JP4037866B2 - POSITION ESTIMATION DEVICE FOR MOBILE BODY, POSITION ESTIMATION METHOD, AND POSITION ESTIMATION PROGRAM - Google Patents

Description

  The present invention relates to a position estimation apparatus, a position estimation method, and a position estimation program for a mobile object that estimates the current position of a pedestrian (pedestrian) or the like in the outdoors or indoors, and particularly uses a satellite positioning system such as GPS. The present invention relates to a position estimation apparatus, a position estimation method, and a position estimation program for a moving body that can accurately and reliably estimate the position of the moving body.

Conventionally, techniques for measuring (estimating) the location of a person or vehicle (moving body) or estimating the position of a moving body for estimating the position at which the moving body moves have been proposed. Such conventional techniques include the following.
(1) A satellite positioning system that measures the position using radio waves from artificial satellites (Global
A GPS (Global Positioning System) -based position estimation system, also called Navigation Satellite System (GNSS) and radio navigation. (2) Self-contained navigation that integrates the distance and direction of movement using direction sensors such as speed sensors and gyro sensors. (3) Hybrid navigation adopted in a car navigation system (commonly called car navigation) combining the above (1) and (2).
(4) In the GPS of (1) above, a network-assisted GPS system that enables higher-sensitivity and higher-speed positioning by supplying information necessary for positioning to a terminal using communication means such as a mobile phone. (5) A base station system that performs positioning using a base station ID and terminal-side electric field strength within a PHS (Personal Handyphone System) or mobile phone base station. This base station system includes a single base station system that uses one base station and a multiple base station system that uses a plurality of base stations to improve accuracy.
(6) A hybrid system using GPS and base station information. (7) A method in which a device is installed in a location whose position is known in advance, and when the device carried by the mobile object approaches the installation location, the position is specified by acting with the device installed in this installation location. . (8) A wireless tag system that measures a position by carrying a wireless tag (radio wave tag) in a positioning target mobile body, transmitting a signal to the wireless tag, and receiving a reaction.
(9) Work in a mobile robot that uses an object (landmark) as a mark existing in the work environment as means for estimating the current position, and estimates the current position based on the arrangement of the landmark detected by the sensor Position estimation device in the environment.

However, the conventional position estimation apparatus described above has the following problems. That is, (1), (3), and (4) employ a satellite positioning system that uses radio waves by GPS. For this reason, there is a problem in that it cannot be used effectively as a device in a room or underground mall where radio waves from an artificial satellite do not reach.
In addition, since (2) and (3) employ the self-contained navigation, a speed sensor and an azimuth sensor for calculating the change in movement are necessary. In other words, in the case of a vehicle, the speed can be detected from the rotational speed of the wheel, but in the case of a pedestrian, since it is not possible to provide a sensor for detecting the speed, it is not possible to execute self-contained navigation for a pedestrian. Can not.

Moreover, although (5) and (6) adopt the base station system position by PHS etc., since the precision of a position measurement is about several dozens-several hundred meters, it is enough for the navigation with respect to a pedestrian is not.
In addition, (7) and (8) are systems in which the device is installed at a location where position measurement is desired. Therefore, in order to enable a wide range of position measurement, it is necessary to install the device over a wide range, which increases costs. There is a problem that causes it.
In the case of the position estimation technology using landmarks in the mobile robot field of (9), the sensor type of the robot (camera, ultrasonic sensor, infrared sensor, etc.) is the angle of the column where the robot originally exists in the work environment. In order to perform position estimation based on landmarks, multiple (three or more) landmarks may be detected simultaneously by a sensor at one point. Since it is necessary, there is a problem that the control related to position estimation becomes complicated.

  Therefore, the present invention has been made to solve the above-described problems caused by the prior art, and does not use a radio wave from a satellite for position measurement like a satellite positioning system such as GPS, and speeds like a self-contained navigation. A position estimation device for a moving body that does not require a sensor or a direction sensor and that does not require a special positioning device and can estimate the user's position based on surrounding landmark information input by the user Another object is to provide a position estimation method and a position estimation program.

In order to solve the above-described problems and achieve the object, the present invention provides a position estimation device for a moving body that estimates a current position and a moving position of the moving body, and a plurality of landmarks that the moving body recognizes as landmarks. Input means for inputting feature information indicating at least one of shape, color, size, and design, and positional relationship information as landmark information, and the position of the landmark information and the feature information are stored in advance. Based on the landmark database and the feature information and positional relationship information of the plurality of landmarks input by the input means , referring to the landmark database, and searching for a plurality of landmarks corresponding to the feature information, Land that is a combination of landmarks corresponding to the input positional relationship information from the extracted landmarks A position estimation means for estimating a current position of the moving body by performing matching processing of the spatial conditions of extracting over click set, the guidance information based on the position information of the movable body estimated by the position estimation means Guide information generating means for generating and output means for outputting the guide information are provided.

  According to the present invention, there is no need to install a special device for positioning at a location such as a base station as in the base station system, and no speed sensor or azimuth sensor is required as in self-contained navigation. On the other hand, there is an effect that it can be used as a navigation system for pedestrians that provides location information and route guidance information.

  According to the present invention, since the surrounding landmark information is input to the position estimation device provided in the server by voice using a PHS or a mobile phone, the estimated position information and the route guidance information to the destination are voiced. There is an effect that it can be used for an output road guidance system, a road guidance system for pedestrians who cannot use a speed sensor or a direction sensor, a sightseeing spot search system, a movie location support system, and the like.

The present invention is also a mobile object position estimating method for estimating a current position and a moving position of a mobile object , wherein the computer recognizes the shape, color, and size of a plurality of landmarks that the mobile object recognizes as landmarks. , Referring to a landmark information input step of inputting feature information indicating at least one of the designs and positional relationship information as landmark information, and a landmark database preliminarily storing the position and feature information of the landmark information Then, based on the inputted feature information and positional relationship information of the plurality of landmarks, a plurality of landmarks corresponding to the feature information are searched, and the extracted positional information is converted into the inputted positional relationship information. Perform a spatial condition matching process to extract a landmark set that is a combination of the corresponding landmarks. A position estimating step for estimating a current position of the moving body; a guide information generating step for generating guide information based on the position information of the moving body estimated by the position estimating step; and And a guide information output step for outputting the information .

