JP2007205948A - Current position estimation method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To estimate a current position of a user without using GPS in a route guide system. <P>SOLUTION: The route guide system searches and guides a route on the basis of the current position of the user. The system inquires relative position relationship between buildings of the user concerning a plurality of the buildings viewed by the user when the current position of the user is unknown. When a certain building B is in a left side of a certain building A, the system demarcates such a viewing range on a map on the basis of the position relationship of the buildings A, B. A region where the current position of the user exists is narrowed by carrying out the processing in combination of the buildings of a plurality of ways, and the current position is estimated at sufficient precision. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a technique for estimating a current position of a user.

  Navigation systems targeting vehicles and pedestrians have become widespread. These navigation systems search for a route to the destination based on the current position of the user, and provide guidance for the searched route. The user's current position is usually detected using GPS (Global Positioning System).

  The detection of the current position by GPS is not perfect because it can be used only under conditions where radio waves from artificial satellites can be received. In order to improve the practicality of the navigation system, it is desirable to be able to specify the current position even in an environment where radio waves cannot be received, or when a GPS is not installed in a terminal owned by the user. In response to such a request, Patent Document 1 discloses a technique for asking the user what is visible on the front and what is visible on the left, and estimating the current position based on the answer. Similarly, Patent Document 2 discloses a technique for estimating the current position by asking the user a question such as “Can you see a restaurant on the left side?”.

JP 9-319991 A JP 2004-340689 A

  In the prior art, a method is adopted in which a building that can be seen from the front and the left is answered based on a pedestrian. However, when the current position is unknown, pedestrians usually look around in all directions and change their posture frequently. In the conventional technology, if the pedestrian changes his / her posture while answering multiple questions, the direction standard such as “front” and “left” will fluctuate, and the current position will be estimated with sufficient accuracy. There was a problem of being unable to.

  In addition, the method of asking whether to approve the viewing as in the prior art is preferable in that the misunderstanding of the user can be reduced by targeting landmarks such as large buildings that are easily identified by the user. However, since such a landmark is usually visible from a distance, even if an answer to the permission of viewing approval for the landmark is obtained, the current position of the user cannot be accurately estimated. In the prior art, there is a problem that it is difficult to select a question target that can sufficiently secure the accuracy of the current position.

  As is apparent from the above-mentioned problems with landmarks, in the prior art, in order to accurately identify the current position, it is necessary to finally make a building in the vicinity of the user a question. Therefore, even if there is a characteristic building that can be specified without misunderstanding by the user, such a building cannot be used for specifying the current position if it is far from a certain range from the user. As a result, a building with few features must be used for the estimation of the current position, and the accuracy of specifying the current position may be reduced due to misunderstanding of the user.

  As described above, there is still room for improvement in practicality in the estimation of the current position in the prior art. The estimation of the current position is not a technique peculiar to the navigation system, but is also a technique required when realizing a service that provides gourmet information, sightseeing information, shopping information, etc. according to the current position of the user. An object of the present invention is to improve the practicality of the present position estimation technique required in various scenes.

  A configuration of the present invention as a current position estimation device for estimating the current position of a user will be described. The present position estimation apparatus of the present invention inputs the result of visual recognition by the user regarding the relative positional relationship between a plurality of features. What is input is not a positional relationship based on the user such as the front or the left as viewed from the user, but a relative positional relationship between a plurality of features. For example, the relative positional relationship such as “Three features can be seen in the order of feature 1, feature 2, feature 3 from the right” or “feature A and feature B appear to overlap each other” Applicable.

  The current position estimation device refers to a map database that records the position of each feature, identifies a range in which the above-described plurality of features can be visually recognized based on the input relative positional relationship, and determines the current position of the user. presume. For example, when the positional relationship that “feature A and feature B appear to overlap each other” is input, the location of such a view is relatively limited, and thus the current position of the user can be narrowed down. It is possible. In the present invention, since the positional relationship between features is used for estimating the current position, there is an advantage that the current position can be estimated without any trouble even if the user changes the posture.

  The feature for which the relative positional relationship is to be input may be determined by the user or the current position estimation device. In the former method, the user inputs the relative positional relationship after designating each feature with a name or the like such that “feature A and feature B appear to overlap each other”. This method is suitable when there are many features such as the center of a big city, and the user can easily know the names of the features using a signboard or the like. However, in this method, it is necessary that the current position estimation apparatus can uniquely identify the feature designated by the user from the map database.

  In the latter method, the current position estimation device presents information that allows the user to identify each feature on the site. The information to be presented includes, for example, features related to appearance such as the image, color, shape, and height of the feature, the name of the feature, and the type of the feature such as a hotel or department store. When the user confirms a feature on the site based on the presented information and inputs a relative positional relationship, the current position estimation device estimates the current position based on this. This method has an advantage that processing after the positional relationship is input can be easily and accurately performed because the current position estimation apparatus already knows the feature for which the relative positional relationship is input.

  The relative positional relationship can be specified in various expression formats. For example, it is preferable that the relative positional relationship is specified by the front-rear relationship or the left-right relationship between a plurality of features when viewed from the user. Even if the positional relationship between the features is oblique, it is relatively easy and accurate to specify that one of the features is seen in front of the other, or is seen on the left or right. Therefore, by specifying the relative positional relationship by such a method, the user can easily use the current position estimation device. In addition, since the positional relationship is specified accurately, it is possible to improve the estimation accuracy of the current position.

  In the present invention, it is possible to narrow the range of the current position to some extent only by using two features, but in order to further improve the estimation accuracy, use three or more features that are not in a straight line. Is preferred.

  As described above, the current position estimation apparatus can use three or more features. At this time, although it is possible to take a method of processing all these features at once, from the viewpoint of standardization and simplification of processing, two features to be processed are selected from a plurality of features. Then, based on the relative positional relationship between the selected features, a process of demarcating a boundary line of an area where the current position of the user can exist (hereinafter referred to as “existing area”) is defined as a processing target. It is preferable to perform the process of narrowing down the existing area by repeatedly executing the feature to be changed.

  As the number of features available for processing increases, the number of combinations of two features to be processed increases. Therefore, by selectively executing a combination useful for defining the boundary of the existence area, it is possible to shorten the processing time and efficiently obtain the existence area. In this case, there is not necessarily a theoretically optimal method for selecting a feature to be targeted, but there are several suitable selection methods.

  In the case of using four or more features as the first selection method, when there is an already defined boundary, the feature in a positional relationship capable of defining the boundary intersecting with the defined boundary is preferentially used. It is preferable to select. For example, an aspect in which two features that are in a positional relationship across a predefined boundary and a mode in which a feature that is on a predefined boundary and a feature that is not on the boundary are applicable. If a feature is selected in this way, the new boundary specified by the selected feature will intersect the already defined boundary. In other words, the new boundary is a form that divides the existing area that has already been defined, so that the existing area can be efficiently narrowed down.

