JP2007213293A - Guiding device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a guiding device for guiding a traveling object by simulating the mechanism of the cognitive map of a living thing. <P>SOLUTION: A guiding device for guiding a traveling object moving according to a moving instruction including a traveling direction and a speed is provided with: an image recognizing part 101 for detecting the peripheral terrain characteristics of the traveling object; a movement target selector 15 for accepting the selection of the terrain characteristics being the movement target; an unique movement instruction part 103 for changing a traveling direction and speed according to the lapse of a time, and for outputting the movement instruction when the movement target is not selected; a space/time information generation part 111 for associating the history of the temporal change of the movement instruction with the detection result of the image recognition part 101 detected during movement, and for recording it as cognitive storage information when the movement target is not selected; and a guidance movement instructing part 113 for outputting the traveling object to the movement target, based on the cognitive storage information when the movement target is selected. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a guiding device that guides a moving body to a moving target.

Conventionally, guidance devices are used for navigation of mobile objects such as airplanes, ships, and automobiles. Conventional guidance devices for these applications perform navigation using a coordinate map in which real space is modeled in an orthogonal coordinate system. Specifically, the navigation using the coordinate map is to sequentially specify the coordinates where the moving body is located on the coordinate map and select a route to the coordinates of the movement target on the coordinate map.
Juan Liu, “Temporal coding of spatial knowledge for mobile robot navigation”, ROBOTICS AND MACHINE PERCEPTION, SPIE International Technical Group Newsletter, September 2003, vol.12, No.2, p.3,11

By the way, with the advance of technology, the field of use of robots is expanding, but depending on the field of use, functions from an unprecedented viewpoint may be required. For example, in behaviors for pets and toys, biological behavior is regarded as important. Therefore, robots for pets, toys, etc. are required to have functions that imitate the ability of living things for navigation.
However, navigation using coordinate maps is not the only mechanism that underlies the ability of living organisms, including humans, to determine the path to a moving target in real space. Rather, it is considered that a mechanism based on a cognitive map is important when an organism such as a human being intuitively determines a route to a movement target.

A cognitive map is an empirical grasp of the space structure in the real world by associating surrounding movements perceived continuously by audio-visual and other movements when moving in real space. It is.
This invention pays attention to this point, and it aims at providing the guidance device which guides a moving body by imitating the mechanism of a cognitive map of a living thing.

  In order to achieve the above object, a mobile object according to the present invention is a guidance device that guides a mobile object that moves based on a movement instruction including a traveling direction and a speed, and detects a terrain feature around the mobile object. Detecting means, selection accepting means for accepting selection of a terrain feature as a moving target, and a moving instruction that outputs a moving instruction by changing the traveling direction and speed according to the passage of time when the moving target is not selected A storage means for associating a history of a change in time of the movement instruction with a detection result of the detection means detected during movement and recording it as cognitive memory information when a movement target is not selected, and a movement target And a guidance movement instruction means for outputting a movement instruction for guiding the moving body to the movement target based on the terrain feature detected by the detection means and the recognition memory information. It is characterized in that to obtain.

  With the above configuration, the guidance device according to the present invention accumulates spatial knowledge as cognitive memory information by temporally associating the detection result of the detection unit such as a sensor with movement, and the moving body based on the accumulated cognitive memory information Can be guided to the moving target. Therefore, it is possible to guide a moving body that mimics the mechanism of a cognitive map used by an organism for intuitive route determination.

The best mode for carrying out the present invention will be described below with reference to the drawings.
(Embodiment)
FIG. 1 is a diagram showing a usage pattern of a moving body 1 according to the present invention. The moving body 1 is a toy that moves by driving wheels in a search space in which a plurality of landmarks are arranged on a plane. In the present embodiment, cones B, R, G, and Y that are separately applied to four colors of blue, red, green, and yellow are arranged as landmarks.

