KR20210009032A - Kit device for automatic truck car and control method - Google Patents

Abstract

Disclosed is an autonomous vehicle kit device and a control method using the same. In the present embodiment, a plurality of participants individually input an algorithm inputted to the autonomous vehicle, and driving is performed along a lane according to an inputted algorithm. The autonomous vehicle kit device includes: a body portion forming an external shape; a first motor and a second motor; an image module unit which provides image information about the front thereof; a sensor which detects the presence of an obstacle in front or the distance to the obstacle; a control unit which extracts data and lane data and provides the same to participants; and an input part for directly inputting an algorithm required for driving.

Description

Kit device for an autonomous vehicle and a control method using the same {Kit device for automatic truck car and control method}

The present invention relates to a kit device for an autonomous vehicle capable of directly and variously inputting and experiencing an algorithm input to an autonomous vehicle by a plurality of participants, and a control method using the same.

In general, an autonomous vehicle refers to an intelligent vehicle in which driving, stopping, turning, acceleration, or deceleration is automatically performed by a computer or the like without directly manipulating the vehicle even if there is no driver or there is a driver. The main tasks of such an intelligent vehicle are maintenance of a driving lane, securing a safe distance with adjacent vehicles, detection of adjacent obstacles, collision avoidance, and vehicle speed control according to traffic conditions or road conditions.

In recent years, due to the development of technology, a Lane Departure Warning System (LDWS), a safe driving assistance system such as lane maintenance, and an automatic vehicle control system have been developed, and the practical use is rapidly progressing. In particular, detection of driving lanes is one of the core technologies to solve major problems in autonomous vehicles, and many studies are actively being conducted amid international interest.

That is, since detection of a driving lane has a great influence on safe driving, accurate driving lanes are detected using various sensors to estimate and determine the position of the lane. For example, various sensors, such as an image sensor, a radar (RADAR), or a lidar (LIDAR) sensor, are used to implement an autonomous vehicle control system in a single or combined form for detection of a lane or object recognition in front of a vehicle.

Apart from the above-described autonomous vehicle, a frame consisting of a simple structure that forms the exterior of the vehicle for the purpose of education or coding education for students, a sensor installed in the frame, a Bluetooth module for communication, and front and rear of the frame A product in the form of an autonomous vehicle kit including a wheel provided in the wheel, a motor that provides driving force to the wheel, and a controller that controls the number of rotations of the motor is being marketed to students and the general public.

The autonomous vehicle kit is used as an experiential type in which a plurality of experienced users variously manipulate or change the logic provided as an open source according to a driving environment, and simply repeatedly experience a previously input driving command.

In this case, the purpose of coding education became meaningless because there was no basis for trying various ideas or driving patterns of users.

For example, the existing autonomous vehicle kit is used for the purpose of driving only from the start position to the end position without simply deviating along a preset road line, so that the user can directly manipulate and change the algorithm according to the movement of the autonomous vehicle. A problem that is difficult to check and reflect the driving result has been caused, so a countermeasure is needed.

Republic of Korea Patent Publication No. 10-1121777

Embodiments of the present invention are intended to provide a kit device for an autonomous vehicle, and a method for controlling the same, in which an individual participant operates a driving algorithm for operating a kit device for an autonomous vehicle differently, and is easy to modify and change according to a result.

A kit device for an autonomous vehicle according to an embodiment of the present invention includes a body part 100 forming the outer shape of the autonomous vehicle 1; A first motor 210 providing driving force to the front wheels 10 provided at the front and lower sides of the body 100, and a second motor 220 providing driving force to the rear wheels 20 provided at the rear lower side. Motor unit 200 including; An image module unit provided in the body unit 100 and providing image information on the front of the body unit 100 when the body unit 100 travels along the track 30 marked with the lane 2 (300); A distance sensing unit 400 for detecting the presence or absence of an obstacle in front and a distance to the obstacle when the body unit 100 moves along the lane 2; After setting the region of interest based on data input from the image module unit 300 and the distance sensing unit 400, and extracting grid data and lane data obtained by dividing the set region of interest into n equal parts on the X and Y axes, The control unit 500 for providing this to the participant; And an input unit 600 for the participant to directly input an algorithm required for driving the autonomous vehicle 1 by using grid data and lane data provided through the control unit 500.

