unit
130... autonomous driving control unit 140... sensing unit
141... IMU sensor 142... steering sensor
143... wheel sensor
150 ... position generation unit 160 ... lane curvature calculation unit
170... Driving direction prediction unit 180... Lane processing unit
181... lane detection unit 182... lane reference pattern unit
183... lane correction unit
A1... 1st area A2... 2nd area
A3... 1st reference area A4... 2nd reference area
S1... first curve S2... second curve
S3... first reference curve S4... second reference curve
S10... first curve S11... first division curve
S12... 2nd split curve S13... 3rd split curve
S20... second curve S21... first division curve
S22... 2nd split curve S23... 3rd split curve
A11... 1st partition A12... 2nd partition
A13... 3rd partition A14... 4th partition
A15... 5th partition A16... 6th partition

Claims (9)

an object detection unit generating information about an object outside the vehicle;
a sensing unit that detects changes in the position and direction of the vehicle;
a location generating unit generating location data of the vehicle;
an autonomous driving control unit that determines whether the vehicle departs from a lane by collecting object information generated through the object detection unit, sensing information detected by the sensing unit, and location data generated by the location generating unit;
a driving control unit controlling a vehicle driving device based on a signal received from the autonomous driving unit;
a lane processing unit that compares current lane data with reference lane data to determine whether an erroneous lane is detected; A lane departure prevention device comprising a.
According to claim 1,
The sensing unit,
an IMU sensor that detects a position and direction change of an autonomous vehicle based on an inertial acceleration force;
Steering sensor for detecting steering angle information of the vehicle;
A wheel sensor for detecting vehicle wheel speed; including,
a driving direction predictor capable of predicting a driving direction of the vehicle using steering angle information sensed by the sensing unit; A lane departure prevention device further comprising a.
According to claim 1,
The lane processing unit,
a lane detection unit that detects a lane based on current lane image information acquired by a camera during driving;
a reference lane data generating unit generating reference lane data;
a lane correcting unit that compares the current lane data by the lane detection unit with the standard lane data generated by the reference lane data generation unit, determines an error detection if the difference value is equal to or greater than a threshold value, and corrects the lane; A lane departure prevention device comprising a.
According to claim 3,
In the reference lane data generating unit,
The lane departure prevention device analyzes lane information received for a predetermined period of time using data generated using the lane information detected through the lane detection unit, and stores the result as reference lane data when a similar pattern is continuously input. According to claim 3,
The lane correction unit,
extracting characteristics of the reference lane data and the current lane data provided at the present time;
The information characteristics of the current lane data include a degree of change in the angle of the lane and a degree of change in the width of the lane over time;
generating and storing lane angles corresponding to reference lane data and current lane data;
Checking lane angles corresponding to the stored current lane data and change values of the lane angles corresponding to the reference lane data;
After extracting the characteristics of the standard lane data and the current lane data provided at the present time, based on this, detecting whether the current lane data is incorrectly detected,
When the lane data is normally detected, a signal for the lane data detected through the lane detection unit is transmitted to the autonomous driving control unit as it is,
If the lane data is incorrectly detected, the past lane data, which is the generated standard lane data, is transmitted to the autonomous driving control unit.
When erroneous lane data is detected, the lane angle change value corresponding to the degree of lane angle change is a difference between the lane angle of the current lane data and the lane angle of the reference lane data;
A lane departure prevention device that compares the lane angle difference value with a threshold value to determine whether currently received lane data is erroneously detected.
According to claim 5,
The reference lane data generation unit deletes the reference lane data and regenerates the reference lane data using the lane data input at the current point in time to prevent lane departure, when the lane correcting unit repeatedly detects lane error detection at least a certain number of times. Device.
According to claim 1,
The lane correcting unit does not immediately compare the current lane data for the curve with the reference lane data, but identifies the correction edge of the curved lane through a correction operation, sets the corrected lane data, and then compares it with the curved lane reference data to determine the difference corrects the part of
During the correction edge operation of the curved lane, the current lane data obtained by subdividing the two first and second curves corresponding to the currently acquired lane data into the first and second regions, respectively, is divided into 2 curves corresponding to the reference lane data Comparing the first reference area and the second reference area for the first reference curve and the second reference curve, which are curves, and reinforcing them so that they do not differ by more than a threshold value,
A lane departure prevention device that compares past lane data for past curves, recognizes them as normal detection if the difference is not greater than a threshold, and transmits the lane data for these curves to the autonomous driving control unit.
According to claim 1,
The lane correcting unit does not immediately compare the current lane data for the curve with the reference lane data, but identifies the correction edge of the curved lane through a correction operation, sets the corrected lane data, and then compares it with the curved lane reference data to determine the difference corrects the part of
If the two first curves and the second curve corresponding to the currently obtained lane data are not recognized as one curve and are discontinuously recognized as being cut,
The differential coefficient, which is the slope of the tangent of the first reference curve and the second reference curve existing in the first reference region and the second reference region, is set to a range greater than the first threshold value and less than the second threshold value,
Autonomous driving that is not used as lane data provided to the autonomous driving control unit when the differential coefficient, which is the slope of the tangent between the first reference curve and the second reference curve, is smaller than the first threshold value or larger than the second threshold value vehicle.
According to claim 1,
In the lane correction unit, when recognizing lane data, a first division curve, a second division curve, and a third division curve of the first curve recognized as being divided into a plurality of first and second curves constituting the lane;
Setting regions for each of the first division curve, the second division curve, and the third division curve of the second curve to form a plurality of division regions;
The first region, the second region, and the third region respectively corresponding to the first division curve, the second division curve, and the third division curve of the first curve detect only regions connected to each other;
The fourth region, the fifth region, and the sixth region respectively corresponding to the first division curve, the second division curve, and the third division curve of the second curve detect only regions connected to each other;
The first to third split curves are not detected as they are, and the average value of the first to third split curves is calculated and adopted,
The first to third split curves of the second curve are also not detected as they are, and the average value of the first to third split curves is calculated and adopted,
A lane departure prevention device that extracts the correction curve lane data corresponding to the average value, compares the reinforcement curve lane data with the reference lane data of the curve, determines whether or not an error is detected according to the difference value, and corrects it.

Images