  As a second selection method, when four or more features are used, when there are already defined boundaries, two features existing in an area where the current position can exist are preferentially used. preferable. The boundary defined by the feature selected in this way will intersect the already defined boundary. In other words, the new boundary is a form that divides the existing area that has already been defined, so that the existing area can be efficiently narrowed down.

  As a third selection method, when three or more features are used, it is preferable to use a feature having a low height preferentially. Considering that it is generally difficult to visually recognize a feature with a low height from a distance, it is considered that a feature with a lower height is closer to the current position of the user. Accordingly, by preferentially using the feature having a low height, the boundary of the existing area can be defined in the vicinity of the current position, and the accuracy of the existing area can be improved.

  Any one of the three selection methods described above may be employed, or may be employed in combination as appropriate. The priority in the case of employing a plurality of selection methods can be arbitrarily set. It is also possible to add a selection method different from the above three selection methods.

  In the present invention, it is preferable to perform pre-processing for roughly grasping the current position of the user prior to the above-described processing for defining the existence area. As pre-processing for this, the following method can be taken, for example. First, the current position estimation apparatus inputs an attribute that can be used to identify a feature according to a user operation. The feature that is the input target of the attribute is a feature that is arbitrarily selected within a range that can be visually recognized by the user. Examples of the attribute include information on the appearance such as the name, type, color, height, and pattern of the feature. The map database stores these attributes in association with each other. The current position estimation device refers to this map database and identifies the feature corresponding to the attribute. In this case, only those that completely match the input attribute may be specified, or the feature may be specified while allowing a mismatch with the input attribute within a certain range. In the latter mode, among the input items, up to a predetermined number may be handled even if they do not match, or each item may be treated as allowing an allowable error.

  When this feature is specified, the range in which the feature can be visually recognized can be specified in general, so that it is possible to obtain a general region where the current position is determined based on the feature. The current position estimation device sets a plurality of features in the approximate area thus obtained as information presentation targets. The feature to be the target of information presentation may be the same as the feature arbitrarily selected by the user or may be different. By doing so, it is possible to execute the process for demarcating the existence area in a state where the current position is substantially narrowed down, so that the existence area can be obtained efficiently and accurately.

  In the pre-processing described above, a single feature corresponding to the attribute cannot be specified only by the attribute input by the user, and there may be a plurality of candidates corresponding to the attribute. In this way, when there are multiple candidate features corresponding to a specific attribute, one of the candidates is specified based on the distance between each candidate and a feature corresponding to another attribute that is input separately. It is preferable to select the feature corresponding to the attribute.

  As an example, let us consider a case where there are two features A and B corresponding to a specific attribute. Assume that the feature C is specified as a feature corresponding to another attribute. In such a case, the distance of the feature AC and the distance of the feature BC are respectively obtained. When the distance between the features AC is a distance that allows the user to visually recognize both buildings, and the distance between the features BC is so far that the two buildings cannot be visually recognized, the location corresponding to the specific attribute is displayed. The feature A can be selected as an object. This is because the user should be present in a range where both the features A and C can be visually recognized. As described above, it is possible to specify the feature corresponding to the attribute with high accuracy by considering not only the correspondence relationship with the input attribute but also the positional relationship with the feature corresponding to the other attribute. .

  In the pre-processing described above, there may be a case where the feature corresponding to the attribute input by the user cannot be specified sufficiently. In such a case, information on a predetermined feature may be presented to the user, and an answer indicating whether the user can visually recognize the feature may be input, and the approximate region may be obtained by reflecting this answer. As such a feature, it is preferable to use a large landmark that is visible from a relatively long distance.

  In this way, when the feature is presented to the user and an answer of permission for viewing is obtained, in order to narrow down the approximate area, it is preferable to repeatedly perform presentation of information about the feature and input of the answer under predetermined conditions. . It is possible to arbitrarily select the feature that asks for permission to view, but if an answer indicating that one feature is visible is obtained, the approximate area is narrowed based on the position of the feature. In the above, it is preferable to select a feature to be the next information presentation target from the approximate region. By doing so, it is possible to omit an unnecessary response about the approval or disapproval of viewing and efficiently narrow down the approximate area.

  On the other hand, if a reply indicating that one feature cannot be seen is obtained, the general area is expanded based on the position of the feature, and then the target area for the next information presentation from the general area is displayed. You may make it select a thing. This is because there is a possibility that a reply indicating that another feature cannot be visually recognized in the existing schematic region is not a little. By expanding the selection range of the features for which the visual approval is rejected, such an unnecessary response can be avoided and the approximate area can be found efficiently.

  As described above, the setting of the approximate area is performed as a pre-process prior to the process of estimating the current position. When the current position estimation apparatus obtains the approximate area by the above-described method, the current position estimation apparatus extracts a plurality of features existing in the area and uses it for the process of defining the existing area. That is, the user is inquired about the relative positional relationship between the plurality of features extracted in this way, and the presence area is defined based on the result.

  In the present invention, the existence area can be defined by various methods. An example of the definition method is shown below. In this method, first, two features to be processed are selected from among a plurality of features. And the convex polygon which envelops the selected feature is calculated | required, and the edge | side which connects two features among the edges which comprise a convex polygon is selected. What is necessary is just to select the side where the both ends of a side overlap with the vertex of a different feature. When the side thus selected is extended, the plane is divided into a plurality of regions. The relative positional relationship when viewing two features from each of these regions is qualitatively different for each region. Therefore, by selecting an area that does not violate the relative positional relationship designated by the user for each area, the boundary line of the existing area can be defined. According to this method, there is an advantage that the boundary line can be defined by a relatively simple geometric process.

  In the present invention, network data representing a road through which a user can pass as a node and a link may be used for estimating the current position. Network data can be thought of as representing a path through which a user can move. Therefore, the current position estimation device sets the current position in the network data in a range where the above-described features can be visually recognized in a relative positional relationship, that is, any node on the existing area and any point on the link. It can be estimated that. For example, when processing for pedestrians, if there are lakes, automobile roads, tracks, etc. in the existing area, refer to the network data for pedestrians to identify those points that pedestrians cannot pass. The current position can be quickly removed from the current position candidates, and the current position can be estimated accurately with a simple process.

  Network data can be used at various stages. For example, after defining the existence area, nodes and links in the existence area may be extracted using network data. In addition, the network data may be used at the stage of defining the existence area, and the existence area may be obtained as a collection of nodes and links that can be visually recognized with respect to a plurality of features in a relative positional relationship inputted by the user. .

  As a method of further narrowing down the current position after obtaining the existence area by the above-described method, for example, an image or attribute of a specific feature in the existence area is presented to the user, and the feature is in front of the eyes A method of allowing the user to input whether or not it can be seen can be adopted.

  In the present invention, the various features described above are not necessarily all provided, and some of them can be omitted or combined as appropriate. Further, the present invention may be configured as a current position estimation method for estimating a user's current position using a computer as well as the above-described aspect of the current position estimation device. Further, the present invention may be configured as a computer program for realizing the current position estimation described above or a recording medium on which such a computer program is recorded. Here, as a recording medium, a flexible disk, a CD-ROM, a magneto-optical disk, an IC card, a ROM cartridge, a punch card, a printed matter on which a code such as a barcode is printed, an internal storage device of a computer (RAM, ROM, etc. Various types of computer-readable media such as a memory) and an external storage device can be used.