An angle adjusting part 700 installed in front of the body part 100 and configured to adjust an angle between the image module part 300 and the distance sensing part 400 according to the inclination of the front track 30 It is equipped.

In the control method of the kit device for an autonomous vehicle according to an embodiment of the present invention, a participant individually directly manipulates an algorithm for driving the autonomous vehicle along a track to experience the driving of the autonomous vehicle. A first step (ST100) of receiving data required for driving of the vehicle through a sensor; A second step (ST200) of setting an ROI required for driving of the autonomous vehicle among the input data; A third step (ST300) of setting a boundary point for extracting a lane formed in the region of interest to provide the participant with lane data necessary for driving of the autonomous vehicle in the set region of interest; A fourth step (ST400) of extracting only straight line data from the set boundary point; A fifth step (ST500) of detecting grid data in an XY axis of a set region of interest after detecting only lane data among the extracted straight line data; A sixth step (ST600) of providing the detected lane data along with the grid data to the participant, and selectively adjusting the detected lane data information by the participant; A seventh step (ST600) in which a driving command is transmitted to the autonomous vehicle according to the command selected by the participant, and the autonomous vehicle is driven along a lane marked on a track; An eighth step (ST700) of determining a driving state by detecting a traffic light while driving along the lane; And a ninth step (ST800) of controlling a driving state when an unexpected situation or an obstacle is found in the autonomous vehicle while driving.

The first step (ST100) includes image information data on the front area of the autonomous vehicle and distance information data on the front area of the autonomous vehicle, and the first step (ST100) includes other than lanes and lanes. A sensitivity adjustment step (ST110) of adjusting a difference in sensitivity for an area of A distortion correction step (ST120) of correcting the distortion after calculating lens distortion generated while passing through a lens of an image module unit providing image information among the sensors is further included.

The third step (ST300) includes changing color image data input through the image information data into gray color image data (ST310);

Blurring to remove noise that does not correspond to the lane from the gray color image data to extract only lane data corresponding to a continuously extended lane among the gray color image data And a first noise filtering step (ST320) of performing noise filtering through.

The fourth step (ST400) includes setting a parameter for extracting straight line data corresponding to a lane from a boundary point through a hop transformation (ST410); And extracting only the last lane data based on the set parameters (ST420).

In the fifth step (ST500), a front lane, a left lane required for driving, and a left lane, required for driving, are extracted from the straight line data extracted from the image coordinate system in which the right direction is the positive x-axis direction and the bottom is the positive y-axis direction with the upper left as the origin. And providing data by selecting only right lane data (ST510).

In the sixth step (ST600), the participant adjusts the layout of the track and the driving mode of the autonomous vehicle to an arbitrary set value together with the lane data information,

It is characterized in that a plurality of participants are adjusted to different settings, respectively.

According to embodiments of the present invention, information about the algorithm may be naturally acquired by a participant individually inputting an algorithm according to the movement of an autonomous vehicle driving along a lane and experiencing whether driving is normally performed along the lane.

Embodiments of the present invention enhance educational effects as participants can understand, implement and experience the operating principle of an autonomous vehicle, and implement them individually while sharing with a plurality of participants using one autonomous vehicle. Education with improved participation becomes possible by experiencing the experience of implementing the algorithm individually.

In the embodiments of the present invention, different algorithms can be input for each participant, and accordingly, an experience in which an optimal algorithm can be implemented through an experience when an autonomous vehicle is driving normally or abnormally along a lane can be experienced. .

1 to 2 are views showing a kit device for an autonomous vehicle according to an embodiment of the present invention.
3 is a block diagram showing a control unit provided in the kit apparatus for an autonomous vehicle according to an embodiment of the present invention and a configuration associated with the control unit.
4 to 5 are operational state diagrams of the angle adjustment unit according to an embodiment of the present invention.
6 to 7 are flowcharts illustrating a method of controlling a kit device for an autonomous vehicle according to an embodiment of the present invention.
8 is a diagram illustrating an example of an image to which the grid data of FIG. 6 is applied.
9 is a view showing an example of an image to which an actual driving image and an algorithm of FIG. 6 are applied.
10 to 11 are flow charts illustrating an embodiment of the third to fourth steps shown in FIG. 6.
12 is a diagram illustrating a state in which a participant adjusts a variable according to a hop transformation according to a fourth step shown in FIG. 6 as an example.
13 is a view showing a state in which a traffic light is installed and detected on a track according to an embodiment of the present invention.

