WO2019194602A1

WO2019194602A1 - Driving control device having improved location correction function, and robot vacuum using same

Info

Publication number: WO2019194602A1
Application number: PCT/KR2019/004003
Authority: WO
Inventors: 홍지혜; 최혁두
Original assignee: 엘지전자 주식회사
Priority date: 2018-04-04
Filing date: 2019-04-04
Publication date: 2019-10-10

Abstract

The present invention relates to a driving control device and a robot vacuum using same. The robot vacuum of the present invention detects current driving location information by checking consistency of a landmark an image of which is captured in a cleaning area, and accurately corrects the driving location information by reflecting a location error due to straight driving and rotational driving, thereby being able to correct the location error which gradually changes due to use environment.

Description

Driving controller with improved position correction function and robot cleaner using the same

The present invention relates to a driving controller and a robot cleaner using the same.

The robot cleaner is an automatic cleaner using a charged battery as a power source to perform cleaning while moving a driving path on the floor according to an input program.

The robot cleaner is designed to perform a cleaning operation while driving a predetermined path according to a built-in program, unlike a conventional cleaning method in which a user performs a cleaning while dragging the cleaner directly. As such, in order to enable the robot cleaner to perform the cleaning operation while automatically driving the predetermined path, various types of sensors are applied to the robot cleaner to detect a driving distance and an obstacle.

In addition, while cleaning along a preset path, if there is an obstacle such as furniture or a wall in the cleaning path or a low land lower than a flat surface such as a stairway, normal driving becomes difficult. In order to detect this, various sensors such as an ultrasonic radiation sensor, an image sensor, and an ultrasonic depth sensor are applied.

In addition to simply detecting obstacles, the sensors of the robot cleaner are used to check the driving route and the position of the robot cleaner. That is, the robot cleaner uses sensors to correct the position by itself when the position recognized by the robot cleaner is separated from the path by colliding with the surrounding objects or sliding while driving.

Korean Laid-Open Patent Publication No. 10-2006-0062607 (June 12, 2006) discloses an optical sensor provided in a robot cleaner main body to recognize a bumper's movement, impact, distortion, and moving distance to recognize the distance and position between obstacles again. Disclosed is a technique to allow. In addition, Korean Unexamined Patent Publication No. 10-2010-0100520 (Sept. 15, 2010) discloses a technique of photographing a divided region of the entire cleaning target to create a map of the entire target region and correcting the position based on the map. .

1 is a view showing a position correction method of the robot cleaner according to the prior art.

Specifically, FIG. 1 is a view provided to explain the position correction technique disclosed in Korean Patent Laid-Open Publication No. 10-2010-0100520 (2010.09.15, published). As shown in FIG. 1, in the related art, the robot cleaner 10 gradually creates a map by matching the feature points Ec of the images photographed in the divided regions of the entire cleaning target. After matching all common feature points Ec of all divided regions, the map of the entire target region was finally completed, and the position information was corrected using the completed map.

However, the robot cleaner 10 according to the related art creates a map and detects a driving position based only on distance information between the feature points Ec photographed in the cleaning space. Therefore, it is difficult to detect a gradually varying position error when the position itself gradually varies due to slip or external force depending on the floor material. And gradually variable error has a problem that can not be corrected by itself.

In addition, the conventional robot cleaner 10 has a difficulty in correcting the position by finding the original feature points again when the position is continuously changed by a floor member such as a carpet, a pet, or the like, or when it is largely changed at one time. As such, if the position of the robot cleaner 10 itself is not reflected and the position is corrected using only the surrounding feature points Ec, the robot cleaner 10 cannot easily cope with the sudden feature point variation.

In addition, the conventional robot cleaner 10 has a problem in that matching accuracy between a feature and a driving route is deteriorated when an obstacle is repeatedly arranged at a specific location at irregular or irregular time intervals. That is, when obstacle detection is frequently caused by anomalously changed surrounding conditions, there is a problem that path management to be avoided increases, making path management difficult, and cleaning efficiency is also lowered.

In addition, since the conventional robot cleaner 10 is set to detect and avoid an object or a foreign object only at the level of a preset reference value range, the object recognition accuracy in the inclined area or the curved groove portion is remarkably degraded. There was. In other words, the object and obstacle detection method applied to the conventional robot cleaner 10 recognizes the foreign matter or the object only based on the background value or the reference value according to the floor plane, and thus the error occurrence rate in specific areas having different heights is increased. There was a growing problem. That is, there is a problem that the reliability is lowered because the error occurrence rate such as recognizing and avoiding specific areas having different heights and heights compared to the entire plane is increased.

In addition, since the conventional object and obstacle detection method uses only one floor planar model, a comparison object for determining a specific object (EI) or a foreign object, etc., as a foreign material or an obstacle may be limited. In fact, since the comparison target is limited only to a reference value, when an object or a foreign object having an ambiguous criterion is detected, there is a problem of simply avoiding or malfunctioning, thereby lowering user satisfaction.

SUMMARY OF THE INVENTION An object of the present invention is to provide a driving controller and a robot cleaner using the same, which accurately corrects driving position information during cleaning by reflecting distance information between characteristic points photographed in a cleaning space and a position change error thereof in real time. .