According to the present invention, for a plurality of landmarks that a user (pedestrian) recognizes as landmarks, feature information indicating at least one of shape, color, size, and design, and positional relationship information are used as landmark information. As a landmark information input process, and in a position estimation process, with reference to a landmark database in which the position and feature information of the landmark information are stored in advance, the feature information and positional relationship information of the plurality of landmarks input On the basis of this, a plurality of landmarks corresponding to the feature information are searched, and a spatial condition matching for extracting a landmark set which is a combination of landmarks corresponding to the input positional relationship information is extracted from the extracted landmarks. estimating the current position of the moving body by performing processing, which is estimated by the guide information generating step position Since the guidance information is generated based on the information and the guidance information generated by the guidance information output process is output, the radio wave from the satellite is not used for position measurement like the satellite positioning system. There is an effect that can be used effectively. In addition, the present invention can be realized as a road guidance system that provides road guidance information to users such as pedestrians in places where GPS signals cannot be received such as in underground shopping centers and buildings.

The present invention also relates to a position estimation program for a moving body that estimates a current position and a moving position of the moving body. The plurality of landmarks that the moving body recognizes as landmarks are identified by the computer as shapes, colors, and sizes. , Referring to landmark information input procedure for inputting feature information indicating at least one of the designs and positional relationship information as landmark information, and a landmark database preliminarily storing the position and feature information of the landmark information Then, based on the inputted feature information and positional relationship information of the plurality of landmarks, a plurality of landmarks corresponding to the feature information are searched, and the extracted positional information is converted into the inputted positional relationship information. Performs a spatial condition matching process to extract the landmark set that is a combination of the corresponding landmarks. Thus, a position estimation procedure for estimating the current position of the mobile body, a guide information generation procedure for generating guide information based on the position information estimated by the position estimation procedure, and the guide information for the mobile body A guide information output procedure to be output is executed .

According to the present invention, feature information indicating at least one of the shape, color, size, and design of a plurality of landmarks recognized by a pedestrian (moving body) as a landmark , and positional relationship information are used as landmark information. Because the position of the moving body is estimated using the feature information and positional relationship information of multiple landmarks that are input, because radio waves from satellites are not used for position measurement unlike satellite positioning systems. Even in places where radio waves cannot be received, such as underground malls and buildings, accurate and reliable position estimation information can be provided to pedestrians (users).

Exemplary embodiments of a position estimation apparatus, a position estimation method, and a position estimation program according to the present invention will be described below in detail with reference to the accompanying drawings. In the first embodiment described below, the mobile position estimation device according to the present invention is incorporated in a mobile phone “PHS (Personal Handyphone System)” 80, an information communication terminal device “PDA (Personal Digital Assistants)” 90, or the like. In the second embodiment, the mobile body position estimation device according to the present invention is divided into an information communication terminal device and a center server, and is combined with each other via a network. To do.
Embodiment 1.
First, the position estimation apparatus for a moving body according to the first embodiment will be described. FIG. 1 is a functional block diagram showing a configuration of a mobile object position estimation apparatus according to the first embodiment. FIG. 2 is a flowchart showing the processing procedure of the moving object position estimation method according to the first embodiment.

  That is, as shown in FIG. 1, the position estimation device 1 stores an input means 10 for inputting a landmark recognized as a landmark by a moving body (pedestrian) as landmark information, and stores the position and attribute of the landmark in advance. And the position estimation means 30 for estimating the position of the moving object based on the landmark information input by the input means 10 and the landmark information stored in the landmark database 20, and the position estimation. Guide information generating means 40 for generating guide information based on the position information estimated by the means 30 and output means 50 for outputting the generated guide information. The control unit 35 has a function as a controller that performs overall control of the position estimation apparatus 1. Reference numeral 60 denotes input / output control means for controlling input / output by the input means 10 and the output means 50. Reference numeral 70 denotes a communication I / F having a function as a communication interface.

As shown in FIG. 2, the position estimation method of the moving body is roughly divided into (1) landmark information input step (step S210), (2) position estimation step (step S211), and (3) guidance information generation step. (Step S212) (4) Guide information output step (Step S213).
That is, a landmark recognized as a landmark by the pedestrian (moving body) is input as landmark information by the landmark information input step, and the pedestrian's position is detected by using the input landmark information and the landmark database by the position estimation step. The location is estimated, guide information is generated based on the estimated position in the guide information generation step, and the guide information generated in the guide information output step is output to the pedestrian. Specifically, guidance information is displayed on the display screen 91 (FIG. 6) of the PDA 90.

  Here, first, what is the landmark used in the present invention is defined. In general, a landmark means “a thing that is visually noticeable in a certain area, such as a mountain or a high-rise building, or a landmark for finding a destination”. In the present invention, the above meaning is expanded and defined as “a person can use to specify (describe) a place or a position and can be seen in a certain place”.

Specifically, the following is called a landmark.
(1) Building: For example, building, residence, tower, chimney, castle, station, fence.
(2) Facilities: Stores, parks, shrines and temples, police boxes, public toilets, amusement parks, vacant lots, etc. (3) Terrain: For example, sea, river, mountain, cliff, slope.
(4) Transportation facilities: intersections, traffic lights, pedestrian crossings, pedestrian bridges, pedestrian crossings, signs, railroad crossings, railroad tracks, elevated bridges, etc. Others include signboards, images, landmark plants, postboxes, telephone boxes, and the like.

  FIG. 3 shows an arrangement example of landmarks specified as landmarks in the field. In FIG. 3, buildings (white buildings, round buildings, etc.), florists, bakers, intersections, parks, parking lots, houses, police stations, etc. are illustrated as landmarks. In the figure, A to D represent estimated positions of pedestrians (users).

