WO2019216673A1

WO2019216673A1 - Object guidance system and method for unmanned moving body

Info

Publication number: WO2019216673A1
Application number: PCT/KR2019/005581
Authority: WO
Inventors: 김경욱; 유근태
Original assignee: 주식회사 아이피엘
Priority date: 2018-05-11
Filing date: 2019-05-09
Publication date: 2019-11-14
Also published as: KR20190129551A

Abstract

An object guidance system is disclosed. More specifically, the present invention relates to: an object guidance system and method provided to a mobile robot device, etc. so as to guide the device to a destination by controlling the moving direction of the device; and an unmanned moving body docking system using same. According to one embodiment of the present invention, by means of a detection device for detecting the location of a marker mounted on a docking station and an unmanned moving body, and a guidance device for displaying the marker, an effect is achieved whereby, when the unmanned moving body is located at a far distance, the unmanned moving body is made to approach within a predetermined distance from the docking station through approximate first location information, and when the unmanned moving body approaches a near distance, precise second location information is generated so that the unmanned moving body may be seated in the docking station.

Description

Object guidance system and method for unmanned moving object

The present invention relates to an object guidance system, and to an object guidance system and method installed in a mobile robot device and the like to control the direction of movement of the device to a destination, and a docking system for an unmanned mobile object using the same.

In general, an unmanned moving object such as a drone or a robot has a built-in battery and operates as a power source of the battery, and thus, a task of charging the battery after a predetermined time after driving is required.

Particularly, in the case of an unmanned moving object such as a robot cleaner, the work is performed while driving a range of working areas within a driving period without a user's operation, so when the battery is discharged below a certain level, the docking station must be found and automatically returned to perform charging. do.

To this end, a docking station for receiving a commercial AC power through the terminal contact to charge the battery on one side of the workspace of the unmanned moving object is installed. The docking station continuously transmits a signal using an infrared ray (IR) transmitter to inform its position, and the unmanned vehicle moves to the infrared ray of the docking station through the infrared receiver when the battery becomes a low battery state. Receiving a signal from the transmitter to the docking station and docked (docking) the docking station according to the docking procedure to charge the battery.

At this time, the existing docking system is configured to measure the position and distance of the docking station to return to the docking station based on the angle of incidence or signal intensity of the infrared signal transmitted from the infrared transmitter of the docking station and received by the infrared receiver of the unmanned moving object. Can be.

However, the prior art controls the moving direction and speed of the unmanned moving object based on the incident angle of the received infrared signal or whether the signal overlaps, but in particular, the moving speed is based on the intensity of the infrared signal detected by the infrared receiver. Since controlling the moving speed of the, there is a problem that detailed control of the moving speed and direction according to the position and distance of the docking station which is the unmanned moving object and the target point is difficult.

In addition, according to the related art, expensive manufacturing means for calculating the moving direction and the speed of the unmanned moving object and a plurality of sensors, including an infrared transceiver, have to be used, thereby increasing the manufacturing cost of the unmanned moving object and the docking station.

The present invention has been made to solve the above-described problems, the present invention provides an unmanned moving object guidance system and method for allowing an unmanned moving object such as a robot cleaner to automatically move and dock to a target point such as a docking station. There is a challenge.

Another object of the present invention is to provide an object guidance system and method for an unmanned moving object by applying a marker recognition technique to calculate an optimal moving direction and speed of the unmanned moving object using only a minimum sensor.

In order to solve the above problems, the object guidance system for an unmanned moving object according to an embodiment of the present invention, the object guidance system mounted on the unmanned moving object and the docking station, an induction apparatus including a marker identified according to the lighting of the light source; And a detector for detecting the marker and calculating a distance and a direction with the induction apparatus at a current position of the unmanned moving object to generate first and second position information with different precisions. The movement of the unmanned moving object may be requested according to the movement route included in the first or second location information.

The induction device, the marker unit including the at least one LED light source, the marker is formed on the light emitting surface of the LED light source and the shape of the rectangle to form a rectangle, and receives the LED control signal transmitted from the detection device, In accordance with the LED control signal may include an infrared receiver for outputting a drive signal to the LED light source.

The detection device may include an infrared transmitter for transmitting an LED control signal to the induction device, an image sensor for acquiring a detection image of the induction device, and when the marker in the detection image is detected, the induction based on the detected marker. A marker extractor for generating first position information by calculating a distance and an angle with a device; and if the marker in the detection image is not detected, the induction based on the induced light emission signal detected by the induction light emission driving of the induction device. The target detector may be configured to calculate distance and angle from the apparatus to generate second position information, and a controller to request movement of the unmanned moving object according to the first or second position information.

