JP2012122760A - Own-vehicle position recognition system, own-vehicle position recognition program and own-vehicle position recognition method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize an own-vehicle position recognition system and an own-vehicle position recognition program capable of adequately correcting an estimated own-vehicle position even when a progressing direction of the vehicle is different from an extension direction of a road.SOLUTION: The own-vehicle position recognition system includes: an image recognition section for processing the image recognition of a target feature M containing in a photographic image by making the feature on a road surface arranged so that a direction parallel to an extension direction ED of the road is a reference direction SD, to be a target feature M; a reference direction specifying section for specifying the reference direction SD of the target feature M based on a relative direction RD between the target feature M recognized by the image recognition section and an own vehicle 100, and a progressing direction MD of the own vehicle 100; and a feature utilizing own-vehicle position correction section for detecting a feature direction coincident position P0 which is a position on the road coinciding the reference direction SD specified by the reference direction specifying section and the extension direction ED based on road information around an estimation own-vehicle position EP to correct the estimation own-vehicle position EP based on the feature direction coincident position P0.

Description

  The present invention includes an estimated own vehicle position acquisition unit that acquires an estimated own vehicle position that is an estimated position of the own vehicle, a traveling direction acquisition unit that acquires a traveling direction of the own vehicle, and a road that includes information related to a road extending direction A vehicle position recognition system including a road information acquisition unit that acquires information and a road surface image acquisition unit that acquires a captured image of the road surface, an estimated vehicle position that acquires an estimated vehicle position that is an estimated position of the vehicle An acquisition function, a traveling direction acquisition function for acquiring the traveling direction of the host vehicle, a road information acquisition function for acquiring road information including information related to the extension direction of the road, and a road surface image acquisition function for acquiring a shot image of the road surface , A vehicle position recognition program for realizing the above, an estimated vehicle position acquisition step for acquiring an estimated vehicle position that is an estimated position of the vehicle, and a traveling direction acquisition step for acquiring a traveling direction of the vehicle, Road A road information obtaining step of obtaining road information including information on orientation, and the road surface image acquiring step of acquiring the captured image of the road, to the vehicle position recognition method comprising a.

  The own vehicle position recognition system for recognizing the position of the own vehicle is provided, for example, in a navigation device for a vehicle. Patent Document 1 below describes a navigation device having a function of correcting an estimated own vehicle position by performing a map matching process and matching a traveling position of the own vehicle on a road of map data. By the way, as described in paragraphs 0015 and 0016 of Patent Document 1, when the traveling direction of the vehicle is different from the extending direction of the road (the link included in the road information), if the map matching process is performed, the own vehicle position May shift. Therefore, in the configuration described in Patent Document 1, when the traveling direction of the vehicle may be different from the extending direction of the road (for example, when entering a roundabout), the configuration of stopping the map matching process is stopped. It is said.

  However, when the execution of the map matching process is stopped in this manner, the estimated own vehicle position cannot be corrected. Therefore, even if the own vehicle is off the road of the map data or on the road, May be displayed at different positions, and route guidance may not be performed properly. Therefore, with the configuration described in Patent Document 1, the convenience of the user of the navigation device may be reduced.

  Patent Document 2 below describes a technique related to the correction of the estimated own vehicle position when the vehicle is traveling on a ring road. However, Patent Document 2 does not mention that the traveling direction of the vehicle may be different from the extending direction of the road, and naturally, means for solving the above-described problem of convenience is shown. Not.

JP 2004-132922 A JP 2006-292928 A

  Therefore, it is desired to realize a host vehicle position recognition system and a host vehicle position recognition program capable of appropriately correcting the estimated host vehicle position even when the traveling direction of the vehicle is different from the extending direction of the road.

  According to the present invention, including an estimated own vehicle position acquisition unit that acquires an estimated own vehicle position that is an estimated position of the own vehicle, a traveling direction acquisition unit that acquires a traveling direction of the own vehicle, and information on an extension direction of the road The characteristic configuration of the vehicle position recognition system including a road information acquisition unit that acquires road information and a road surface image acquisition unit that acquires a captured image of the road surface is based on an orientation parallel to the extension direction of the road as a reference orientation. An image recognition unit for performing image recognition processing of the target feature included in the captured image, using the feature on the provided road surface as the target feature, and the target feature and the vehicle recognized by the image recognition unit Based on the relative azimuth of the vehicle and the traveling azimuth of the host vehicle, the reference azimuth specifying unit that specifies the reference azimuth of the target feature, and the reference azimuth specifying based on the road information around the estimated host vehicle position. The reference orientation specified by the part A feature use own vehicle position correcting unit that detects a feature direction coincidence position that is a position on the road having the same extension direction and corrects the estimated own vehicle position based on the feature direction coincidence position; In the point.

  In the present application, “coincidence” with respect to the azimuths is not limited to the case where they completely coincide, but the two azimuths are within the range corresponding to the respective derivation methods (specific methods) of the two azimuths to be compared. It is used as a concept including the case where the

  According to the above characteristic configuration, the reference azimuth of the target feature specified by the reference azimuth specifying unit matches the extended azimuth of the road at the point where the target feature is provided. Therefore, even if the traveling direction of the host vehicle differs from the extending direction of the road due to lane changes etc., the road extension in the vicinity where the vehicle is located by specifying the reference direction of the target feature The direction can be specified, and the estimated own vehicle position can be appropriately corrected based on the feature direction coincidence position detected based on the extended direction and the map information.

  Here, the vehicle further includes a normal vehicle position correction unit that corrects the estimated vehicle position based on an extension direction of the road included in the road information around the estimated vehicle position and a traveling direction of the vehicle, It is preferable that the estimated vehicle position is corrected by the feature-use vehicle position correction unit when the estimated vehicle position is within a roundabout that is an intersection having a ring road.

  According to this configuration, the estimated own vehicle position can be appropriately corrected at a roundabout where the possibility that the traveling direction of the own vehicle is different from the extending direction of the road may be higher than that of other roads.

  In addition, it is preferable that a feature on a road surface having a shape having no rotational symmetry of less than 180 degrees with an axis orthogonal to the road surface as a center is the target feature.