  In FIG. 3, landmarks are given as examples in the outdoors, but can be used to specify the position of a person (moving body) in an underground mall or building (for example, stairs, Stores, doors, elevators, entrances, etc.) can be used as landmarks. FIG. 4 shows an example of an underground shopping street instead of the outdoors, and stores (P, Q, R,...), Stairs, entrances, etc. can be specified and used as landmarks. In the figure, X and Y represent estimated positions of pedestrians (users).

  Hereinafter, each component in the functional block diagram of the position estimation apparatus will be described with reference to FIG. The input means 10 has an input function for inputting landmark information by a moving body (hereinafter, the moving body is referred to as “user”). For the input of the landmark information, an input method used in a normal information device such as voice input, button input, keyboard input, pen input, touch panel input, and image input can be used.

  Also, the landmark information input by the user that is input by the input means 10 is information regarding a certain landmark that is visible to the user, and is information having the following items (1) to (4). is there. However, it is not necessary to input all items for one landmark, and only items that can be input are input.

(1) Landmark classification (category) is an item, such as a building, a store, a house, a police box, a station, a bridge, a pedestrian bridge, a telephone box, a mail box, a traffic light, a mountain, a river, and the like. This classification can also be configured to have a hierarchical structure such as a large classification, a medium classification, and a small classification. In this case, the classification can be simplified, so that convenience can be improved.
(2) An item with a unique name of a landmark, such as Tokyo Tower, Makuhari Laboratory Building, Makuhari Station intersection, etc.
(3) Features of landmarks are items, and features such as the color, shape, and size of the landmarks can be used as items. For example, landmarks can be white in color, elongated / round in shape, large in size, red in the roof, many windows, and the like.
(4) The positional relationship is an item, and this positional relationship is the positional information of landmarks viewed from the user and the arrangement information between the landmarks. The landmark position information viewed from the user is near / far / next / front / opposite / opposite / distance information such as 10 m with respect to the landmark, and azimuth information such that the landmark is north. .

  Here, the arrangement information between landmarks is information related to the arrangement between landmarks, for example, a certain landmark is next to another landmark, or on the opposite side of the road. In addition, as other items, for example, a prefecture name / city name / area name or a shopping street name / underground street name can be input even when the user does not know the current position accurately. Information. These are not information of individual landmarks, but can be regarded as common information of surrounding landmarks.

  Further, as the input format of the landmark information, there are the following (1) to (3). (1) The input format is a free text format, which includes voice input and text input. Voice input is an input method using spoken language. For example, “a large white building is in front of you, north and there is a florist next to it” can be input by voice. In the case of text input, a free sentence can be input as text using a keyboard or a pen.

  FIG. 5 shows an example in which landmark information is input by voice using the PHS 80 (free text format). In this example, it is shown that a voice is inputted by a person, “There is a white building in front of you, and there is a florist next to it. There is a post box in front of the florist.”

  FIG. 6 shows an example of a screen in which landmark information is input by text input (free text format) using the PDA 90. This example shows that a text entry such as "There is a large white building in front of you and a florist next to it. There is a postbox in front of the florist." . Here, in the following description, 91 is a display screen of the PDA 90, 92 is an OK key for setting input, and 93 is a cancel key for canceling the setting.

  (2) This is an input format in which an answer is freely input for each item with respect to a landmark by an input method using an item-specific input format. Here, FIG. 7 shows an item-specific input screen displayed on the display screen 91 of the PDA 90, and this item-specific input screen includes five fields of category 1, category 2, name, feature, and positional relationship. I have. This item-specific input screen is a screen for inputting landmark information that can be seen around, and information corresponding to each item can be input using keyboard input, pen input, or the like.

  (3) An input method in which an answer to the above item is selected from a candidate list by an input method using a candidate selection input format. FIG. 8 shows an item-by-item screen having a pull-down setting function (candidate selection input format), and classification 2 input can be selected from items in the pull-down menu. In the pull-down menu, there are items such as a convenience store, a gas station, a flower shop, a bookstore, etc., and input is made by selecting from these items. In this example, a bookstore is set as classification 2.

  Further, when the user can use prior information on the current position of the user, the input unit 10 can display a list of nearby landmarks using the prior information and select an item from the list. . FIG. 9 is a display example in the case where the prior information regarding the position can be used, and the prior information setting screen is displayed. Here, when the prior information on the user's current location can be used, when the location is estimated in advance and the previous location is known, the user can obtain rough information on the current location based on the area name, etc. Is given. In this example, a bakery with a landmark nearby is set on the opposite side of the road.

  The landmark data base 20 is a database that stores landmark information that can be recognized as landmarks in advance. FIG. 10 is a configuration example of the landmark database 20. In the landmark data shown in FIG. 10, there are nine items such as ID, name, classification (1, 2), address, longitude / latitude, arrangement, features, etc. as items of management data contents.

  The data items of the landmark database 20 are as follows. That is, (1) ID is a number / symbol that can identify landmark data, and indicates, for example, “1234”. (2) The name is a unique name of the landmark. For example, the proper names are Tokyo Tower, Makuhari Laboratory Building, Makuhari Station Crossing, etc. (3) Classification is a landmark type (category). For example, a building, a store, a house, a police box, a station, a bridge, a footbridge, a railroad crossing, a telephone box, a post box, an intersection, a traffic light, a mountain, a river, and the like. Here, as a classification method, for example, when classification is performed by category so as to have a hierarchical structure such as classification 1 and classification 2, or large classification, medium classification, and small classification, versatility as data is improved. be able to.

  (4) The address is a landmark address. For example, Chiba Prefecture Chiba ○ △. (5) The longitude and latitude are the longitude and latitude of the landmark location. For example, east longitude 139.1234, north latitude 38.1234, and the like. (6) Arrangement is an arrangement (positional relationship) between landmarks. For example, landmark A is 10 meters northeast of landmark B. Details of this arrangement will be described later. (7) A feature is a feature representing the color, shape, size, etc. of a landmark. For example, it shows white, large, many windows, and paintings on the wall.

  Here, among the data items described above, (6) arrangement data will be described in detail. FIG. 11 and FIG. 12 are networked by connecting the landmarks in the example shown in FIG. 3 and FIG. 4 with the landmarks as nodes and connecting nearby landmarks with links. Is. The arrangement data is data related to the link of the landmark network, and the content includes the distance between the landmarks connected by the link, the direction, and the state between the landmarks (information on what exists between them. For example, Road presence / absence, road type, etc.).

