KR100882944B1 - Parking control method of an automated guided vechicle - Google Patents

Abstract

A stop control method for a manless transfer apparatus is provided to recognize a stop indicator accurately and stop a manless transfer apparatus at an exact stop location, thereby minimizing necessary time for loading or unloading containers. A stop control method for a manless transfer apparatus comprises the following steps. A camera installed on a manless transfer apparatus photographs images to find a reference point of a stop indicator(S100). RGB image information of the photographed images is converted into HIS image information(S110). The stop indicator is found by comparing the HIS image information to previously inputted HIS image information about the surroundings of the stop indicator(S130). The exact stop location is determined and the manless transfer apparatus is stopped at the exact stop location(S170).

Description

PARKING CONTROL METHOD OF AN AUTOMATED GUIDED VECHICLE}

The present invention relates to a stop control method of an unmanned transport device, and more particularly, to a stop control method of an unmanned transport device that can stop the unmanned transport device to a desired target point more accurately and efficiently.

As global demand increases, there is a growing demand to reduce the time and cost of loading and unloading containers by crane at the port. Many researches are being conducted at home and abroad on the unmanned transfer device.

Advanced container terminals, such as the European Terminal, have built and operated an unmanned automation system to make container unloading and shipping more efficient. The most important component of such port unmanned automation system is moving within the terminal and transporting containers. Automated Guided Vechicle (AGV).

In the port unmanned automation system, when the unmanned transportation equipment stops under the crane, the spreader of the crane descends and lifts the container loaded on the unmanned transportation equipment to load it on the cargo ship or to load the container on the yard on the unmanned transportation equipment. Do it.

In this case, there should be little error in determining the stop position of the unmanned transportation system so that the spreader of the crane can lower or lift the container precisely in the unmanned transportation device, and if the unmanned transportation device does not stop at the correct stop position, There is a problem that the time and cost burden increases to correct the position.

Accordingly, there is an increasing demand for a system and method for recognizing a stop position of an unmanned transportation device more accurately and quickly so that the unmanned transportation device can be positioned at a desired stop position.

In the present invention, in operating the unmanned transportation device, prior to unloading the cargo loaded on the unmanned transportation device or before loading the cargo on the unmanned transportation device to determine the exact stop position of the unmanned transportation device quickly and virtually no error. have.

In addition, there is a technical problem that the operation can be made more quickly and efficiently when loading or unloading cargo in the unmanned transportation device by using a crane by quickly and accurately grasp the exact stop position of the unmanned transportation device.

The present invention for performing the technical problem, and the camera provided in the unmanned transportation device; A stop display unit provided outside the unmanned transportation device, the RGB image information being provided, and having a reference point which is a reference of whether the target stop position of the unmanned transportation device is deviated; An HSI converting unit converting RGB image information of the still display unit recognized by the camera into HSI image information; The position of the reference point is found by comparing the information recognized by the HSI conversion unit with the HSI image information of the reference point, and it is determined whether the unmanned transportation apparatus reaches the target stop position by using the position of the reference point. It provides a unmanned transportation apparatus stop control system comprising a control unit for controlling the operation of the unmanned transportation apparatus.

The RGB image information included in the still display unit may be provided in a plurality of colors provided to be partitioned with each other and having different color values.

The HSI image information may include brightness image information and color image information.

The controller may search for a corner using the Harris corner detector from the brightness image information, and determine the reference point by determining whether the color image information around the found corner and the HSI image information of the reference point coincide with each other.

The stop display unit may be provided in a disc shape, and the intersection may be disposed at the center of the stop display unit.

When the control point is recognized in the frame of the image photographed by the camera, the control unit continuously tracks the control point by the Lucas-Kanade method.

When the reference point is recognized within the frame of the image photographed by the camera, the controller calculates a change in the relative position of one corner of the frame of the image photographed by the camera and the reference point to the target stop position and the unmanned transportation device. It is characterized by calculating the distance from the current position.

The controller may be configured to stop the unmanned transportation apparatus when the one point specified in the frame of the image photographed by the camera coincides with the reference point.

The unmanned vehicle stop control system includes a signal generator configured to generate a signal for guiding movement of the unmanned vehicle, and a signal disposed in the unmanned vehicle so as to recognize the signal. And a recognition unit, wherein the signal generation unit is arranged in such a way that the unmanned transportation device can circulate around the stop display unit.