A kit device for an autonomous vehicle according to an embodiment of the present invention will be described with reference to the drawings.

For reference, in this embodiment, a participant including a plurality of students or ordinary people who are interested in the autonomous driving vehicle, not a vehicle that a participant directly boards and drives along a road, is a straight section and a curved section. Individually input algorithms for various driving of the autonomous vehicle (1) driving along a track marked with a lane having a path of, but the participant inputs different algorithms to move the autonomous vehicle moving along the lane We want to be able to experience various routes.

In this case, the present embodiment promotes independent participation of the participant, and even when an error occurs while the autonomous vehicle 1 is driving along a lane, the cause of the error is directly found and the cause is solved. It is possible to induce a solution to be solved directly, and to naturally acquire and experience the thinking ability for implementing an autonomous driving algorithm.

1 to 3, the kit device for an autonomous vehicle according to the present embodiment includes a body part 100 forming the outer shape of an autonomous vehicle 1, and provided at the front and lower sides of the body part 100. A motor unit 200 including a first motor 210 providing driving force to the front wheel 10 and a second motor 220 providing driving force to the rear wheel 20 provided at the rear lower side, and the body An image module unit 300 provided in the unit 100 and providing image information on the front of the body unit 100 when the body unit 100 travels along the track 30 marked with the lane 2 ) And, when the body part 100 moves along the lane 2, a distance detection part 400 that detects the presence or absence of an obstacle in front or the distance to the obstacle, and the image module part 300 and the distance detection A control unit 500 configured to set a region of interest based on data input from the unit 400, extract grid data and lane data obtained by dividing the set region of interest into n equal parts on the X-axis and Y-axis, and provide it to the participant. And an input unit 600 for a participant to directly input an algorithm required for driving the autonomous vehicle 1 using grid data and lane data provided through the control unit 500.

The body part 100 is made of plastic for weight reduction, and the motor part 200 includes a first motor 210 made of a servo motor to drive the front wheel 10 and a DC motor to drive the rear wheel 20 It consists of a second motor 220 made of.

The first motor 210 is provided for steering, and the forward and backward movement of the body part 100 is performed by the second motor 220 provided in the rear wheel 20. In addition, the second motor 220 is provided on each of the rear wheels, and the rotation speed may be variable.

The body unit 100 may receive data input from the input unit 600 through a transmission/reception module provided on the front side.

The participant inputs a control command through the input unit 600, and the controller 500 moves the autonomous vehicle 1 along the lane 2 and the curved section according to the control command input by the participant. In this case, the first mode and the second mode are controlled.

The first mode controls the second motor 220 to rotate at the same speed, and is used when moving a straight section.

In the second mode, when driving a curved section such as a curved section, the number of revolutions per minute of the first motor 210 and the second motor 220 are controlled differently, and through this, stable driving according to the degree of bending of the lane 2 This is done.

The image module unit 300 may be, for example, a camera module, installed in front of the body unit 100, and a wide-angle lens may be used, but other lenses may be used.

The image module unit 300 is located in the center with respect to the front of the body unit 100, and a camera having a view angle of 150 degrees or more than 150 degrees is used. The resolution has a resolution of 720P or 720P or higher.

As an example, the distance sensing unit 400 uses a laser sensor, and both the image module unit 300 and the distance sensing unit 400 transmit corresponding data to the control unit 500.

The image module unit 300 may be installed at each of the front and rear sides of the body unit 100, and the distance sensing unit 400 is also installed at the front and rear sides of the body unit 100. It could also be possible.

The present embodiment is for a participant to directly manipulate and experience the algorithm required for driving of the autonomous vehicle 1, and the control unit 500 is inputted from the image module unit 300 and the distance detection unit 400. Lane data and grid data corresponding to the region of interest S set based on the data are provided to participants.

The region of interest S corresponds to an area in front of the body 100, and the front area corresponds to an area necessary for accurately recognizing lane information required for the autonomous vehicle 1 to travel along a lane.