The marker extracting unit obtains three-dimensional coordinates of the marker through homography operation using four-point vertex coordinates and a preset actual marker size included in the marker image, and calculates the distance and angle corresponding to the three-dimensional coordinates. can do.

The target detector may be configured to transmit an LED control signal for requesting the inductive light emission driving through the infrared transmitter, and the distance to the target region set based on a plurality of images continuously photographed for a predetermined period of time through the image sensor. The second position information may be generated by calculating an angle.

The target detector, when the image sensor acquires a plurality of first images of the surroundings during the output period of the induction light emission signal, applies one or more change regions in the plurality of first images by applying a difference image processing technique. The number of change for each change area may be detected, and a registration procedure for registering one or more change areas whose change number exceeds a threshold may be performed as a target area.

The induced light emission signal may be a signal that causes the LED light source to repeatedly turn on and blink in a cycle of 1 second or less.

The target detection unit calculates a distance and an angle with the induction apparatus in response to the size and the center coordinate of the target area when the target area is one, and weights each target area when the target area is two or more. When the image sensor acquires a plurality of second images photographing the surroundings during the non-output period of the induced light emission signal, detects one or more change regions in the second image by applying a difference image processing technique. The target area overlapping the changed change area is deleted, and the process of detecting the change area using the first image and the second image is repeatedly repeated a predetermined number of times until the target area becomes one. By increasing, one target area having the highest weight can be set.

The angle A with the induction apparatus is represented by the following equation,

(Where x is the center abscissa of the image for the target area, wh is 1/2 of the horizontal size of the image for the target area, and a is the angle of view of the image sensor).

Distance (L) with the induction apparatus is the following equation,

Where tl is the diagonal length of the image relative to the target area, Real_M is the maximum detectable distance, Real_m is the minimum detectable distance, Img_M is the diagonal length of the image at the maximum detectable distance, and Img_m is the detection Image diagonal length at the minimum possible distance).

In addition, the object induction method according to an embodiment of another aspect of the present invention for solving the above problems, the object induction method by the object induction system for an unmanned moving object, comprising the steps of (a) controlling the lighting of the light source, (b) ) Detecting the marker by executing the marker extracting unit, (c) calculating a distance and an angle between the detection apparatus and the marker when the marker is detected, outputting a search completion, and (d) not detecting the marker; Or (e) calculating the distance and angle between the detection device and the induction device, requesting to move the unmanned moving object in the direction of the induction device, if the induction device is detected. a) proceeding to the step, (f) if the induction apparatus is not detected, extinguishing the light source, outputting a search failure, (g) requesting the unmanned moving object to change the search direction, and It may include the step of proceeding to step (a).

The step (b) includes (b1) acquiring a marker image from an image sensor, and (b2) extracting a marker, and the step (c) includes: (c1) if a marker is not detected, searching Outputting a failure, proceeding to step (d), (c2) if a marker is detected, acquiring four-point coordinates from the detected marker, (c3) obtaining four-point coordinates and a preset actual marker; Obtaining a three-dimensional coordinate of the marker through a homography operation using a size, (c4) calculating a distance and an angle with the marker, and (c5) outputting the distance and an angle can do.

The step (d) may include: (d1) turning on and controlling the light source; and (d2) acquiring a first image from the image sensor for a predetermined period of time and searching for a change area in the first image to set a target area. (D3) controlling the flashing of the light source, (d4) acquiring a second image from the image sensor for a predetermined period of time, searching for a change region according to the second image in the first image, and at least one target region. In the following steps, deleting a region in which the change continuously occurs in the target region, (d5) assigning a weight to the target region, (d6) steps (d2) and (d4), If the target area does not exist, outputting a target search failure, and if the step (f) and the step (d7) (d2) and (d4), there is only one target area, Calculate the distance and angle with the induction device, Outputting the output distance and angle, and proceeding to the step (e) and (d8) in the step (d7), if the target area is two or more and does not exceed the designated number of times, the current target area is weighted. In step (d1) and (d9), in step (d7), if the target area is two or more and the specified number of times is exceeded, one of the current target areas has the highest weight. It may include selecting a target area and proceeding to the step (d7).

According to an embodiment of the present invention, the detection apparatus for detecting the position of the marker mounted on the unmanned moving object and the docking station and the guide device for displaying the marker through the first position information of the docking station at the remote position of the unmanned moving object By approaching within a certain distance, and generating a second position information accurate in close proximity, there is an effect that the unmanned moving object can be seated in the docking station.