The present invention also includes the steps of photographing the image to find the reference point of the still display unit by a camera provided in the unmanned transportation device; Converting RGB image information of the captured image into HSI image information; Finding the still display unit by comparing the HSI image information with HSI image information about the reference point previously input; And determining a target stop position from the position of the stop display unit and stopping the unmanned transportation apparatus at the target stop position.

The HSI image information of the photographed image may include brightness image information and color image information.

The step of finding the still display unit by comparing the HSI image information and the HSI image information around the reference point previously inputted may include searching a corner using a Harris corner detector from the brightness image information, and searching the color image information around the found corner. The reference point is found by determining whether the HSI information around the reference point matches.

And comparing the HSI image information with the previously input HSI image information around the reference point to find the still display unit, continuously updating the photographed image frame to continuously track the position of the still display unit. It is characterized by.

The step of continuously tracking the position of the stop indicator is characterized by continuously tracking the position of the stop indicator using the Lucas-Kanade method.

And calculating a distance between the stop display unit and the unmanned transportation device.

The step of stopping the unmanned transportation device is characterized in that it comprises a step of reducing the speed fed back to the proportional controller while increasing the gain value of the proportional controller.

According to the present invention, it is possible to reduce the time required for unloading or shipping a container by accurately determining a stop display indicating a stop position, and continuously tracking it to stop an unmanned transportation device at a target stop position within a short time. There is an advantage.

Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings.

FIG. 1 schematically shows a crane 10 constituting a part of a container terminal and an unmanned transportation device 80 operated below the crane 10. As disclosed in FIG. 1, the crane 10 includes a frame 20 made of an iron structure, a rail 30 provided on an upper portion of the frame 20, and guided by the rail 30. It consists of a trolley 40 to be moved.

The trolley 40 is provided with a spreader 50 connected to a wire 60 to move the container C up and down while holding the container C.

In addition, the unmanned transportation device 80 moves around the crane 10 according to a predetermined signal. For this purpose, a circulation track 90 is provided around the crane 10, and the circulation track 90 A signal generator 100 for generating a signal for guiding the movement of the unmanned transportation device 80 is provided. Here, the signal generator 100 may include an RFID or a magnetic device, but is not limited thereto.

On the other hand, the unmanned transportation device 80 is provided with a body portion 110 capable of loading a container, drive wheels 120 provided on both sides of the body portion 110 and the body portion 110, the signal generator It includes a signal receiving unit 130 for receiving the guide signal generated in the (100).

The unmanned transportation device 80 should be in the target stop position to allow the spreader 50 of the crane 10 to lift or lower the container C from the unmanned transportation device 80. do. Here, the target stop position is a stop position of the unmanned transportation apparatus 80 in which the container C should be located at the bottom of the trajectory when the spreader 50 moves up and down.

In order to display the target stop position, a predetermined stop display unit 200 is provided in the frame 20 of the crane 10, and the stop display unit 200 is recognized on one side of the unmanned transportation device 80. A camera (not shown) is provided. The camera continuously photographs an image to recognize the still display. Here, the camera may be composed of, for example, a CCD.

As shown in FIG. 2, the unmanned transportation device 80 controls the stopping and movement of the device, and includes the signal recognition unit 130, the camera 150, and the driving unit 170 of the unmanned transportation device. The controller 160 to be connected is provided.

On the other hand, the camera 150 is connected to the HSI conversion unit 150 for converting the RGB image information of the still display unit 200 to the HSI image information, this HSI conversion unit 150 is also connected to the control unit 160. . As will be described later, it is possible to determine whether the target stop position of the unmanned transportation apparatus is reached by comparing the HSI image information with preset data.

As such, the reason for replacing the RGB image information of the captured image with the HSI image information is that in the case of the RGB image information, since the wrong information may be recognized by the light, that is, the gloss, in addition to the color, the adverse effect caused by the light may be excluded. This is to read image information based on Hue, Saturation and Intensity.

As shown in FIG. 3, when the stop display unit 200 is configured, the overall shape is a circular shape having a predetermined radius R, and the inside thereof has a predetermined horizontal line 220 and a vertical line 210. It is equally divided by, and the intersection point 230 is formed to intersect the horizontal line 220 and the vertical line 210, the intersection point 230 is a reference point for determining whether the target stop position of the unmanned transportation device is reached. (230).

The reason for the circular configuration of the stop display unit 200 is that the reference point can be clearly recognized regardless of the direction of approach to the stop display unit, the rotation of the stop display unit, and the reduction of other sizes.

