JP2015141580A - Mobile device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a mobile device which requires a small amount of power consumption during self-position estimation.SOLUTION: A mobile device comprises: a map storage part 101 for storing map information; a distance sensor 102 for creating range data; a camera part 110 for creating image data; first self-position estimation means for estimating a self-position by collating the range data to shape data included in the map information; second self-position estimation means for estimating the self-position by detecting a landmark from the image data, and using the positional information on the image of the detected landmark, and the positional information of the landmark included in the map information; a detection area storage part 105 for storing detection area information where the landmark can be detected with desired estimation accuracy by the second self-position estimation means; and a control part 106 for deciding whether self-position estimation can be performed or not by the second self-position estimation means based on a result collating the self-position estimated by the first self-position estimation means and the detection area information.

Description

  The present invention relates to a mobile device, and more particularly to a mobile device that autonomously moves within a work environment.

  2. Description of the Related Art A mobile device that performs two types of self-position estimation at all times and performs self-position estimation by fusing two estimation results is known (see, for example, Patent Document 1). The moving device described in Patent Document 1 estimates a self-position using both image data captured by a camera unit and range data acquired by a distance sensor. Furthermore, the mobile device described in Patent Document 1 determines the reliability of self-position estimation based on image data and self-position estimation based on range data according to the surrounding environment, and reflects the determined reliability to determine two self-position estimations. The result of means is integrated to determine the final self-position.

JP 2007-322138 A

  The mobile device described in Patent Document 1 always executes two types of self-position estimation, ie self-position estimation using image data captured by a camera unit and self-position estimation using range data acquired by a distance sensor. ing. Therefore, compared to the case where only one type of self-position estimation is performed, there is a problem that the load due to the calculation process increases and the power consumption of a calculation processing unit such as a CPU (Central Processing Unit) increases.

  The present invention has been made to solve such a problem, and an object of the present invention is to provide a mobile device that consumes less power at the time of self-position estimation.

The mobile device of the present invention
A mobile device that moves autonomously in a work environment,
A map storage unit storing map information of the work environment;
A distance sensor that measures the distance to an object present in the work environment and generates range data;
A camera unit for generating image data obtained by imaging an object existing in the work environment;
First self-position estimation means for estimating the self-position of the mobile device by collating the range data with the shape data included in the map information;
By detecting the landmark from the image data and calculating the position of the detected landmark on the image and the position information of the landmark in the work environment included in the map information, the position of the mobile device is determined. Second self-position estimating means for estimating;
A detection area storage unit storing detection area information capable of detecting a landmark with a desired estimation accuracy by the second self-position estimation unit;
Control means for determining whether or not to execute self-position estimation by the second self-position estimation means based on a result of collating the self-position estimated by the first self-position estimation means and the detection area information;
Is provided.

  ADVANTAGE OF THE INVENTION According to this invention, the mobile device with little power consumption at the time of self-position estimation can be provided.

1 is a block diagram illustrating a configuration of an autonomous mobile robot according to a first embodiment. It is a figure which shows a mode that the autonomous mobile robot concerning Embodiment 1 operate | moves. It is a figure explaining the concept of the detection area concerning Embodiment 1. FIG. 3 is a flowchart showing a method for measuring a detection area according to the first embodiment; 4 is a flowchart illustrating processing in which the control unit controls the range data self-position estimation unit and the landmark self-position estimation unit in the autonomous mobile robot according to the first embodiment. In the autonomous mobile robot concerning Embodiment 1, it is a figure which shows notionally the process in which a control part controls a range data self-position estimation part and a landmark self-position estimation part. It is a figure which shows the power consumption of the autonomous mobile robot concerning Embodiment 1. FIG. In the autonomous mobile robot concerning Embodiment 2, it is a figure which shows the detection area corresponding to the landmark in map information. 9 is a flowchart illustrating a process in which a control unit controls a range data self-position estimation unit and a landmark self-position estimation unit in the autonomous mobile robot according to the second embodiment. In the autonomous mobile robot concerning Embodiment 2, it is a figure which shows notionally the process in which a control part controls a range data self-position estimation part and a landmark self-position estimation part. It is a block diagram which shows the structure of the autonomous mobile robot concerning other embodiment.