In addition, the present invention allows to correct its position by accurately reflecting the position change and the external impact amount according to the slip or external force. Accordingly, an object of the present invention is to provide a traveling control device and a robot cleaner using the same that can quickly cope with changes in the external environment.

In addition, the present invention matches and monitors the current driving position with the position information where anomalous operation such as collision or slippage occurs. Accordingly, an object of the present invention is to provide a driving controller and a robot cleaner using the same, which can easily cope with an anomalous situation by changing a traveling speed and a cleaning method at a position where an anomalous operation occurs.

Another object of the present invention is to provide a robot cleaner capable of clearly detecting and distinguishing foreign objects and objects even in an inclined or curved grooved area.

In addition, the present invention divides the sensing space detected by the depth map into a plurality of areas, and enables to accurately analyze and determine the bending and inclination of the floor surface for each of the plurality of divided areas. Accordingly, an object of the present invention is to provide a robot cleaner capable of clearly distinguishing a foreign object and an object, a lowland area, and an highland area according to the floor shape of each divided area.

In addition, when a specific object or a foreign object is recognized for each divided area of the depth map, the present invention may be analyzed by comparing with other divided area analysis information of the surrounding area or by expanding the specific divided area. Accordingly, an object of the present invention is to provide a robot cleaner capable of further increasing the detection accuracy in a divided area in which foreign objects, objects, obstacles, etc. are detected among the plurality of divided areas.

The driving control apparatus and the robot cleaner using the same of the present invention detect the current driving position information by checking the consistency of the landmarks photographed in the cleaning space, and accurately reflect the driving position information by reflecting the position error due to the straight running and the rotating driving. Correct.

In addition, uncertainty position error information that may occur as the robot cleaner moves is detected based on the rotation driving motion and the straight driving motion. Accordingly, the driving position information may be corrected by reflecting the uncertainty position information detected based on the motion in the driving position information. In addition, it is possible to correct collision acceleration information due to collision or slip in real time, and reflect the detected collision acceleration information to the driving position information to correct the driving position information.

Also, based on the degree and position of the correction of the driving position information, anomalous motion occurrence position generated during cleaning is checked and stored in real time. By checking the driving position information corrected in real time, when the cleaner is adjacent to the anomalous motion occurrence position, the driving speed and the cleaning method are changed.

In addition, the robot cleaner of the present invention generates a depth map according to an imaging range through at least one plane distance sensing unit provided with an imaging module. The depth map of the imaging range is divided into a plurality of areas so that the analysis can be performed for each divided area.

In addition, the robot cleaner of the present invention analyzes and determines the degree of plane, non-plane, and inclination using algorithms such as planar equations, non-planar and inclination analysis algorithms, and non-planar height analysis algorithms for each divided area divided into a plurality of areas. . And based on the determined plane, non-plane, slope information, etc., it is possible to detect the non-plane height, obstacles, low ground, high-altitude position information.

In addition, when the robot cleaner of the present invention recognizes a specific object or a foreign object for each divided region of the depth map, the robot cleaner compares and confirms the analysis information of the surrounding divided region with the expanded values using the resolution and depth value expansion algorithm. In addition, the position coordinate information determined as a foreign object, an object, an obstacle, etc. is calculated in real time to control the driving route to be variable or maintained.

The driving control apparatus and the robot cleaner using the same of the present invention may correct the driving position information of the robot cleaner by reflecting the error according to the straight and rotational driving in real time, thereby correcting the position error gradually changed by the use environment.

In addition, the driving control apparatus of the present invention and the robot cleaner using the same may quickly return to the original position even with a large external force or sudden impact by correcting the driving position information by reflecting the uncertainty position information and the collision acceleration information detected on the basis of motion.

In addition, the driving control apparatus of the present invention and the robot cleaner using the same may monitor in real time by matching the position information in which the anomalous operation such as a collision, slip, and the like currently running position. Thus, by changing the running speed and the cleaning method at the position where the anomalous operation occurs, the user's satisfaction and reliability can be improved by easily coping with the anomalous situation.

In addition, the robot cleaner of the present invention divides the depth map into a plurality of regions and compares them with each other. Thus, there is an effect that can improve the object detection function, such as to be able to clearly distinguish between the foreign object and the object in the inclined region or the curved groove region.

In addition, the robot cleaner of the present invention accurately analyzes and determines the bending and inclination of the bottom surface for each divided area. Therefore, it is possible to accurately recognize foreign objects, objects, lowlands and highlands according to the floor shape of each divided area, thereby reducing the error occurrence rate and improving customer reliability.

In addition, the sensing device of the present invention and the robot cleaner using the same may compare or expand the analysis information of the surrounding divided area when a specific object or a foreign object is recognized for each divided area of the depth map. Accordingly, it is possible to accurately determine foreign objects, objects, obstacles, and the like, thereby reducing malfunctions such as simple avoidance, thereby further improving customer satisfaction.

2 is a view showing in detail a robot cleaner equipped with a driving control apparatus according to an embodiment of the present invention.

3 is a block diagram showing in detail the driving control apparatus of the present invention.

4 is a block diagram illustrating in detail the position corrector of FIG. 3.