The embodiment of the present invention will be described in the following order, taking as an example a case where it is realized as a current position estimating function in a route guidance system.
A. System configuration:
B. Route search / guidance processing:
C. Current position estimation processing:
D. Standard building setting process:
E. Existence area definition processing:
F. Reference building selection:
G. Variation:

A. System configuration:
FIG. 1 is an explanatory diagram showing a configuration of a route guidance system as an embodiment. In the present embodiment, a configuration example as a route guidance system for pedestrians is shown, but it can also be configured as a route guidance system for vehicles. The route guidance system is configured by connecting a terminal 100 using a mobile phone and a server 200 via a network INT. The network INT is a network using wireless communication, and may be limited such as a LAN or an intranet, or may be a wide area such as the Internet. As the terminal 100, a mobile phone, a so-called vehicle navigation device, a PDA, a personal computer having a network communication function, and the like can be used. When configured as a route guidance system for pedestrians, it is desirable that the vehicle navigation device be removed from the vehicle and portable.

  The terminal 100 has a function for transmitting instructions necessary for route search and route guidance to the server 200 in accordance with user operations. In the figure, functional blocks of the terminal 100 are also shown. The terminal 100 incorporates a microcomputer having a CPU, a RAM, and a ROM as a control device, and this CPU executes software stored in the ROM to constitute each functional block shown in the figure. These functional blocks can be configured in hardware as well as in software.

  The communication unit 120 has a function of communicating with the server 200 via the network INT. The GPS 140 detects the latitude and longitude of the current position of the terminal 100 using a global positioning system. The command input unit 130 inputs a destination designation, a search condition designation, information for current position estimation, and other commands based on a user operation on a switch or the like. The display control unit 150 displays these command input menus on the display of the terminal 100 or displays a map based on the route guidance data received from the server 200. The main control unit 110 controls the overall functions of the terminal 100 by performing overall control of the functional blocks described above.

  Based on commands from the terminal 100, the server 200 performs route search and route guidance while referring to various databases. In addition, when the GPS 140 of the terminal 100 is in a state where it cannot receive radio waves during these processes, the server 200 performs predetermined information exchange with the user when the current position of the user is unknown. It also has a function of estimating the current position, that is, a function as a current position estimating device of the present invention. In the figure, examples of functional blocks and databases for realizing these functions are shown. Each functional block is configured by software by a computer program executed by the CPU of the server 200, but may be configured by hardware.

  The server 200 executes each process of route search, route guidance, and current position estimation while referring to the map DB 250. The map DB 250 includes network data 252 in which paths are represented by nodes and links, and drawing data 254 made up of map drawing polygons and character data. In the drawing data 254, attributes of various features such as buildings are also recorded. As attributes, for example, the type, name, shape, height, color, and appearance (for example, brick construction, glass lining, etc.) of the building are recorded. You may record the photograph of a building as an attribute.

  The route guidance DB 260 stores various data input from the terminal 100 or generated by the server 200 in the course of route search and guidance. Such data includes, for example, search conditions specified by the user operating the terminal 100 during route search, the current location of the user, route guidance data indicating the result of the route search, and the like.

  The communication unit 210 controls communication with the terminal 100 via the network INT. The route search unit 220 searches for a route between the departure point and the destination designated by the user by referring mainly to the network data 252 in the map DB 250 in accordance with the search conditions stored in the route guidance DB 260. The searched result is stored as route guidance data in the route guidance DB 260 as described above. The route guidance unit 240 generates display data that guides and displays the route on which the user should travel while referring to the route guidance data in the route guidance DB 260 and the current position, and transmits the display data to the terminal 100.

  The estimation engine 230 performs a process for estimating the current position when a situation in which the current position of the user cannot be acquired from the terminal 100 occurs in the course of the route search and the route guidance, and the process with the user. Send and receive information necessary for The reference building setting unit 232 has a function of setting a reference building to be used for processing through interaction with the user as preprocessing for estimating the current position. The boundary demarcating unit 234 has a function of demarcating the boundary of the existing area, that is, the area where the “current position” may exist, by exchanging with the user using the above-described reference building. In the boundary definition, the selection unit 231 has a function of determining a reference building to be used by the boundary definition unit 234 according to a predetermined rule when there are many reference buildings. The information providing unit 232 presents information such as images and attributes in order for the user to specify the reference building locally. The boundary demarcation unit 234 performs the above-described boundary demarcation using information input by the user regarding the positional relationship of the reference building thus selected. The processing contents of the reference building setting unit 232 and the estimation engine 230 boundary demarcation unit will be described later.

B. Route search / guidance processing:
FIG. 2 is a flowchart of route search processing and route guidance processing. This process is executed by the route search unit 220 and the route guide unit 240 of the server 200, and is executed by the CPU of the server 200 in terms of hardware.

  In the route search process, the server 200 sets the starting point of the route search by a method according to the designation from the user. When designation is made to use the current position as the departure place (step S10), current position estimation processing is executed (step S12) to acquire the current position. When it is designated that the current position is not set as the departure place, the designation of the departure place is input from the terminal 100 (step S11). For the designation of the departure place, for example, a method in which the user clicks one point on the map, a method of inputting an address, or the like can be employed.

  Next, the server 200 inputs a route search destination and search conditions (step S14). As the destination, a method similar to the designation of the departure point, or a method of searching from the destination name, the type of restaurant, shop, or the like can be adopted. As the search condition, it is possible to specify that the route with the shortest distance is searched, specify that the route with the shortest required time is searched, specify that the route that avoids the uphill or the stairs is searched, and the like.

  The server 200 executes a route search process based on the above-described conditions (step S16). The route search can be performed by the Dijkstra method. Since the Dijkstra method is a well-known method, a detailed description is omitted. However, the Dijkstra method searches for a route that minimizes the cost attached to each node and link representing a road. When a search condition for avoiding an uphill or a staircase is attached, a route satisfying the search condition can be searched by setting the cost of the link corresponding to the uphill or the like to a large value. Similarly, when other search conditions are specified, it is possible to search for a route that satisfies each search condition by setting the cost of each link according to the search condition. When the route search is completed, the server 200 outputs the route guidance data to the route guidance DB 260 and ends the route search process (step S18).

  The server 200 continues the route guidance process with reference to the route guidance data. Although the route search process and the route guidance process can be executed sequentially as a series of processes, in the present embodiment, both are configured as separate processes. In this way, route search and route guidance can be executed at different timings, and the convenience of the route guidance system can be improved. For example, it is possible to search for a route from the station after getting off the train to the destination while getting on the train, and start route guidance after arriving at the station.