On the other hand, a color having a different color value is arranged in each zone divided by the horizontal line 220 and the vertical line 210, so that the recognized stop display unit 200 is a true stop display unit showing the actual target stop position. In the case of cognitive judgment, it may be a more accurate criterion.

That is, when a plurality of colors meet the criteria at the same time than when only one color exists, the result of determining that the criteria meets the relative reliability is relatively high.

4 shows a control flow of reading the target stop position and stopping the unmanned transport apparatus in the stop control method of the unmanned transport apparatus of the present invention.

As shown in FIG. 4, first, an image is photographed by a camera installed in an unmanned transportation apparatus (S100), and after converting RGB image information of the image into HSI image information, the still display unit is input based on the converted information. It is determined whether the HSI image information is recognized (S110).

The conversion of the RGB image information into HSI image information will be described with reference to FIG. 5. FIG. 5 discloses a process of converting an RGB image into HSI image information. FIG. 5A relates to RGB image information of a captured image. 5B shows that the RGB image information is replaced with brightness image information, that is, gray-scale image information. The reason for switching to gray-scale image information is to easily determine the reference point provided on the still display unit by using a Harris corner detector. 5C and 5D replace RGB image information with color image information (Hue) and chroma image information (saturation), that is, color and chroma channel images, respectively.

The corner is searched from the brightness information using a Harris Corner detector, and the searched corners become candidates for the reference point of the stop display unit. Next, color image information around these corners is obtained, and the color image information and the HSI image information of the still display unit, that is, the HSI image information around the reference point, are compared. do.

However, when only the color image information is used, it is difficult to make an accurate judgment based on meaningless color image information (information on an unsaturated image such as achromatic color), so that binary image information (based on a predetermined threshold value in the chroma image) After obtaining FIG. 5E), a destination image is obtained by combining the binary image (FIG. 5E) with a color image, and comparing the destination image information with the HSI image information of the reference point of the still display unit. Find the reference point. By using the destination image obtained by combining the binary image and the color image as above, the color information of the low saturation pixel below the threshold value can be made zero. The threshold of saturation for obtaining binary image information is not particularly limited.

Meanwhile, the step of recognizing corners from the HSI image information converted as described above and determining whether each corner is a true still display unit is implemented by the following algorithm.

That is, a plurality of spaces separated from the brightness image information by using a Harris Corner dector to recognize a corner (I _gray (x, y)) and separated from each other by a predetermined distance d as shown in Table 1 and FIG. 6 below the corner. Obtain the destination image information obtained by using color image information or binary image information as a mask at the point of I _dest (x ± d, y ± d), and then compare the information with the HSI image information of the reference point to determine the corner. Is a true stop display indicating the target point of the unmanned transportation system?

If the still display unit is not recognized in the image photographed through the above process, the image is continuously recorded with the movement of the unmanned transportation device and the process of searching for the still display unit is repeated.

Meanwhile, when the HSI image information of the specific point of the captured image and the image information of the still display unit coincide with each other, the still display unit is found (S130).

The tracking method is as follows. That is, when the unmanned transportation device moves and the camera provided in the unmanned transportation device recognizes the stop display unit, the position of the stop display unit continuously changes within an image frame in which the stop display unit is photographed. We use the Lucas-Kanade method which continuously tracks the optical flow of the reference point.

After the circular stop indicator is detected, the trace of the stop indicator in the continuous image frame is traced using the Lucas-Kanade optical flow method. After locating the moving position of the reference point of the stop display unit, in order to obtain the exact coordinates of the reference point, the Harris corner detector is applied from the estimated reference point position to a predetermined radius to obtain the exact reference point coordinates.

In this case, it is preferable that the range of obtaining the exact coordinates of the reference point by applying the harris corner detector is the range of the radius of the stop display part. In other words, by applying the Harris corner detector from the estimated position of the reference point to the range outside the radius of the stop display, the exact coordinates of the reference point are obtained.

If you look with reference to Figure 7 and Table 2 below for how to obtain the radius of the stop display portion, around the reference point of the stop display of the circle by a plurality of portions spaced apart by the predetermined distance _{(d) (I dest (x} ± d , y ± d)), and then, by increasing the distance (r) and checking the difference in the color values of the plurality of parts (I _dest (x ± d ± r, y ± d ± r)), The radius R of is obtained.