[Embodiment 1]
Embodiments of the present invention will be described below with reference to the drawings.
FIG. 1 is a diagram showing a configuration of an autonomous mobile robot 100 according to the present embodiment.
The autonomous mobile robot 100 includes a map storage unit 101, a distance sensor 102, a camera unit 110, a range data self-position estimation unit 103, a landmark self-position estimation unit 104, a detection area storage unit 105, and a control unit 106. And comprising.

  FIG. 2 is a diagram illustrating how the autonomous mobile robot 100 operates. The autonomous mobile robot 100 acquires information on the outside world using two sensors, the distance sensor 102 and the camera unit 110. The autonomous mobile robot 100 autonomously moves within the work environment while performing self-position estimation based on each of the image data acquired by the camera unit 110 and the range data acquired by the distance sensor 102.

  The map storage unit 101 stores map information of the work environment. The map information includes shape data of objects existing in the work environment. As map information, a grid map and landmark position and orientation information are stored in the map storage unit 101. The grid map is a map in which the work environment is divided into grid-like areas and each area has attributes such as walls and empty spaces. The landmark position / posture information is a template image indicating how the landmark looks at a certain position in the work environment.

  The distance sensor 102 measures the distance to an object existing in the work environment and generates range data. The distance sensor 102 is preferably capable of measuring a distance as a surface, and for example, a laser range finder can be used.

The camera unit 110 generates image data obtained by imaging an object existing in the work environment.
The camera unit 110 includes two cameras 111a and 111b. Each of the cameras 111 a and 111 b includes an image sensor such as a CCD image sensor or a CMOS image sensor, and outputs digital image data obtained by imaging the outside world to the landmark self-position estimation unit 104. The camera unit 110 is not limited to a stereo camera using two cameras, and may be a monocular camera using one camera.

  The range data self-position estimation unit 103 (first self-position estimation means) estimates the self-position of the autonomous mobile robot 100 by collating the range data with the shape data included in the map information.

  The landmark self-position estimating unit 104 (second self-position estimating means) detects the landmark from the image data, and the position of the landmark in the work environment included in the position information on the image of the detected landmark and the map information. The self-position of the autonomous mobile robot 100 is estimated by calculation using information.

  A landmark is a reference point that is fixed in an environment in which the autonomous mobile robot 100 moves. In addition to natural features such as walls that originally exist in the work environment, landmarks are artificially placed in the work environment. Also good. Information indicating the position of the landmark and the shape of the landmark present in the work environment is stored in advance in the map storage unit 101 as map information. As a specific example, template images showing various appearances of landmarks are stored in the map storage unit 101, and landmarks are detected by correlation detection, pattern matching, etc. between these template images and captured image data. Can be detected.

  The landmark self-position estimation unit 104 calculates the three-dimensional position of the landmark in the camera coordinate system using the landmark position on the image data. Here, the camera coordinate system is a coordinate system fixed to the camera unit 110. Note that the three-dimensional position of the landmark in the camera coordinate system can be calculated by acquiring the distance to the landmark by, for example, stereo viewing of two images taken by the two cameras 111a and 111b.

  The landmark self-position estimation unit 104 uses the position of the landmark calculated in stereo view in the camera coordinate system and the landmark position in a coordinate system fixed to the work environment (hereinafter referred to as the world coordinate system). The robot position in the world coordinate system is calculated. Note that the position of the landmark in the world coordinate system is information stored in advance in the map storage unit 101 as map information.