  In the route guidance process, the server 200 reads route guidance data stored in the route guidance DB 260 (step S50). Then, the current position of the user is acquired by the current position estimation process (step S52), and the route that the user should travel is displayed on the map by the guidance display process (step S54). The server 200 repeatedly executes the above processing until the user arrives at the destination (step S56), and completes the route guidance processing.

C. Current position estimation processing:
FIG. 3 is an explanatory diagram showing an example of the current position estimation process. In this embodiment, the server 200 determines a plurality of buildings to be used for processing, inquires about the relative positional relationship between the plurality of buildings as seen from the user, and based on the result, the presence of the current position A method of defining a region is taken. In the following description, a process when a relative positional relationship with respect to the buildings A, B, C, and D shown on the map of FIG. The buildings A to D used for defining the existence area are hereinafter referred to as reference buildings.

  The user inputs the positional relationship between the left and right of these buildings. FIG. 3B shows the existence area when the “building B is visible on the left side of the building A” for the user. In the figure, the user can see the buildings A and B in the above-described positional relationship as long as the side is indicated by the arrow rather than the boundary indicated by the bold line. A method of defining the existence area in the form shown in FIG. 3B when the relative positional relationship between the two buildings is given will be described later.

  The reason why the boundary line of the existence area is bent between the buildings A and B in FIG. 3B is that the area between the buildings A and B is included as the existence area that meets the above-described conditions. . In the area between the buildings A and B, the relative positional relationship between the buildings A and B changes according to the position and posture of the user. Building A may appear to the left of building B, or vice versa. Also, building A may be visible in front of the user and building B may be visible behind the user, and vice versa. In the present embodiment, in consideration of the possibility that “building B can be seen on the left side of building A”, the region between buildings A and B is also included in the existing region that matches the condition. In the present embodiment, the positional relationship between the buildings is inquired in the alternative of being visible on either the left or right side, but in addition to this, an option of “I am between buildings” may be added. Moreover, you may add the option that the buildings A and B seem to overlap, or only one of them can be seen. These options represent the relative positional relationship of the three parties that the building and the user are substantially aligned. Since the range in which such a positional relationship is established is relatively limited, the user's current position can be quickly narrowed down by adding the above-described options.

  Similarly, when the user inputs the positional relationship “building C can be seen on the left side of building B”, the existence area shown in FIG. 3C can be defined. When the positional relationship “building D is visible on the left side of building C” is input, the existence area shown in FIG. 3D can be defined. When the positional relationship “building A is visible on the left side of building D” is input, the existence area shown in FIG. 3E can be defined. By overlapping the existence areas defined in FIG. 3B to FIG. 3E, the existence areas satisfying all the relative positional relationships described above can be obtained as shown in FIG.

  FIG. 4 is an explanatory diagram showing an example of processing for demarcating boundaries of existing areas. FIG. 3 shows a processing example when the relative positional relationship between the buildings A and B is designated, that is, a processing example for demarcating the boundary of FIG. Assume that buildings A and B exist at the positions shown in FIG. In order to avoid the complexity of the figure, the background map is omitted in FIGS. 4B to 4D and the positions and shapes of the buildings A and B are shown.

  In the process of demarcating the boundary, for the buildings A and B, as shown in FIG. Next, among the sides of the convex polygon, those that extend over different buildings are extracted. That is, the side where one end of the side overlaps with any vertex of the polygon representing the building A and the other end overlaps with any vertex of the polygon representing the building B is extracted. In the example of FIG. 4B, the sides s1 and s2 correspond to these sides.

  As shown in FIG. 4C, the extracted sides s1 and s2 are extended. Thus, the plane is divided into regions α0 to α5. The appearance of buildings A and B is different in each region. The region α0 is a region where the relative positional relationship between the buildings A and B varies depending on the posture of the user. The area α1 is an area where only the building A can be visually recognized and the building B cannot be visually recognized. In contrast, the area α2 is an area where only the building B can be visually recognized and the building A cannot be visually recognized. However, in the regions α1 and α2, depending on the height relationship between the buildings A and B, the buildings A and B may be seen side by side. The region α3 is a region where the building B can be seen behind the building A, and the building B can be seen on the left side of the building A depending on the posture of the user.

  The region α4 is a region where the building A can be seen on the left side of the building B. The region α5 is a region where the building B can be seen on the left side of the building A. As described above, when the positional relationship “the building B is visible on the left side of the building A” is designated by the user, the regions α0, α3, and α5 correspond to this condition. Therefore, for this condition, as shown in FIG. 4D, a range including the regions α0, α3, and α5 is defined as an existing region, and the boundary line is specified. The boundary line shown in FIG. 3B is obtained in this way. Each boundary shown in FIG. 3C to FIG. 3E can also be obtained by a similar method.

  Although the existence area defined by the method of FIG. 4 theoretically extends infinitely, the range in which the building can be visually recognized is limited. Therefore, it is preferable to consider the viewable range set according to the height of the building and treat the existence region as a finite region within the viewable range.

  FIG. 5 is a flowchart of the current position estimation process. The estimation engine 230 (see FIG. 1) is a flowchart for realizing the processing described with reference to FIGS. 3 and 4 and is executed by the CPU of the server 200 in terms of hardware.

  In the current position estimation process, the server 200 first determines whether or not the GPS 140 of the terminal 100 is functioning effectively (step S102). If the GPS 140 is functioning effectively, the current position is input from the terminal 100 (step S104), and the current position estimation process is terminated.

  When it is determined that the GPS 140 is not functioning, processing for estimating the current position is executed by the method shown in FIGS. The server 200 first performs reference building setting processing (S110). The reference building setting process is a process for setting a reference building used for demarcating the existence area as in the buildings A to D in FIG. The details of the reference building setting process will be described later.

  The server 200 executes the existing area defining process using the set reference building (step S150). The process described with reference to FIGS. 3 and 4 receives an input of the relative positional relationship between the reference buildings from the user and demarcates the corresponding existence area. The contents of the existing area defining process will also be described later. The reference building setting process and the existing area defining process continue until the existing areas are defined. For example, depending on the positional relationship of the reference buildings set in the reference building setting process, the existing area may not be sufficiently narrowed down. In such a case, the server 200 determines that the existing area is not defined, and retries the reference building setting process and the existing area defining process.

  The determination as to whether or not the existence area has been defined can be made by various methods. For example, it may be determined that the existing area is defined when it is a closed figure, or it may be determined that the area is defined when the area of the existing area or the radius of the circumscribed circle is a predetermined value or less. Good.

  When the presence area is defined, the server 200 performs a process for specifying the current position with higher accuracy. As a process for this, the server 200 extracts a sidewalk link in the existence area (step S182), and extracts a group of buildings around the sidewalk link (step S184). Then, the building information of the extracted building is presented on the terminal 100, and an answer from the user is input (step S186).

  The interface screen DISP for this input is illustrated in the figure. In the example of the figure, a method of presenting an image of the extracted building and confirming with the user whether or not the building is approved for viewing is adopted. In this example, the user is asked whether he / she can see “in front of”, so when the user answers “YES”, the current position can be specified almost pinpointed. If “NO” is replied, the current position is not specified, and therefore the server 200 confirms whether or not the view authorization is the same for other buildings (step S188).