  For example, in FIG. 11, “white building (id = 1) is 5 meters west of florist (id = 2) and there is no road between landmarks”, “white building (id = 1) is to the north of the white building (id = 3), the distance is 20 meters, and there is a road (National Road No. △) between the landmarks ”. In FIG. 11, solid lines and broken lines are shown in the lines of links between landmarks. The solid line indicates that there is no road between the landmarks (the landmarks are adjacent to each other), and the broken line indicates that there is a road between the landmarks and the landmarks face each other across the road. It shows that it is.

  In FIGS. 3 and 11, the intersection is also a landmark, and in FIG. 11, a link is established with a landmark near the intersection. In this link (arrangement data), data indicating where the landmark linked to the intersection is located is also described. For example, in FIG. 11, it is memorized that the florist (id = 2) exists in the “north-west corner” of the intersection 1 on the link between the florist (id = 2) and the intersection 1. )

  Furthermore, a landmark three-dimensional model (occupation space data in a three-dimensional space) can be stored in the landmark database 20. By using this 3D model to reconstruct the townscape and so on, and by virtually placing the user's viewpoint in the reconstructed townscape, a certain landmark can be seen from the user at a certain position. , It can be determined that it is no longer visible in the shadow of other landmarks. As described above, when the landmark three-dimensional model is employed, the accuracy of position estimation for the pedestrian based on the landmark information can be improved.

  The position estimation unit 30 has a processing function for estimating the position of the user based on the landmark information input by the input unit 10 and the landmark information stored in the landmark database 20. The position estimation process of the moving body by the position estimation means 30 includes (1) a landmark search range limiting process based on prior information on the position of the user (moving body), and (2) a land based on the inputted landmark attribute. The process is divided into a search process of the mark database, (3) a matching process of the spatial condition of the landmark of the search result, and (4) a determination process of the user position.

  Details of these processes (1) to (4) will be described below with reference to the flowchart of FIG. FIG. 13 is a flowchart showing a procedure for position estimation processing of a moving object by the position estimation apparatus according to the present invention. First, information (landmark information) related to landmarks present in the surroundings recognized by the user is input (step S310), and then a limiting process is performed on the search range of the landmark database 20 based on prior information on the user position. (Step S311), the landmark database 20 is searched by the inputted landmark attribute (Step S312).

  Then, a spatial condition matching process is performed on the search result landmark (step S313), and then a user position determination process is performed (step S314). The steps from (Step S311) to (Step S314) are the position estimation process. Then, guidance information generation processing is performed (step S315), and finally the generated guidance information is output (step S316). Details of each process in the above-described flowchart will be described below.

(1) Landmark Database Search Range Limiting Process The landmark database search range limiting process is based on prior user location information, and is position information processing for a user that can be used prior to the position estimation process. Here, the user location advance information can be used when the result of the previously performed location estimation process is held. For example, if the position estimation process is performed 10 minutes ago and the result is held, the search range can be limited to within the area centered on the estimated position of the previous result. The size of this area can be adjusted according to the user's moving means and the like. It can also be used when the user inputs rough position information.

  In other words, even when the user does not know the current location accurately, in many cases, the location information such as the name of the prefecture or the name of the municipality, or the location information such as the name of the shopping district or the name of the underground street is input in advance. be able to. This prior position information can also be set in the input means 10 as one of the input items. Here, when the position estimation device is incorporated in the PHS 80 or the PDA 90 as in the first embodiment, the base station information extracted by the PHS 80 or the PDA 90 can be used as the prior information of the user position. . In the case of an underground shopping center, landmark information can be received using an infrastructure hotspot for Internet connection.

(2) Landmark Database Search Processing Using Landmark Attributes Landmark database search processing using landmark attributes is a landmark processing method using landmark attribute information existing around the user input by the input means 10 as a search key. This is a process of searching for landmarks stored in the database 20. Here, the landmark attribute is data other than the arrangement data among the data items stored in the landmark database 20, and includes a name, classification, address, longitude and latitude, characteristics, and the like. In this landmark search process, the landmark database 20 is searched using the inputted landmark attributes such as name, classification, address, longitude and latitude, and features as keys. Further, when attributes are input for a plurality of landmarks, a search is performed for each landmark.

  For example, when the landmark attribute as shown in FIG. 6 is input, the search process in the landmark database 20 is performed according to the search conditions as shown in the following (a), (b), and (c). Is done. That is, (a) the search condition for landmark 1 is (classification = building) And (feature = white) And (feature = large), and (b) the search condition for landmark 2 is (classification = store) And (classification) (= Florist) and (c) the search condition of the landmark 3 is (classification = postal post). Here, when a name is input as the landmark attribute, the landmark can be effectively searched by the search by this name. For this reason, landmark search by name may be executed in preference to search processing using values of other attributes as keys.

  FIG. 14 is a table summarizing search results based on the above search condition examples. According to FIG. 14, the landmark IDs hit under the search conditions for landmarks 1, 2, and 3 are “1a, 1b, 1c,...”, “2a, 2b, 2c,. 3b, 3c, ... ".

(3) Spatial Condition Matching Processing This spatial condition matching processing is performed when the user inputs information on a plurality of landmarks, with respect to the landmark of the search processing result, the spatial condition of the landmark. Is a process for checking whether or not a match occurs. In this matching process, landmarks [(1a, 1b, 1c,...), (2a, 2b, 2c,...), (3a, 3b, 3c,. Whether or not the condition is met is checked. If the condition is not met, it is deleted from the search processing result, and processing is performed so that only the combination of the landmarks remains.

  Here, the combination of landmarks can be expressed as a set such as {1a, 2b, 3b}, {1c, 2a, 3c}. Hereinafter, in the present specification, these combinations are referred to as landmark sets. It should be noted that the landmark set includes those having only one landmark such as {2a}.