5 is a view for explaining a driving position detection method of the current position detection unit shown in FIG.

FIG. 6 is a diagram for describing a method of correcting position information in the speed detection correcting unit and the collision acceleration applying unit illustrated in FIG. 4.

FIG. 7 is a block diagram illustrating in detail a robot cleaner equipped with a sensing device according to another exemplary embodiment. Referring to FIG.

FIG. 8 is a diagram illustrating a bottom surface of a robot cleaner configured with at least one planar distance detector illustrated in FIG. 7.

FIG. 9 is a diagram for describing a method of dividing depth map information of an image processor illustrated in FIG. 8.

FIG. 10 is a block diagram illustrating in detail the image analyzer illustrated in FIG. 8.

FIG. 11 is a diagram illustrating a segmentation area comparison and analysis method of the image analyzer illustrated in FIG. 10.

The above objects, features, and advantages will be described in detail with reference to the accompanying drawings, whereby those skilled in the art to which the present invention pertains may easily implement the technical idea of the present invention. In describing the present invention, when it is determined that the detailed description of the known technology related to the present invention may unnecessarily obscure the gist of the present invention, the detailed description will be omitted. Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the drawings, the same reference numerals are used to indicate the same or similar components.

Hereinafter, the driving control apparatus and the robot cleaner using the improved position correction function according to an embodiment of the present invention will be described in detail.

Referring to FIG. 2, the robot cleaner RC is disposed so as to correspond to a case 140 and an outer surface of the case 140 forming an appearance, and includes a distance detecting module 120 including at least one transmitter and a receiver. In this case, the case 140 may be formed in a circular disk shape having a predetermined height.

In addition, the robot cleaner RC is connected to the display unit 160 displaying an operation state, the interface unit 150 to which a control command from a user is input, the display unit 160 and the interface unit 150, and the robot cleaner RC. It includes a travel control device for controlling the travel path of the.

The distance detection module 120 includes a sensor for detecting a distance from the robot cleaner RC to a wall or an obstacle in the room. In the present invention, an example in which at least one ultrasonic sensor is configured in the distance detection module 120 will be described.

The ultrasonic sensor of the distance detection module 120 detects distance information for each of the plurality of split points by transmitting and receiving an infrared or ultrasonic signal to a plurality of split points preset in the sensing space. That is, the distance detection module 120 may calculate the distance information of the cleaning space that is detected by transmitting and receiving the ultrasonic signal.

At least one direction of the case 140 is provided with a bumper 130 that buffers an impact during an external collision. The bumper 130 is configured with a shock sensor for generating a collision acceleration signal corresponding to the external shock. The impact sensor generates collision acceleration information by digitally processing a collision acceleration signal according to an external impact amount, and transmits the collision acceleration information to the driving controller.

In addition, the robot cleaner RC is provided with an image detection module 110 in the upper front portion so as to capture an image of the cleaning target region or the driving region. The image detection module 110 is installed to capture images of the front and the upper side where the robot cleaner moves. The image detection module 110 may be fixedly installed forward and upward, and may be installed to move in left, right, and up and down directions.

The image detection module 110 calculates depth value information of the sensing space using at least one depth sensor. The image detection module 110 may digitally process the depth value information of the sensing space detected by the at least one depth sensor and transmit the digital signal to the driving controller. Here, the depth sensor detects distance information between the depth sensor and the sensing space according to the resolution of the depth sensor itself, and an image sensor or a 3D camera may be used.

For example, a CCD image sensor, a CMOS image sensor, and a 3D camera convert an image formed by light reflected from an imaging object through a lens to the imaging device and converts the image into an electrical signal according to the intensity of light, and outputs it as an analog electric signal. Done. Thus, the image detection module 110 may digitally process the depth value information of the sensing space detected by the at least one depth sensor and sequentially transmit the digital signal to the driving control apparatus.

The spaces sensed by the image detection module 110 may be three-dimensional spaces on a screen having a preset resolution. That is, as a space for detecting distance information from the depth sensor according to the resolution of the depth sensor, the sensing space may be set by the resolution and the detectable range of the depth sensor. In this example, the sensing space may be configured in the form of a rectangular parallelepiped corresponding to the rectangular screen.

As described above, the travel control device is embedded in the robot cleaner RC in association with the display unit 160 and the interface unit 150. The driving control device controls the driving path and the speed in real time according to the preset driving path and the position information detected in real time, and controls the overall cleaning option according to the user setting.

The driving control device corrects its own location information in real time and resets the map of the cleaning space according to the corrected location information when the vehicle moves away from the preset driving path or suddenly changes position due to an external impact or slippery flooring. Then, based on the reset map information and the location information, a smooth cleaning operation is performed.

To this end, the driving control apparatus of the present invention includes a landmark detector 102, a position correction unit 170, and a driving control module 180.

The landmark detection unit 102 processes the depth value information from the image detection module 110 to generate a depth map of the sensing space, splits the depth map, and extracts at least one landmark for each divided area.

