KR101998298B1 - Vehicle Autonomous Driving Method Using Camera and LiDAR Sensor - Google Patents

Abstract

The present invention relates to an autonomous vehicle driving method using a camera and LiDAR sensors comprising the steps of: obtaining sensor values from a left LiDAR sensor and a right LiDAR sensor installed at a front side of a vehicle; allowing the left LiDAR sensor or the right LiDAR sensor to sense an object and checking the distance to an object sensed by one or two LiDAR sensors to determine whether an emergency occurs; checking the distance to the object to transmit a signal for performing an emergency stop when the distance is less than a set value; when performing the emergency stop is not necessary, checking whether an obstacle exists around the vehicle through a rotary LiDAR sensor; recognizing the position of an object through the obtained rotary LiDAR sensor value to determine the means to avoid the obstacle; when an object around the vehicle is recognized, transmitting a signal to perform a maneuver to avoid the obstacle; when no obstacle exists around the vehicle, obtaining data through a webcam camera; sensing a lane through the obtained data and tracking the lane along the sensed lane; and transmitting the speed and steering commands while making an emergency stop signal the highest priority, transmitting the speed and steering commands while making an obstacle avoidance signal the next priority, and transmitting the speed and steering commands through the lane tracking as the next priority to drive the vehicle. The present invention can ensure safe vehicle driving.

Description

{Vehicle Autonomous Driving Method Using Camera and LiDAR Sensor}

More particularly, the present invention relates to a vehicle autonomous traveling method using a camera and a Lidar sensor. More specifically, the present invention relates to a Lidar sensor for detecting an emergency situation, It is used to detect obstacles by 360 degree environment information through a rotary lidar sensor. It is used for lane recognition and tracking through a WebCam camera installed in the front of the vehicle. It can be used for emergency stop situations, obstacle detection, lane recognition and tracking And a method for autonomous driving of a vehicle using the camera and the Lidar sensor for autonomous driving of the vehicle through the vehicle.

In recent years, much research has been conducted in relation to the autonomous navigation system, especially regarding autonomous driving in the automobile field.

In general, the autonomous driving system or advanced driver assistance system automatically controls the driving of the vehicle from the starting point to the ending point on the road by using the GPS position information and signals acquired from various sensors based on the road map information, It enables safe driving.

Especially, the autonomous navigation system needs the help of sensors and graphic processing devices that can recognize nearby objects in order to recognize and judge the driving environment of a moving vehicle at high speed in real time.

The sensor measures the distance of objects and objects, detects danger and helps to see all areas without blind spots. The graphic processing device analyzes the image of the surrounding environment of the vehicle through several cameras and car safely Help you to go.

Conventionally, a method for reducing an error caused by disturbance due to an obstacle or an obstacle element for autonomous driving has been mainly studied. However, as the precision is increased by using a plurality of sensors, the number of components increases, This is a weighted disadvantage.

Therefore, an autonomous driving method that can reduce the burden of calculation processing for route tracking while ensuring reliability of position determination and route tracking to a target point is steadily progressing.

As a related prior art, Korean Patent Laid-Open Publication No. 10-2018-0084463, published on August 27, 2018 (hereinafter referred to as "Patent Document 1") discloses an apparatus and method for changing a lane of an autonomous vehicle using inter- It is open to the public. Patent Document 1 discloses an autonomous driving vehicle lane changing apparatus and method using inter-vehicle communication in which lane change is performed using V2V communication between autonomous driving vehicles.

In addition, Korean Patent Publication No. 10-2018-0117754, Laid-open Oct. 30, 2018 (hereinafter referred to as "Patent Document 2") discloses a vehicle autonomous-running control device. In the above-described Patent Document 2, the rear-end running vehicle is photographed through the AVB camera module and the autonomous-running control of the own vehicle is controlled on the basis of the photographed image, thereby reducing the accident between vehicles and the traffic congestion To the vehicle.