  A feature on a road surface having a shape having at least one of a linear component parallel to the road extending direction and a linear component orthogonal to the road extending direction is also suitable as the target feature. is there.

  According to these configurations, since it becomes relatively easy to detect the reference orientation of the target feature by image recognition, the calculation load of the image recognition processing can be reduced, and the reference orientation of the target feature can be further increased. It becomes possible to detect accurately.

  In addition, it is preferable that the feature-use own vehicle position correcting unit corrects the estimated own vehicle position based on a distance between the target feature recognized by the image recognition unit and the own vehicle.

  According to this configuration, even if the time required for the correction process of the estimated host vehicle position or the distance between the host vehicle and the target feature at the time of shooting the captured image of the road surface is long, the estimated host vehicle position is It becomes possible to correct appropriately.

  The technical features of the vehicle position recognition system according to the present invention having the above-described configurations can also be applied to a vehicle position recognition program and a vehicle position recognition method. Therefore, the present invention provides such a program. And methods can also be subject to rights.

  In that case, a road including an estimated own vehicle position obtaining function for obtaining an estimated own vehicle position that is an estimated position of the own vehicle, a traveling direction obtaining function for obtaining a traveling direction of the own vehicle, and information on an extension direction of the road The feature configuration of the vehicle position recognition program for realizing the road information acquisition function for acquiring information and the road surface image acquisition function for acquiring a captured image of the road surface is based on the direction parallel to the road extension direction as the reference direction. An image recognition function for performing image recognition processing of the target feature included in the captured image, and a target feature recognized by the image recognition function as a target feature on a road surface provided as a target feature. Based on the reference direction specifying function for specifying the reference direction of the target feature based on the relative direction with respect to the vehicle and the traveling direction of the host vehicle, and based on the road information around the estimated host vehicle position, the reference direction specific The feature use vehicle that detects the feature orientation matching position that is the position on the road where the extended orientation matches the reference orientation specified by the performance, and corrects the estimated vehicle position based on the feature orientation matching position. The vehicle position correction function is realized by a computer.

  Further, road information including an estimated own vehicle position acquisition step for acquiring an estimated own vehicle position that is an estimated position of the own vehicle, a traveling direction acquisition step for acquiring a traveling direction of the own vehicle, and road information including information on the extending direction of the road The characteristic configuration of the vehicle position recognition method including a road information acquisition step to acquire and a road surface image acquisition step to acquire a captured image of the road surface is provided with a direction parallel to the extension direction of the road as a reference direction. An image recognition step for performing image recognition processing of the target feature included in the captured image, using the feature on the road surface as the target feature, and the target feature and the vehicle recognized in the image recognition step A reference direction specifying step for specifying the reference direction of the target feature based on a relative direction and a traveling direction of the host vehicle, and the reference direction specifying step based on the road information around the estimated host vehicle position. A feature direction matching position detection step for detecting a feature direction matching position that is a position on a road where the extension direction matches the reference direction specified in the step, and the feature direction matching position detection step And a feature utilizing own vehicle position correcting step for correcting the estimated own vehicle position based on the detected feature direction coincidence position.

  Naturally, these vehicle position recognition programs and vehicle position recognition methods can also obtain the effects of the vehicle position recognition system described above, and some additional examples given as examples of suitable configurations thereof. It is possible to incorporate technology.

It is a schematic diagram which shows the concept of the own vehicle position recognition system which concerns on embodiment of this invention. It is a mimetic diagram showing a schematic structure of a navigation device concerning an embodiment of the present invention. It is a schematic diagram which shows the example of the roundabout which concerns on embodiment of this invention. It is a schematic diagram which shows the example of the road information which concerns on embodiment of this invention. It is a schematic diagram which shows the example of the target terrestrial feature which concerns on embodiment of this invention. It is a schematic diagram which shows the example of the target terrestrial feature which concerns on embodiment of this invention. It is a flowchart which shows the procedure of the estimation own vehicle position correction process which concerns on embodiment of this invention. It is a flowchart which shows the procedure of the feature utilization own vehicle position correction process which concerns on embodiment of this invention.

  Embodiments of the present invention will be described with reference to the drawings. Here, the case where the vehicle position recognition system according to the present invention is applied to an in-vehicle navigation device will be described as an example. The navigation apparatus 1 according to the present embodiment includes a host vehicle position recognition unit 3 (see FIG. 2), and a host vehicle position recognition system is constructed with the host vehicle position recognition unit 3 as a core. Hereinafter, regarding the configuration of the vehicle position recognition system (navigation device 1) according to the present embodiment, “schematic configuration of the vehicle position recognition system”, “schematic configuration of the navigation device”, “configuration of the vehicle position recognition unit”, This will be described in detail in the order of “operation processing procedure”.

1. First, a schematic configuration of a vehicle position recognition system according to the present embodiment will be described with reference to FIG. In FIG. 1, it is assumed that the vehicle (hereinafter sometimes referred to as “own vehicle”) 100 is changing lanes. The vehicle 100 includes a navigation device 1 (see FIG. 2) having a vehicle position recognition unit 3. As will be described later, the navigation device 1 has a function of calculating an estimated own vehicle position EP that is an estimated position of the own vehicle 100. The vehicle 100 is provided with a camera 16 (see FIG. 2). In this example, the camera 16 is a back camera (rear camera) that captures the road surface behind the vehicle 100.

  The own vehicle position recognizing unit 3 is configured so that the camera 16 can be used when the function of correcting the estimated own vehicle position EP based on the direction of the feature on the road surface (feature using own vehicle position correcting process) is set to be effective. An image of the target feature included in the photographed image is obtained by acquiring a photographed image of the photographed road surface and using a feature on the road surface provided with the direction parallel to the road extending direction ED as the reference direction SD as the target feature. Recognition processing is performed (step # 01). In the present embodiment, the road marking M is the target feature. Further, in the present embodiment, when the estimated own vehicle position EP is in the roundabout 5 (an example is shown in FIG. 3), the feature-use own vehicle position correction processing function is set to be effective. ing.