The region of interest S corresponds to a range sensed by the image module unit 300 and uses image data for the lower region based on the center of the front region of interest for lane recognition.

The lane data includes lane data for the front left side, the front side, and the front right side of the body 100 in order to accurately recognize the lane displayed in the region of interest S to be described later.

The lane data will be described in more detail in an embodiment to be described later.

Since the autonomous vehicle 1 according to the present embodiment is not a vehicle on which the participant directly boards, the autonomous vehicle 1 is displayed on the track 30 even when only data is acquired for the front area of the body 100. It is easy to move along the lane 2.

The input unit 600 may be a PC or a laptop computer, or a keyboard associated with a terminal capable of arithmetic processing, but may be replaced with other components capable of input.

In addition, the input unit 600 can visually check the data input by the participant in connection with the monitor, and check the image information on the region of interest corresponding to the front of the autonomous vehicle 1 sensed through the image module unit 300. can see.

3 to 5, the present embodiment is installed in front of the body part 100, and the image module part 300 and the distance sensing part 400 are installed according to the inclination of the front track 30. ) Is provided with an angle adjustment unit 700 for adjusting the angle. The body part 100 is provided with a tilt sensor 800, and the controller 500 receives the current tilt value of the body part 100 from the tilt sensor 800, and when it exceeds a preset tilt, a tilting plate ( The angle of the image module unit 300 and the distance sensing unit 400 are simultaneously adjusted by rotating the angle of 710 toward the upper or lower side of the body unit 100.

In this case, the present embodiment can accurately recognize the lane 2 regardless of whether the track 3 is inclined or not and minimize the occurrence of errors, thereby achieving stable movement.

The angle adjustment unit 700 is provided with a plate-shaped tilting plate 710 on the front of the body portion 100, and the image module unit 300 and the distance sensing unit 400 are provided on the front of the tilting plate 710 Installed.

The angle adjustment unit 700 has a first gear 720 having one end connected to the rear surface of the tilting plate 710, the other end extending rearward, and having a tooth shape G formed in some sections, and the first gear 720 It includes a tilting motor 730 that provides a rotational force in a clockwise or counterclockwise direction as a spur gear meshed with the teeth (G) of.

The tilting plate 710 is coupled to the body portion 100 via a hinge H provided in the center of the rear surface.

For example, when the autonomous vehicle 1 travels in an inclined section, tilt data of the body part 100 detected by the tilt sensor 800 is transmitted to the control unit 500, and the control unit 500 is When it exceeds the inclination data, the tilting motor 730 is operated to control the tilting plate 710 to move toward either the upper side or the lower side.

For example, when the autonomous vehicle 1 is driven in a section in which the track 30 is horizontal, the tilting plate 710 is maintained in a vertical state with respect to the front surface of the body part 100.

As shown in FIG. 4, when there is a section in which the track 30 is inclined upward, the data of the forward lane detected through the image module unit 300 provided in the front of the body unit 100 is The region of interest S corresponding to the front of the unit 100 may not be accurately detected, and a region outside the region of interest S may be detected.

In this embodiment, the tilting plate 710 is tilted and moved according to the presence or absence of the inclination of the track 30 by the inclination data sensed by the inclination sensor 800 for accurate detection of the region of interest (S). Data on the area S can be transmitted to the control unit 500 so that the autonomous vehicle 1 recognizes and drives the lane 2 at all times.

For reference, when the track 30 is inclined upward, the tilting plate 710 is moved from the solid line position to the front upward side shown by the dotted line.

As shown in FIG. 5, when the track 30 is inclined downward, the tilting plate 710 is moved from the solid line position to the front and lower side shown by the dotted line to accurately detect the region of interest (S). You can do it.

The control unit 500 controls the operation of the autonomous vehicle 1 to be stably operated based on algorithm data input by the participant, and the subject according to the movement of the autonomous vehicle 1 is mostly performed by the participant.

A method of controlling a kit for an autonomous vehicle according to an embodiment of the present invention will be described with reference to the drawings.