  The moving body 1 is provided with a camera 11, a sensor 12, a search instruction button 13, a reset button 14, and a movement target selector 15, while detecting the terrain features around the moving body 1 with the camera 11 and the sensor 12. In response to a user operation on the search instruction button 13, the reset button 14, and the movement target selector 15, the search space is searched and the movement target is moved.

FIG. 2 is a diagram illustrating a configuration for realizing search space search and movement to a movement target in the mobile body 1.
The moving body 1 includes an image recognition unit 101, an obstacle detection unit 102, a specific movement instruction unit 103, an obstacle avoidance instruction unit 104, a time series processing unit 105, a switching unit 106, a driving unit 107, and a direction sensor 108. .

The image recognition unit 101 analyzes the image data from the camera 11 and outputs a detection signal indicating the color of the landmark when the landmark is reflected in the image data.
The obstacle detection unit 102 outputs an obstacle detection signal when contact with the obstacle is detected by the sensor 12.
The unique movement instructing unit 103 outputs a movement instruction to the time-series processing unit 105 when the search instruction button 13 is in an ON state (hereinafter referred to as “haze mode”). In this embodiment, the movement instruction is information for instructing the direction change and the straight movement of the moving body 1, and the specific movement instruction unit 103 randomly determines the traveling direction α and the speed v to change the direction, and Directly go straight ahead.

The obstacle avoidance instructing unit 104 sets a traveling direction α and a speed v so as to avoid an obstacle in accordance with an obstacle detection signal indicating that an obstacle has been detected from the obstacle detecting unit 102. Is output. Specifically, after outputting the direction change instruction that sets the traveling direction α changed by 180 ° from the current course, the straight line instruction that sets the predetermined speed is output for a predetermined time.
The time-series processing unit 105 accumulates recognition memory information that can empirically obtain terrain features around the moving body by moving on a plane, and reproduces a mechanism for recognizing and guiding a space based on the accumulated information It is a block.
As shown in FIG. 3, the time series processing unit 105 includes a space-time information generation unit 111, a space-time information holding unit 112, a guided movement instruction unit 113, and a selection unit 114, and includes space-time information. And navigation based on space-time information.

The space-time information generation unit 111 is a functional block that generates space-time information that is recognition storage information, and is based on a detection signal from the image recognition unit 101 and an input from the specific movement instruction unit 103. The movement instruction history from when the movement is started to when another landmark is detected is recorded in the space-time information holding unit 112 as space-time information.
The space-time information holding unit 112 is specifically a recording device such as an HD, and records the space-time information generated by the space-time information generation unit 111. In the space-time information holding unit 112, when a plurality of pieces of space-time information are recorded, the space-time information holding unit 112 manages them in association with the order generated by the space-time information generation unit 111. Also, the space-time information recorded in the space-time information holding unit 112 is deleted when the reset button 14 is pressed.

  FIG. 4 is a diagram illustrating a data structure of space-time information. The space-time information includes information indicating the start and end landmarks and a movement instruction history. The number of movement instructions included in each space-time information is the same as the number of times that the specific movement instruction unit 103 changes the traveling direction α and the speed v while the moving body moves from the landmark at the start point to the landmark at the end point. It is. Here, as the traveling direction α, an azimuth angle in which the value is increased clockwise with the north as 0 ° is used. Among the movement instructions, an instruction to move forward is recorded in association with the speed v and the time when the constant speed v is continuously input to the space-time information generation unit 111.

When the search instruction button 13 is in the OFF state (hereinafter referred to as guidance mode), the guidance movement instruction unit 113 guides the moving body 1 to the landmark of the movement target selected by the movement target selector 15. , A movement instruction in which the speed v is set is output.
The selection unit 114 outputs a movement instruction output from the specific movement instruction unit 103 to the driving unit 107 in the guide mode and a movement instruction output from the guidance movement instruction unit 113 in the guidance mode.

The switching unit 106 in FIG. 2 includes a timer and a single-pole single-throw switch. When an obstacle detection signal is output from the obstacle detection unit 102, the switching direction 106 and the speed v that are input to the operation unit 107 are set to the timer. During the predetermined time set to T, the time series processing unit 105 is switched to the one output from the obstacle avoidance instruction unit 104.
The driving unit 107 is a functional block that drives a wheel and moves a moving body based on a movement instruction output from the obstacle detection unit 102 or the time-series processing unit 105.