(4) User position determination process The user position determination process is performed by performing a search process from the landmark database 20 and a spatial condition matching process for the landmark of the search result, and conforming to the input surrounding landmark information. This is a process for enabling the position of the user to be estimated more accurately after the landmark set candidate to be acquired is acquired. Details of the user position determination process will be described later.

  The guide information generating unit 40 has a function of generating various types of guide information based on the user position estimated by the position estimating unit 30. The guidance information generated by the guidance information generating means 40 includes (1) guidance information for teaching the estimated current position (for example, “You are in the vicinity of □”), (2) The guidance information for teaching the route so that the user can go to the place known to the user who has become, (3) guidance information for telling the way to the destination.

  This guidance information includes a direction to go to the destination, distance information (reach distance information), and the like. That is, the guide information generating means 40 can display the current position (position having a certain area) on the map as one piece of guide information. Specifically, when the current position of the user can be estimated by positive position estimation, the map range including the current position of the user is selected based on the longitude and latitude of the position, and displayed on the terminal. The current position of the user is displayed on the map. The same applies when downloading map data from a server using the communication function.

  Here, in many cases, since the estimated user position is an area, the position display of the user on the map is performed by displaying the user estimated position area on the map. Further, the user may be requested to move so that navigation can be easily performed to determine the current position of the user. Specifically, a movement instruction such as “Please move to the place of XX” is given.

  The output unit 50 has a function of outputting the estimated position estimated by the position estimation unit 30 and the guidance information generated by the guidance information generation unit 40. The output method by the output means 50 includes text output on a screen using liquid crystal or the like, output using speech such as speech synthesis, and output using a map displayed on the screen. The output means 50 corresponds to the display means 91 of the PDA 90. The output means 50 can also display information on the way about the position estimation process by the position estimation means 30 by output.

  The operation processing procedure in the position estimation processing method will be described below with reference to the flowcharts of FIGS. FIG. 15 shows an operation flowchart using screen display for input and output. That is, first, it is determined whether or not there is prior information on the user position in advance (step S510). If there is prior information (Yes in step S510), the prior information is used to determine the vicinity from the landmark database 20. A list of landmarks recognized in step S511 is acquired. Next, an input screen is displayed using the acquired landmark list (step S513). On the other hand, if it is determined in (Step S510) that there is no prior information on the user position (No in Step S510), a landmark input screen indicating that there is no prior information is displayed (Step S512).

  Next, landmark information is input to the input screen by the user (step S514), and position estimation processing is performed based on the landmark information input (step S515). Next, it is determined whether or not the location has been estimated in (Step S516). If the location is estimated by this determination (Yes in Step S516), the estimated location is displayed (Step S516). Step S517).

  Next, it is determined whether or not guidance information is to be generated (step S518). When the guidance information is generated by this determination, the content of the guidance information is displayed (step S519). On the other hand, when the position (location) cannot be estimated by the determination in (Step S516) (No in Step S516), the failure of the position estimation is displayed (Step S520), and when the position estimation is stopped as it is. (Yes at step S521), the process is terminated as it is (end). On the other hand, when the position estimation is performed again (No at Step S521), the process returns to (Step S510) and the processing procedure is performed again from the beginning.

  Next, FIG. 16 shows a flowchart of an operation process using voice for input / output. That is, first, it is determined whether or not there is prior information on the user position in advance (step S610). If there is prior information (Yes in step S610), the prior information is used to determine the vicinity from the landmark database 20. A list of landmarks recognized in step S611 is acquired. Next, the acquired landmark list is registered as a recognition target vocabulary (step S612).

Next, landmark information is input by voice by the user (step S613), and landmark information is extracted by voice recognition using the recognition target vocabulary by the landmark list (step S614). Then, position estimation processing is performed based on the extracted landmark information (step S615). Next, it is determined whether or not the position of the place has been estimated in (Step S616). If the position is estimated by this determination (Yes in Step S616), the output of the estimated position is sounded. (Step S617). Specifically, this output is a question of a confirmation landmark.
Next, it is determined whether or not guidance information is to be generated (step S618). If the determination is to generate guidance information (Yes at step S618), the content of the guidance information is output by voice (step S619). ). On the other hand, when the position (location) cannot be estimated by the determination in (Step S616) (No in Step S616), the failure of the position estimation is displayed (Step S620), and the position estimation process is stopped as it is. (Yes at step S621), the process ends (END). On the other hand, when the position estimation process is performed again (No at Step S621), the process returns to (Step S610) and the process procedure is performed again from the beginning.

  Hereinafter, the details of the position estimation processing step will be described. FIG. 17 is a flow chart showing a spatial condition matching process in the position estimation process shown in FIG. That is, this spatial condition matching processing includes (1) distance matching processing (step S700), (2) azimuth matching processing (step S711), and (3) arrangement matching processing (step S712). It consists of each processing step. Actually, matching of spatial conditions can be performed by combining these processes.

(1) Matching process by distance This matching process by distance checks whether or not the landmarks of the search results obtained for each of the plurality of landmarks are close in distance. Is a process of deleting from the search processing result. Specifically, only the combination of landmarks that remain is left. The criteria for determining the proximity in distance can be calculated using the height of the landmark or the shape of the landmark (2D or 3D model).

  If this proximity criterion is calculated by the height of the landmark, for example, a high landmark such as Tokyo Tower can be used as a landmark even from a distance. Further, by using the shape of the landmark, it is possible to use the landmark in consideration of the overlapping state of the buildings or a wide open section.

  Further, the distance calculation between the landmarks of the search result can be performed by using longitude and latitude or positional relationship data among the data items in the landmark database 20. In the case of longitude and latitude, a distance calculation formula based on longitude and latitude is used. In the case of positional relationship data, it is checked whether or not landmarks connected by links are included in the search processing result, starting from one of the landmarks of the search result. For example, a process of checking whether or not the landmark “2a, 2b, 2c...” Exists in the landmark connected by a link from the landmark 1a is performed.

  In addition, when the user inputs information about the distance of the landmark (distance information) as one of the surrounding landmark information, a process of checking whether the landmark of the search processing result matches the distance information is performed. . When the user inputs the expression “near, far, etc.” as the distance information, the distance is converted into a numerical value (which may have a certain range), and the matching process is performed based on the distance.