3 or 6 to 9, the present embodiment uses the above-described control unit 500 to provide basic data required for direct driving by a participant in order to stably drive an autonomous vehicle along a lane. After receiving through the input unit 600 and inputting through the input unit 600, it is visually checked whether the driving is stably performed, and if driving is unstable along the lane 2 indicated on the track 30, or the lane 2 If it exceeds, other data may be input through the input unit 600 to promote normal driving.

To this end, in this embodiment, a participant individually directly manipulates an algorithm for driving the autonomous vehicle along a track to experience the driving of the autonomous vehicle, and transmits data necessary for driving the autonomous vehicle through a sensor. A first step (ST100) of receiving an input, a second step (ST200) of setting a region of interest necessary for driving of the autonomous vehicle among the input data, and a lane required for driving of the autonomous vehicle in the set region of interest A third step (ST300) of setting a boundary point for extracting a lane formed in the region of interest to provide data to the participant, a fourth step (ST400) of extracting only straight line data from the set boundary point, and the extracted straight line A fifth step (ST500) of detecting grid data in the XY axis of a set region of interest after detecting only lane data among data, and the detected lane data along with the grid data are provided to the participant, and the participant is detected. A sixth step of selectively adjusting lane data information (ST600), a driving command is transmitted to the autonomous vehicle according to the command selected by the participant, and the autonomous driving vehicle is driven along a lane displayed on the track. Step (ST600), an eighth step (ST700) of determining a driving state by detecting the color of a traffic light while driving along the lane, and controlling a driving state when an unexpected situation or an obstacle is found while the autonomous vehicle is driving. And a ninth step (ST800).

The first step ST100 includes image information data on a front area of the autonomous vehicle and distance information data on a front area of the autonomous vehicle.

The image information data acquires data necessary for lane recognition by using image information on an ROI S corresponding to the front of the autonomous vehicle 1.

For example, the image module unit photographs the region of interest S and transmits image information data to the control unit 500.

The participant adjusts the difference in sensitivity for areas other than the lane in the lane data input as image information data (ST110).

In order to accurately detect the boundary between a road and a lane, the difference in sensitivity is adjusted to adjust the difference in sensitivity according to the black displayed as the lane and the white pixel representing the track.

The reason for adjusting the difference in sensitivity between the above lanes and areas other than the lanes is to accurately detect the boundary between the road and the lane, and differentiates whether a foreign substance or a specific structure on the road corresponds to a lane or another part of the track. The algorithm can be adjusted by the participant to ensure that the driving vehicle stably moves along the lane.

The lane 2 is displayed on the track 30 as an example with a black line, and the track 30 is displayed in white. In the case of light, when all three primary colors of red, green, and blue are mixed, the characteristic is brightened. And the mixture of the three primary colors is displayed in white.

When the white color is detected by the image module unit, the pixel is relatively high compared to the lane 2 displayed in black, so that the pixel and the sensitivity of the lane are different from each other.

The present embodiment detects a difference in pixels between the black indicated by the lane 2 and the white indicated by the track 30 by using this characteristic, and detects the boundary between the lane and the lane.

In addition, the control unit calculates lens distortion generated while passing through the lens of the image module unit, and then performs a distortion correction (ST120) that corrects the distortion to enable accurate matching when the image data and the actual track are matched with each other.

For example, due to lens distortion, the lanes of the image data and the lanes of the actual track may appear different from each other, and since distortion increases toward the edge of the grid data to be described later, correction is performed to obtain accurate lane data. .

For example, in order to set the region of interest after removing image distortion, the region of interest is set after removing the distortion of the entire region based on V4 (see FIG. 8), and the region of interest is provided to the user.

That is, in order to correct lens distortion generated by the lens provided in the image module unit 300, accurate image information data by removing distortion for the entire area based on the center position of the grid data (position where V4 and H2 intersect). To the participant.

The reason why lens distortion occurs is due to the inherent characteristics of the lens provided in the image module unit. In the correction of the lens distortion, the distortion is removed in a radial form centered on the center. When the distortion is corrected in this way, image data that is flattened without bending the edge portion is obtained.

In addition, the region of interest S is set. Since the region of interest S has already been described above, a detailed description thereof will be omitted.

Referring to FIGS. 10 to 12, in the third step ST300, only accurate lane data required for lane extraction may be extracted by finding a boundary point corresponding to a lane and noise.