In addition, according to an embodiment of the present invention, by using a low-cost image sensor that is impossible to identify markers on a distant subject, the first positional information is generated until the near-field approach, and then the second positional information is generated precisely, thereby reducing the low resolution. Even low-cost image sensors have the effect of precisely inducing docking stations.

1 is a view showing an application example of the object guidance system for an unmanned moving object according to an embodiment of the present invention.

2 is a view showing the structure of a thing guidance system according to an embodiment of the present invention.

3 to 5 are diagrams illustrating a method of inducing objects according to an embodiment of the present invention.

Prior to the description, when any part of the present specification is said to "include" or "include" any component, unless otherwise stated, this will not include other components, but will further include other components. It means you can.

The term "embodiments" herein is meant to serve as illustrations, examples or illustrations, but the subject matter of the invention is not limited by such examples. And other similar terms such as "comprising," "including," and "having" are used, but as used in the claims, as an open transition word that does not exclude any additional or other component. It is used generically in a similar way to the term "comprising".

Also, terms such as "... unit," ... module "," ... device "and" ... system "described throughout this specification. Means a unit for processing an operation in which one or more functions are combined, which may be implemented in hardware, software, or a combination of hardware and software.

The various techniques described herein may be implemented with hardware or software, or where appropriate, with a combination of both. As used herein, terms such as "part", "module", "device" and "system" likewise refer to computer-related entities, such as hardware, a combination of hardware and software, software or software at run time. Can be treated equivalently. In addition, in the present invention, the application program executed in the user terminal may be configured in "units" and may be recorded in one physical memory in a form that can be read, written, and erased or distributed between two or more memories or recording media. Can be recorded.

Hereinafter, an object guidance system and method for an unmanned moving object according to an embodiment of the present invention will be described with reference to the drawings.

Referring to FIG. 1, an object guidance system for an unmanned moving object according to an exemplary embodiment of the present invention may be applied to an unmanned moving object 10 such as a drone or a robot and a docking station 20 to which the unmanned moving object 10 is docked.

In particular, the unmanned mobile body 10 may be a mobile robot or the like moving on the ground, and a motherboard equipped with circuits such as a controller and a power supply unit may be built into the unmanned mobile body 10. The lower portion of the unmanned moving object 10 may be provided with a moving part 11 including a motor, and a display 12 for displaying a driving screen may be provided on a front surface thereof.

The docking station 20 is a device that allows the unmanned moving object to be docked to perform charging and the like, and may be fixedly installed in one region. The front side of the talking station 20 may be provided with an output terminal 21 for connecting to the charging terminal installed in the unmanned moving object (10).

In particular, the unmanned moving object 10 and the docking station 20 to which the object guidance system according to the embodiment of the present invention is applied are equipped with means for detecting and inducing movement of each other. The image sensor 220 photographs the surroundings and generates an image on the front or rear surface, and the infrared transmitter 210 that transmits the LED control signal for driving the marker unit 320 to the docking station 20 is exposed to the outside. Can be.

In addition, the docking station 20 includes a marker unit 320 including an infrared receiver 310 for receiving an LED control signal transmitted from the unmanned moving object 10 and a marker which is an identification means for guiding the unmanned moving object 10. May be exposed.

According to this structure, the unmanned moving object 10 and the docking station 20 to which the object guidance system is applied according to the embodiment of the present invention transmits the LED control signal from the infrared transmitter 210 of the unmanned moving object 10 when the docking is induced. As a result, the infrared receiver 310 of the docking station 20 receives the light and turns on the marker 320.

In addition, as the image sensor 220 of the unmanned moving object 10 photographs the surroundings, an image of the marker unit 320 is acquired, and the distance and angle with the docking station 20 are calculated based on the image. The movement path can be obtained. The unmanned moving object 10 attempts to approach and dock the docking station 20 automatically by controlling the moving part 11 according to the obtained movement path.

In this case, the object guidance system according to the embodiment of the present invention may selectively execute the position coordinate calculation algorithm adaptively according to the current separation distance between the unmanned moving object 10 and the docking station 20.

Specifically, the object guidance system of the present invention is within 10 m, which is a distance capable of accurately recognizing the shape of the marker 320 in terms of the performance of the image sensor between the unmanned moving object 10 and the docking station 20, and the recognition performance is At a distance of 10 m or more, which is significantly lowered, a separate algorithm is applied to generate first or second location information, and the movement of the unmanned moving object 10 is controlled.