The detection area storage unit 105 stores detection area information by which the landmark self-position estimation unit 104 can detect a landmark with a desired estimation accuracy.
When the camera unit 110 recognizes a landmark, it is not always detected if the landmark is included in the angle of view of the camera. Due to the influence of lens distortion and the relationship between the number of pixels of the image sensor and the size of the landmark, the range in which the landmark can be detected accurately is a region narrower than the angle of view of the camera. An area where a landmark can be detected with a desired estimation accuracy is measured in advance, and this area is stored in the detection area storage unit 105 as a detection area.

FIG. 3 is a diagram for explaining the concept of the detection area. The detection area is represented with reference to a camera coordinate system fixed to the autonomous mobile robot 100. The detection area is an area in which the landmark 210 can be detected by the landmark self-position estimation unit 104 with desired estimation accuracy.
In FIG. 3, a reference line 220 for measuring the recognition accuracy of the landmark 210 is provided in a lattice shape, and the distance is measured by the camera unit 110 by moving the landmark 210 to various places. At this time, an area in which the landmark 210 can be detected with desired estimation accuracy by the landmark self-position estimation unit 104 is set as a detection area. In FIG. 3, the angle of view of the cameras 111a and 111b is indicated by broken lines 201a and 201b, the angle of view at which the camera unit 110 can be viewed in stereo is indicated by a one-dot chain line 202, and the detection area is indicated by a thick two-dot chain line 203.

  The control unit 106 determines whether or not the landmark self-position estimation unit 104 can execute self-position estimation based on the result of collating the self-position estimated by the range data self-position estimation unit 103 with the detection area information. The control unit 106 controls the range data self-position estimation unit 103 and the landmark self-position estimation unit 104.

FIG. 4 is a flowchart showing a method for creating detection area information.
A measurement environment is prepared so that the accurate positions of two distant points can be measured (ST401). As shown in FIG. 3, as the measurement environment, graph paper with reference lines 220 drawn at equal intervals, an XY stage that understands the amount of movement in a plane, and the like are used.

  Next, the camera unit 110 is installed at a specific point in the measurement environment (ST402). Then, the landmark 210 is set at an arbitrary point in the measurement environment (ST403). After installing the landmark 210, the position of the landmark 210 viewed from the camera unit 110 is measured using the reference line 220 in the measurement environment (ST404).

Next, landmark self-position estimating section 104 calculates the position of landmark 210 in the camera coordinate system by stereo viewing using the image acquired by camera section 110 (ST405).
Next, the difference between the position (measurement position) of the landmark 210 measured using the reference line 220 and the position (calculation position) of the landmark 210 calculated by stereo viewing using the image acquired by the camera unit 110. Is less than or equal to the threshold value (ST406).

When it is determined that the difference between the measurement position and the calculation position is equal to or less than the threshold (YES in ST406), the detection area storage unit 105 stores the measurement position (ST407). If it is determined that the difference between the measured position and the calculated position is greater than the threshold (NO in ST406), the detection area storage unit 105 does not store the measured position, changes the position of the landmark 210 (ST403), and changes the landmark 210. The position is measured again (ST404). Information on the measurement position where the difference between the measurement position and the calculated position is less than or equal to the threshold value is detection area information.
In measurement of the detection area, the angle of the landmark 210 with respect to the camera unit 110 may be changed, and the landmark position may be calculated by stereo viewing multiple times.

FIG. 5 is a flowchart showing processing in which the control unit 106 controls the range data self-position estimation unit 103 and the landmark self-position estimation unit 104. FIG. 6 is a diagram conceptually illustrating a process in which the control unit 106 controls the range data self-position estimation unit 103 and the landmark self-position estimation unit 104.
A method in which the control unit 106 controls the range data self-position estimation unit 103 and the landmark self-position estimation unit 104 will be described with reference to FIGS. 5 and 6.

First, control section 106 acquires the self-position (first self-position estimation value) estimated by range data self-position estimation section 103 and the reliability of the first self-position estimation value (ST501). In FIG. 6, an area with a high probability that the autonomous mobile robot 100 exists, which is calculated from the first self-position estimation value and its reliability, is indicated by a broken-line circle 610.
Next, control unit 106 converts the detection area information stored in detection area storage unit 105 into a first self-position estimate reference (ST502).