  And when the image recognition of the road marking M is successful, the own vehicle position recognition unit 3 determines the relative orientation RD between the road marking M acquired by the image recognition and the own vehicle 100 and the traveling direction MD of the own vehicle. Based on this, the reference direction SD of the road marking M is specified (step # 02). Since the reference direction SD of the road sign M identified here matches the extended direction ED of the road at the point where the road sign M is provided, the vicinity where the vehicle 100 is located can be determined by specifying the reference direction SD. It is possible to specify the extending direction ED of the road (the portion on the road where the vehicle 100 is located). If the vehicle position recognition unit 3 succeeds in recognizing the image of the road marking M, the vehicle position recognition unit 3 acquires road information (road map data) RI around the estimated vehicle position EP from the road map database 12 (step #). 03).

  The own vehicle position recognition unit 3 is based on the road information RI around the estimated own car position EP and is on the road (on the road map data) on which the reference direction SD and the extended direction ED identified as described above coincide. The feature orientation matching position P0, which is the position, is detected (step # 04). Then, the vehicle position recognition unit 3 corrects the estimated vehicle position EP based on the feature direction matching position P0 obtained as described above (step # 05). In this example, the own vehicle position recognition unit 3 corrects the estimated own vehicle position EP by replacing the feature direction matching position P0 with the estimated own vehicle position EP. In the present invention, the estimated vehicle position EP is corrected as described above, so that the traveling direction MD of the vehicle 100 differs from the road extending direction ED due to, for example, a lane change. However, the extended direction ED of the road in the vicinity where the vehicle 100 is located can be specified based on the direction of the target feature (the road marking M in this example), and the estimated own vehicle position based on the extended direction ED It is possible to correct EP appropriately.

  On the other hand, if the vehicle position recognition unit 3 does not execute the feature-use vehicle position correction process (in this example, the estimated vehicle position EP is not within the roundabout 5), the vehicle position recognition unit 3 A correction process (normal vehicle position correction process) of the estimated host vehicle position EP that is not based on the direction is executed.

2. Schematic Configuration of Navigation Device Next, a schematic configuration of the navigation device 1 that realizes the vehicle position recognition system described above will be described with reference to FIG. As shown in FIG. 2, the navigation device 1 includes a navigation control unit 2, a vehicle position recognition unit 3, an estimated vehicle position calculation unit 11, and a road map database 12. The navigation control unit 2 and the own vehicle position recognition unit 3 include a plurality of functional units described later.

  Each functional unit included in the navigation device 1 is a hardware unit or a functional unit for performing various processes on input data, with a common processing unit such as a CPU as a core member. It is implemented by software (program) or both. Each of these functional units is configured to be able to exchange information with each other via a communication line such as a digital transfer bus, and can extract data (road information RI) from the road map database 12. It is configured. Here, when each function part is comprised with software (program), the said software is memorize | stored in memory | storage means, such as RAM and ROM which an arithmetic processing unit can refer.

  The road map database 12 is configured by a storage device. This storage device includes, as a hardware configuration, a recording medium that can store and rewrite information, such as a hard disk drive and a flash memory.

  The navigation control unit 2 includes a route setting unit 21, a route search unit 22, and a route guide unit 23. The route setting unit 21 is a functional unit that sets, for example, a departure place such as the vehicle position, an input destination, a passing point, and traveling conditions (whether or not an expressway is used). The route search unit 22 is a functional unit that performs calculation processing (route search processing) for searching for a guidance route from the departure point to the destination based on the conditions set by the route setting unit 21. The route guidance unit 23 operates according to the route from the departure point to the destination searched by the route search unit 22 by guidance display on a display screen of a monitor (not shown) or voice guidance by a speaker (not shown). It is a functional part which performs the calculation process (route guidance process) for performing suitable route guidance with respect to a person.

  The road map database 12 stores road information (road map data) RI that is referred to when the above route search processing and route guidance processing are executed, when map display processing on a monitor is executed, and the like. The road information RI includes information related to the road extending direction ED. In this example, the road information RI is information that expresses the road by the node N and the link K as shown in FIG. The node N is set at, for example, a road intersection, a turning point, a branch point, or the like. FIG. 4 conceptually shows road information RI corresponding to the roundabout 5 shown in FIG. 3 as an example.

  As a supplementary explanation, the road information RI includes information on a large number of nodes N having position information on a map expressed by latitude and longitude, and a large number of links that connect the two nodes N to form a route corresponding to the road. K information. That is, in this example, the information on the link K represents information on the road extending direction ED. The road information RI includes information such as road type (type of highway, toll road, national road, prefectural road, etc.) and link length as link information (link attribute) of each link K. In the present embodiment, the link information includes information on whether or not the link K is a circular intersection link that forms the circular intersection 5. The road information RI further includes information on a shape interpolation point S that is arranged between two nodes N (on the link K) and represents the detailed shape of the link K, and information on the road width.

  The road information RI as described above is made into a database and stored in a form according to the specifications of the road map database 12. For example, the road information RI may be divided into a plurality of layers associated with each other and stored in the road map database 12. Specifically, a layer (road network layer) indicating connection information between roads by the node N and the link K, a layer (road shape layer) indicating information on the shape interpolation point S and road width, and detailed information on the road The road information RI can be divided into at least three layers including a layer (road attribute layer) indicating (information on various features provided on or around the road).

  The estimated own vehicle position calculation unit 11 calculates the estimated own vehicle position EP, which is the estimated position of the own vehicle 100, based on the outputs of the GPS measurement unit 13, the distance sensor 14, and the direction sensor 15 provided in the vehicle 100. Part. The GPS measurement unit 13 receives a GPS signal from a GPS satellite and outputs the GPS signal to the estimated own vehicle position calculation unit 11. Further, the distance sensor 14 detects the vehicle speed and the moving distance of the vehicle 100 and outputs information on the vehicle speed and the moving distance as a detection result to the estimated own vehicle position calculating unit 11. The azimuth sensor 15 is constituted by, for example, a gyro sensor, a geomagnetic sensor (electronic compass), and the like, and outputs azimuth information as a detection result to the estimated vehicle position calculation unit 11.