In addition, Korean Patent Registration No. 10-1532320 (hereinafter referred to as "Patent Document 3") discloses a dynamic object detection method using a binocular camera mounted on an autonomous vehicle. Patent Document 3 discloses a method for detecting a dynamic object using a binocular camera mounted on an autonomous traveling unmanned vehicle that detects a dynamic object in front by using a camera provided in a self-running unmanned vehicle so as to be spaced apart from each other. A data input step of inputting driving information of an autonomous driving unmanned vehicle and images photographed by a binocular camera mounted on the autonomous driving unmanned vehicle; and performing stereo matching from a pair of images simultaneously photographed from the binocular camera to extract feature points A super-pixel conversion step of dividing the image into a plurality of super-pixels so that an area having a value of an interval defined in the image is one super-pixel; and a high-order Markov random field for each of the plurality of super- A high-dimensional Marco that computes and detects superpixels in which dynamic objects exist. A method for detecting a dynamic object using a binocular camera mounted on an autonomous navigation unmanned vehicle, comprising: a random field calculation step; and an optimum information selection step of determining that there is a dynamic object in a super pixel having a dynamic object in the high dimensional Marcoff random field calculation step .

However, in the existing autonomous navigation apparatus and method, the autonomous navigation apparatus and the method used in the above-described Patent Document 1 perform lane change by using V2V communication between autonomous vehicles, and another autonomous vehicle must exist around the autonomous vehicle. V2V communication is required to be carried out. [Patent Document 2] requires that V2V communication and V2I communication be supported in order to collect the rear vehicle information, and the rear vehicle information is recognized by recognizing the vehicle number to calculate the relative speed of the vehicle However, when the car number is difficult to recognize due to the foreign substance on the license plate, or when the car number is not recognized like a bike, the car number, which is the rear car information, can not be recognized and the relative speed can not be calculated.

[Patent Document 3] discloses a technique of detecting a dynamic object located in front of an autonomous vehicle so that it can not detect the surrounding environment of the vehicle and avoids a stopped object in autonomous driving, There is a problem that it is difficult to carry out the autonomous traveling.

[Patent Document 1] Korean Patent Laid-open No. 10-2018-0084463 Disclosure Date 2018. 07. 25. [Patent Document 2] Korean Published Patent Application No. 10-2018-0117754 Publication Date Oct. 30, [Patent Document 3] Korean Patent Registration No. 10-1532320 Publication Date May 2015. 07 22.

The present invention has been developed in order to solve such conventional problems, and a controller for controlling the speed and steering of a vehicle and a sensor for measuring physical data is provided as a self-driving method using a camera and a Lidar sensor The vehicle is stopped in the emergency situation through the Raidasensor, 360 degrees environment information is obtained through the rotary Raidas sensor, and obstacles are detected to avoid obstacles. The lane recognition and tracking are performed through the webcam camera of the vehicle. The present invention provides a vehicle autonomous driving method using a camera and a lidar sensor capable of traveling safely and autonomously driving the vehicle.

(A) obtaining a sensor value from a left Lidar sensor and a right Lidar sensor installed in front of a vehicle, the method comprising the steps of: (a) S410); (b) determining whether the left or right Lidar sensor senses an object and determine the distance to the object detected by the one or two Lidar sensors to determine whether the object is in an emergency (S420); (c) transmitting (S421) a signal to proceed with the emergency stop when the distance to the object in step (b) is less than the set value; (d) if it is not necessary to proceed with the emergency stop in step (b), checking whether there is an obstacle around the vehicle through the rotary sensor (S430); (e) recognizing the position of the object through the rotary sensor value obtained in the step (d) and determining obstacle avoidance (S440); (f) transmitting (S441) a signal for avoiding an obstacle when an object is recognized around the vehicle in the step (e); (g) acquiring data through the webcam when there is no obstacle around the vehicle in the step (e) (S450); (h) detecting a lane through the data obtained in the step (g) (S460); (i) tracking the lane along the lane sensed in the step (h) (S470); (j) sending a speed and steering command with the emergency stop signal of the step (c) as a top priority, sending a speed and steering command to the signal of the obstacle avoidance in the next priority (f) (S480) by transmitting a speed and steering command through lane-tracking.

More preferably, (k) it is checked whether or not the target point has been reached at the time of the autonomous running through the steps (a) to (j), and if the target point is reached, the autonomous running is stopped; (S490) repeatedly to return to the step (a) until reaching the step (a).

Preferably, in the step (b), the left Lidar sensor or the right Lidar sensor senses the object and confirms the distance to the object detected by the Lidar sensor. If the distance to the object is less than 50 cm, .

Preferably, the step (e) recognizes the position of the object sensed using the rotary sensor and recognizes that there is an object to be avoided when there are a certain number of points in the set range, And transmits the command.