(2) Orientation matching processing In orientation matching processing, information on the orientation of the landmark (azimuth information) is input as surrounding landmark information, and the landmark of the search processing result matches the distance information. If it does not match, it is a process of deleting from the search processing result. Specifically, this is a process that leaves only the matching landmark set. For example, when the azimuth information “Landmark 1 is north of Landmark 2” is input, among the Landmark 1 search results (1a, 1b,...), The Landmark 2 search result (2a 2b...) To check whether there is a combination in the north.

  In addition, when the direction information “Landmark 1 is north and Landmark 2 is west” is input, the south area of the search result (1a, 1b,...) Of the landmark 1 It is checked whether there is a combination in which the region on the east side of the search result (2a, 2b,...) Of the mark 2 overlaps. If there are overlapping combinations of the south and east regions by this check, it can be estimated that the pedestrian may exist in that region.

(3) Matching process in arrangement In the matching process in this arrangement, when information about arrangement between landmarks (arrangement information) is input as surrounding landmark information, the landmark of the search process result is the arrangement information. In this process, if it does not match the arrangement information, it is deleted from the search processing result and only the matching landmark set is left. For example, when arrangement information “landmark 1 is next to landmark 2” is input, among the search results (1a, 1b,...) Of landmark 1, the search result (2a of landmark 2). Check if there is a combination next to 2b,. Here, the meaning of “adjacent” indicates that the landmarks are truly adjacent to each other or are lined up nearby. Since the landmark database 20 stores the distance between landmarks connected by links as position information data (links) between landmarks, the input arrangement information matches the arrangement information stored by the links. You can check if you want to.

  Further, when the arrangement information “Landmark 1 is across the road of Landmark 2” is input, Landmark 2 is included in the search results (1a, 1b,...) Of Landmark 1. It is checked whether there is a combination that sandwiches the road between the search results (2a, 2b,...). In the landmark database 20, information on what exists between landmarks connected by links, for example, the presence or absence of roads, is described as position information data (links) between landmarks. It will be checked whether the arrangement information described in the link matches.

  In addition, when the arrangement information “Landmark 1 is at the corner of the intersection” is input, the information is linked to the intersection by a link based on the search result (1a, 1b,...) Of landmark 1. Check if there is any. As described above, in the landmark database 20, the intersection is set as one of the landmarks. Therefore, in the positional relationship data (link) connecting the intersection and the nearby landmark, where the landmark is located at the intersection. Is stored (stored). Thereby, it is possible to check the matching between the input arrangement information and the arrangement information described in the link.

(4) User position determination process The user position determination process is performed by performing a search process from the landmark database 20 and a spatial condition matching process for the landmark of the search result, and conforming to the input surrounding landmark information. This is a process for enabling the position of the user to be estimated more accurately after the landmark set candidate to be acquired is acquired. FIG. 18 is a flowchart showing a user position determination process in the position estimation process shown in FIG.

  That is, the user position determination processing includes (1) inquiry processing for the user for narrowing down the landmark set (step S810), and (2) inquiry processing for the user for determination of the landmark set (step S811). ), (3) Inquiry process to the user for limiting the estimated position (step S812), and (4) Position estimation process using the landmark set (step S813). In practice, the user position can be determined by combining these processes.

(1) Inquiry process to user for narrowing down landmark groups This is performed when there are a plurality of remaining landmark groups as a result of performing a retrieval process based on the landmark database 20 and a matching process of retrieval results based on positional relationship information. It is processing. Specifically, this is a process of narrowing down a landmark set for a plurality of landmark sets. The process of narrowing down the landmark group is not included in the landmark group by following the link (representing the positional relationship) of the landmark database 20 from the landmark included in the landmark group for each landmark group. Search landmarks, extract landmarks that can be distinguished from other landmark sets, and ask the user whether the extracted landmarks are nearby. This process narrows down the set.

  FIG. 19 is a display example of a user inquiry screen for narrowing down a set of landmarks. Specifically, a plurality of estimated position candidates are displayed, and a landmark that is closer to the user as viewed from the user is selected from the plurality of candidates. The example of FIG. 19 shows an example in which a red roof house is checked (selected) by a user from among a plurality of estimated position candidates such as a post box, a traffic light, a Makuhari bakery, and a red roof house. The selection of landmarks can be narrowed down by this selection.

(2) Inquiry processing to the user for the determination of the landmark set Among the search processing from the landmark database 20, the matching processing of the search result based on the positional relationship information, and the inquiry processing to the user for narrowing down the landmark set This is a process that is performed when only one landmark set remains after performing some processes. By performing the landmark group determination inquiry process, the accuracy of the landmark group, that is, the position estimation process accuracy can be improved.

  In the landmark group determination inquiry process, the landmark database 20 is traced from the landmarks included in the landmark group to extract landmarks not included in the landmark group, and whether or not there is an extracted landmark. A question is asked to the user, and a landmark set is determined based on the answer. Here, landmarks are extracted in consideration of the fact that they are characteristic landmarks and that the landmarks included in the landmark set and the extracted landmarks can form a polygon that surrounds the user. It is.

  FIG. 20 is a display example showing a landmark inquiry confirmation screen. That is, based on the check contents (answers) for the inquiry displayed in FIG. 19, the landmark set is determined by the inquiry screen shown in FIG. In the example of FIG. 21, a confirmation message is displayed as to whether or not there is a “red roof house” that the user checked last time. If a negative answer is input to this confirmation inquiry, check again with another landmark (postbox, traffic light, Makuhari Bakery, etc.) linked from the landmark set, or position estimation processing Is re-executed, or the position estimation failure is reported to the user.

(3) Inquiry process to the user for limiting the estimated position The inquiry process to the user for limiting the estimated position constitutes the landmark set in order to limit the user position that can be estimated from one landmark set. This is a process of inquiring how the landmark looks from the user. If the landmark information input first is sufficient to limit the estimated position, this process can be omitted.