The direction sensor 108 is a magnetic compass and acquires the direction of the path of the moving body 1.
The above is the description of the configuration of the moving body 1.
Next, the operation of the moving body 1 configured as described above will be described. First, the operation of each functional block in the heel mode will be described.
FIG. 5 is a flowchart illustrating a process flow of the specific movement instruction unit 103 in the heel mode.

The specific movement instructing unit 103 repeatedly executes the loop processing of Step S11 to Step S15 during the period when the heel mode is set (Step S11: Yes).
In the process of step S12, the specific movement instructing unit 103 determines the traveling direction α and outputs a movement instruction for instructing a direction change. Subsequently, after determining the duration t (step S13), the speed v is determined and a movement instruction for instructing forward is output (step S14). After that, until the continuation time t determined in step S13 elapses or until the detection signal for detecting any one of the landmarks is output from the image recognition unit 101, the output of the movement instruction for instructing the forward movement is continued and continued. If time t has elapsed or a detection signal is output (step S15: Yes), the processing is repeated from step S11. Through the above process, in the heel mode, the movement instruction instructing the direction change and the forward movement is alternately output from the specific movement instruction unit 103.

FIG. 6 is a flowchart showing a processing flow of the space-time information generation unit 111 in the heel mode.
The space-time information generation unit 111 repeatedly executes the loop processing from step S21 to step S27 during the period when the heel mode is set (step S21: Yes).
First, the space-time information generation unit 111 waits for the output of a detection signal in which any landmark is detected from the image recognition unit 101 in step S22, and outputs the detection signal (step S22: Yes), space-time information in which the landmark indicated by the received detection signal is registered as the start point is generated (step S23).

Thereafter, until the detection signal indicating that the next landmark has been detected is received from the image recognition unit 101, the loop processing of step S24 to step S26 is repeated, and is output from the specific movement instruction unit 103 during that period. The movement instruction is registered in order in the space-time information (steps S24 and S25).
When a detection signal indicating that the next landmark is detected is output (step S26: Yes), the space-time information generation unit 111 registers the detected landmark as an end point of the space-time information, and this space -Time information is recorded on the space-time information holding part 112 (step S27). Through the above processing, the movement instruction history output while the moving body 1 moves between two landmarks is accumulated as space-time information.

Next, the operation of the guidance movement instruction unit 113 in the guidance mode will be described.
FIG. 7 is a flowchart showing a process flow of the guided movement instruction unit 113.
When the guidance mode is set (step S31: Yes), the guidance movement instruction unit 113 determines whether any landmark is currently detected based on the detection signal from the image recognition unit 101 ( Step S32).

If no landmark is currently detected (step S32: No), a movement instruction is output to the specific movement instruction unit 103 (step S33). Thereafter, the processes in steps S31 to S33 are repeatedly executed until some landmark is detected.
If it is determined in step S32 that any one of the landmarks is currently detected (step S32: Yes), a plurality of space-time information stored in the space-time information holding unit 112 is stored. Among them, the one whose start point or end point coincides with the landmark selected by the movement target selector 15 (hereinafter referred to as a movement target) is selected (step S34).

Next, among the space-time information recorded in the space-time information holding unit 112, either the start point or the end point of the space-time information recorded before or after the one selected in step S34 is currently It is determined whether or not there is a match with the detected landmark (step S35, step S36).
If the space-time information in which either the start point or the end point matches the currently detected landmark is recorded before the one selected in step S34 (step S35: Yes), step S37 and step The process of S38 is executed.