Further, preferably, the step (h) includes: (h1) acquiring image information through a webcam camera; (h2) adjusting a resolution of an image for image processing; And (h3) setting an RGB image as an 8-bit image.

Also preferably, step (i) comprises: (i1) removing small particles from the image through a particle filter; (i2) recognizing a lane by applying a gray morphology filter; (i3) an edge finding step of finding five points having the same distance between two adjacent points in the lane; (i4) converting pixel coordinates into actual coordinates; And (i5) executing an algorithm at the incoming lane tracking point.

Also preferably, the vehicle control system mounted for the autonomous vehicle traveling method comprises: a platform controller (100) for controlling the speed of the vehicle and the steering device; Left and right Lidar sensors 111 and 112 used for stopping the emergency state in the platform controller 100; A GPS 120 connected to the platform controller 100 to calculate a current position of the user; A wireless receiver (130) connected to the platform controller (100) to enable radio control; A steering control system 140 connected to the platform controller 100 to control steering; A braking control system 150 connected to the platform controller 100 to control braking; A drive control system 160 connected to the platform controller 100 to control the drive; And a sensor controller (200) connected to the platform controller (100) and controlling the rotation rider (210) and the webcam camera (220); The left and right Lidar sensors 111 and 112 are installed in front of the vehicle to determine an emergency stop state by checking the distance to the object. (210) the obstacle is avoided by checking an object around the vehicle through a sensor (210); And detects and tracks the lane through the web camera camera 220 so as to autonomously run.

Preferably, the platform controller 100 is connected to the steering control system 140, the braking control system 150, and the drive control system 160 through a PWM port.

Preferably, the platform controller 100 is connected to a GPS via a USB, and the sensor controller 200 is connected to a rotation sensor 210 and a webcam camera 220 through a USB .

More preferably, the platform controller 100 controls the wireless receiver 130 and the wireless receiver 130 communicates with the radio control system via Bluetooth.

According to the autonomous vehicle driving method using the camera and the Lydia sensor according to the present invention, in order to stop the emergency situation, two Raidas sensors mounted on the front of the vehicle and a 360 degree environment information are mounted in the center of the vehicle for detecting obstacles And the web camera camera installed at the front of the vehicle for lane recognition and lane-tracking can drive the vehicle autonomously, thereby enabling safe driving.

Other objects and advantages of the present invention will become apparent from the detailed description of the embodiments with reference to the attached drawings.

BRIEF DESCRIPTION OF DRAWINGS FIG. 1 is an exemplary diagram showing coordinates of a vehicle modeling of the present invention; FIG.
2 is an exemplary diagram showing coordinates of a vehicle system according to the present invention;
3 is an exemplary diagram showing a kinematic model of a vehicle of the present invention.
4 shows a vehicle control system configuration diagram
Fig. 5 is a block diagram of the automatic driving control algorithm obstacle avoidance first embodiment
6 is a block diagram of an automatic driving control algorithm obstacle avoidance second embodiment
Figure 7 shows a lane detection and tracking flow chart
Figure 8 shows a lane-
Fig. 9 is a flowchart of an automatic driving control algorithm

The present invention will now be described in detail with reference to the accompanying drawings.

First, the present invention relates to a vehicle autonomous traveling method using a camera and a Lydia sensor, which comprises three sensors and two controllers, one controller is used for vehicle speed and steering control, Is used for sensor control. The autonomous navigation platform consists of three sensors: a WebCam camera, a LiDAR Lite V3 sensor, and a Slamtec RPLiDAR. Web cameras are mainly used for lane identification and tracking, The Rida sensor is used to stop the emergency situation and the RPLiDAR sensor is used to obtain 360 degree environment information and it is able to avoid obstacles based on the data of the rotary sensor.

At this time, LiDAR is a device that fires laser pulses and reflects the light from the object to be reflected and measures the distance to the object to accurately depict the surroundings. Lada is an abbreviation of Light Detection And Ranging in English. Lada is a radar that uses light instead of radio waves. Lada has the same principle as a conventional radar. However, since the wavelength of the electromagnetic wave used is different, Lida is used not only for the distance to the object but also for analyzing the atmospheric matter and the concentration of the moving speed, direction, temperature, and surrounding. In recent years, Lada has been a core technology of sensor that acquires necessary information to realize 3D image .