  FIG. 21 is a display example showing an inquiry screen to the user for limiting the estimated position. Here, description will be made on the assumption that the position is estimated at a place as shown in FIG. First, it is assumed that the user inputs “a white building and a florist next to it” as surrounding landmark information. At this time, by limiting the estimated position, the landmark set that matches the landmark information is {white building (id = 1), florist (id = 2)}. The position of the user that can be estimated based on the information of the landmark set can be estimated as being in any of the vicinity of the point A, the vicinity of the point B, and the vicinity of the point D in FIG. it can.

  At this time, the process of limiting the estimated position is a process of inquiring where the white building can be seen (whether it is south or north, the opposite side of the road, or this side) as shown in FIG. In the example of FIG. 21, it is assumed that the white building is in the north of the user and on the opposite side of the road. Hereinafter, based on this answer, the current position of the user can be estimated by the position estimation process using the next (4) landmark set.

(4) Position Estimation Processing Using Landmark Set Position estimation processing using a landmark set is a process of estimating a user position by using a determined landmark set and a user's positional relationship with the set. It is a process to perform. Here, the position estimated by the position estimation apparatus of the present invention is not a single point position represented by longitude and latitude, for example, so as to be measured by GPS, but is an estimation region where a user exists.

  The position estimation process using the landmark set is performed using information on the positional relationship of the landmarks constituting the determined landmark set as viewed from the user. Information on the positional relationship of the landmarks as viewed from the pedestrian includes the value of positional relationship data (such as the distance and direction of the landmarks as viewed from the pedestrian) in the input landmark information, and the above (3) estimated position. The value obtained as a result of the inquiry process to the pedestrian for the limitation of can be used.

  FIG. 22 shows an example of position estimation processing using the user's positional relationship with respect to the landmark. In this example, a white building can be identified by determining (white building, florist, building, intersection) as a landmark set. It is assumed that the positional relationship information “is in” is input. At this time, it can be estimated from the positional relationship “in front of you” that the user is in the area centered on the white building (circular area in the figure), and from the positional relationship “northeast” It can be estimated that there is a user in the area. From the positional relationship between “in front of” and “northeast”, it can be estimated that there is a user at the overlapping portion of the two regions (near point A in FIG. 11).

  In the case of the present invention, since the location of the user can be estimated based on the surrounding landmark information input by the user, the location of the user is not a single point expressed by the longitude and latitude by the conventional GPS positioning. It is possible to easily and accurately estimate the area to be estimated (for example, in front of a △ station, □ × in front of a building, etc.). It can also be used as a search system for a place based on landmarks that can be seen around, for example, a search system that searches for a place that meets the condition “Tokyo Tower is visible and close to the station”.

Embodiment 2
Next, a mobile object position estimation apparatus according to the second embodiment will be described. FIG. 23 is a functional block diagram showing the configuration of the mobile position estimation apparatus 1a according to the second embodiment. The components of the position estimation apparatus 1a are divided into an information communication terminal apparatus 81 and a center server 82. It is configured. That is, as shown in FIG. 23, the information communication terminal 81 includes an input unit 11 for inputting landmark information, an output unit 51 for outputting guide information generated by the guide information generating unit 67, an input / output control unit 16, a control unit. Part 65.

  On the other hand, the center server 82 is based on the landmark information input by the input unit 11, the landmark database 66 in which the positions and attributes of the landmarks are stored in advance, and the landmark information stored in the landmark database 66. Position estimation means 68 for estimating the position and guidance information generation means 67 for generating guidance information based on the position information estimated by the position estimation means 68.

  The center server 82 includes an input unit 62, an output unit 63, and an input / output control unit 64 as input / output units of a normal personal computer. As the input means 62, a keyboard, a trackball, a mouse or the like is used. As the output means 63, a CRT display of a personal computer or the like is used. The control unit 65 has a function as a controller that performs overall control of the center server 82. Reference numeral 72 denotes a communication I / F having a function as a communication interface. The communication I / F 72 can communicate with the communication I / F 71 of the information communication terminal device 81 through the communication medium 73. Here, the functional units that play the same roles as the respective units shown in FIG. 1 are denoted by the same reference numerals, and detailed description thereof is omitted.

  In the second embodiment, similarly to the first embodiment described above, surrounding landmark information is input by the input unit 11, and the input landmark information and the landmark database 66 are used by the position estimation unit 68. Thus, the location of the user is estimated, guidance information is generated by the guidance information generation means based on the estimated position, and the guidance information generated by the output means 51 is output.

  Further, a program for realizing the configuration shown in the first and second embodiments is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read and executed by the computer. Thus, the same effects as those of the mobile object position estimation devices 1 and 1a shown in the first and second embodiments can be realized by using a general computer system.

  Here, the storage medium is a portable storage medium such as a semiconductor memory, a CD-ROM, a floppy disk, a DVD disk, a magneto-optical disk, an IC card, or a storage device such as a hard disk, or an installation connected via a LAN. It includes a database that holds the original model building program and a transmission medium on the public network.

  As described above, the position estimation apparatus, the position estimation method, and the position estimation program for a moving body according to the present invention estimate the current position of a pedestrian or the like using an information communication terminal device such as a PHS, a portable terminal, or a PDA. This is suitable for a position estimation device for a moving body.

FIG. 1 is a functional block diagram showing a configuration of a mobile object position estimation apparatus according to the first embodiment. FIG. 2 is a flowchart showing a processing procedure of a method for estimating the position of a moving object. FIG. 3 is a diagram (outdoor) showing an example of landmark arrangement. FIG. 4 is a diagram (underground shopping mall) showing an example of landmark arrangement. FIG. 5 is a diagram for inputting landmark information by voice input and free text format. FIG. 6 is a diagram for inputting landmark information in text input and free text format. FIG. 7 is a diagram showing an example of an item-specific input format screen. FIG. 8 is a diagram showing an example of a candidate selection input format screen. FIG. 9 is a diagram illustrating a display example of the screen when the prior information regarding the position is available. FIG. 10 is a diagram showing the configuration contents of the landmark database. FIG. 11 is a diagram showing a network structure of landmarks (in the case of outdoors). FIG. 12 shows a landmark network structure (in the case of an underground shopping mall). FIG. 13 is a flowchart showing the processing procedure of the position estimation operation of the moving object by the position estimation device. FIG. 14 is a diagram illustrating an example of a search processing result. FIG. 15 is a flowchart showing a processing procedure using screen display for input and output. FIG. 16 is a follow chart showing a processing procedure using voice for input and output. FIG. 17 is a follow chart showing the spatial condition matching processing procedure. FIG. 18 is a follow chart showing the user position determination processing procedure. FIG. 19 is a diagram showing a display example of a user inquiry screen for narrowing down a set of landmarks. FIG. 20 is a diagram showing a display example of a user inquiry screen for determining a landmark set. FIG. 21 is a diagram showing a display example of an inquiry screen for the user for limiting the estimated position. FIG. 22 is a diagram showing a position estimation process using the positional relationship. FIG. 23 is a functional block diagram showing the configuration of the moving object position estimating apparatus according to the second embodiment.