FIG. 8 is a photograph showing that such an algorithm is actually applied. In FIG. 8, a circle of green grass that surrounds a still display plate inside a red rectangle shows a position of a still display plate present in an image frame at the shooting moment.

As such, the reason why the green-colored torus display, the horizontal division line, and the vertical division line are drawn is to finally determine the area of the still display panel to align the position of the still display part on the frame (S140).

On the other hand, when the unmanned transportation device continuously moves to calculate the distance between the unmanned transportation device and the target stop position (S150), wherein the control unit is a corner of the frame of the image recognized by the camera and the reference point The relative position change is calculated to calculate the distance between the target stop position and the current position of the unmanned vehicle.

The result of the calculation is shown in FIG. The distance represented by Dis in FIG. 9 is distance.

Then, the above-described comparison value reaches a predetermined initial setting value, and it is determined whether or not the reference point on the still display panel coincides with a specific point indicated in the image frame (in the present invention, a lime green reference line) (S160). If this is the case (when the distance between the target stop position indicated by the stop display unit and the unmanned transport device becomes 0), the unmanned transport device is stopped, as shown in FIG.

In the control configuration for stopping the unmanned transportation device, a proportional controller 400 is attached in front of the process 300 to control, where the input value e is a speed and the output value y is a target stop position and an unmanned transport device. As the distance between the unmanned transportation system and the target stop position increases, the feedback value is increased by increasing the gain value of the proportional controller 400, and the speed (e) fed back toward the proportional controller 400 gradually increases to 0. As it nears, it eventually stops at the target stop position.

FIG. 11 shows the experimental results of the present invention. As a result of reviewing over 50 experiments, the average deviation range of about 0.25 cm from the desired stop position reference point and the maximum deviation range of 0.48 cm are substantially accurate. It can be seen that the vehicle can be stopped at the target stop position.

1 is a perspective view schematically showing an unmanned transportation apparatus and a crane to which the present invention is applied.

2 is a control block diagram of a stop control system for an unmanned transportation apparatus according to the present invention.

3 is a front view of the stop display according to the present invention.

4 is a control flowchart of a stop control method for an unmanned transportation apparatus according to the present invention.

5 is a photograph showing a process of converting RGB image information into HSI image information.

FIG. 6 illustrates a coordinate system for collecting image information data based on a corner in HSI image information.

7 illustrates a coordinate system for obtaining a radius of the stop display unit.

8 is a photograph showing a process of tracking the stop display unit.

9 is a table showing a result of calculating the distance between the stop display unit and the unmanned transport apparatus.

10 is a control block diagram for stopping the unmanned transportation apparatus in the present invention.

11 is a table showing a deviation range between the target stop position and the actual position when the unmanned transportation apparatus of the present invention is stopped.

<Description of main parts of drawing>

10: crane 20: frame

30: rail 40: trolley

50: spreader 60: wire

80: unmanned transportation device 90: circulation track

100: signal generator 110 body

120: drive wheel 130: signal receiver

140: camera unit 150: HSI conversion unit

160: control unit 170: unmanned transportation device driving unit

200: stop display portion 230: reference point

Claims (7)

Photographing an image to find a reference point of the still display unit by a camera provided in the unmanned transportation device; Converting RGB image information of the captured image into HSI image information; Finding the still display unit by comparing the HSI image information with HSI image information about the reference point previously input; And determining a target stop position from the stop display unit and stopping the unmanned transportation apparatus at the target stop position. The method of claim 1, The HSI image information of the captured image includes a brightness image information and a color image information unmanned transportation apparatus stop control method. The method of claim 2, The step of finding the still display unit by comparing the HSI image information and the previously input HSI image information around the reference point, The corner is searched using the Harris corner detector from the brightness image information, and the stop display unit is searched by determining whether the color image information around the found corner and the HSI image information around the reference point coincide with each other. Control method. The method of claim 1, After the step of finding the still display unit by comparing the HSI image information and the pre-input HSI image information around the reference point, And continuously tracking the position of the still display by continuously updating the photographed image frame. The method of claim 4, wherein And continuously tracking the position of the stop indicator by using the Lucas-Kanade method. The method of claim 1, And calculating a distance between the stop display unit and the unmanned transportation device. The method of claim 1, The step of stopping the unmanned transportation device is a step of controlling the unmanned transportation device, characterized in that the step of increasing the gain value of the proportional controller while reducing the speed fed back to the proportional controller.

Images