That is, at a short distance, since the marker shape of the marker part 320 can be recognized relatively accurately, the position information can be calculated by precisely calculating the distance and angle with the docking station 20 of the unmanned moving object 10 based on the marker shape. It generates and moves the unmanned moving object 10.

In addition, although the distance corresponding to the marker of the marker 320 may be determined at a long distance, since it is difficult to calculate the exact distance and angle, the distance and angle are roughly calculated in the direction of the area estimated as the marker. The information is generated, the unmanned moving object 10 is moved to a short distance of the docking station 20, and the search of the marker part 320 is performed again to precisely move the unmanned moving object 10.

According to the current separation distance between the unmanned moving object 10 and the docking station 20, a detailed description of selectively executing the position coordinate calculation algorithm will be described later.

As described above, the object guidance system according to the embodiment of the present invention uses an inexpensive image sensor having a recognition accuracy of 10 m or less by being mounted on an unmanned moving object such as a mobile robot and a docking station and applying an algorithm differently according to its distance. Nevertheless, it is possible to precisely move the unmanned vehicle to a short distance as well as to a remote destination.

Hereinafter, the structure of the thing guidance system according to an embodiment of the present invention with reference to the drawings in detail.

2, the object guidance system 100 according to an embodiment of the present invention, by detecting the marker 340 to calculate the distance and direction with the guidance device 300 at the current position of the unmanned moving object to achieve different precision It may include a detection device 200 for generating the first and second position information having, and the induction device 300 including a marker 340 displayed according to the lighting of the light source 330, the detection device 200 ) May request the movement of the unmanned moving object according to the movement path included in the first or second location information according to different distances.

Here, the detection apparatus 200 and the induction apparatus 300 may be implemented as independent circuit members, respectively, and may be mounted on the unmanned moving object (10 of FIG. 1) and the docking station (20 of FIG. 1), respectively.

In detail, the detection apparatus 200 is mounted on an unmanned moving object to control the docking station, that is, the induction apparatus, to display or not display the marker 340 which is a reference for determining the moving direction, and based on the marker 340. By determining the relative position of the current unmanned moving object from the docking station can determine the movement path.

In order to implement such a function, the detection apparatus 200 detects an infrared ray transmitter 210 for transmitting an LED control signal to the induction apparatus 300, an image sensor 220 for acquiring a detection image for the induction apparatus 300, and detection. When the marker 340 in the image is detected, the marker extractor 230 generating the first position information by calculating the distance and angle with the induction apparatus 300 based on the detected marker 340, the marker in the detected image ( If the detection unit 340 is not detected, the target detector 240 generates the second position information by calculating a distance and an angle with the induction device 300 based on the induction light emission signal detected by the induction light emission driving of the induction device 300. ) And a controller 250 for requesting movement of the unmanned moving object according to the first or second location information.

The infrared transmitting unit 210 may be exposed on the surface of the unmanned moving object, and under the control of the control unit 250, the infrared transmitting unit 210 transmits an infrared waveform signal to the surroundings to the infrared receiving unit 310 of the induction apparatus 300 mounted in the docking station. Can transmit

As the signal transmitted through the infrared transmitter 210 is an LED control signal for controlling the lighting and blinking of the LED light source 330 provided in the induction apparatus 300, the detection apparatus 200 is the induction apparatus 300. When you want to detect the location of the, via the infrared transmitter 210 will output the LED control signal.

The image sensor 220 is a predetermined camera member, which is exposed to the surface of the unmanned moving object and photographs a surrounding to generate an image.

In particular, the image sensor 220 may detect the marker 340 according to the lighting of the LED light source 330 when photographing the direction of the induction apparatus 300, and may generate a detection image including the marker 340. have.

The low-resolution low-cost image sensing member may be used as the image sensor 220. However, at least 10 m away from each other, a sensor capable of implementing the object shape relatively clearly may be used.

Here, the aforementioned separation distance may be changed according to conditions such as the performance and size of the image sensor 220 and the size of the marker 340.

The marker extractor 230 and the target detector 240 may calculate the position information of the unmanned moving object by analyzing the photographing result of the image sensor 220 according to the distance between the unmanned moving object and the docking station.

First, the marker extractor 230 may determine whether the marker 340 in the detection image generated by the image sensor 220 appears. In the present invention, if the detection apparatus 200 enters within 10 m of the induction apparatus 300, an area corresponding to the marker will necessarily exist even if noise is taken into account.

Accordingly, when the marker extractor 230 extracts the marker 340 from the detection image, the marker extractor 230 may be regarded as located in the near field, and the distance from the docking station through image analysis of the marker 340. And the angle can be calculated.