Claims (11)

A position estimation device for a moving body that estimates a current position and a moving position of the moving body,
For a plurality of landmarks that the mobile body recognizes as landmarks, input information that inputs at least one of shape, color, size, and design, and feature information indicating positional relationship information as landmark information;
A landmark database pre-stored with the location of the landmark information and the feature information ;
Based on feature information and positional relationship information of a plurality of landmarks input by the input means , the landmark database is searched for a plurality of landmarks corresponding to the feature information, and the extracted landmarks From the position estimation means for estimating the current position of the moving body by performing a matching process of a spatial condition to extract a landmark set that is a combination of landmarks corresponding to the input positional relationship information ,
Guidance information generating means for generating guide information based on the position information of the moving body estimated by the position estimating means;
An apparatus for estimating the position of a moving body, comprising: output means for outputting the guidance information. A position estimation device for a moving body that estimates a current position and a moving position of the moving body,
For a plurality of landmarks that the mobile body recognizes as landmarks, input information that inputs at least one of shape, color, size, and design, and feature information indicating positional relationship information as landmark information;
A landmark database pre-stored with the location of the landmark information and the feature information ;
Based on feature information and positional relationship information of a plurality of landmarks input by the input means , the landmark database is searched for a plurality of landmarks corresponding to the feature information, and the extracted landmarks From the position estimation means for estimating the current position of the moving body by performing a matching process of a spatial condition to extract a landmark set that is a combination of landmarks corresponding to the input positional relationship information ,
Guidance information generating means for generating guide information based on the position information of the moving body estimated by the position estimating means;
Output means for outputting the guidance information,
The input means and the output means are provided in an information communication terminal device,
The position estimating means, the landmark database, and the guidance information generating means are provided in a center server,
The information communication terminal device and the center server transmit / receive information to / from each other via a network. The landmark database stores, as arrangement item data indicating a positional relation between the landmarks, a direction, a distance between adjacent landmarks, and object information existing between the landmarks. Item 3. The moving object position estimation apparatus according to Item 1 or 2. The input means further includes a landmark list display function for displaying a list of landmark information located in the vicinity of the moving body, and a landmark is displayed from the landmark list displayed based on the landmark list display function. The apparatus for estimating a position of a moving body according to claim 1 or 2, wherein selection of information and input of arrangement are performed . 3. The position estimation of a moving object according to claim 1, wherein the input unit answers an item specifying the landmark information and has an input function by item for inputting the information by item. apparatus. The said input means is provided with the function to select and input from the several candidate list | wrist further specifying the landmark information corresponding to the said item input by the said input function according to item. The position estimation apparatus of the moving body as described in 2. A method for estimating a position of a moving body for estimating a current position and a moving position of the moving body,
Computer
Landmark information input step of inputting, as landmark information, feature information indicating at least one of shape, color, size, and design, and positional relationship information for a plurality of landmarks recognized by the mobile object as landmarks When,
With reference to a landmark database in which the position and feature information of the landmark information is stored in advance, a plurality of landmarks corresponding to the feature information are obtained based on the inputted feature information and positional relationship information of the plurality of landmarks. The current position of the moving object is estimated by performing a spatial condition matching process that searches and extracts a landmark set that is a combination of landmarks corresponding to the input positional relationship information from the extracted landmarks. A position estimation step to perform,
A guide information generating step for generating guide information based on the position information of the moving body estimated by the position estimating step;
A guidance information output step for outputting the guidance information to the mobile body;
A method for estimating the position of a moving object comprising : The position estimating step performs a search range limiting process for landmarks based on prior information of a moving body position that is either the current position estimated in the past, an administrative division name, or a shopping street name. The mobile object position estimation method according to claim 7 , further comprising: performing matching processing of the spatial condition, and further performing mobile object position determination processing . The spatial condition matching process includes a matching process based on a distance based on the landmark distance information, a matching process based on an orientation based on the landmark orientation information, and an arrangement based on the landmark layout information. The method for estimating a position of a moving object according to claim 7 , further comprising a matching process . The user position determining process includes a landmark group narrowing inquiry process for extracting nearby landmarks, a landmark group finalizing query process for limiting the landmark group, and a position of the landmark group limited. 9. A moving object according to claim 8 , comprising: a limited inquiry process of estimated positions to be performed; and a position estimation process using landmark sets based on the positional relationship of landmarks constituting the determined landmark groups. Location estimation method. A moving object position estimation program for estimating a current position and a moving position of a moving object,
On the computer,
Landmark information input procedure for inputting, as landmark information, feature information indicating at least one of shape, color, size, and design, and positional relationship information for a plurality of landmarks recognized by the moving object as landmarks When,
With reference to a landmark database in which the position and feature information of the landmark information is stored in advance, a plurality of landmarks corresponding to the feature information are obtained based on the inputted feature information and positional relationship information of the plurality of landmarks. The current position of the moving object is estimated by performing a spatial condition matching process that searches and extracts a landmark set that is a combination of landmarks corresponding to the input positional relationship information from the extracted landmarks. Position estimation procedure to
A guide information generating procedure for generating guide information based on the position information estimated by the position estimating procedure;
A program for estimating the position of a moving body, which executes a guide information output procedure for outputting the guide information to the mobile body.

Images