  In the process of step S37, the space-time information recorded between the space-time information including the landmark at the current position detected in step S35 and the space-time information including the movement target selected in step S34 is stored. Among them, the one excluding the space-time information corresponding to the section where the movement route is overlapped is selected as a reading target. In the process of step S38, the space-time information to be read after excluding the overlapping sections is read in the order recorded in the space-time information holding unit 112, and the movement instruction registered in each space-time information is recorded. Are output to the operation unit 107 in the order registered in the space-time information. At this time, for the movement instruction for instructing the forward movement, the value of the velocity v recorded in the space-time information is output for the period of time t recorded in the space-time information.

If space-time information in which either the start point or the end point matches the currently detected landmark is recorded after the selection in step S34 (step S36: Yes), step S39 and step S40 are performed. The process is executed.
In the process of step S39, the space-time information recorded between the space-time information including the movement target selected in step S34 and the space-time information including the landmark of the current position detected in step S36 is stored. Among them, the one excluding the space-time information corresponding to the section where the movement route is overlapped is selected as a reading target. In the process of step S40, the space-time information to be read after excluding overlapping sections is read in the reverse order of the order recorded in the space-time information holding unit 112, and is registered in the read space-time information. The movement instruction that has been made is output to the driving unit 107 so as to go back along the movement route.

  To output the movement instruction registered in the space-time information so that the movement instruction is traced back to the movement route, specifically, the movement instruction registered in the space-time information is 2n−1, 2n. The movement instruction is output in the order of the 2nd, 2n-3rd, 2n-2th,..., 3rd, 4th, 1st, 2nd. Here, in the present embodiment, the movement instruction registered in the odd-numbered space-time information instructs to change direction, and the movement history registered in the even-numbered instruction instructs to move forward. is there. At this time, the guided movement instructing unit 113 sets and outputs a value obtained by changing the direction of the traveling direction α registered in the space-time information by 180 ° for the direction instructing the direction change. Then, with respect to the one instructing the forward movement, the value of the velocity v recorded in the space-time information is output as it is for the period of time t recorded in the space-time information.

Through the above processing, a movement instruction in which the traveling direction α and the speed v are set so as to guide the moving body 1 from the currently detected landmark to the landmark selected by the movement target selector 15 is output.
Below, the operation example of the moving body 1 comprised as mentioned above is demonstrated using FIGS.
First, when the heel mode is set, the moving body 1 crawls the search space based on the movement instruction output by the unique movement instructing unit 103 setting a value at random. Fig.8 (a) is a figure which shows an example of the movement locus | trajectory of the mobile body 1 to which the heel mode was set. In the example of FIG. 8A, the moving body 1 receives “STOP” from the “START” position via P 1, P 2, P 3, and P 4 on the basis of a movement instruction randomly output by the specific movement instruction unit 103. Move to the position.

  During this movement, the image recognition unit 101 detects a green (G) landmark at the position P1, a blue (B) landmark at the positions P2 and P3, and a yellow (Y) landmark at the position P4. And outputs a detection signal. As a result, while the moving body 1 moves as shown in FIG. 8A, the space-time information holding unit 112 records the space-time information shown in FIG. 9 in the order of d1 to d3. The space-time information of d1 is space-time information generated while the moving body 1 moves from the position of P1 to the position of P2, and d2 is generated while the moving body 1 moves from P2 to P3. D3 is generated while the moving body 1 moves from P3 to P4. In the movement instruction history included in the space-time information of d1 to d3, a movement instruction for instructing a change in direction at the landmark of the starting point is registered at the head, and the movement instruction for instructing a forward movement. As a result of the movement based on, those that reach the position where the end landmark is detected are registered at the end of the history. In addition, when the moving body 1 comes to the “STOP” position, the space-time information generation unit 111 holds the space-time information in the state of d4 in the middle of generation. In the space-time information of d4, the end point is not registered.

Next, the operation of the moving body 1 when the operation mode is switched from the heel mode to the guidance mode when the “STOP” position in FIG.
When the mobile body 1 is switched to the guidance mode, the mobile body 1 starts moving to the movement target set by the movement target selector 15. However, when no landmarks are detected around the mobile body 1 at the position switched to the guidance mode, the mobile body 1 moves the movement instruction output by the specific movement instruction unit 103 even after switching to the guidance mode. The movement based on the guidance is started from the time when any landmark is detected, and the movement based on the guidance by the guidance movement instruction unit 113 is started from that landmark.