  The estimated vehicle position calculation unit 11 analyzes the GPS signal, calculates the current position (latitude and longitude) of the vehicle 100, and obtains GPS position data. Further, the estimated host vehicle position calculation unit 11 calculates the dead reckoning position based on the moving distance information and the direction information sent every moment, and obtains dead reckoning position data. Then, the estimated own vehicle position calculation unit 11 performs an operation for calculating the estimated own vehicle position EP that is the estimated position of the vehicle 100 from the GPS position data and the dead reckoning position data by a known method. The estimated vehicle position calculation unit 11 may be configured to calculate the estimated vehicle position EP using only one of GPS position data and dead reckoning position data.

  The vehicle position recognition unit 3 includes an estimated vehicle position correction unit 40 for correcting the estimated vehicle position EP calculated by the estimated vehicle position calculation unit 11, which will be described in detail later. The navigation apparatus 1 according to the present embodiment includes the own vehicle position recognition unit 3 to appropriately correct the estimated own vehicle position EP that may include a measurement error so that the estimated own vehicle position EP deviates from the road. Even in such a case, the vehicle position can be appropriately corrected (map matching) on the road shown on the road map.

3. Next, the configuration of the vehicle position recognition unit 3 that is a main part of the present invention will be described. As shown in FIG. 2, the vehicle position recognition unit 3 includes an estimated vehicle position acquisition unit 31, a traveling direction acquisition unit 32, a road surface image acquisition unit 33, a road information acquisition unit 34, and an estimated vehicle position correction unit 40. I have.

3-1. Configuration of Estimated Own Vehicle Position Acquisition Unit The estimated own vehicle position acquisition unit 31 is a functional unit that acquires an estimated own vehicle position EP that is an estimated position of the own vehicle 100. In the present embodiment, the estimated own vehicle position acquisition unit 31 acquires the estimated own vehicle position EP by inputting the estimated own vehicle position EP calculated by the estimated own vehicle position calculation unit 11 to the estimated own vehicle position acquisition unit 31. Is configured to do.

  Information on the process of deriving the estimated vehicle position EP is input to the estimated vehicle position acquisition unit 31, and the estimated vehicle position acquisition unit 31 derives the estimated vehicle position EP based on the information, and the estimated vehicle position EP It can also be set as the structure which acquires. For example, the GPS position data and dead reckoning position data can be input to the estimated vehicle position acquisition unit 31. The estimated vehicle position acquisition unit 31 may be configured to receive detection results of the GPS measurement unit 13, the distance sensor 14, and the direction sensor 15.

3-2. Configuration of Traveling Direction Acquisition Unit The traveling direction acquisition unit 32 is a functional unit that acquires the traveling direction MD of the host vehicle 100. Specifically, the traveling direction acquisition unit 32 acquires the traveling direction MD of the host vehicle 100 in the absolute direction. In this example, the azimuth sensor 15 is a geomagnetic sensor (electronic compass) capable of detecting an absolute azimuth, and the detection result (absolute angle) of the azimuth sensor 15 is input to the traveling azimuth acquisition unit 32. Then, the traveling direction acquisition unit 32 acquires the traveling direction MD by using the detection result input from the direction sensor 15 as the traveling direction MD of the host vehicle 100. “Absolute orientation” means an orientation that is uniquely determined based on a predetermined orientation (for example, the direction of geomagnetism) in an absolute coordinate system (for example, a geographic coordinate system). As described above, in the present embodiment, the road information RI includes information on the node N expressed by latitude and longitude. Therefore, in the present embodiment, the information related to the extended direction ED of the road included in the road information RI is information related to the extended direction ED in the absolute direction.

  When the azimuth sensor 15 is a gyro sensor that detects a relative angle, the traveling azimuth acquisition unit 32 determines the current traveling azimuth based on the detection result (relative angle) of the azimuth sensor 15 and the previous traveling azimuth MD. It can be set as the structure which derives and acquires MD. Further, when the direction sensor 15 is a gyro sensor that detects a relative angle, the estimated vehicle position calculation unit 11 derives the current traveling direction MD based on the detection result of the direction sensor 15 and the previous traveling direction MD. The traveling direction acquisition unit 32 may acquire the derived traveling direction MD.

3-3. Configuration of Road Surface Image Acquisition Unit The road surface image acquisition unit 33 is a functional unit that acquires a captured image of a road surface. As described above, the vehicle 100 is equipped with the camera 16 for photographing the road surface, and the road surface image acquisition unit 33 acquires a captured image of the road surface photographed by the camera 16. In this example, the camera 16 is a back camera, and the road surface image acquisition unit 33 acquires a captured image of the road surface behind the vehicle 100.

  The captured image acquired by the road surface image acquisition unit 33 is used for the feature-use vehicle position correction process by the feature-use vehicle position correction unit 42 described later. Thus, in the present embodiment, the road surface image acquisition unit 33 acquires a captured image of the road surface at a predetermined timing when the feature-use vehicle position correction processing function is set to be effective. This predetermined timing can be set, for example, every predetermined time interval (for example, every second) or every time the vehicle 100 travels a predetermined distance (for example, 1 meter). The captured image acquired by the road surface image acquisition unit 33 is output to the image recognition unit 43 provided in the feature-use vehicle position correction unit 42.

3-4. Configuration of Road Information Acquisition Unit The road information acquisition unit 34 is a functional unit that acquires road information RI including information related to the road extending direction ED. As described above, in this embodiment, the road map database 12 storing the road information RI is provided in the navigation device 1, and the road information acquisition unit 34 extracts and acquires the road information RI from the road map database 12. .

  The road information RI acquired by the road information acquisition unit 34 is used for an estimated vehicle position correction process by an estimated vehicle position correction unit 40 (a normal vehicle position correction unit 41 and a feature-use vehicle position correction unit 42) described later. It is done. Therefore, in this embodiment, the road information acquisition unit 34 acquires the road information RI from the road map database 12 in response to a request from the normal vehicle position correction unit 41 or the feature-use vehicle position correction unit 42.