Prior to extracting the correct lane, the lane extends in a straight line and there is a boundary area adjacent to the lane.

In the boundary area, boundary points other than the lanes and lanes exist in order to extract accurate lane data, and when the boundary points are first removed prior to extraction of the lanes, extraction of the lanes is advantageous and accuracy is improved at the same time. .

In particular, unnecessary noise irrelevant to the lane can be removed in advance by changing the color of the part corresponding to the noise or blurring the image.

To this end, in the present embodiment, color image data input through image information data is changed to gray color image data (ST310).

According to the present embodiment, image processing such as image enhancement may be performed more easily by performing edge detection by changing image data to gray color in order to more accurately detect only lanes. . Participants can adjust the sensitivity to the optimum state by increasing or decreasing the sensitivity according to the change through the input unit.

And blurring to remove noise not corresponding to the lane from the gray color image data in order to extract only lane data corresponding to the continuously extended lane among gray color image data. Through the first noise filtering (ST320) is performed.

The first noise filtering (ST320) removes unnecessary noise by blurring the entire image in order to filter noise that looks like a lane although it is not a lane among images obtained from image data.

In this case, the lane is continuously extracted from the front of the free-running vehicle 1, but in the case of noise other than the lane, the non-continuous characteristic is maintained. The participant numerically adjusts the boundary points for lane extraction so that they can accurately find them.

In this case, the sensitivity and result values for lane extraction are slightly different, and through this, each participant can directly change or modify the input value.

Referring to FIGS. 11 to 12, since the boundary points corresponding to the lane and noise become clear after the lane is extracted, hop transformation for extracting straight line data corresponding to the lane is performed (ST400).

The hop transform (ST400) sets a parameter for extracting straight line data corresponding to the lane at the boundary point (ST410), and the final lane data is extracted (ST420) by the set parameter.

In the hop transformation, a parameter for detecting a case where a length of a certain length or more is maintained as a straight line and a parameter for detecting a case in which a certain number or more points are included as a straight line may be input. In the present invention, the optimal parameter for lane recognition The user's convenience was sought by inputting in advance.

The user sets the Min threshold parameter and Max threshold value of the Canny Edge Detector to arbitrary values so that the boundary points of the lanes can be recognized.The white line shown in FIG. 8 is the boundary point of the lane extracted from the Canny Edge Detector, and above the lane. The drawn red line corresponds to the straight line obtained through the hop transformation.

Boundary point extraction is performed by expressing the boundary surface with color change (e.g., the boundary surface changing from white to black) in white and the non-changing part in black. A straight line with coordinate values is extracted.

In this case, two boundary surfaces (white) are extracted for one lane due to the width of the lane, and when the hop transformation is performed, one lane is recognized as two straight lines. In order not to step on the lane while driving, the second noise Filtering (S330) is performed.

Since the second noise filtering (ST330) needs to know the boundary surface near the vehicle, a straight line to be recognized as a left lane and a right lane is selected from all extracted candidate lane lines.

For reference, the boundary point is set according to the color change in the Canny Edge Detector, and the part where the intensity of the image changes rapidly is defined as the edge.

The position of the boundary is determined through a differential value, and two locations, maximum and minimum, can be found.

The parameters of the Canny Edge Detector that can be changed by the user can find the boundary by changing the maximum and minimum values.

For example, when the lane is in a dark environment such as at night, white is also gray or has a value close to black. In this case, in order to distinguish between the black floor and the lane, the maximum and minimum values can be entered larger than the previously entered values.

Therefore, the user can variously adjust and input the Min threshold and Max threshold values according to the driving environment of the current autonomous vehicle, check the driving state of the autonomous vehicle, and then derive the optimal driving.

In the HOPE transformation, the direction and angle for the region of interest can be set to a specific angle, and the participant can change input values in various ways to make adjustments for lane extraction.

The direction and angle for the region of interest are searched in the range of pi/180 degrees.

The fifth step (ST500) extracts linear data located on the image coordinate system.

For example, from the straight line data extracted from the image coordinate system with the upper left as the origin, the right direction as the positive x-axis, and the downwards as the positive y-axis, only data on the front lane, left lane, and right lane required for driving are selected. Thus, data is provided (ST510).