  In step S186, it is not limited to “in front of the eyes”, and it may be possible to simply visit the permission to view the building. The visible range of the building can be set as a circular area drawn with a radius corresponding to the height of the building, centered on the position of the building. Therefore, when the user replies with viewing authorization, it is determined that the current position of the user is within the above-described visible region. By confirming with the user whether or not viewing is permitted for a plurality of buildings, the locations where the viewable regions overlap are limited, and the current position of the user can be specified. In the present embodiment, since the existence area is sufficiently narrowed down to the process up to the user step S184, the question and answer input of the step S186 can be performed on the building near the user, and the current position is specified with sufficient accuracy. It becomes possible to do.

  In the above-described embodiment, the processes in steps S182 to S188 may be omitted. When this process is omitted, for example, an arbitrary point in the existing area, the center of gravity of the existing area, and the like can be specified as the current position. In the process up to step S180, when the existing area is sufficiently narrowed down, the current position is specified with an accuracy that does not hinder the route search process and the route guidance process even if the current position is specified by the above-described method. Is possible.

D. Standard building setting process:
FIG. 6 is a flowchart of the reference building setting process (first half). This is a process corresponding to step S110 of the current position estimation process (FIG. 5). This is a process corresponding to the function of the reference building setting unit (see FIG. 1), and can also be referred to as pre-processing of the existing area defining process in that the reference building used for the existing area defining process is set.

  When the reference building setting process is started, the server 200 inputs a building attribute for a building arbitrarily selected within a range that can be visually recognized by the user (step S111). An interface screen example DISP1 displayed on the terminal 100 for inputting building attributes is shown in the figure. In this example, the name, type, height, and color can be input. If the name of the building is known, enter it in the “Name” field. In “Type”, a corresponding item is selected from hospitals, department stores, condominiums, hotels, and the like. In the “Height” column, enter the number of floors of the building. For the color, enter the color of the exterior of the building. It may be possible to input the appearance such as brickwork or glass. The attribute is not limited to the items exemplified here, and various other items may be input.

  The server 200 extracts candidate buildings corresponding to the input attribute from the map DB 250 (step S112). In this case, only those that completely match the input attribute may be specified, or the feature may be specified while allowing a mismatch with the input attribute within a certain range. In the latter mode, it can be handled that even a predetermined number of input items are not matched even if they do not match. For example, if one item out of the four items shown in the figure is allowed to be inconsistent, a “non-hospital” five-story white building or a “four-story” white hospital is a candidate. Will be extracted. As another aspect in which the mismatch is allowed, it may be handled that an allowable error is allowed for each item. In the example in the figure, for example, if an error of “± 1 floor” is allowed with respect to a height of 5 stories, a white hospital with 4 stories or 6 stories is extracted as a candidate. In the standard building setting process, there is a possibility that the user may input an attribute for a building that is a little far away. Thus, by allowing a certain error in this way, an input error of the attribute due to a user's mistake or the like is supplemented. It is possible to increase the possibility of extracting candidate buildings according to the user's intention.

  In the process of step S112, one extracted candidate may not be identified. In such a case, the server 200 narrows down the candidates to one according to the distance from the already set reference building and other candidate buildings (steps S113 and S114).

  The method of narrowing down is shown in the figure. The figure on the left shows an example in which when three buildings a, b, and c are candidates, they are narrowed down to any one of them based on the distance from the reference building BB. Circular area | region Ca-Cc is an area | region which can visually recognize each of the buildings ac. The circular area Cbb is a range in which the reference building BB can be visually recognized. The radii of the areas Ca to Cc and Cbb can be set according to the heights of the buildings a to c and BB, and the radius increases as the height increases. When the reference building BB that has already been set exists, the user exists in a position where the reference building BB can be visually recognized, that is, in the region Cbb. On the other hand, when the buildings a to c are extracted as candidates, the user is present in any of the ranges in which the buildings a to c can be visually recognized, that is, the regions Ca to Cc. In the illustrated positional relationship, only the region Ca is at least partially overlapped with the region Cbb. Therefore, it is considered that the user exists in the range where the areas Cbb and Ca overlap, and it is possible to specify that the building a corresponds to the attribute input by the user.

  An example of narrowing down based on the distance to other candidates is shown on the right side of the figure. Assume that the candidates corresponding to the first attribute input by the user are the buildings d1 and d2, and the candidates corresponding to the second attribute are the buildings e1 and e2. In the figure, the visible area of each building is also shown. The user is considered to be at a position where the building d1 or d2 can be visually recognized and at a position where the building e1 or e2 can be visually recognized. In the illustrated positional relationship, the visible areas of the candidate d2 and the candidate e1 overlap. From this, it is possible to specify that the candidate corresponding to the first attribute is the building d2, and the candidate corresponding to the second attribute is the building e1.

  When the candidate corresponding to the attribute is narrowed down to one as a result of the above-described processing, the server 200 registers the candidate as a reference building (step S115). If the candidate in step S114 cannot be narrowed down to one, the plurality of extracted buildings are stored as candidates. This is because, in the process of narrowing down other candidate buildings as in the processing example of the above candidates d1, d2 and candidates e1, e2, it may be possible to narrow down to one candidate at the same time. In addition, when one candidate is specified at the time of extraction from the map DB (step S113), the server 200 skips narrowing down in step S114 and registers the extracted candidate as a reference building ( Step S115).

  When the user gives an instruction to input another attribute (step S116), the server 200 executes the processes from steps S111 to S115 again. If there is an instruction not to input another attribute (step S116), the server 200 proceeds to the latter half of the process.

  FIG. 7 is a flowchart of the reference building setting process (second half). When the number of reference buildings is 3 or more by the process described above, the server 200 determines that it is not necessary to set more reference buildings (step S120), and ends the reference building setting process. If the number is less than three, processing for further increasing the number of reference buildings is executed. This is because it is preferable to use three or more reference buildings in order to narrow down the existence area with high accuracy. Of course, when not so high accuracy is required for the definition of the existence area, the reference building setting process may be terminated when two reference buildings are set.

  If the number of reference buildings is less than three, the server 200 sets the approximate existence area based on the already registered reference buildings (step S121). The approximate existence area is an existence area at the current position determined from the visible range of the reference building. The method for setting the approximate existence area is illustrated in the figure. Here, it is assumed that buildings B1 and B2 are set as reference buildings. For each building, a viewable area can be obtained as described above. As shown in the figure, the visible areas for the buildings B1 and B2 are circular areas with radii r1 and r2. The radii r1 and r2 can be determined according to the heights of the buildings B1 and B2. The approximate existence region can be specified as a portion where the visually recognizable regions of the respective reference buildings overlap, and in the example of the figure, the region is a hatched region.

  The server 200 extracts a candidate building that should become the reference building from the approximate existence area thus identified (step S122). Candidate buildings can be extracted in various ways. For example, a building located near the center of the approximate presence area may be extracted, or a building associated with many attributes may be extracted. In consideration of ease of visual recognition by the user, a building having a height higher than a predetermined level may be extracted.