To this end, the marker extractor 230 uses the four-point vertex coordinates included in the marker appearing in the detection image and the size of the preset actual marker 340 to perform three-dimensional coordinates of the marker through a homography operation. Obtaining the distance and the angle corresponding to the three-dimensional coordinates.

Here, the marker 340 is assumed to be installed in the same direction as the front of the docking station and its output terminal (21 in Figure 1). In addition, the marker 340 may be a figure in which a specific mark is surrounded by a white border, and a shape in which the edge of the marker forms a closed curve in a rectangular shape by a binarization process may be recognized. The marker extractor 230 may recognize the marker 340 by binarizing the detected image by a built-in binarizer and detecting a region corresponding to the figure.

In addition, the marker extractor 230 extracts the coordinates on the two-dimensional plane as the marker has a quadrangular shape having four vertices, and projects the coordinates on the three-dimensional plane through a homography operation, thereby the three-dimensional coordinates of the markers. Can be extracted.

Here, the three-dimensional coordinates of the marker is proportional to the distance from the detection device 200 to the actual marker 340 of the induction device 300 and the distance of the actual marker 340 viewed from the detection device 200. The direction and location of the detection apparatus 200 may be specified on three-dimensional coordinates, and first position information may be generated. Using this, the marker extractor 230 calculates the distance and angle with the induction apparatus 300 to obtain a movement route.

The target detector 240 may be driven when the marker 340 is not extracted by the marker extractor 230 as the detector 200 is out of the guide device 300 by more than 10 m. have.

In this case, even if there is an area corresponding to the marker 340 in the detection image by the image sensor 220, it is difficult to distinguish from the noise. In order to overcome this limitation, the controller 250 of the present invention can control the infrared transmitter 210 to transmit an LED control signal for driving the LED light source 330 in a specific lighting pattern. According to the LED control signal, the LED light source 330 is driven with light having a specific lighting pattern-hereinafter, referred to as an 'induction light emitting signal'.

Preferably, the induced light emission signal may be a lighting pattern in which the LED light source 330 repeats lighting and blinking at one second intervals.

The shape of the marker 340 is not clearly displayed in the detection image detected by the image sensor 220 according to the induced light emission signal, but at least includes an area that repeats the appearance and exit of the detection image. It can be used to calculate the distance and angle of the induction device.

In detail, the target detector 240 may extract a region in which the induced light emission signal is detected in the detection image by photographing the surrounding area for a predetermined period of time, preferably 5 seconds, through the image sensor 220.

In this case, when the area estimated as the guided emission signal is not extracted by the target detector 240, the controller 250 may control the moving unit (11 of FIG. 1) to change the direction of the unmanned moving object at an angle. .

When the image sensor 220 acquires a plurality of detection images (hereinafter, referred to as 'first images') of the surroundings during the output period of the induced light emission signal, one or more images of the first image may be applied by applying a difference image processing technique. The area where the change has occurred-hereinafter referred to as 'change area'-and the number of change for each change area are detected, and the registration procedure of registering one or more change areas whose change number exceeds the threshold as a 'target area' is performed. Can be.

In this case, when there is only one registered target area, the target detector 240 may generate second position information by calculating a distance and an angle with the induction apparatus corresponding to the size and the center coordinate of the target area, respectively. . The target detector 240 obtains the movement path by calculating the distance and the angle with the induction apparatus 300 using the target detector 240.

In addition, when the target detection unit 240 has two or more registered target areas, the target weighting unit 240 weights each target area and detects the image captured by the image sensor 220 during the non-output period of the induced light emission signal. When the second image is obtained, one or more change areas in the second image are detected by applying a difference image processing technique.

At this time, the target detector 240 deletes the target region overlapping the detected change region, and repeats the process of detecting the change region using the first image and the second image for a predetermined number of times until the target region becomes one. Can be done.

According to the repetition of the detection process, the weight is increased when the target area that is not deleted is duplicated. When the repetition of the detection process is completed, the weight is increased by using one target area having the highest weight. The movement path is obtained by calculating the distance and the angle.

The angle A between the detection device 200 and the induction device 300 calculated by the target detection unit 240 may satisfy the following Equation 1.

In Equation 1, 'x' is the center abscissa of the image for the target area, 'wh' is 1/2 of the horizontal size of the image for the target area, and 'a' indicates the angle of view of the image sensor.

In addition, the distance L between the detection device 200 and the induction device 300 calculated by the target detection unit 240 may satisfy the following equation (2).