  Next, control section 106 expands detection area 601 converted into the first self-position estimation value reference according to the reliability of the first self-position estimation value (ST503). Thereafter, control unit 106 compares the landmark location information estimated by range data self-position estimation unit 103 with detection area 602 after expansion (ST504). Then, control unit 106 determines whether or not landmark 210 is included in detection area 602 after the expansion (ST505).

  When control unit 106 determines that landmark 210 is included in detection area 602 after expansion (YES in ST505), it instructs landmark self-position estimation unit 104 to execute self-position estimation (ST506). Thereby, the self-position estimation value (second self-position estimation value) by the landmark self-position estimation unit 104 is calculated. On the other hand, when it is determined that the landmark 210 is not included in the detection area 602 after expansion (NO in ST505), the control unit 106 instructs the landmark self-position estimation unit 104 to stop self-position estimation (ST507).

  In the autonomous mobile robot 100 according to the present embodiment described above, the other components other than the camera unit 110 and the distance sensor 102 may be configured by a computer system having a CPU (Central Processing Unit) that executes a program. Is possible. Specifically, the range data self-position estimating unit 103, the landmark self-position estimating unit 104, and the control unit 106 are performed in a storage unit such as a ROM or a flash memory provided in the CPU or connected to the outside of the CPU. One or a plurality of programs for causing the computer system to execute the process may be stored, and the programs may be executed by the CPU. The map storage unit 101 and the detection area storage unit 105 may be non-volatile storage units such as a hard disk and a flash memory provided in the computer system.

  The autonomous mobile robot 100 according to the present embodiment and the range data self-position estimating unit 103 and the land data only when the landmark self-position estimating unit 104 is located within a detection area where the landmark 210 can be detected with a desired estimation accuracy. The mark self-position estimation unit 104 performs self-position estimation. If it is located outside the detection area, the self-position estimation by the landmark self-position estimation unit 104 is stopped, and only the self-position estimation by the range data self-position estimation unit 103 is performed.

  With this configuration, the autonomous mobile robot 100 stops the self-position estimation by the landmark self-position estimation unit 104 when the landmark 210 cannot be detected with the desired estimation accuracy. Therefore, according to the present embodiment, it is possible to provide an autonomous mobile robot that consumes less power at the time of self-position estimation.

  FIG. 7 is a diagram illustrating the power consumption of the range data self-position estimation unit 103 and the landmark self-position estimation unit 104 in the autonomous mobile robot 100. In FIG. 7, the power consumption when the self-position estimation by the landmark self-position estimation unit 104 is always performed is indicated by a broken line. On the other hand, the power consumption when the self-position estimation by the landmark self-position estimation unit 104 is performed only within the detection area is represented as a solid line in FIG. From FIG. 7, it can be seen that the power consumption is reduced while the self-position estimation by the landmark self-position estimation unit 104 is stopped.

[Embodiment 2]
An autonomous mobile robot 300 according to the second embodiment will be described.
Similar to the autonomous mobile robot 100 according to the first embodiment, the autonomous mobile robot 300 includes a map storage unit 101, a distance sensor 102, a camera unit 110, a range data self-position estimation unit 103, and a landmark self-position estimation. Unit 104, detection area storage unit 105, and control unit.

  In the autonomous mobile robot 100 according to the first embodiment, the detection area is stored in the detection area storage unit 105 as an area based on the camera coordinate system. The autonomous mobile robot 300 according to the second embodiment is different from the autonomous mobile robot 100 according to the first embodiment in that the map information stored in the map storage unit 101 includes a detection area.

  FIG. 8 is a diagram showing a detection area corresponding to the landmark 210 in the map information. In FIG. 8, the detection area is represented as a region 810 surrounded by a two-dot chain line. When the camera unit 110 of the autonomous mobile robot 300 is located within the detection area (area 810), the landmark self-position estimation unit 104 can detect the landmark 210 with desired estimation accuracy.