In the image coordinate system, the curve appears to be curved in the left or right direction when viewed from a distance, but when the curve is enlarged and divided into a specific length, it corresponds to a straight line, so it can be accurately determined whether it corresponds to the left or right lane.

The above-described grid data shown in FIG. 8 corresponds to grid data obtained by dividing image data corresponding to the front area into n equal parts on the X-axis and Y-axis. For example, in this embodiment, as shown in FIG. 8, the X-axis of the front region is divided into eight, the Y-axis is divided into four, and all of the X-axis are divided into eight, so the position V4 corresponds to the center of the drawing, and the V4 is the reference. It is divided into left and right areas.

When the image data is input through the image module unit 300, the front of the body unit 100 is set as a reference, so H3 corresponds to the bottom of the track 30, which is in front of the body unit, and as the image data goes to H2 and H1, the body unit ( It corresponds to the front upper area of 100).

For example, a line marked in white corresponds to a lane, a line marked in red corresponds to the inside of the lane based on the driving direction of the autonomous vehicle 1, and the intersection point where the left lane and horizontal lines (H1 to H3) meet is blue. It corresponds to a point, and the intersection of the right lane and the horizontal line corresponds to the red dot.

When the blue dot and the red dot are connected to each other in the traveling direction of the autonomous vehicle 1, grid data having lane data capable of stably driving is generated based on the center position V4 (ST200).

In addition, the curved section connected to the straight section may be linearized as much as possible, and the lane data corresponding to the curved section can be finally extracted and recognized through the extraction and transformation of the above-described lane data.

As an example, in the original image input through the image module unit, as described above, the track 30 is formed in white and the lane 2 is displayed in black, but it may be changed to a color contrasting with each other.

The image module unit transmits image data to the control unit 500 to provide information on the region of interest, and the control unit 500 divides the region of interest into upper and lower regions, and then extracts information on the lower region and detects data. It is used for detection, and the upper side is used for detecting structures such as traffic lights.

In particular, in this embodiment, both the lower image data and the upper image data of the region of interest are input to the control unit 500, but if a traffic light is not found in the upper image data for rapid image data processing, It enables image data processing.

Therefore, since lane extraction and detection are performed more quickly and stably, the autonomous vehicle 1 can be accurately moved without leaving the lane.

In the sixth step (ST600), the participant adjusts the layout of the track and the driving mode of the autonomous vehicle together with the lane data information to an arbitrary input value, but the plurality of participants may adjust each to different input values. Because it is possible, it is possible to directly enter and implement various input values for each participant, improving the training effect.

In this way, a driving command is transmitted to the autonomous vehicle according to a command selected by the participant, and the autonomous vehicle travels along a lane marked on a track.

Referring to FIG. 13, the eighth step ST700 determines a driving state by detecting a traffic light 50 while driving along a lane. The detection of the traffic light 50 first detects the position of the traffic light 50 when it is located in the region of interest through the image module unit (ST710).

The traffic lights 50 are positioned in the order of red and yellow or orange and green, for example, and the image module unit determines brightness to accurately determine at which position of the current red, yellow, and green lights (ST720).

For example, when it is determined that the red color is currently lit, the control unit 500 controls the operation of the motor unit 200 to be stopped so that the autonomous vehicle 1 can be stopped before the traffic light.

In addition, the control unit 500 may control the motor unit 200 so that the autonomous vehicle 1 is unconditionally stopped when the lighting signal of the traffic light 50 is not input.

In the ninth step (ST800), when a pedestrian or an obstacle is found in the image data transmitted by the image module unit, the distance sensing unit 400 determines a position separated from the autonomous vehicle 1.

In addition, the operation state of the motor unit 200 is controlled so as not to collide with the pedestrian or obstacle. For example, a collision may be avoided by decelerating the speed or changing the driving position.

As described above, one embodiment of the present invention has been described, but those of ordinary skill in the relevant technical field add, change, delete or add components within the scope not departing from the spirit of the present invention described in the claims. Various modifications and changes can be made to the present invention by means of the like, and it will be said that this is also included within the scope of the present invention.