In Equation 2, 'tl' is the diagonal length of the image with respect to the target area, 'Real_M' is the maximum detectable distance, 'Real_m' is the minimum detectable distance, and 'Img_M' is the maximum detectable distance of the image. Diagonal length, 'Img_m' indicates the diagonal length of the image at the minimum detectable distance.

The controller 250 may control the overall driving of the detector 200 and the unmanned moving body on which the detector 200 is mounted.

In particular, the controller 250 controls the infrared transmitter 210 and the image sensor 220 to control the transmission of the LED control signal and the generation of the detection image, and according to the detection image marker extractor 230 or the target detection unit ( By analyzing the detection image through 240 to generate the first or second position information, and thereby obtain the movement path of the unmanned moving object to control the moving unit (11 in Fig. 1), the distance between the unmanned moving object and the docking station According to the algorithm, it is possible to calculate the movement path and control the movement through an appropriate algorithm.

Hereinafter, the structure of the induction apparatus 300 of the object guidance system of the present invention corresponding to the detection apparatus 200 will be described.

The induction apparatus 300 is mounted on the docking station and, under the control of the detection apparatus 200, displays or hides the marker 340 which is a moving direction reference, and moves the unmanned moving object based on the marker 340. Make decisions and move them around.

In order to implement such a function, the induction apparatus 300 receives an LED control signal transmitted from the detection apparatus 200, and outputs a driving signal to the LED light source 330 according to the LED control signal, and an infrared receiver 310. It may include a marker unit 320 including at least one LED light source 330, and a marker 340 is formed on the light emitting surface of the LED light source 330 to form a rectangular border.

The infrared receiver 310 may receive the LED control signal transmitted from the detection apparatus 200, and control the lighting and blinking of the LED light source 330 of the marker 320 in response thereto.

The marker unit 320 outputs an induced light emission signal in the form of an optical signal, and may include an LED light source 330 that emits light and a marker 340 in which a specific figure is formed.

The LED light source 330 may be composed of one or more LED lamps, and may emit light having a predetermined brightness toward the marker 340 attached to the front surface.

The marker 340 may be formed of a substrate formed of a transparent resin material and a figure formed on one surface of the substrate. The figure may have a shape in which a specific mark is surrounded by a white border, and an edge thereof may be displayed in a rectangular shape forming a closed curve in a specific shape. The marker 340 may be displayed to the outside as the light incident from the rear surface by the LED light source 330 is transmitted through the transparent region.

According to the above-described structure, the induction apparatus 300 displays the marker 340 under the control of the detection apparatus 200 so that the detection apparatus 200 calculates the current position information based on the marker 340, and accordingly, A reference is made to approach the guidance device 300.

Hereinafter, an object induction method by an object induction system according to an exemplary embodiment of the present invention will be described with reference to the accompanying drawings.

3 to 5 are diagrams illustrating a method of inducing objects according to an embodiment of the present invention. In the following description, each step execution subject may be the above-described detection apparatus and induction apparatus and their components even if there is no separate description.

First, referring to FIG. 3, according to the object guidance method of the present invention, when an unmanned moving object equipped with a detection device attempts to dock to a docking station equipped with an induction device, the controller of the induction device may control an LED control signal through an infrared transmitter. By controlling the lighting source is controlled (S100).

Accordingly, the infrared receiver of the induction apparatus receives the LED control signal and outputs the induced light emission signal through the marker unit, and the image sensor of the detection apparatus captures the same to generate a detection image. When the detection image is generated, the controller executes the marker extraction unit (S200) and attempts to detect the marker (S300).

When the marker is detected in step S300, the precise distance and angle between the detection device and the marker are calculated (S400), and the search completion (S800) is output to the controller (S800).

In operation S300, when the marker is not detected, the controller executes the target detector (S500) and attempts to detect the induction apparatus (S600). At this time, attempting to detect the guidance device substantially corresponds to attempting to detect the marker.

Next, when the guidance device is detected in step S600, the target detection unit calculates an approximate distance and angle between the detection device and the guidance device (S700), and the controller requests the moving unit to move the unmanned vehicle in the direction of the guidance device (S1100). In step S100, the procedure is resumed.

In addition, if the induction apparatus is not detected in step S600, the control unit turns off the LED light source (S900), and receives the search failure by the target detection unit (S1000). In addition, the control unit requests the moving unit to change the direction of the current unmanned moving object (S1200), and proceeds the procedure again from the step S100.

Meanwhile, when the unmanned moving object is located at a short distance from the docking station by the marker detection unit of the present invention in the above-described step, the step of calculating the precise distance and angle can be subdivided into the following procedure.