3-5. Configuration of Estimated Vehicle Position Correction Unit The estimated vehicle position correction unit 40 is a functional unit that corrects the estimated vehicle position EP acquired by the estimated vehicle position acquisition unit 31, that is, a function that executes estimated vehicle position correction processing. Part. The estimated own vehicle position correction unit 40 includes a normal own vehicle position correction unit 41 and a feature-use own vehicle position correction unit 42. In the present embodiment, the estimated own vehicle position correcting unit 40 determines the feature when the estimated own vehicle position EP is within the circular intersection 5 (see FIG. 3) that is an intersection having the circular road (circumferential road) 51. The function of the use vehicle position correction process is set to be effective, and the feature use vehicle position correction unit 42 corrects the estimated vehicle position EP. In addition, when the estimated vehicle position EP is not within the roundabout 5, the estimated vehicle position correction unit 40 sets the normal vehicle position correction processing function to be effective, and the normal vehicle position correction unit 41 performs estimation. The own vehicle position EP is corrected.

  In the present embodiment, as described above, the link information (link attribute) includes information on whether or not the link K is a roundabout link. Therefore, the estimated vehicle position correction unit 40 determines whether or not the estimated vehicle position EP is within the roundabout 5 based on the link information. Specifically, the estimated vehicle position EP is determined to be within the circular intersection 5 when the estimated vehicle position EP enters the roundabout link or the coordinate range in the vicinity of the roundabout link. can do. In addition, when the coordinate range including the circular road 51 included in the roundabout 5 is determined in advance for each roundabout 5, and the estimated own vehicle position EP is within the coordinate range, the estimated own vehicle position EP is within the roundabout 5 It can also be set as the structure determined to exist in.

3-5-1. Configuration of a normal vehicle position correction unit The normal vehicle position correction unit 41 is based on the road extending direction ED and the traveling direction MD of the vehicle 100 included in the road information RI around the estimated vehicle position EP. This is a functional unit that corrects the vehicle position EP, that is, a functional unit that executes normal vehicle position correction processing. At this time, the normal vehicle position correction unit 41 acquires the information of the estimated vehicle position EP acquired by the estimated vehicle position acquisition unit 31 and the vicinity of the estimated vehicle position EP acquired by the road information acquisition unit 34. The road information RI is acquired.

  In the present embodiment, the normal vehicle position correction unit 41 executes a correction process (normal vehicle position correction process) of the estimated vehicle position EP that is not based on the direction of the feature on the road surface. As such normal vehicle position correction processing, a known method (map matching method) for performing map matching (correction for adjusting the vehicle position on the road shown on the road map) can be employed. . Although a detailed description is omitted, for example, the normal vehicle position correction unit 41 corrects the estimated vehicle position EP by employing the pattern matching method and the projection method described in paragraphs 0009 to 0013 of Patent Document 1 above. It can be configured.

  The range of the road information RI required when the normal vehicle position correction unit 41 executes the normal vehicle position correction process differs depending on the map matching method, and the normal vehicle position correction unit 41 uses the map matching to be adopted. Road information RI in a range necessary for executing the law is acquired from the road map database 12 via the road information acquisition unit 34.

  The correction of the estimated own vehicle position EP by the normal own vehicle position correcting unit 41 is performed when the function of the normal own vehicle position correcting process is set to be effective (in this example, the estimated own vehicle position EP is not within the circular intersection 5). Case) at a predetermined timing. This predetermined timing is set every predetermined time interval (for example, every 1 second, 2 seconds, every 5 seconds, etc.) or every time the vehicle 100 travels a predetermined distance (for example, 1 meter, 2 meters, 5 meters, etc.). Can be set. Further, a predetermined timing is set every time the estimated own vehicle position EP enters a new link K or a coordinate range near the new link K, or the estimated vehicle position EP and the road indicated on the road map It can be set every time the deviation exceeds a predetermined threshold.

3-5-2. Configuration of feature-use own vehicle position correction unit The feature-use own vehicle position correction unit 42 includes a road extension direction ED included in the road information RI around the estimated vehicle position EP and a reference direction of the feature on the road surface. It is a functional unit that corrects the estimated vehicle position EP based on SD, that is, a functional unit that executes a feature-based vehicle position correction process. As shown in FIG. 2, the feature-use own vehicle position correcting unit 42 includes an image recognition unit 43, a reference orientation specifying unit 44, and a feature orientation matching position detection unit 45. The feature using own vehicle position correcting unit 42 is configured to correct the estimated own vehicle position EP based on the feature direction matching position P0 detected by the feature direction matching position detection unit 45.

3-5-2-1. Configuration of Image Recognizing Unit The image recognizing unit 43 uses a feature on the road surface provided with a direction parallel to the extending direction ED of the road as a reference direction SD as a target feature, and an image of the target feature included in the photographed image. It is a functional unit that performs recognition processing. Specifically, the image recognition unit 43 uses the captured image of the road surface acquired by the road surface image acquisition unit 33 and performs binarization processing, edge detection processing, and the like on the captured image, so that the captured image is displayed. Extract contour information of contained features. Thereafter, the image recognition unit 43 performs pattern matching between the extracted contour information of the feature and the feature quantity of the target feature, and performs processing for extracting contour information that matches the shape of the target feature. When the contour information that matches the form of the target feature is extracted, the contour information is recognized as the target feature, and it is determined that the image recognition of the target feature is successful. In the present embodiment, such recognition processing of the road marking M by the image recognition unit 43 is executed every time a new captured image is acquired from the road surface image acquisition unit 33.

  When the image recognition unit 43 succeeds in image recognition of the target feature, the image recognition unit 43 derives the relative direction RD between the recognized target feature and the host vehicle 100. That is, the image recognizing unit 43 derives the relative reference direction SD (in the relative direction) of the target feature with the traveling direction MD of the host vehicle 100 as a reference. Here, the direction (that is, the traveling direction MD) of the host vehicle 100 in the captured image is a fixed direction (vertical direction in the example shown in FIG. 1) in the captured image, regardless of the traveling direction MD of the vehicle 100. In addition, the reference orientation SD of the target feature in the captured image can be acquired from the contour information recognized as the target feature. Therefore, when image recognition of the target feature is successful, as conceptually shown in FIG. 1, a line segment representing the direction (traveling direction MD) of the host vehicle 100 in the captured image and the reference direction of the target feature Based on the angle formed by the line segment representing SD, the relative orientation RD can be derived. Although detailed description is omitted, since the camera 16 is installed at a position higher than the road surface, the photographic angle with respect to the road surface is also taken into consideration when deriving the relative direction RD. The derived relative orientation RD is output to the reference orientation specifying unit 44.