In detail, the landmark detector 102 filters depth value information input through the image detection module 110, samples every frame, and converts the AD to generate a depth map for each frame of the sensing space. The depth map of the sensing space may be generated by signal processing depth value information of the sensing space input through the distance detection module 120. Thus, the landmark detector 102 maps the depth map for each frame generated using the image detection module 110 and the depth map for each frame generated using the distance detection module 120 to the same resolution. The data may be divided and transmitted to the position corrector 170. That is, the landmark detector 102 may process and arrange the depth value information of the sensing space in every frame unit according to a preset resolution, and generate the depth map of the sensing space by matching the distance information of the plurality of divided points. have.

The landmark detector 102 compares depth value information of adjacent positions with each other according to a preset resolution, and sets an object or an edge at a position where a difference in depth value is greater than a preset reference as a landmark. In detail, when the depth value information of the sensing space is arranged according to a preset resolution, the landmark detector 102 compares the depth value information of the positions adjacent to each other and compares the depth value difference with a preset reference. In addition, the objects of the locations where the detected difference in depth value is greater than or equal to a predetermined reference may be set as a landmark. In this case, the difference in depth value is large in an object, an edge, or a window frame that protrudes from the wall, and an object, an edge, or a window frame that protrudes from the wall may be set as a landmark.

The position correction unit 170 generates driving position information in real time using distance information with at least one landmark, and corrects driving position information by using rotation and driving error information, uncertainty position error information, and collision acceleration information. .

Specifically, the position correction unit 170 detects the driving position information of the robot cleaner RC by checking the consistency of the landmark photographed in the sensing space, and accurately travels by reflecting the position error due to the straight running and the rotating driving. Correct the position information. To this end, the position correction unit 170 detects the uncertainty position error information that may be generated as the robot cleaner moves based on the rotation driving motion and the straight driving motion. The driving position information is corrected by reflecting the uncertainty position information detected based on the motion to the driving position information. In addition, the position correction unit 170 detects collision acceleration information according to a collision or slip in real time, and corrects the driving position information by reflecting the detected collision acceleration information in the driving position information. The driving position information correction structure and method of the position correction unit 170 will be described in more detail with reference to the accompanying drawings.

The driving control module 180 changes the driving path of the robot cleaner and controls the cleaning option and driving speed by using the depth map of the sensing space reset according to the driving position information corrected by the position corrector 170. At this time, the driving control module 180 checks and stores in real time an anomalous motion occurrence position generated during cleaning based on the degree and position of the driving position information correction. In addition, by checking the driving position information corrected in real time, when the cleaner is adjacent to the anomalous motion occurrence position, the driving speed and the cleaning method may be changed.

4 is a block diagram illustrating in detail the position corrector of FIG. 3.

The position correcting unit 170 illustrated in FIG. 4 includes a landmark matching unit 171, a current position detecting unit 172, a speed detecting correcting unit 173, and a collision acceleration applying unit 174.

In detail, the landmark matching unit 171 registers landmarks of each divided area and calculates driving position information in real time according to distance information of the matched plurality of landmarks. In detail, the landmark matching unit 171 matches and matches landmarks of each divided area, and the current position detection unit 172 may determine the location of each landmark in the sensing space according to the depth value information between the current position and each landmark. The driving position information can be detected.

The current position detector 172 corrects the position information in the stationary state by reflecting the position information and the error information in the stationary state to the driving position information detected in real time.

Referring to FIG. 5, the current position detector 172 reflects the driving position information detected in real time using the position information of the stationary state and the error information. here,

Robot cleaner location information at the time point

Is corrected through Equation 1 below.

[Equation 1]

Vs applied to Equation 1 is the speed of the robot cleaner, vt is the rotation speed of the robot cleaner. And

Is the rotational speed, including the uncertainty error,

Is the speed straight forward including the uncertainty error.

Rotational speed with uncertainty error

Straightness including, and uncertainty errors

Is defined as in Equation 2 below.

[Equation 2]

here,

Rotation speed error

Including,

Is

It includes.

Also,

Is the error of velocity with mean zero and variance k,

Is the rotation speed weight for the rotation speed uncertainty. And

Is the straight velocity weight for the rotational speed uncertainty,

Is the rotational speed weight for the straight-forward uncertainty. Also,

Is the straight velocity weight for the straight velocity uncertainty.

Referring to FIG. 6, the speed detection correcting unit 173 corrects the position information in the moving state by reflecting the rotation and driving error information and the uncertainty position error information in the detected driving position information in real time.

In detail, the speed detection correcting unit 173 includes rotation and driving error information (

,

) And uncertainty location error information (

,

) May be reflected in driving position information detected in real time using Equation 2 above.

In FIG. 6, VB represents a position correction state in a static range in a static state. In addition, the VS arrow indicates a change in position according to the collision and a conversion width of driving position information according to the change.

The collision acceleration application unit 174 corrects the position information in the collision and slip situation by reflecting the collision acceleration information from the bumper to the position information corrected by the speed detection correcting unit 173.

Specifically, the collision acceleration application unit 174 is a rotation and driving error information (

,

) Uncertainty location error information (

,

) The collision acceleration information is further reflected in the detection equation 2, and as shown in Equation 3 below, the driving position information is further corrected according to the collision acceleration information.