The autonomous vehicle driving method using the camera and the Lidar sensor of the present invention will be described in detail with reference to FIG. 1 to FIG.

Fig. 1 is an exemplary view showing the coordinates of the vehicle modeling of the present invention, Fig. 2 is an exemplary view showing a vehicle system of the present invention, and Fig. 3 is an exemplary view showing a kinematic model of the vehicle of the present invention.

Referring to FIG. 1, X _l MY _l denotes a local coordinate frame, and X _g OY _g denotes a global coordinate frame. Also, M _r is the center point of the rear axis and is the origin of the local coordinate frame. M _f represents the center point of the front axis. The angular difference between the global and local coordinates is given by the value of [theta], and the rotation transformation matrix equation (1) represents the direction of the local coordinates with respect to the global coordinates.

는 휠 스티어링 각도를 나타내며, d는 두 개의 뒷바퀴 사이의 거리를 나타내며, ICR은 순간 회전 중심이며, d_r과 d_f는 각각 M_r과 M_f의 순간 회전 반지름을 나타낸다. 차량의 전체적인 속도 V_r은 로컬 좌표 프레임의 X_l 축 방향과 함께 후륜의 속도를 나타내며, V_f는 앞쪽 휠의 속도를 나타낸다. 이때 V_r과 V_f의 관계는 수학식 2와 같다.Next, referring to FIG. 2, 1 denotes an inter-axis distance,

D is the distance between the two rear wheels, ICR is the instantaneous center of rotation, and d _r and d _f represent the instantaneous radius of rotation of M _r and M _f , respectively. The overall speed V _r of the vehicle, along with the X ₁ axis direction of the local coordinate frame, represents the speed of the rear wheels, and V _f represents the speed of the front wheels. The relationship between V _r and V _f is expressed by Equation (2).

3, the point W represents the center of gravity of the vehicle, the configuration of the vehicle for the point W is defined by the global coordinates [x, y, delta] ^T , and x, Coordinate, y coordinate and direction, V represents the speed of the vehicle, and β represents the angle between the vehicle's speed direction and the vehicle direction.

Next, W _f and W _r represent the respective distances between the center gravity point and the front wheel and the rear wheel.

Also, the kinematic model used in connection with the navigation device is used to determine the coordinates of a point moving at a high speed. The data obtained at this time is not used in the real time processing method as in the navigation device, When processed by calculation, processed data can be used as digital map data. The method of the subsequent calculation can reduce the error by the real-time processing method, enables detailed analysis of the residuals, and has a merit that the model can be constructed. In the case of roads, a series of linear data acquired by this method in a vehicle traveling at high speed is processed by subsequent calculation, and the characteristics of the road network can be produced at once by assigning attributes of roads to the data. The kinematic model of Equation (3) is calculated as Equation (3).

Here,? Is calculated by the following equation (4).

Next, the vehicle control system configuration and algorithm proposed in the present invention will be described in detail with reference to Figs. 4 to 9. Fig. 4 is a configuration diagram of the vehicle control system, Fig. 5 is the first embodiment of the auto driving control algorithm obstacle avoidance, Fig. 6 is a second embodiment of an auto driving control algorithm obstacle avoidance, Fig. 7 is a lane detection and tracking flowchart, Fig. 8 is a lane-tracking algorithm, and Fig. 9 is an automatic driving control algorithm.

4, the vehicle control system includes a platform controller 100, a steering control system 140, a braking control system 150, a drive control system 150, A control system 160, a radio control system 131, and front sensing rider sensors 111 and 112, and an interface for connection with the sensor system. The platform controller 100 controls the speed of the vehicle and the steering device, and acquires the value of the LR sensor 110 to be used for stopping the emergency state. The platform controller 100 preferably uses an MI-myRIO controller. However, it is needless to say that the platform controller 100 may be any platform controller as long as it can control the speed of the vehicle and the steering apparatus.

The platform controller 100 is connected to the rider sensors 111 and 112 via the PWM port with the steering control system 140, the braking control system 150 and the drive control system 160, And the wireless receiver 130 communicates with the radio control system 131 via Bluetooth.