  As described above, in this embodiment, the road marking M provided with the direction parallel to the road extending direction ED as the reference direction SD is the target feature. Here, the “road marking M” is a symbol or character for regulation or instruction provided on the road surface by painting or the like. For example, an arrow-shaped sign indicating the traveling direction (arrow sign, see FIG. 5), a character sign representing characters such as speed (for example, maximum speed) (see FIG. 6), and a sign indicating a pedestrian crossing (not shown) ), Markings (not shown) indicating lane markings (center line, lane boundary line (lane mark), etc.) and the like are included in the road marking M.

  Since the feature (road marking M in this example) provided on the road surface has a predetermined orientation with respect to the road, the reference orientation SD is derived from the shape and orientation represented by the contour information obtained by image recognition. Can be derived. In the linear arrow marking as shown in FIG. 5, the direction indicated by the arrow is generally the reference orientation SD. In the character marking as shown in FIG. 6, the direction from the bottom to the top of the character is generally set as the reference direction SD.

  The target features to be used in the present invention include objects of any shape, but it is preferable to use a feature having a shape that makes it easy to detect the reference orientation SD of the target feature by image recognition. . Specifically, it is preferable that a feature on a road surface having a rotational symmetry of less than 180 degrees around an axis orthogonal to the road surface is a target feature. For example, as shown in FIG. 5 and FIG. 6, a feature having a shape that does not have rotational symmetry about an axis orthogonal to the road surface (an axis orthogonal to the paper surface) or a shape that does not have a rotational symmetry is omitted, but is orthogonal to the road surface. A feature having a shape (for example, I-shape) having a rotational symmetry of 180 degrees around the axis to be processed can be set as the target feature.

  A feature on the road surface having at least one of a straight line component M1 parallel to the road extending direction ED and a straight line component M2 orthogonal to the road extending direction ED is set as a target feature. Is also suitable. The feature illustrated in FIG. 5 has a shape having both a linear component M1 parallel to the road extending direction ED and a linear component M2 orthogonal to the road extending direction ED. Even with such a configuration, it becomes easy to detect the reference orientation SD of the target feature by image recognition.

  In the example shown in FIG. 5, a shape that satisfies both of the above-described requirements regarding the shape, that is, a shape that does not have rotational symmetry of less than 180 degrees about an axis orthogonal to the road surface, The shape has at least one of a linear component M1 parallel to the extending direction ED and a linear component M2 orthogonal to the extending direction ED of the road. On the other hand, in the example shown in FIG. 6, it is a shape that satisfies only one of the above-described requirements regarding the shape, that is, a shape that does not have a rotational symmetry of less than 180 degrees around an axis orthogonal to the road surface. The shape does not significantly include a straight line component M1 parallel to the road extending direction ED and a straight line component M2 orthogonal to the road extending direction ED. Although not shown in the drawings, the shape satisfies only the other of the above-mentioned requirements regarding the shape, that is, has rotational symmetry of less than 180 degrees (for example, 90 degrees or 45 degrees) about an axis orthogonal to the road surface. Although it is a shape, a feature having a shape having at least one of a linear component M1 parallel to the road extending direction ED and a linear component M2 orthogonal to the road extending direction ED is defined as a target feature. You can also

  By the way, in the present invention, it is not always necessary that the information on the target feature (type, arrangement position, etc.) is provided in the road map database 12, but the information on the target feature is provided in the road map database 12. If the image recognition unit 43 executes image recognition processing of the target feature, the information on the target feature can be referred to. With such a configuration, it is possible to more easily perform image recognition of the target feature and derivation of the relative orientation RD, and the configuration of the image recognition unit 43 can be simplified.

3-5-2-2. Configuration of reference azimuth specifying unit The reference azimuth specifying unit 44 is a relative azimuth RD between the target feature recognized by the image recognition unit 43 and the host vehicle 100 (that is, a relative azimuth based on the traveling direction MD of the host vehicle). This is a functional unit that specifies the reference direction SD of the target feature recognized by the image recognition unit 43 based on the reference direction SD) and the traveling direction MD of the host vehicle 100. As described above, when the image recognition unit 43 succeeds in image recognition of the target feature (in this example, the road sign M), information on the relative direction RD of the target feature is input to the reference orientation specifying unit 44. Is done.

  When the relative orientation RD of the target feature is acquired, the reference orientation specifying unit 44 acquires the traveling direction MD of the vehicle 100 from the traveling direction acquisition unit 32. In the present embodiment, the reference orientation specifying unit 44 acquires the traveling direction MD of the vehicle 100 when capturing a captured image including the target feature or when acquiring the captured image by the road surface image acquiring unit 33. Then, the reference orientation specifying unit 44 specifies the reference orientation SD of the target feature by adding or subtracting the traveling orientation MD and the relative orientation RD in consideration of the sign (direction) of the relative orientation RD. That is, the reference orientation SD specified here is an orientation in the same coordinate system as the traveling orientation MD. Then, the identified reference orientation SD is output to the feature orientation matching position detection unit 45. As described above, since the traveling direction MD of the vehicle 100 acquired by the traveling direction acquisition unit 32 is the traveling direction MD in the absolute direction, the reference direction of the target feature specified by the reference direction specifying unit 44 SD is also a reference orientation SD in absolute orientation. That is, the reference azimuth specifying unit 44 converts the reference azimuth SD in the relative azimuth based on the traveling azimuth MD into the reference azimuth SD in the absolute azimuth. Thereby, it becomes possible to compare the reference azimuth SD with the extended azimuth ED of the road obtained from the road information RI also expressed in the absolute azimuth.

  By the way, the reference azimuth SD matches the extended azimuth ED of the road at the point where the target feature is provided. That is, according to the present invention, even if the traveling direction MD of the host vehicle 100 is different from the extended direction ED of the road due to, for example, a lane change, the target feature (the road marking M in this example) By specifying the reference direction SD, it is possible to specify the road extending direction ED in the vicinity where the vehicle 100 is located.