  The server 200 presents building information for the extracted candidate building, and inputs an answer from the user regarding viewing approval / disapproval (step S123). The example of the interface screen for this is shown in the figure. In this example, an image of a candidate building is displayed on the terminal 100 and a viewing authorization rejection is input. In order to avoid misunderstanding of the user, it is preferable to present the image of the candidate building in this way, but the image is not necessarily required, and information is presented with attributes such as the name, type, floor number, and color of the building. It may be.

  When the user replies viewing authorization to the presented building information (step S124), the server 200 registers this building as a reference building and adds one reference building number (step S125). As a result, if the number of reference buildings is 3 or more (step S120), the reference building setting process is terminated, and if it is less than 3, the processes after step S121 are executed again. When it is executed again, the approximate existence area is set in consideration of the reference building newly registered in step S125 (step S121). Accordingly, the approximate existence region gradually narrows.

On the other hand, when the user replies to the presented building information that the user cannot visually recognize (step S124), the server 200 extends the approximate existence area until the number of trials reaches a preset upper limit (step S126). (Step S127), the processing after Step S122 is executed again. The existing approximate existing area may be expanded with a certain similarity ratio, or may be expanded by adding a band-shaped area having a constant width around the approximate existing area. The reason why the approximate existence area is expanded is that, if it is answered that the candidate building extracted in step S122 cannot be viewed, the approximate existence area set in step S121 may be incorrect.
If the number of trials reaches the upper limit, the reference building setting process is terminated without performing this process.

  In the present embodiment, in the first half of the process (FIG. 6), a process example is shown in which the building in which the user inputs the attribute is registered as the reference building. However, the setting of the reference building is not limited to this method. For example, the building whose attribute is input by the user in the first half of the process (FIG. 6) is used only for the setting of the approximate existence area in FIG. 7, and the reference building used in the existence area defining process (step S150 in FIG. 5) is Only those set in the latter half (FIG. 7) may be used.

E. Existence area definition processing:
FIG. 8 is a flowchart of the existing area defining process. This process corresponds to step S150 in FIG. 5 and corresponds to the function of the boundary demarcation unit 234 (see FIG. 1). In this process, the existence area is defined by the method described with reference to FIGS.

  First, the server 200 inputs the position, shape and height data of the reference building (step S151). In this embodiment, three or more reference buildings are input. As shown in FIG. 4, the existing area is defined by repeatedly executing the boundary defining process (FIG. 4) for two reference buildings.

The server 200 selects two reference buildings to be subjected to boundary demarcation processing from the input reference buildings (step S152). This selection can be made by any method, but in this embodiment, the selection is made based on the following three criteria.
Priority 1: in a positional relationship that intersects a defined boundary;
Priority 2: Two buildings that exist in the presence area;
Priority 3: low building height;

  Consider the case where there are a building A (10th floor), a building B (12th floor), and a building C (14th floor) shown in the figure as reference buildings. It is assumed that the boundary L is already defined by the processing between the buildings A and B, and the region including C is defined as an existing region as illustrated. In this state, the server 200 next selects either one of the buildings A and C or the buildings B and C and executes the boundary defining process. In any of these combinations, since a boundary that intersects the defined boundary L can be defined, either one of the combinations cannot be selected under the above-described conditions of “priority 1” and “priority 2”. Therefore, according to the priority 3 condition, the building having a low height, that is, the building A is preferentially used. As a result, the boundary demarcation process is executed with the buildings A and C as reference buildings. When the boundary demarcation process is executed with the buildings A and C as the reference buildings, the boundary demarcation process is executed with the buildings B and C as the reference buildings if the existence area is not yet demarcated.

  The server 200 presents to the user the attributes of the reference building selected in step S152 and inputs the relative positional relationship between the two (step S153). The example of the interface screen displayed on the terminal 100 for inputting the positional relationship is shown in the figure. In this example, images of two reference buildings are displayed, and the positional relationship between the left and right is input. In the example of the figure, the upper building is seen from the user on the left side of the lower building. In this interface screen, when left or right is selected for an upper or lower building, the input state of the other building is changed in conjunction with the other building. For example, in the state shown in the figure, when the user selects “right” for the upper building, the input state of the lower building is automatically switched to “left”. By doing so, input errors can be reduced.

  In the present embodiment, the left-right relationship is input, but instead of this, the front-rear relationship between buildings may be input. However, when specifying the positional relationship between buildings as viewed from the user, there is an advantage that the left-right relationship can be specified more easily and accurately than the front-rear relationship.

  When the server 200 inputs the left-right relationship of the reference building, the server 200 executes boundary demarcation processing based on the result (step S154). The boundary demarcation process is a process of specifying an existing area that matches the left-right relationship designated by the user by the method shown in FIG.

  The server 200 repeatedly executes the above processing until a predetermined completion condition is satisfied (step S154). In the present embodiment, “the existing region becomes a closed region” is set as the completion condition. In the example of FIG. 3 shown above, when the four boundaries of FIGS. 3B to 3E are defined, the existing region becomes a closed region, and the completion condition is satisfied. If the completion condition is not satisfied, the combination of the reference buildings is changed (step S152), the relative positional relationship input (step S153), and the boundary demarcation (step S154) are executed again. Become. The server 200 outputs the existence area defined by the above processing (step S156) and ends the existence area definition processing.

  The completion condition (step S155) is not limited to the above-described condition, and can be arbitrarily set. For example, in addition to a closed figure, a condition in which the area of the existing region or the radius of the circumscribed circle is a predetermined value or less may be used. Further, when the area of the existing region and the radius of the circumscribed circle are used as the completion conditions, the condition that only the boundary defined by the boundary definition process (step S154) is a closed figure is relaxed, and the outline described in FIG. You may make it use the conditions that a closed figure is comprised with the boundary of an existing area | region. As the completion condition, in addition to the condition based on the shape of the existence area itself, a condition that all reference buildings are used once for the boundary demarcation process may be used. In the case of performing the most strict processing, it may be a condition that the boundary definition processing is completed for all combinations of the reference buildings.

F. Reference building selection:
As described above in step S152 of FIG. 8, in this embodiment, the reference building is selected according to the following three priorities.
Priority 1: in a positional relationship that intersects a defined boundary;
Priority 2: Two buildings that exist in the presence area;
Priority 3: low building height;
Below, the significance of these conditions is demonstrated using a specific example.

  FIG. 9 is an explanatory diagram showing an example of selecting a reference building. The procedure of the boundary demarcation process and the existence area when the buildings A to D exist at the positions illustrated as the reference buildings are illustrated.