The GPS (Global Positioning System) 120 is a satellite navigation system that receives a signal transmitted from a GPS satellite and calculates a current position of the user. The GPS is used mainly for a navigation device such as an aircraft, a ship, ) Is connected to the platform controller 100 via the first USB 121. [

It is preferable to use the PC or the NVIDA Jetson TX2 in connection with the rotation sensor 210 and the sensor controller 200 connected to the webcam camera 220. However, as long as the sensed data can be analyzed and transmitted, Of course not.

The sensor controller 200 analyzes and transmits the sensed data through the rotary sensor 210 and the web camera 220 and avoids an obstacle in accordance with the steering command received by the platform controller 100, Recognizes and traces the vehicle.

The sensor controller 200 communicates with the platform controller via the RS232 cable 230 and the RPLiDAR sensor 210 and the webcam camera 220 are respectively connected to the sensor platform controller via USB.

Also, the autonomous driving algorithm is mainly handled in the sensor platform controller, and the speed and steering commands are transmitted from the sensor controller to the platform controller based on the driving algorithm.

The autonomous driving algorithm detects the presence of an emergency stop through two Lidar sensors located on the front of the vehicle, and d _lidar-l and d _lidar-r represent the sensing distances of the two sensors, respectively. And the distance to the object is less than 50cm, it indicates that the vehicle is in an emergency stop and the pseudo control code is as follows.

That is, when the distance between the left side sensor or the right side sensor is less than 50 cm, the speed and steering of the vehicle are controlled to be zero, and an emergency stop is performed.

Next, referring to FIG. 5, regarding the obstacle avoidance of the auto driving control algorithm, the RRLiDAR sensor is located at the L-point of the center of the vehicle (zero point of the RPLiDAR sensor).

When the RPLiDAR sensor detects the object point P, it feeds back the distance value d _ls and the bearing angle α. La is rotated based on the equation (5), Equation 6 (RPLiDAR) shows the conversion of a distance and bearing angle, the distance l _ro feedback between the sensor and the zero point of the local coordinate frame.

Referring to FIG. 6, a RPLiDAR sensor rotates while scanning 360 degrees in all directions, generates data about a peripheral contour, and the sensor controller obtains a detection point from a RPLiDAR sensor, Analyzes the data and recognizes the position of the vehicle and the sensed object in the coordinate frame.

x range [0.5m, 1m] and the y range [0.5m, 2m] and the y range [-0.4m, 0.4m] [0.2m, 1m], there is an object on the left side of the vehicle to be avoided.

If there is more than the set point in the x range [0.2m, 1m] and the y range [-1m, -0.2m], the sensor controller recognizes that there is an object on the right side of the vehicle to be avoided, Indicates that the command should be sent.

Further, the steering command is determined according to Table 1 below.

The steering command in Table 1 is that the vehicle advances if there is no obstacle in Case 0, the vehicle advances if the obstacle is in the right side in Case 1, the vehicle turns right if there is an obstacle ahead in Case 2, The vehicle will turn left if there is no obstacle on the left side of the vehicle, the vehicle will advance if there is an obstacle on the left or right side in cases 4 and 5 and no obstacle ahead, If there are obstacles in the front, left, and right sides of the vehicle, it indicates that the vehicle moves backward.

Accordingly, the present invention is characterized in that data sensed by the RL PDAR sensor is analyzed and transmitted through a sensor controller, and an obstacle is avoided according to a steering command received from the platform controller.

Next, the vehicle is equipped with a WebCam camera in front of the vehicle, and the camera is used to detect the lane and track the lane. 7, the first three blocks are used for lane detection 300 through image detection, and then five blocks are used for lane tracking 310. The lane detection and tracking flow diagram of FIG.

(a) acquiring (301) image information through a camera;

(b) adjusting an image resolution for image processing (302);

(c) Color Threshold is a step 303 of setting an RGB image as an 8-bit image;

(d) the particle filter for lane tracing includes removing (311) small particles from the image;

(e) recognizing a lane by applying a filter to obtain a desired data image in the image data with a gray morphology technique (312);

(f) a corner searching step (313) of finding five points having the same distance between two adjacent points in the lane;

(g) transforming the pixel to the real world, transforming the pixel coordinate to the real coordinate (314);

(h) executing an algorithm at an incoming lane tracking point (315);

Hereinafter, the lane detection and tracking step will be described in detail.

(b), the image size is adjusted to 640 * 480 resolution for image processing, and (c) the color threshold is set as an 8-bit image in (c) A device that displays an image such as a monitor, a projector, or a TV usually represents a color using three color combinations of RGB (Red, Green, and Blue), and expresses all colors using the three primary colors of light, The colors that can be represented by the bits of RGB are determined.