FIG. 9 is a flowchart illustrating a process in which the control unit 106 controls the range data self-position estimation unit 103 and the landmark self-position estimation unit 104.
First, control section 106 acquires the self position (first self position estimated value) estimated by range data self position estimating section 103 and the reliability of the first self position estimated value (ST701).

  Next, control section 106 compares the first self-position estimation value considering the reliability with the detection area included in the map information (ST702). FIG. 10 is a diagram illustrating an example of map information including a detection area. Then, control unit 106 determines whether or not the first self-position estimation value considering the reliability is included in the detection area (ST703). The first self-position estimation value considering the reliability is an area where the probability that an autonomous mobile robot is present is high, and is indicated by an area surrounded by a broken line 610 in FIG.

  When it is determined that the first self-position estimation value considering the reliability is included in the detection area (YES in ST703), control section 106 commands landmark self-position estimation section 104 to execute self-position estimation (ST704). Thereby, the self-position estimation value (second self-position estimation value) by the landmark self-position estimation unit 104 is calculated. On the other hand, when it is determined that the first self-position estimation value considering the reliability is included in the detection area (NO in ST703), the control unit 106 instructs the landmark self-position estimation unit 104 to stop the self-position estimation ( ST705)

[Other embodiments]
FIG. 11 is a block diagram showing a configuration of an autonomous mobile robot 900 according to another embodiment.
In the first and second embodiments, the control unit 106 stops the self-position estimation processing by the landmark self-position estimation unit 104 having a high calculation load when the expected estimation accuracy is low. Autonomous mobile robot 900 according to other embodiments is different from autonomous mobile robot 100 according to the first and second embodiments in that control unit 106 can control power on and off of camera unit 110. ing.

  Further, in the autonomous mobile robot 900, the power consumption of the camera unit 110 is also reduced by turning on and off the power of the camera unit 110 in accordance with the execution and stop of the self-position estimation process by the landmark self-position estimation unit 104. can do. Thereby, the power consumption at the time of self-position estimation can be reduced more.

  In Embodiments 1 and 2, the self-position estimation using the range data acquired by the distance sensor 102 is given as the first self-position estimation means, but other self-position estimation means may be used. For example, self-position estimation by wheel odometry may be used.

  Note that the present invention is not limited to the above-described embodiment, and can be changed as appropriate without departing from the spirit of the present invention. For example, the present invention is not limited to an autonomous mobile robot, and information on the outside world is acquired by two sensors, the distance sensor 102 and the camera unit 110, and image data and range data acquired by these two sensors, respectively. The present invention can be applied to a mobile device that autonomously moves in the work environment while performing self-position estimation based on the above.

100, 300, 900 Autonomous mobile robot 101 Map storage unit 102 Distance sensor 103 Range data self-position estimation unit 104 Landmark self-position estimation unit 105 Detection area storage unit 106 Control unit 110 Camera unit 111a, 111b Camera 210 Landmark 220 Reference line 601 Detection area 602 converted to first self-position estimation value reference Detection area 610 after expansion Area 810 where there is a high probability that an autonomous mobile robot exists 810 Detection area included in map information

Claims (1)

A mobile device that moves autonomously in a work environment,
A map storage unit storing map information of the work environment;
A distance sensor that measures the distance to an object present in the work environment and generates range data;
A camera unit for generating image data obtained by imaging an object existing in the work environment;
First self-position estimation means for estimating the self-position of the mobile device by collating the range data with the shape data included in the map information;
By detecting the landmark from the image data and calculating the position of the detected landmark on the image and the position information of the landmark in the work environment included in the map information, the position of the mobile device is determined. Second self-position estimating means for estimating;
A detection area storage unit storing detection area information capable of detecting a landmark with a desired estimation accuracy by the second self-position estimation unit;
Control means for determining whether or not to execute self-position estimation by the second self-position estimation means based on a result of collating the self-position estimated by the first self-position estimation means and the detection area information;
A mobile device comprising:

Images