  FIG. 9A will be described. When the boundary demarcation process is performed for the first time, since there is no demarcated boundary, there is no building that meets the conditions of priority 1 and priority 2. Therefore, a building with a low height is selected according to the priority 3 condition. In the example of FIG. 9A, one of the 10th floor building A and the 12th floor buildings B and C is selected. Either building B or building C may be selected. In the illustrated example, it is assumed that building B is selected. As a result, a boundary line L [1] passing through the buildings A and B is defined. The hatched side of the boundary line L [1] represents the existence area. The same applies to the following examples. As shown in FIG. 4, the boundary line has a complicated shape between the buildings A and B, but is simplified here and shown as a straight line.

  Next, a building that can demarcate a boundary that intersects the boundary L [1] is selected according to the priority 1 condition. In the example of FIG. 9A, four combinations of a building AC, a building AD, a building BC, and a building BD are possible as combinations that meet this condition. These combinations are two buildings existing in the existing area defined by the boundary line L [1], that is, buildings that define the boundary line in the existing area. There is no superiority or inferiority. However, according to the priority 3 condition, since a combination passing through the building A having a low height is selected, the combination of the buildings AC and AD remains as a candidate. Furthermore, the building C having a low height is selected from the buildings C and D according to the condition of the priority 3. As a result, the boundary L [2] is defined using the building AC.

  Next, again, according to the priority 1 condition, first, a building that can demarcate a boundary that intersects the boundary L [2] is selected. In this example, four conditions of building BD, building AD, building CD, and building BC satisfy this condition. However, only the building BD can delimit completely intersecting boundaries. This is because the combinations of the buildings AD and CD use the buildings on the boundary L [2]. As described above, in the present embodiment, when the priority 1 condition is taken into consideration, a building in a positional relationship that intersects the defined boundary “completely” is used with higher priority. As a result, in the example of FIG. 9A, the building BD is selected and the boundary L [3] is defined.

  As described above, in the case of FIG. 9A, the existence area APa is efficiently defined by selecting the reference building according to the priorities 1 to 3 of the present embodiment and performing the boundary line defining process three times. Is possible.

  FIG. 9B will be described. When processing is performed for the first time, since there are no buildings corresponding to the priorities 1 and 2, the building AC is selected according to the priority 3, and the boundary L [1] is defined. In the next process, a building BD having a positional relationship that completely intersects the boundary L [1] is selected, and the boundary L [2] is defined. In the third processing, since there is no building having a positional relationship that completely intersects the boundaries L [1] and L [2], the priority 2 condition is effective, the building AB is selected, and the boundary L [3 ] Is defined. As a result, even in FIG. 9B, the existence area APb can be defined by three processes.

  FIG. 9C will be described. When processing is performed for the first time, since there are no buildings corresponding to the priorities 1 and 2, the building AC is selected according to the priority 3, and the boundary L [1] is defined. In the next process, a building BD having a positional relationship that completely intersects the boundary L [1] is selected, and the boundary L [2] is defined. In the third processing, since there is no building having a positional relationship that completely intersects the boundaries L [1] and L [2], the priority 2 condition is effective, the building AD is selected, and the boundary L [3 ] Is defined. The reason why the selected building is different from the case of FIG. 9B is that the existing areas that are narrowed down when the second processing is completed are different. In the case of FIG. 9C as well, the existence area APc can be defined by three processes.

  When there are four reference buildings, if the boundary line defining process is executed for all combinations, six processes are required. On the other hand, if the priority of the present embodiment is used, as shown in FIG. 9, the existence area can be defined by three processes. Although FIG. 9 shows an example in which four reference buildings exist, even when there are five or more reference buildings, the number of processes for defining the existence area can be reduced in the same manner, and the existence is efficient. Regions can be defined.

  According to the route guidance system of the present embodiment described above, the current position estimation function is provided, so that the current position can be estimated with sufficient accuracy even when the current position of the user is unknown. As a result, the adverse effect that the route search and route guidance cannot be used due to the unknown current position can be alleviated, and the convenience of the route search and route guidance can be improved. In the above-described embodiment, an example in which the terminal 100 having the GPS 140 is used has been described. However, this embodiment can also be used in the terminal 100 that does not include the GPS 140.

G. Variation:
In the embodiment, as shown in FIGS. 8 and 9, an example in which the existence area is obtained as a plane has been shown. The existence area is not necessarily obtained as a plane. For example, a network area as an aggregate of nodes and links included in network data for pedestrians may be used as the existence area. In the modification, a process for obtaining such a net-like existence area will be described.

  FIG. 10 is a flowchart of a presence area defining process as a modification. This is an alternative to the processing shown in FIG. 8 of the embodiment. As in the embodiment, the server 200 inputs the reference building position, shape, and height data (step S151), selects the reference building (step S152), inputs the relative positional relationship between the buildings (step S153), and defines the boundary (step S153). S154) is executed. Each processing is the same as in the embodiment (FIG. 8).

  Next, the server 200 demarcates a range in which the reference building can be visually recognized (step S154a), and extracts nodes and links existing in the area surrounded by the boundary set in step S154 and the visible range (step S154). S154b). Examples of these processes are shown on the right side of the figure. The network data stores nodes and links corresponding to each path on the map shown in the figure. It is assumed that the boundary Lab is obtained using the reference buildings Ba and Bb in the processing up to step S154. The existence area is assumed to be the hatched side in the drawing, that is, the lower side of the boundary Lab.

  Circles Ca and Cb in the figure are circles centered on the reference buildings Ba and Bb, respectively, and represent visible ranges. The radius of the circle can be set according to the height of the reference buildings Ba and Bb. In the process of step S154b, nodes and links Lnk in the area surrounded by the boundary Lab, circles Ca and Cb are extracted. In the figure, the extracted nodes and links are indicated by bold lines. In the modification, not the hatched area in the figure but the aggregate of extracted nodes and links becomes the existence area. In the modification, nodes and links are extracted in consideration of the visible range (circles Ca and Cb in the figure are taken into account), but nodes and links may be extracted in consideration of only the boundary Lab.

  The server 200 repeatedly executes the above processing until a predetermined completion condition is satisfied (step S155A). As the completion condition, for example, a condition that the total number of extracted nodes and links is a predetermined value or less can be used. Also in the example of the figure, if more boundaries are obtained in addition to the boundary Lab by executing the repetitive processing, the area surrounded by the boundary and the reference building can be narrowed, and the extracted nodes, The total number of links is also reduced. Then, it is determined that the completion condition is satisfied when the total number is equal to or less than a predetermined value. After completion, the server 200 outputs the obtained existence area (step S156) and ends this process.

  After the above-described existence area defining process is completed, the processes after step S182 in FIG. 5 are performed as in the embodiment. However, in the modification, since the existence area itself is configured by a sidewalk link, the process of step S182 can be omitted. In the embodiment, there is a possibility that a very large number of sidewalk links may be extracted by executing the process of step S182, whereas in the modification, the sidewalk links are sufficiently narrowed down at the time of the existing area defining process. Is possible. Therefore, in the modified example, there is an advantage that the current position can be efficiently identified (steps S184 to S188) and the current position can be obtained with high accuracy.