2: lane
30: track
50: traffic light
100: body
200: motor part
300: image module unit
400: distance detection unit
500: control unit
600: input
700: angle adjustment unit
710: tilting plate
720: first gear
730: tilting motor
800: tilt sensor

Claims (8)

A body portion 100 constituting the outer shape of the autonomous vehicle 1;
A motor unit 200 including a first motor 210 provided at a lower front side of the body 100 and a second motor 220 providing driving force to a rear wheel 20 provided at a lower rear side;
An image module unit provided in the body unit 100 and providing image information on the front of the body unit 100 when the body unit 100 travels along the track 30 marked with the lane 2 (300);
A distance sensing unit 400 for sensing the presence or absence of a front obstacle or a distance to the obstacle when the body unit 100 moves along the lane 2;
After setting the region of interest based on data input from the image module unit 300 and the distance sensing unit 400, and extracting grid data and lane data obtained by dividing the set region of interest into n equal parts on the X and Y axes, The control unit 500 for providing this to the participant; And
Kit apparatus for an autonomous vehicle comprising an input unit 600 for the participant to directly input an algorithm required for driving the autonomous vehicle 1 by using grid data and lane data provided through the control unit 500. The method of claim 1,
An angle adjusting part 700 installed in front of the body part 100 and configured to adjust an angle between the image module part 300 and the distance sensing part 400 according to the inclination of the front track 30 Kit equipment for self-driving vehicles equipped. The first step of receiving data necessary for driving of the autonomous vehicle through a sensor in order to experience the driving of the autonomous vehicle by individually directly manipulating the algorithm for the autonomous vehicle to travel along the track ( ST100);
A second step (ST200) of setting an ROI required for driving of the autonomous vehicle among the input data;
A third step (ST300) of setting a boundary point for extracting a lane formed in the region of interest to provide the participant with lane data necessary for driving of the autonomous vehicle in the set region of interest;
A fourth step (ST400) of extracting only straight line data from the set boundary point;
A fifth step (ST500) of detecting grid data in the XY axis of a set region of interest after detecting only lane data among the extracted straight line data (ST500)
A sixth step (ST600) of providing the detected lane data along with the grid data to the participant, and selectively adjusting the detected lane data information by the participant;
A seventh step (ST600) in which a driving command is transmitted to the autonomous vehicle according to the command selected by the participant, and the autonomous vehicle is driven along a lane marked on a track;
An eighth step (ST700) of determining a driving state by detecting a traffic light while driving along the lane; And
A control method of a kit apparatus for an autonomous vehicle comprising a ninth step (ST800) of controlling a driving state when an unexpected situation or an obstacle is found in the autonomous vehicle while driving. The method of claim 3,
The first step (ST100) includes image information data on a front area of the autonomous vehicle and distance information data on a front area of the autonomous vehicle,
The first step (ST100) includes a sensitivity adjustment step (ST110) of adjusting a difference in sensitivity between a lane and an area other than the lane;
A method of controlling an autonomous vehicle kit apparatus further comprising a distortion correction step (ST120) of correcting the distortion after calculating lens distortion generated while passing through a lens of an image module unit providing image information among the sensors. The method of claim 4,
The third step (ST300) includes changing color image data input through the image information data into gray color image data (ST310);
Blurring to remove noise that does not correspond to the lane from the gray color image data to extract only lane data corresponding to a continuously extended lane among the gray color image data Control method of a kit device for an autonomous vehicle comprising a first noise filtering step (ST320) performing noise filtering through. The method of claim 3,
The fourth step (ST400) includes setting a parameter for extracting straight line data corresponding to a lane from a boundary point through a hop transformation (ST410);
A method of controlling a kit for an autonomous vehicle comprising the step of extracting only the final lane data based on the set parameter (ST420). The method of claim 3,
In the fifth step (ST500), a front lane, a left lane required for driving, and a left lane, required for driving, are extracted from the straight line data extracted from the image coordinate system in which the right direction is the positive x-axis direction and the bottom is the positive y-axis direction with the upper left as the origin Control method of a kit for an autonomous vehicle comprising the step of selecting only right lane data and providing data (ST510). The method of claim 3,
In the sixth step (ST600), the participant adjusts the layout of the track and the driving mode of the autonomous vehicle to an arbitrary set value together with the lane data information,
A method of controlling a kit for an autonomous vehicle, characterized in that a plurality of participants are each adjusted to different set values.

Images