[Equation 3]

here,

Is the time duration of the collision time relative to the magnitude of the collision acceleration acceleration (> 0), and is the amount of change in the collision speed in the rotation direction. And

Is the time duration of the collision time compared to the magnitude of the straight collision acceleration (> 0), and is the amount of change in the collision velocity in the straight direction.

The map / position correction unit 175 resets the depth map with respect to the current position according to the driving position information corrected by the speed detection correction unit 173 or the collision acceleration application unit 174, and transmits it to the driving control module 180.

In addition, the map / position correction unit 175 sets the position where the position information is corrected according to the collision acceleration information in the collision acceleration application unit 174 as a position where the anomalous operation occurs, and transmits it to the driving control module 180.

Thus, the driving control module 180 is the driving position information corrected by the speed detection correction unit 173 or the collision acceleration application unit 174 and the position information (CC) where the anomaly operation set by the map / position correction unit 175 has occurred. Can be monitored in real time. In addition, the driving mode of the robot cleaner may be controlled such that the traveling speed and the cleaning method are changed at the position where the anomalous operation occurs.

As described above, the driving control apparatus and the robot cleaner RC using the same may correct the driving position information of the robot cleaner in real time by reflecting an error resulting from the straight and the rotation driving. Accordingly, the positional error that is gradually changed by the influence of the surrounding environment in which the robot cleaner RC is being used may be corrected in real time.

In addition, the driving control apparatus of the present invention and the robot cleaner using the same may quickly return to the original position even with a large external force or sudden impact by correcting the driving position information by reflecting the uncertainty position information and the collision acceleration information detected on the basis of motion. In addition, the driving control apparatus and the robot cleaner using the same of the present invention can monitor in real time by matching the position (CC) information in which anomalous operation such as collision, slip, and the like currently running position. Thus, by changing the running speed and the cleaning method at the position (CC) where the anomalous operation has occurred, it is possible to easily cope with the anomalous situation, thereby improving user satisfaction and reliability.

In the above, reference numerals shown in the description of FIGS. 2 to 6 and detailed descriptions thereof are limited to the descriptions of FIGS. 2 to 6 and detailed descriptions thereof.

Reference numerals shown in FIGS. 7 to 11 and detailed descriptions according to the following embodiments are limited only to descriptions of FIGS. 7 to 11 and detailed descriptions according to the following embodiments.

FIG. 7 is a block diagram illustrating in detail a robot cleaner equipped with a sensing device according to another exemplary embodiment. Referring to FIG. 8 is a block diagram illustrating a bottom surface of the robot cleaner configured with at least one planar distance detector illustrated in FIG. 7.

First, referring to FIG. 7, the sensing apparatus includes at least one planar distance sensor 100, an image processor 200, an image analyzer 300, and an evasion supporter 400.

The at least one plane distance detector 100 illustrated in FIGS. 7 and 8 calculates depth value information of the sensing space using at least one depth sensor. To this end, each planar distance sensing unit 100 may further include at least one imaging module 110 and a transmission / reception module 120.

The imaging module 110 may digitally process the depth value information of the sensing space sensed by the at least one depth sensor and transmit the digital signal to the image processor 200. Here, the depth sensor detects distance information between the depth sensor and the sensing space according to the resolution of the depth sensor itself, and an image sensor or a 3D camera may be used.

For example, a CCD image sensor, a CMOS image sensor, and a 3D camera convert an image formed by light reflected from an imaging object through a lens to the imaging device and converts the image into an electrical signal according to the intensity of light, and outputs it as an analog electric signal. Done. The imaging module 110 may digitally process the depth value information of the sensing space sensed by the at least one depth sensor and sequentially transmit the digital signal to the image processor 200.

The spaces sensed through each imaging module 110 may be three-dimensional spaces on a screen having a preset resolution. That is, as a space for detecting distance information from the depth sensor according to the resolution of the depth sensor, the sensing space may be set by the resolution and the detectable range of the depth sensor. In this example, the sensing space may be configured in the form of a rectangular parallelepiped corresponding to the rectangular screen.

The transmission / reception module 120 detects distance information for each of the plurality of division points by transmitting and receiving an infrared or ultrasonic signal to a plurality of division points preset in the sensing space. That is, unlike the method of detecting depth information of the sensing space through the imaging module 110, the transmission / reception module 120 may transmit and receive infrared or ultrasonic signals to calculate depth information of the sensing space.

The transmission / reception module 120 detects the distance information for each of the plurality of split points, and then digitally processes the distance information for each of the divided points and transmits the digital signal to the image processor 200. To this end, the transmission / reception module 120 may include an infrared transmission / reception module or an ultrasonic transmission / reception module, and the plurality of splitting points may be divided into N × M points according to the location and resolution of transmission / reception point of each transmission / reception module. Where N and M are the same or different natural numbers.

The transmission / reception module 120 may be selectively configured unlike the imaging module 110. However, in the present invention, in order to increase the accuracy of the depth map information, the depth information of the sensing space generated through the imaging module 110 and the depth information of the sensing space generated through the transceiving module 120 are mapped to the same resolution. It can be used.

The image processor 200 generates depth maps of the sensing space by signal processing depth value information of the sensing space input through the imaging module 110. In this case, the image processor 200 filters depth value information input through the imaging module 110, samples every frame, and converts the AD to generate a depth map for each frame of the sensing space. The image processor 200 may directly transmit the depth map for each frame to the image analyzer 300.