  As mentioned above, although the various Example of this invention was described, it cannot be overemphasized that this invention is not limited to these Examples, and can take a various structure in the range which does not deviate from the meaning. In the embodiment, the case where the current position estimation function is provided as one function of the route guidance process and the route search process is illustrated. However, the current position estimation function of the present embodiment can be used for other purposes. For example, you may implement | achieve as a present position estimation function in the service which searches information, such as a surrounding restaurant and a shop. Regardless of these services, it can also be configured as a unique device for estimating the current position.

It is explanatory drawing which shows the structure of the route guidance system as an Example. It is a flowchart of a route search process and a route guidance process. It is explanatory drawing which shows the example of a present position estimation process. It is explanatory drawing which shows the process example which demarcates the boundary of an existing area. It is a flowchart of a current position estimation process. It is a flowchart of a standard building setting process (first half). It is a flowchart of a standard building setting process (second half). It is a flowchart of a presence area demarcation process. It is explanatory drawing which shows the example of selection of a reference | standard building. It is a flowchart of a presence area demarcation process as a modification.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Terminal 110 ... Main control part 120 ... Communication part 130 ... Command input part 140 ... GPS
DESCRIPTION OF SYMBOLS 150 ... Display control part 200 ... Server 210 ... Communication part 220 ... Route search part 230 ... Estimation engine 232 ... Reference | standard building setting part 234 ... Boundary demarcation part 240 ... Route guidance part 250 ... Map DB
252 ... Network data 254 ... Drawing data 260 ... Route guidance DB

Claims (16)

A current position estimation device for estimating a user's current position,
A positional relationship input unit for inputting a result visually recognized by the user with respect to a relative positional relationship between a plurality of features;
A map database reference section that references a map database that records the location of each feature;
A current position estimation apparatus comprising: an estimation engine that estimates a current position of the user by specifying a range that can be visually recognized in the relative positional relationship with respect to the plurality of features with reference to the map database. A current position estimating apparatus according to claim 1, wherein
An information presenting unit configured to set information on a feature to be input for the relative positional relationship and to present information for the user to specify the feature locally;
The positional relationship input unit is a current position estimation device that inputs the relative positional relationship with respect to the feature to be input.   The current position estimation device according to claim 1, wherein the relative positional relationship is a front-rear relationship or a left-right relationship between the plurality of features when viewed from the user.   The current position estimation apparatus according to any one of claims 1 to 3, wherein the plurality of features are three or more features that are not on a straight line. A current position estimation device according to any one of claims 1 to 3,
The plurality of features are four or more features;
The estimation engine is
A selection unit that selects two features to be processed among the plurality of features;
A boundary demarcating section that demarcates a boundary line of an area where the current position of the user may exist based on the relative positional relationship between the selected features;
The selection unit is a current position estimation device that preferentially selects a feature capable of demarcating a boundary that intersects the demarcated boundary as the processing target when a demarcated boundary exists. A current position estimation device according to any one of claims 1 to 3,
The plurality of features are four or more features;
The estimation engine is
A selection unit that selects two features to be processed among the plurality of features;
A boundary demarcating section that demarcates a boundary line of an area where the current position of the user may exist based on the relative positional relationship between the selected features;
The selection unit is a current position estimation device that performs the estimation by preferentially using two features existing in an area where the current position may exist when a boundary that has already been defined exists. A current position estimation device according to any one of claims 1 to 3,
The plurality of features are three or more features;
The estimation engine is
A selection unit that selects two features to be processed among the plurality of features;
A boundary demarcating section that demarcates a boundary line of an area where the current position of the user may exist based on the relative positional relationship between the selected features;
The selection unit is a current position estimation device that performs the estimation by preferentially using a feature having a low height among the plurality of features. The current position estimation device according to any one of claims 2 to 7,
The map database is a database that further stores data relating to attributes that can be used to identify features in association with each feature,
Prior to the information presentation by the information presenting unit, an attribute input unit that inputs the attribute of the feature arbitrarily selected within a range visible to the user according to the operation of the user;
A feature corresponding to the attribute is identified with reference to the map database, a general region where the current position is determined based on the feature is obtained, and a plurality of features in the general region are determined as the information. A current position estimation device having a reference feature setting unit that is set as an object to be presented. A current position estimating apparatus according to claim 8, wherein
The reference feature setting unit, when there are a plurality of feature candidates corresponding to a specific attribute, based on the distance between each candidate and a feature corresponding to another attribute input separately, A current position estimation apparatus including a candidate feature selection unit that selects any one as a feature corresponding to the specific attribute. The current position estimation device according to claim 8 or 9,
The reference feature setting unit further presents information on the predetermined feature to the user, inputs an answer as to whether or not the user can visually recognize the feature, and reflects the answer to the outline area. A current position estimation device for obtaining. A current position estimating apparatus according to claim 10, wherein
The reference feature setting unit is configured to repeatedly execute presentation of information about the feature and input of a response under a predetermined condition,
If an answer indicating that a feature can be visually recognized is obtained, the outline area is narrowed based on the position of the feature, and then the next area where information is to be presented from the outline area. Current position estimation device for selecting an object. A current position estimation device according to claim 10 or 11,
The reference feature setting unit is configured to repeatedly execute presentation of information about the feature and input of a response under a predetermined condition,
When an answer indicating that the feature cannot be visually recognized is obtained, the outline area is expanded based on the position of the feature, and then the next area where information is to be presented from the outline area. Current position estimation device for selecting an object. A current position estimation device according to any one of claims 1 to 12,
The estimation engine is
A selection unit that selects two features to be processed among the plurality of features;
A boundary demarcating section that demarcates a boundary line of an area where the current position of the user may exist based on the relative positional relationship between the selected features;
The boundary delimiter is
Obtaining a convex polygon enveloping the plurality of features;
Of the sides constituting the convex polygon, select a side connecting the two features,
Extend the selected side,
A current position estimating device that demarcates the boundary line by selecting a region that does not violate the relative positional relationship for each region having the extended side as a boundary. A current position estimation apparatus according to any one of claims 1 to 13,
A network data reference section for referring to network data in which roads that a user can pass are represented by nodes and links;
The estimation engine estimates a current position as a current position on a node or a link existing within a range where the plurality of features are visible in the relative positional relationship in the network data. apparatus. A current position estimation method for estimating a user's current position by a computer,
As a process executed by the computer,
A positional relationship input step for inputting a result visually recognized by the user with respect to a relative positional relationship between a plurality of features;
A map database reference process that references a map database that records the location of each feature;
A current position estimation method comprising: an estimation step of estimating a current position of the user by specifying a range that can be visually recognized by the relative positional relationship with respect to the plurality of features with reference to the map database. A computer program for estimating the current position of a user by a computer,
For a plurality of features, an information presentation subprogram that presents information that can identify each feature on-site to the user;
A positional relationship input subprogram for inputting a result visually recognized by the user with respect to a relative positional relationship between the plurality of features;
A map database reference subprogram that references a map database that records the location of each feature;
A computer program comprising: an estimation subprogram that estimates a current position of the user by specifying a range that can be visually recognized by the relative positional relationship for the plurality of features with reference to the map database.