3-5-2-3. Configuration of Feature Direction Matching Position Detection Unit The feature direction matching position detection unit 45 is a position on the road where the reference direction SD and the extension direction ED match based on the road information RI around the estimated vehicle position EP. This is a functional unit that detects the feature orientation matching position P0. As described above, the information on the reference direction SD specified by the reference direction specifying unit 44 is input to the feature direction matching position detection unit 45. Then, when the feature azimuth coincidence position detection unit 45 acquires the reference azimuth SD, the feature azimuth coincidence position P0 is derived based on the reference azimuth SD and the road information RI around the estimated vehicle position EP. As described above, this reference direction SD coincides with the road extending direction ED at the point where the target feature is provided.

  Specifically, the feature direction matching position detection unit 45 receives road information about the estimated vehicle position EP acquired by the estimated vehicle position acquisition unit 31 when the reference direction SD is input from the reference direction specifying unit 44. The RI is acquired from the road map database 12 via the road information acquisition unit 34. In the present embodiment, the feature direction coincidence position detection unit 45 is the road information RI around the estimated vehicle position EP at the time of shooting a shot image including the target feature or when the road surface image acquisition unit 33 acquires the shot image. To get. The range of road information RI to be acquired can be set based on the maximum value of errors that can be included in the estimated host vehicle position EP, for example. Specifically, the range of the road information RI to be acquired can be set inside the circular area whose center is the estimated own vehicle position EP and whose radius matches the maximum value of the error.

  Further, in the present embodiment, as described above, when the estimated own vehicle position EP is within the circular intersection 5 that is the intersection having the ring road 51, the estimated own vehicle position EP by the feature-use own vehicle position correcting unit 42 is used. It is comprised so that correction | amendment may be performed. Therefore, the range of the road information RI to be acquired can be a range including the whole or a part of the ring road 51 constituting the roundabout 5. The “part” in this case is, for example, a part of the annular road 51 (for example, a semicircular arc part) in which roads of equal length (distance) are included on both sides centering on the own vehicle position. be able to.

  The feature azimuth coincidence position detection unit 45 specifies the extended azimuth ED of the road included in the acquired road information RI. In the present embodiment, as described above, the road information RI includes information on the link K (see FIG. 4) corresponding to each road, and the feature direction matching position detection unit 45 determines the direction of the link K. Based on the above, the road extending direction ED is specified. Then, the feature orientation matching position detection unit 45 detects the position on the road (on the road map data) where the reference orientation SD and the extended orientation ED match, and sets the position as the feature orientation matching position P0. If the feature direction matching position detection unit 45 cannot detect the position on the road (on the road map data) where the reference direction SD and the extended direction ED match, the estimated own vehicle based on the current captured image The position EP is not corrected.

  When the road represented by the link K is a one-way road (for example, the ring road 51 shown in FIG. 3), the link information (link attribute) may include information on the traveling direction of the road. In such a case, the feature azimuth coincidence position detecting unit 45 obtains the road extending azimuth ED included in the acquired road information RI as a vector, and the reference azimuth SD and the extending azimuth ED coincide with each other in consideration of the direction. It is preferable that the position on the road to be detected (on the road map data) is detected.

  By the way, in this embodiment, as shown in FIG. 4, the shape of the link K linearly connects between two adjacent nodes N and the shape interpolation point S along the road extending direction ED. Has been. Therefore, even when the road extending direction ED changes continuously as in the ring road 51 shown in FIG. 3, the shape of the link K changes stepwise as shown in FIG. In such a case, the road extending direction ED acquired by the feature direction matching position detection unit 45 from the road information RI also changes stepwise.

  In view of the method for deriving the road extending direction ED as described above, in the present embodiment, the feature direction coincidence position detection unit 45 detects the feature direction coincidence position P0 and the reference direction SD and the road Even if the extended direction ED does not completely match, if the deviation between the two directions is within a predetermined range, it is determined that the reference direction SD and the extended direction ED of the road match. Is configured to do. In addition, when there are a plurality of positions (points) on the road having the extended azimuth ED within a predetermined range, the position (point) with the smallest deviation is selected. Can do. At this time, the distance from the estimated vehicle position EP is also taken into consideration, and the priority can be increased even if the distance is short.

  By the way, the feature azimuth coincidence position P0 obtained as described above is a straight line (that is, the whole of the link K or the link K) between two adjacent nodes N and the shape interpolation point S along the road extending direction ED. (Part) It was not specified to the upper position. When the feature direction matching position P0 is further specified, the midpoint of the straight line (line segment) can be set as the feature direction matching position P0. Further, when the position moved from the estimated vehicle position EP obtained last time along the link K by the movement distance of the vehicle 100 from the time when the last estimated vehicle position EP was obtained to the present is on the straight line. The position can be set as the feature direction coincidence position P0. Alternatively, according to the degree of coincidence between the reference direction SD and the road extending direction ED, the midpoint of the straight line is set as the feature direction coincidence position P0 when there is a perfect coincidence, and when there is a deviation, the degree of coincidence is A point moved from the midpoint of the straight line toward the end of the straight line according to the degree of decrease may be set as the feature direction matching position P0.

  When the road information RI includes information on the curvature of the road, a curve connecting two adjacent nodes N and shape interpolation points S along the road extending direction ED is based on the curvature. Assuming that the tangential direction of the curve is the road extending direction ED, the road extending direction ED acquired from the road information RI can be continuously changed. Information regarding the curvature of the road can be information regarding the diameter of the circular intersection 5 when the target road is the circular intersection 5. In such a case, the predetermined range related to the azimuth deviation can be reduced. For example, the reference direction SD and the road extending direction ED completely coincide with each other with the predetermined range as one point. In this case, it may be configured that it is determined that the reference direction SD and the road extending direction ED match.

  Specifically, for example, when the target road is a road that curves with a certain curvature, the road information RI includes information on the radius of curvature and the center of curvature (the center of the virtual circle along the curve) for the target road. If provided, the tangential direction of the curve (arc) determined from the radius of curvature and the center of curvature can be set as the road extending direction ED. Further, for example, when the target road is a curved road with a certain curvature, if the road information RI includes information on the curvature and distance (the length of the road) about the target road, these curvatures The tangential direction of the curve (arc) determined from the distance can be set as the road extending direction ED. In such a case, the road information RI may be configured not to include information on the node N and the link K for at least the target road.