In addition, the color threshold distinguishes the important structure of the image from the rest of the image, and the threshold sets all pixels in the threshold range to 1 and sets all other pixels in the image to zero, Expresses what is expressed.

In other words, bit is also referred to as bit-depth (bit density), which indicates how precisely the brightness of light can be adjusted. Here, the number of colors = 2 ⁿ (bits).

Thus, if the RGB is 8 bits, then 2 ⁸ = 256, which means that the image is expressed in 256 colors.

(d), the small particles are removed through a particle filter for lane tracking. In (e), gray morphology indicates that the shape of the feature of the image is modified by changing the pixel value, Outputs the result image with the brightness value size. It is useful when the brightness difference between the object and the background is large, and recognizes it by eroding the lane.

In step (h), the process proceeds as shown in FIG. 8, where B _i (x _bi , y _bi ) and C _i (x _ci , y _ci ) are applied to edge detection and conversion of the pixel to the actual block.

Also, the tracking point A _i (x _ai , y _ai ) is calculated by Equation (7) and Equation (8).

9, two Lidar sensors, that is, a left Lidar sensor value and a right Lidar sensor value, which are installed in front of the vehicle during driving, are obtained (S410)

In step S410, the left or right Lidar sensor senses an object and determines whether the distance from the object detected by the one or two Ladid sensors is less than 50 cm (S420)

If the distance to the object is less than 50 cm, a signal is transmitted to proceed to the emergency stop (S421)

If it is not necessary to proceed with the emergency stop in S420, the controller checks whether there is an obstacle around the vehicle through the rotary sensor. (S430)

The position of the object is recognized through the rotary sensor value obtained in S430 and the obstacle avoidance is determined (S440)

If an object is recognized in the vicinity of the vehicle in step S440, a signal is transmitted so as to proceed with obstacle avoidance (S441)

If there is no obstacle around the vehicle in S440, data is acquired through the webcam (S450)

The lane is detected through the data obtained in the step S450 (S460)

The lane is tracked along the lane sensed in step S460 (S470)

In step S421, the signal of the emergency stop phase is transmitted with priority to the speed and steering command, and the speed and steering command are transmitted to the emergency stop signal in the obstacle avoidance phase in the next priority order (S441) And transmits the speed and steering command through the lane-tracking (S480)

In step S480, it is checked whether the target point has been reached in the autonomous running. When the target point is reached, the autonomous running is stopped. If the target point is not reached, the procedure is repeatedly performed until the target point is reached in step S410 (S490)

The autonomous driving control algorithm combines the emergency stop algorithm, the obstacle avoidance algorithm, the lane detection and tracking algorithm, and the emergency stop algorithm has the highest priority.

The vehicle must first determine whether to stop the emergency, and if the vehicle is in an emergency stop, it executes the algorithm with the pseudo control code and proceeds to an obstacle avoidance algorithm. If the vehicle does not require an emergency stop, it checks to see if an obstacle should be avoided, and if the vehicle should avoid an obstacle, proceed to the step of avoiding the obstacle according to the steering command, and proceed to the lane detection and tracking algorithm step.

If it is determined that the vehicle is not traveling, the step S420 is automatically selected as N, and the process proceeds to step S430.

Therefore, if the vehicle stops or avoids obstacles, the vehicle traces points based on the lane detection and tracking flow chart.

The present invention can be applied to a vehicle autonomous driving method using a camera and a Lidar sensor according to various modifications by a person having ordinary skill in the art, .

100: Platform Control Unit (PCU)
110:
111: Left Lidar sensor
112: right side sensor
120: GPS (Global Positioning System)
121: First Universal Serial Bus (USB)
130: Wireless receiver
131: Radio control system
140: Steering control system
150: Braking control system
151: Brake motor
160: drive control system
161: Drive motor
170: Encoder
200: Sensor Controller (SCU: Sensor Control Unit)
210: Rotary Lidar sensor
211: Second Universal Serial Bus (USB)
220: Webcam camera
221: Third Universal Serial Bus (USB)
230: RS232 cable
300: Lane detection
301: Image acquisition
302: Adjust image resolution
303: Color Threshold
310: Lane Tracking
311: Particle removal filter
312: Gray Morphology
313: Finding a corner
314: Pixel Conversion
315: Lane Tracking Obtained