Here, first, an operation when a yellow landmark is designated as a movement target by the movement target selector 15 will be described.
FIG. 8B is a diagram showing an example of a movement locus when a yellow landmark is designated as a movement target in the guidance mode. In the example of FIG. 8B, after switching to the guidance mode, based on a random movement instruction output from the specific movement instruction unit 103, the position switched to the guidance mode is changed to a position P5 in the vicinity of the blue landmark. Moving.

  Thereafter, the guidance movement instruction unit 113 searches the space-time information recorded in the space-time information holding unit 112 at the position P5, and the yellow target which is the movement target is detected from the blue landmark detected at the current position. D3 space-time information, which is a history of movement to the landmark, is detected, and according to the movement instruction registered in the d3 space-time information, as shown in FIG. Are output in this order. As a result, the moving body 1 moves from the position P5 to the vicinity of the yellow landmark.

Next, an operation when a green landmark is designated as a movement target by the movement target selector 15 at the position “STOP” in FIG. 8A will be described.
FIG. 8C is a diagram illustrating an example of a movement locus when a green landmark is designated as a movement target in the guidance mode. First, similarly to the example of FIG. 8B, the moving body 1 changes P5 in the vicinity of the blue landmark from the position switched to the guidance mode based on the random movement instruction output by the specific movement instruction unit 103. Move to position.

  Thereafter, the guidance movement instruction unit 113 searches the space-time information recorded in the space-time information holding unit 112 at the position P5, and the blue color detected at the current position from the green landmark that is the movement target. The space-time information of d1, which is the history of movement to the landmark, is detected. Since the detected space-time information of d1 is information recorded based on the experience of movement from the movement target to the current position, the guided movement instructing unit 113 traces back the movement route of d1 as shown in FIG. The movement instruction is output in the order of 1 to 4 in (b). Here, the movement instruction 1 in FIG. 10B is obtained by reversing the traveling direction of the movement instruction 3 included in d1 by 180 °, and the movement instruction 2 in FIG. Based on the included movement instruction 4, the movement instruction 3 in FIG. 10B is obtained by setting the traveling direction of the movement instruction 1 included in d1 to be 180 ° opposite to that shown in FIG. 10B. The movement instruction No. 4 is based on the movement instruction 2 included in d1. Based on these movement instructions, the moving body 1 moves from the position P5 to the vicinity of the green landmark.

As described above, according to the present embodiment, the ability of a human to convert a spatial-temporal experience in real space into cognitive memory information and store it and to determine a movement route based on the cognitive memory information is simulated. I can do it.
In the present embodiment, if no landmark is detected around the moving object in the guidance mode, a random movement instruction is output to the specific movement instruction unit 103 to detect any landmark. It moves to the landmark that is the starting point of guidance. According to such a configuration, it is possible to move to the moving target even when the moving body after storing the space-time information is set to the guidance mode and placed at an arbitrary position in the search space. There are features.

However, when switching to the guidance mode by operating the search instruction button 13 and the movement target selector 15 in the middle of the movement in the heel mode, it is possible to return to the landmark that passed immediately before by the movement in the heel mode. There is no need for random movement based on the movement instruction output by the movement instruction unit 103, and the movement target can be moved more reliably.
FIG. 8D shows a modified example in which, when a landmark is not detected in the vicinity when the mode is switched to the guidance mode in the middle of the movement in the heel mode, the landmark is passed to the immediately preceding landmark and then moved to the moving target. It is a figure which shows a movement locus | trajectory. Here, it is assumed that a green landmark is set as a movement target.