In detail, referring to FIG. 4, the step of executing the marker extracting unit (S200) is a step of extracting a marker from the marker extracting unit by the image sensor (S210) and extracting the marker from the detected image (S220). Can be broken down.

Subsequently, when a marker is not detected in the marker detection step S300, a search failure is output (S310). When the marker is detected, the marker extractor obtains four-point coordinates corresponding to each vertex in the detected marker. In operation S320, the two-dimensional marker is projected on the three-dimensional coordinates through a homography operation (S330).

Subsequently, the marker extracting unit calculates the distance and angle with the induction apparatus based on the three-dimensional coordinates of the marker (S340), and outputs the distance and angle (S350). Thereafter, the process proceeds to step S800 described above.

In addition, referring to FIG. 5, the step S500 of executing the target detector may be divided into the following steps S510 to S550.

In detail, when the controller transmits the LED control signal through the image sensor and controls the lighting to output the induced light emission signal (S510), the image sensor of the detection device receives the induced light emission signal to generate a detection image, that is, the first image. One image is generated, and the target detector searches for a change region of the detected image for a predetermined period of time (S520).

Subsequently, the control unit transmits the LED control signal to control the marker unit to be turned off (S530). Accordingly, the target detection unit maintains the change compared to the second image on the change region of the first image by using the detection image for the predetermined time, that is, the second image. The region is deleted (S540). Here, step S540 is a noise removing step. When two or more target areas set by the first image are less than three times in the same region in the second image, the region is not a region corresponding to the marker, but is a simple noise. It is determined to be removed from the set target area.

That is, since the second image is a period in which the marker is not displayed and the first image is an image in which the marker for 5 seconds is displayed in a flashing cycle of 1 second, the second image is changed at least three times. This is because the area below it is obvious that the noise.

Next, weights are assigned to two or more target areas (S550).

Subsequently, as a step (S600) of detecting an induction apparatus, the target detector determines whether a target area exists (S610), and if the target area does not exist, outputs a search failure to the controller (S100). Otherwise, the target detector determines whether there is one currently set target area (S620), and if so, proceeds with the rough distance and angle calculation step S700 and the distance and angle output step S7010. .

In addition, if it is determined in step S620 that two or more target areas exist, it is determined whether the preset number of times is exceeded (S630). When the number of times is exceeded, the maximum weight is determined according to the weights assigned to the plurality of target areas. By selecting a target area having a (S640), and proceeds to the steps S700 and S710 based on this.

In addition, if the specified number of times has not yet exceeded in step S630, if any one of the currently set target areas is redetected, the weight is increased to the target area (S650), and the process is performed again from step S510. The above-described step S650 is a step for removing through the above-described noise removing steps by repetitive search procedure, since all but one of two or more set target areas are noise, and setting only one of them as the target area.

Many details are set forth in the foregoing description but should be construed as illustrative of preferred embodiments rather than to limit the scope of the invention. Therefore, the invention should not be defined by the described embodiments, but should be defined by the claims and their equivalents.

Claims

An object guidance system mounted on an unmanned moving object and a docking station,

An induction device including a marker identified according to the lighting of the light source; And

Detecting the marker to calculate the distance and direction with respect to the induction apparatus at the current position of the unmanned moving object includes a detection device for generating first and second position information with different precision,

The detection device,

According to the movement path included in the first or second location information, the object guidance system for an unmanned moving object requesting the movement of the unmanned moving object.
The method of claim 1,

The induction device,

A marker unit including one or more LED light sources and the markers disposed on a light emitting surface of the LED light source and having a rectangular border; And

An infrared receiver for receiving the LED control signal transmitted from the detection device, and outputs a drive signal to the LED light source according to the LED control signal.

Object guidance system for an unmanned moving object comprising a.
The method of claim 1,

The detection device,

An infrared transmitter for transmitting an LED control signal to the induction apparatus;

An image sensor acquiring a detection image of the induction apparatus;

A marker extractor configured to generate first position information by calculating a distance and an angle with the induction apparatus based on the detected marker when the marker in the detected image is detected;

A target detector configured to generate second position information by calculating a distance and an angle with the induction apparatus based on the induction emission signal detected by the induction emission driving of the induction apparatus when the marker in the detection image is not detected; And

Control unit for requesting the movement of the unmanned moving object according to the first or second location information

Object guidance system for an unmanned moving object comprising a.
The method of claim 3, wherein