Claims (10)

(a) acquiring a sensor value from a left Lidar sensor and a right Lidar sensor installed in front of the vehicle (S410);
(b) determining whether the left or right Lidar sensor senses an object and determine the distance to the object detected by the one or two Lidar sensors to determine whether the object is in an emergency (S420);
(c) transmitting (S421) a signal to proceed with the emergency stop when the distance to the object in step (b) is less than the set value;
(d) if it is not necessary to proceed with the emergency stop in step (b), checking whether there is an obstacle around the vehicle through the rotary sensor (S430);
(e) recognizing the position of the object through the rotary sensor value obtained in the step (d) and determining obstacle avoidance (S440);
(f) transmitting (S441) a signal for avoiding an obstacle when an object is recognized around the vehicle in the step (e);
(g) acquiring data through the webcam when there is no obstacle around the vehicle in the step (e) (S450);
(h) acquiring image information from the data acquired through the webcam camera in step (g), adjusting the resolution of the image for image processing, and detecting the lane by setting an RGB image as an 8-bit image );
(i) removing small particles from the image through the particle filter in the lane detected in step (h), recognizing the lane by applying a gray morphology filter, and calculating five points having the same distance between two adjacent points in the lane Converting the pixel coordinates into actual coordinates, and executing the algorithm at the incoming lane tracking point to track the lane (S470);
(j) sending a speed and steering command with the emergency stop signal of the step (c) as a top priority, sending a speed and steering command to the signal of the obstacle avoidance in the next priority (f) A step S480 of traveling by transmitting a speed and steering command via lane-tracking; And
(k) It is checked whether or not the target point has been reached at the time of the autonomous running through the steps (a) to (j), and if the target point is reached, the autonomous running is stopped. The method of claim 1, further comprising the steps of: (a) returning to step (a); and repeating the step (S490). delete 2. The method of claim 1, wherein step (b)
The left or right Lidar sensor senses an object and determines the distance from the object detected by the Lidar sensor to determine an emergency stop when the distance to the object is less than 50 cm. Autonomous vehicle driving method. 2. The method of claim 1, wherein step (f)
Recognizing the presence of an object to be avoided and transmitting a steering command for avoiding an obstacle when a predetermined number of points exist in the set range, Autonomous vehicle driving method using sensor. delete delete A vehicle control system mounted for an autonomous vehicle traveling method comprises:
A platform controller (100) for controlling the speed of the vehicle and the steering device;
Left and right Lidar sensors 111 and 112 used for stopping the emergency state in the platform controller 100;
A GPS 120 connected to the platform controller 100 to calculate a current position of the user;
A wireless receiver (130) connected to the platform controller (100) to enable radio control;
A steering control system 140 connected to the platform controller 100 to control steering;
A braking control system 150 connected to the platform controller 100 to control braking;
A drive control system 160 connected to the platform controller 100 to control the drive; And
And a sensor controller (200) connected to the platform controller (100) and controlling the rotation rider (210) and the webcam camera (220);
The left and right Lidar sensors 111 and 112 are installed in front of the vehicle to determine an emergency stop state by checking the distance to the object.
(210) the obstacle is avoided by checking an object around the vehicle through a sensor (210);
The image information is acquired from the data obtained through the web camera camera 220, the resolution of the image is adjusted for image processing, the lane is detected by setting the RGB image as an 8-bit image, In this paper, we propose an algorithm to remove the small particles from the image, apply the gray morphology filter to recognize the lane, convert the pixel coordinates to the actual coordinates by finding five points with the same distance between two adjacent points in the lane, To detect and track lanes;
And the vehicle is driven by sending a speed and a steering command by setting priority in the order of the emergency stop signal, obstacle avoidance, and lane tracing. 8. The method of claim 7,
Wherein the platform controller (100) is connected to the steering control system (140), the braking control system (150), and the drive control system (160) by a PWM port. 8. The method of claim 7,
The platform controller 100 is connected to a GPS via a USB and the sensor controller 200 is connected to a sensor 210 and a webcam camera 220 via a USB. Autonomous vehicle driving method using. 8. The method of claim 7,
Wherein the platform controller 100 controls the wireless receiver 130 and the wireless receiver 130 communicates with the radio control system via Bluetooth.

Images