  In such a modified example, when the landmark is not detected around the moving body when the mode is switched to the guidance mode, the guidance movement instruction unit 113 first displays the space shown in FIG. Based on the d4 space-time information being generated by the time information generation unit 111, the movement instructions 1 and 2 in FIG. The movement instruction 1 in FIG. 10C is obtained by setting the traveling direction of the movement instruction 1 included in d4 to be reversed by 180 °, and the movement instruction 2 in FIG. 10C is a movement included in d4. This is based on the instruction 2. Based on these two movement instructions, the mobile body according to the modification can move to the vicinity of the yellow landmark.

  Thereafter, the guide movement instruction unit 113 searches the space-time information recorded in the space-time information holding unit 112 at a position near the yellow landmark, and the yellow landmark detected at the current position from the blue landmark. The d3 space-time information that is the history of movement to the landmark and the d1 space-time information that is the movement history from the green landmark that is the movement target to the blue landmark are sequentially read.

  The guided movement instruction unit 113 outputs a movement instruction in the order of 3 to 6 in FIG. 10C so as to trace back the movement path of d3 based on the read space-time information. The movement instruction is output in the order of 7 to 10 in FIG. Based on these movement instructions, the moving body 1 moves from the vicinity of the yellow landmark to the vicinity of the green landmark.

Here, since the end point of d1 and the start point of d3 are both blue landmarks, the space-time information recorded between d1 and d3, that is, the space-time information of d2, is a section in which movement paths overlap. Is excluded from the reading target by the guided movement instruction unit 113.
(Other variations)
Although the present invention has been described based on the above embodiment, it is needless to say that the present invention is not limited to the above embodiment. The following cases are also included in the present invention.

(1) The present invention may be a content recording method disclosed by the processing procedure of the flowchart described in each embodiment. Further, it may be a computer program including a program code that causes a computer to operate according to the processing procedure, or may be a digital signal composed of the computer program.
The present invention also provides a computer-readable recording medium such as a flexible disk, hard disk, CD-ROM, MO, DVD, DVD-ROM, DVD-RAM, BD (Blu-ray Disc). ), Recorded in a semiconductor memory or the like.

In the present invention, the computer program or the digital signal may be transmitted via an electric communication line, a wireless or wired communication line, a network represented by the Internet, or the like.
Further, by recording and transferring the computer program or the digital signal on the recording medium, or by transferring the computer program or the digital signal via the network or the like, by another independent computer system It may be carried out.

  (2) The terrain feature recorded in the recognition memory information corresponding to the time change of the movement instruction is not limited to a landmark such as a cone, and any other terrain feature may be used as long as it is detected by the detection means. . For example, in addition to the camera and the image recognition unit 101 as a detection means, a gravity sensor or the like is added to the moving body, the inclination angle of the road surface is regarded as a terrain feature, and the change in the detected inclination angle is made to correspond to the time change of the movement history. By recording, the present invention can be applied.

  (3) The effective range of the detection means can be determined from the size and speed of the moving body. The effective range may be a distance that allows the moving body to brake its movement without collision with the obstacle after the obstacle is detected. However, if the effective range is too wide, detection of landmarks separated from the moving object is also recorded as the recognition memory information, so when guiding the moving object using this, the accuracy of guidance decreases. It will be. Therefore, even if a plurality of sensors with different effective ranges are provided, detection results with a wide effective range are used for obstacle avoidance, and detection results with a narrow effective range are used for accumulating recognition memory information and guiding a moving body. Good. Thereby, the precision of guidance can be raised while avoiding an obstacle appropriately. Here, the effective range indicates the detection distance and direction of the sensor. In the above embodiment, a camera provided in front of the moving body is used for accumulating and guiding recognition memory information, and a sensor for detecting the side surface of the moving body is used for avoiding an obstacle.

(4) In the above embodiment, the movement instruction history is divided and recorded as space-time information for each section in which the landmark is detected. However, the movement instruction history continuously records all actions of the moving object. One record may be used.
(5) In the above embodiment, the azimuth angle is used to specify the traveling direction in the movement instruction. However, it is also possible to use an angle or steering angle that changes the traveling direction from the previous course to specify the traveling direction. It is.