The marker extracting unit,

An object for an unmanned moving object that obtains three-dimensional coordinates of a marker through a homography operation using four-point vertex coordinates included in a marker image and a preset actual marker size, and calculates the distance and angle corresponding to the three-dimensional coordinates. Induction system.
The method of claim 3, wherein

The target detection unit,

The LED transmitter transmits the LED control signal for requesting the induction light emission driving through the infrared transmitter, and calculates the distance and angle of the set target area based on a plurality of images continuously photographed for a predetermined period of time through the image sensor. An object guidance system for an unmanned moving object generating second position information.
The method of claim 5,

The target detection unit,

When the image sensor acquires a plurality of first images of the surroundings during the output period of the induced light emission signal, one or more change regions in the plurality of first images and the number of changes for each change region are applied by applying a difference image processing technique. And registering the at least one change area whose threshold number of changes exceeds a threshold as a target area.
The method of claim 6,

The induced light emission signal is an object guidance system for an unmanned moving object, which is a signal that causes the LED light source to repeat lighting and blinking at a cycle of 1 second or less.
The method of claim 7, wherein

The target detection unit,

If there is only one target region, the distance and angle with the induction apparatus are calculated corresponding to the size and the center coordinate of the target region, respectively.

If the target area is more than one, weighting is applied to each target area, and when the image sensor acquires a plurality of second images of the surroundings during the non-output period of the induced light emission signal, the difference image processing method is applied. Detect one or more change regions in the second image,

If the target area overlapping with the detected change area is deleted, the change area detection process using the first image and the second image is repeatedly performed a predetermined number of times until the target area becomes one, and the remaining target area is repeatedly detected. An object guidance system for an unmanned moving object which increases the weight and sets one target area having the highest weight.
The method of claim 8,

The angle A with the induction apparatus is represented by the following equation,

The object guidance system for an unmanned moving object calculated by (where x is the center abscissa of the image for the target area, wh is 1/2 of the horizontal size of the image for the target area, and a is the angle of view of the image sensor).
The method of claim 8,

Distance (L) with the induction apparatus is the following equation,

The object guidance system for an unmanned moving object calculated by (where tl is the diagonal length of the image with respect to the target area, Real_M is the maximum detectable distance, Real_m is the minimum detectable distance, and Img_M is the diagonal length of the image at the maximum detectable distance). , Img_m is the diagonal image length at the minimum detectable distance).
An object induction method by the object induction system for an unmanned moving object according to claim 1,

(a) controlling the lighting of the light source;

(b) detecting the marker by executing the marker extracting unit;

(c) if the marker is detected, calculating a distance and an angle between the detection device and the marker and outputting a search completion;

(d) if the marker is not detected, detecting a guide device by executing a target detector;

(e) if the induction apparatus is detected, calculating a distance and an angle between the detection apparatus and the induction apparatus, requesting to move the unmanned moving object in the direction of the induction apparatus, and performing the step (a);

(f) if the induction apparatus is not detected, turning off the light source and outputting a search failure; And

(g) requesting the unmanned moving object to change the search direction and proceeding to step (a)

Object induction method comprising a.
The method of claim 11,

In step (b),

(b1) obtaining a marker image from the image sensor;

(b2) extracting the marker,

In step (c),

(c1) if no marker is detected, outputting a search failure and proceeding to step (d);

(c2) if the marker is detected, obtaining four-point coordinates from the detected marker;

(c3) acquiring three-dimensional coordinates of the marker using a homography operation using the acquired four-point coordinates and a preset actual marker size;

(c4) calculating a distance and an angle with the marker; And

(c5) outputting the distance and angle

Object induction method comprising a.
The method of claim 11,

In step (d),

(d1) controlling lighting of the light source;

(d2) acquiring a first image from the image sensor for a predetermined period of time, searching for a change region in the first image, and setting a target region;

(d3) controlling the flashing of the light source;

(d4) a second image is acquired from the image sensor for a predetermined period, one or more target regions are set by searching for a change region according to the second image in the first image, and a region continuously changing among the target regions; Deleting;

(d5) assigning a weight to the target area;

(d6) outputting a target search failure if the target area does not exist in steps (d2) and (d4), and proceeding to step (f);

(d7) In steps (d2) and (d4), if there is only one target region, the distance and angle with the induction apparatus are calculated, the calculated distance and angle are output, and the step (e) Proceeding with;

(d8) in step (d7), if the target area is two or more and does not exceed a predetermined number of times, assigning a weight to the current target area, and performing the step (d1); And

(d9) in step (d7), if the target area is two or more and exceeds a specified number of times, selecting one target area having the highest weight among the current target areas and proceeding to step (d7).

Object induction method comprising a.