In addition, the image processor 200 may generate depth maps of the sensing space by signal processing depth value information of the sensing space input through the transmission / reception module 120. In this case, similarly, depth value information input through the transmission / reception module 120 is filtered, sampled every frame, and then AD converted to generate a depth map for each frame of the sensing space.

The image processor 200 maps and splits the depth map for each frame generated using the imaging module 110 and the depth map for each frame generated using the transmission / reception module 120 at the same resolution. 300 can be sent.

In detail, the image processor 200 may process and arrange the depth value information of the sensing space in every frame unit according to a preset resolution, and generate the depth map of the sensing space by matching the distance information for each of the plurality of split points. . In this case, the image processing unit 200 arranges depth value information of the sensing space according to N × M divided region sizes according to a preset resolution. In addition, the image processor 200 generates a depth map in units of frames by arranging them according to the size of n × m pixel arrays set for each divided region. Where n and m are also the same or different natural numbers.

Subsequently, the image processor 200 maps the depth map generated through the imaging module 110 and the depth map generated through the transceiver module 120 to the same resolution, divides the image by the divided regions, and transmits the image to the image analyzer 300. do.

The image analyzer 300 divides and stores the depth map of the sensing space input through the image processor 200 into a plurality of preset regions. Then, at least one of flat, non-planar, slope, lowland, highlands, foreign objects, objects, and obstacle information is generated using an average value for each divided area.

In detail, the image analyzer 300 reads and stores depth value information of each divided region according to N × M divided region sizes or a plurality of divided points. Depth values corresponding to n x m pixels are averaged for each of the plurality of divided regions, and average value information for each divided region is calculated.

The image analyzer 300 may average the average value information for each divided area in every frame unit, calculate a background value of the sensing space, and determine whether the plane is flat for each divided area. The image analyzer 300 compares the background value of the sensing space with the average value information of each divided area and determines the non-planar height value and the slope of the divided area determined as the non-planar surface. Accordingly, the image analyzer 300 may determine an object, a foreign material, a low land, an high land, and an obstacle in each divided area based on at least one of a background value, a non-planar height value, and a slope of the sensing space. The detailed configuration and operation features of the image analyzer 300 will be described in more detail with reference to the accompanying drawings.

The avoidance support unit 400 receives at least one piece of information of a plane, a non-plane, a slope, a low land, a high land, a foreign object, an object, and obstacle information detected by the image analyzer 300. And generating and outputting position coordinate information corresponding to at least one of plane, non-plane, slope, low land, high ground, foreign object, object, and obstacle information. Accordingly, the robot cleaner of the present invention may change the driving route based on the position coordinate information corresponding to at least one of the plane, non-plane, tilt, low ground, high ground, foreign object, object, and obstacle information input to the path driving controller. have.

Referring to FIG. 9, the image processing unit 200 generates a depth map of a sensing space by signal processing depth value information of the sensing space input through the imaging module 110.

In detail, the image processor 200 filters depth value information input through the imaging module 110, samples every frame, and performs AD conversion. The depth value information of the sensing space is signal processed and arranged in units of frames according to a preset resolution. Although the image processor 200 may directly transmit the depth map for each frame to the image analyzer 300, the depth map of the sensing space may be further matched by further matching the distance information for each of the divided points received from the transmission / reception module 120. You can also create

To this end, the image processor 200 arranges the depth values of the sensing spaces according to the N × M divided region sizes and the n × m pixel array sizes of each divided region according to a preset resolution. Create a depth map with. In this case, the image processor 200 separately processes depth value information of the sensing space input through the transmission / reception module 120 to generate a depth map of the sensing space.

Subsequently, the image processor 200 maps the depth map generated through the imaging module 110 and the depth map generated through the transmission / reception module 120 to the same resolution, divides the image into the divided regions, and transmits the image to the image analyzer 300. do.

Referring to FIG. 10, the image analyzer 300 may include an area divider 310, a plane equation calculator 320, a plane determiner 330, a non-plane height determiner 340, and an object and foreign material information determiner ( 350).

In detail, the region dividing unit 310 reads and stores depth value information of each divided region to correspond to N × M divided region sizes or a plurality of divided points, respectively. In this case, the area dividing unit 310 reads and stores depth values corresponding to n ×× m pixel array sizes so as to correspond to a plurality of division points for each N × M division area. At the same time, depth value information for each partition point included in the N × M partition areas is read and stored.

The plane equation calculator 320 calculates average value information for each divided region by averaging depth values corresponding to n × m pixels for each of the N × M divided regions. Here, the planar equation may be an equation for obtaining an average of depth values included in each of the N × M divided regions. The plane equation calculator 320 transmits the average depth values of the plurality of divided regions to the plane determiner 330.

The plane determination unit 330 calculates the background value of the sensing space by averaging the depth average value information of each divided area in the whole frame unit. In addition, it is determined whether or not each plane is flat for the background value of the sensing space.