(6) The search for space-time information in the guidance mode may be started from the movement target side or from the current position side. Furthermore, it is also possible to start searching the recorded contents of the space-time information holding unit 112 from both the movement target and the current position, and to guide the mobile body using a route that reaches the same space-time information from both.
(7) In the above embodiment, the guidance of the vehicle that horizontally moves on the plane on which the landmark is arranged has been described. However, the present invention is not limited to the vehicle, and the vehicle moves on the plane by 2-speed walking, 4-speed walking, and the like. It may be applied to a moving body. Furthermore, the present invention can also be applied to guidance of an aircraft or the like that moves in a three-dimensional space based on a movement instruction that specifies horizontal and vertical angles as traveling directions.

  (8) The above embodiment and the above modifications may be combined.

  INDUSTRIAL APPLICABILITY The present invention is useful as a technique for guiding a moving object to a moving target. In particular, the present invention is useful for navigation of pet robots, toy robots, etc. by imitating the mechanism of a biological cognitive map and guiding the moving object. Can be applied.

The figure which shows the usage type of the moving body 1 which concerns on 1st Embodiment. The figure which shows the structure which implement | achieves the search of search space, and the movement to a movement target in the mobile body 1. The figure which shows the detailed structure of the time series processing part 105. The figure which shows the data structure of space-time information DB The flowchart which shows the flow of a process of the specific movement instruction | indication part 103. The flowchart which shows the flow of a process of the space-time information generation part 111. The flowchart which shows the flow of a process of the guidance movement instruction | indication part 113. (A) is a schematic diagram showing an example of a movement trajectory in the heel mode (b) is a schematic diagram showing an example of a movement trajectory when yellow is designated as the movement target in the guidance mode (c) is a movement target in the guidance mode (D) is a schematic diagram showing a movement trajectory according to a modification when green is designated as a movement target in the guidance mode. The figure which shows the space-time information produced | generated as a result of the movement shown to Fig.8 (a) FIG. 6A is a diagram showing a movement instruction output by the guidance movement instruction unit 113 when yellow is designated as the movement target. FIG. 5B is a movement outputted by the guidance movement instruction unit 113 when green is designated as the movement target. FIG. 8C is a diagram showing a movement instruction output by the guided movement instruction unit 113 according to the modification when green is designated as the movement target.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Mobile body 11 Camera 12 Sensor 13 Search instruction button 14 Reset button 15 Movement target selector 101 Image recognition part 102 Obstacle detection part 103 Specific movement instruction part 104 Obstacle avoidance instruction part 105 Time series process part 106 Switching part 107 Driving part 108 Direction sensor 111 Space-time information generation unit 112 Space-time information holding unit 113 Guide movement instruction unit 114 Selection unit

Claims (3)

A guidance device for guiding a moving body that moves based on a movement instruction including a traveling direction and a speed,
Detection means for detecting the terrain features around the moving body;
Selection accepting means for accepting selection of a terrain feature as a movement target;
徘徊 movement instructing means for outputting a movement instruction by changing the traveling direction and speed according to the passage of time when a movement target is not selected;
A storage means for associating a history of a change in time of the movement instruction with a detection result of the detection means detected during movement and recording it as cognitive storage information when a movement target is not selected;
And a guidance movement instruction means for outputting a movement instruction for guiding the moving body to the movement target based on the terrain feature detected by the detection means and the recognition memory information when a movement target is selected. A guidance device characterized by.   The guided movement instruction means recognizes and stores a movement instruction history of a section from a point where a terrain feature matching the terrain feature detected at the current position to a point where the terrain feature as the movement target is detected. The guidance device according to claim 1, wherein the guidance device is obtained from information and outputs a movement instruction indicated in the acquired history in the order of the history.   The guided movement instruction means recognizes and stores a movement instruction history of a section from a point where the terrain feature as the movement target is detected to a point where a terrain feature that matches the terrain feature detected at the current position is detected. The guidance device according to claim 1, wherein the guidance device is obtained from information and outputs a movement instruction indicated by the acquired history and a movement instruction whose traveling direction is opposite to the history.

Images