The non-planar height determination unit 340 determines the height value and the slope of the divided area determined as the non-plane by the plane determination unit 330. In detail, the non-planar height determining unit 340 compares the background value of the sensing space with the average value information for each divided area. The non-planar height value and the background value of the divided region determined as the non-planar are compared in units of N × M pixels, and the height value and the slope of the divided region determined as the non-planar are detected.

The object and foreign material information determiner 350 compares the depth average value of each divided area with the background value, the non-planar height value, and the slope of the sensing space. Objects, foreign objects, lowlands, highlands, and obstacles in each divided region are determined based on at least one of a depth average value of each divided region, a background value of the sensing space, a non-planar height value, and a slope.

Referring to FIG. 11, first, the non-planar height determination unit 340 compares the background value of the sensing space with the average value information for each divided region. The non-planar height value and the background value of the divided region determined as the non-planar are compared in units of N × M pixels, and the height value and the slope of the divided region determined as the non-planar are detected.

The non-planar height determiner 340 may compare average value information for each divided area, depth values corresponding to n × m pixels included in each divided area, or depth values for n × m divided points. Thus, when the result of comparing the at least one depth value preset with respect to the average value is less than the preset reference distance, it may be determined as a plane, and when it is above the preset value, it may be determined as non-planar.

In addition, the object and foreign material information determiner 350 compares the non-planar height value of each divided area, the slope, and the non-planar height value of at least one other divided area adjacent to each other, and the slope (tilt in arrow direction) value. . In addition, objects EI, foreign objects, lowlands, highlands, and obstacles in each divided region may be determined based on the difference result compared with each other.

At this time, the object and foreign material information determination unit 350 selects at least two or more divided areas of the plurality of divided areas, and sets the average value, the non-planar height value, and the inclination value for the two or more selected divided areas at a preset ratio. You can zoom in. Accordingly, the object and foreign material information determination unit 350 may convert the average value, the non-planar height value, and the slope value of the two or more divided regions that are selectively enlarged into the enlarged non-planar height value and the slope value of the other divided region. Can be compared. According to the result of the enlarged and compared difference, the object and the foreign material information determination unit 350 may accurately determine the object EI, the foreign material, the low land, the high land, and the obstacle in each divided area.

As described above, the sensing device of the present invention and the robot cleaner using the same may compare and analyze the depth map by dividing the depth map into a plurality of regions. Accordingly, the object sensing function may be improved by clearly distinguishing the foreign material from the object EI even in an inclined area or a curved groove area.

In addition, the sensing device of the present invention and the robot cleaner using the same can accurately analyze and determine the curvature and the inclination of the floor for each divided area. Accordingly, the foreign material and the object EI, the lowland and the highland can be accurately recognized according to the floor shape of each divided area, thereby reducing the error occurrence rate and improving the customer's reliability.

In addition, the sensing device of the present invention and the robot cleaner using the same may analyze or expand and analyze the analysis information of the surrounding divided area when a specific object or foreign object is recognized for each divided area of the depth map. Accordingly, it is possible to accurately determine foreign objects, objects (EI), obstacles, etc., thereby reducing malfunctions such as simple avoidance, thereby further improving customer satisfaction.

The present invention described above is capable of various substitutions, modifications, and changes without departing from the spirit of the present invention for those skilled in the art to which the present invention pertains. It is not limited by.

Claims

A landmark detector for signal processing depth value information from an image detection module to generate a depth map of a sensing space, and extracting at least one landmark for each divided region by dividing the depth map;

A position correction unit for generating driving position information in real time using distance information with at least one landmark and correcting driving position information using rotation and driving error information, uncertainty position error information, and collision acceleration information; And

And a driving control module configured to change a driving path of the robot cleaner and control a cleaning option and a driving speed by using the depth map of the sensing space reset according to the corrected driving position information.

Travel control.
The method of claim 1,

The landmark detection unit

Signal processing and arrangement of the depth value information of the sensing space in every frame unit according to a preset resolution,

Generating a depth map of the sensing space by matching the distance information for each of the plurality of split points;

Travel control.
The method of claim 2,

The landmark detection unit

Compare depth value information of adjacent positions with each other according to a preset resolution, and set an object or corner at a position where a difference in depth value is larger than a preset reference as a landmark,

Travel control.
The method of claim 3, wherein

The position correction unit

A landmark matching unit that matches the landmarks of each of the divided regions and calculates driving position information in real time according to distance information with the matched plurality of landmarks;

A current position detection unit correcting the position information in the stationary state by reflecting the position information and the error information in the stationary state to the driving position information detected in real time;

A speed detection correction unit correcting the positional information in the moving state by reflecting the rotational and driving error information and the uncertainty position error information in the detected driving position information in real time; And

It includes a collision acceleration application unit for correcting the position information in the collision and slip situation by reflecting the collision acceleration information from the bumper to the position information corrected by the speed detection correction unit,

Travel control.
The method of claim 4, wherein

The position correction unit

Resets the depth map relative to the current position according to the driving position information corrected by the speed detection correcting unit or the collision acceleration applying unit, and transmits the depth map to the driving control module;

The collision acceleration applying unit further includes a map / position correction unit for setting the position where the position information is corrected according to the collision acceleration information as the position where the anomalous motion occurs and transmitting it to the driving control module.

Travel control.
The method of claim 5,

The driving control module

Monitors by matching in real time the driving position information corrected by the speed detection correcting unit or the collision acceleration applying unit and the position information in which the anomaly motion set in the map / position correcting unit occurs,

Controlling the driving mode of the robot cleaner so that the traveling speed and the cleaning method is changed at the location where the anomaly operation occurs,

Travel control.
The method of claim 4, wherein

The current position detector

By using the position information of the stop state, and the error information is reflected in the driving position information detected in real time,

Robot cleaner location information at the time
Is corrected through Equation 1 below,

[Equation 1]

Where vs is the straight velocity of the robot cleaner, vt is the rotational speed of the robot cleaner,
Is the rotational speed, including the uncertainty error
Is the straight velocity that includes the uncertainty error,

Travel control.
The method of claim 7, wherein

The speed detection correction unit

The rotation and driving error information and the uncertainty position error information is reflected in the driving position information detected in real time using Equation 2 below.

[Equation 2]

Where rotational speed includes uncertainty error
Rotation speed error
Speed, including an uncertainty error
Is
Including;

Is the error of velocity with mean zero and variance k,

Is the rotational speed weight for rotational uncertainty,

Is the straight velocity weight for the rotational speed uncertainty,

Is the rotation speed weight for the straight line uncertainty,

Is the straight velocity weight for straight velocity uncertainty,

Travel control.
A case forming an appearance;

A distance detection module composed of at least one transmitter and receiver;

An image detection module for capturing an image of the detection area including the cleaning area;

An interface unit to which a control command from a user is input;

A driving control device connected to the interface unit to control a driving route;

The driving control device

Receiving depth value information from the image detection module and processing the signal to generate a depth map of the sensing space, and generating and correcting driving position information in real time, thereby changing the driving route, controlling cleaning options and driving speed,

robotic vacuum.
The method of claim 9,

The driving control device

A landmark detector configured to process depth value information from the image detection module to generate a depth map of a sensing space, and divide the depth map to extract at least one landmark for each divided region;

A position correction unit generating driving position information in real time using distance information with the at least one landmark and correcting driving position information using rotation and driving error information, uncertainty position error information, and collision acceleration information; And

And a driving control module configured to change a driving path of the robot cleaner and control a cleaning option and a driving speed by using the depth map of the sensing space reset according to the corrected driving position information.

robotic vacuum.
The method of claim 10,

The landmark detection unit

Signal processing and arrangement of the depth value information of the sensing space in every frame unit according to a preset resolution,

Generating a depth map of the sensing space by matching the distance information for each of the plurality of split points;

robotic vacuum.
The method of claim 11,

The landmark detection unit

Compare depth value information of adjacent positions with each other according to a preset resolution, and set an object or corner at a position where a difference in depth value is larger than a preset reference as a landmark,

robotic vacuum.
The method of claim 12,

The position correction unit

A landmark matching unit that matches the landmarks of each of the divided regions and calculates driving position information in real time according to distance information with the matched plurality of landmarks;

A current position detection unit correcting the position information in the stationary state by reflecting the position information and the error information in the stationary state to the driving position information detected in real time;

A speed detection correction unit correcting the positional information in the moving state by reflecting the rotational and driving error information and the uncertainty position error information in the detected driving position information in real time; And

It includes a collision acceleration application unit for correcting the position information in the collision and slip situation by reflecting the collision acceleration information from the bumper to the position information corrected by the speed detection correction unit,

robotic vacuum.
The method of claim 13,

The position correction unit

Resets the depth map relative to the current position according to the driving position information corrected by the speed detection correcting unit or the collision acceleration applying unit, and transmits the depth map to the driving control module;

The collision acceleration applying unit further includes a map / position correction unit for setting the position where the position information is corrected according to the collision acceleration information as the position where the anomalous motion occurs and transmitting it to the driving control module.

robotic vacuum.
The method of claim 14,

The driving control module

Monitors by matching in real time the driving position information corrected by the speed detection correcting unit or the collision acceleration applying unit and the position information in which the anomaly motion set in the map / position correcting unit occurs,

Controlling the driving mode of the robot cleaner so that the traveling speed and the cleaning method is changed at the location where the anomaly operation occurs,

robotic vacuum.
A sensing device generating a depth map with depth value information of a sensing space to detect at least one of planar, non-planar, tilting, lowland, highlands, foreign objects, objects, and obstacle information; and

And a path driving controller for maintaining or changing a driving path according to the position coordinate information corresponding to at least one of the plane, the non-plane, the slope, the low land, the high land, the foreign object, the object, and the obstacle information from the sensing device.

robotic vacuum.
The method of claim 16,

The sensing device

At least one planar distance detector configured to calculate depth value information of a sensing space using at least one depth sensor;

An image processor configured to signal depth value information of the sensing space to generate a depth map of the sensing space;

An image of dividing the depth map of the sensing space into a plurality of preset areas and generating at least one piece of information of plane, non-plane, slope, low land, highlands, foreign objects, objects, and obstacle information by using an average value for each divided area. An analysis unit;

And a avoidance support unit for generating and outputting position coordinate information corresponding to at least one of the plane, non-plane, slope, low land, high ground, foreign object, object, and obstacle information.

robotic vacuum.