  Then, based on the feature direction coincidence position P0 detected by the feature direction coincidence position detection unit 45, the feature using own vehicle position correction unit 42 corrects the estimated own vehicle position EP acquired by the estimated own vehicle position acquisition unit 31. To do. In the present embodiment, the feature direction matching position detection unit 45 corrects the estimated own vehicle position EP by replacing the feature direction matching position P0 with the estimated own vehicle position EP. Thereby, for example, even when the traveling direction MD of the host vehicle 100 is different from the road extending direction ED due to a lane change or the like, the reference direction SD of the target feature (the road marking M in this example) is specified. By doing so, it is possible to identify the extension ED of the road in the vicinity where the vehicle 100 is located, and to appropriately correct the estimated own vehicle position EP based on the extension ED.

  In the example shown in FIG. 3, the circular road 51 included in the circular intersection 5 is circular. However, the circular road 51 includes objects other than circular (such as an oval). Further, the ring road 51 includes roads of all shapes formed in a ring shape as a whole. For example, roads formed in a polygonal shape are also included in the ring road 51. In the example shown in FIG. 3, the ring road 51 is a one-way road, but the ring road 51 may be a road that allows two-way traffic.

4). Next, referring to FIG. 7 and FIG. 8, a procedure of estimated vehicle position correction processing (vehicle position recognition processing) executed in the vehicle position recognition unit 3 according to the present embodiment, that is, estimation. The own vehicle position correction method (own vehicle position recognition method) will be described. The procedure of the estimated vehicle position correction process described below is executed by hardware and / or software (program) or both constituting each functional unit of the navigation device 1 (vehicle position recognition unit 3). When each of the above function units is configured by a program, the arithmetic processing device included in the navigation device 1 operates as a computer that executes the program that configures each of the above function units. FIG. 8 is a flowchart showing the procedure of the feature-use vehicle position correction process in step # 12 of FIG.

4-1. Overall Procedure of Estimated Vehicle Position Correction Process As shown in FIG. 7, when the estimated vehicle position acquisition unit 31 acquires the estimated vehicle position EP (step # 10), the estimated vehicle position EP is within the roundabout 5 Is determined (step # 11). When the estimated own vehicle position EP is within the roundabout 5 (step # 11: Yes), the feature using own vehicle position correcting process is executed by the feature using own vehicle position correcting unit 42 (step #). 12). Note that the feature-use vehicle position correction process will be described in detail later based on the flowchart of FIG.

  On the other hand, when the estimated own vehicle position EP is not within the roundabout 5 (step # 11: No), the normal own vehicle position correcting process by the normal own vehicle position correcting unit 41 is executed (step # 13). In this normal vehicle position correction process, a publicly known vehicle position correction unit 34 is known based on the road extending direction ED and the traveling direction MD of the vehicle 100 included in the road information RI around the estimated vehicle position EP acquired by the road information acquisition unit 34. The estimated vehicle position EP is corrected based on the map matching method.

4-2. Procedure of feature-use vehicle position correction processing Next, a procedure of feature-use vehicle position correction processing will be described with reference to FIG. In the feature-use vehicle position correction process, first, a road surface photographed image is acquired by the road surface image acquisition unit 33 (step # 20). Then, the image recognition unit 43 performs image recognition processing of the target feature (in this example, road marking M) on the acquired captured image (step # 21). As a result of this image recognition process, when the image recognition of the target feature has not succeeded (step # 22: No), the process ends.

  On the other hand, when the image recognition of the target feature is successful (step # 22: Yes), the image recognition unit 43 derives the relative orientation RD between the recognized target feature and the host vehicle 100 (step #). 23). Next, the traveling direction acquisition unit 32 acquires the traveling direction MD of the vehicle 100 (step # 24). The reference azimuth specifying unit 44 then recognizes the target feature recognized by the image recognition unit 43 based on the relative azimuth RD derived by the image recognition unit 43 and the traveling direction MD of the vehicle 100 acquired by the traveling direction acquisition unit 32. Is determined (step # 25).

  When the reference orientation SD is specified, the feature orientation matching position detection unit 45 acquires the road information RI around the estimated vehicle position EP from the road map database 12 via the road information acquisition unit 34 (step # 26). ). Then, the feature direction matching position detection unit 45 determines whether or not the position where the reference direction SD and the extension direction ED match is on the road included in the acquired road information RI (step # 27). If there is a position where the reference direction SD and the extended direction ED match (step # 27: Yes), the position is specified (detected) as the feature direction matching position P0 (step # 28). The feature using own vehicle position correcting unit 42 corrects the estimated own vehicle position EP based on the feature direction matching position P0 (step # 29). In this example, the estimated vehicle position EP is corrected by replacing the feature direction matching position P0 with the estimated vehicle position EP.

  On the other hand, if there is no position where the reference azimuth SD and the extended azimuth ED coincide in step # 27 (step # 27: No), the process ends.

5. Other Embodiments Finally, other embodiments according to the present invention will be described. Note that the features disclosed in each of the following embodiments can be used only in that embodiment, and can be applied to other embodiments as long as no contradiction arises.

(1) In the above embodiment, the configuration in which the feature direction matching position detection unit 45 corrects the estimated own vehicle position EP by replacing the feature direction matching position P0 with the estimated own vehicle position EP will be described as an example. did. However, the embodiment of the present invention is not limited to this, and based on the distance between the target feature recognized by the image recognition unit 43 and the host vehicle 100, that is, the feature orientation matching position P0, A configuration may be adopted in which the estimated host vehicle position EP is corrected based on the distance between the target feature recognized by the image recognition unit 43 and the host vehicle 100. Here, the distance between the target feature recognized by the image recognition unit 43 and the subject vehicle 100 is the distance between the subject vehicle 100 and the subject feature at the time of photographing the road surface photographed image, and the correction processing is completed from the time of photographing. It is possible to include at least one of the travel distances of the host vehicle 100 up to. Note that the distance between the subject vehicle 100 and the target feature at the time of shooting the shot image of the road surface can be calculated from the position of the target feature in the shot image, the azimuth of the camera 16, the in-vehicle position, and the like. Further, the travel distance of the host vehicle 100 from the time of shooting to the completion of the correction process can be acquired by the distance sensor 14 or the like.

(2) In the above embodiment, the image recognition unit 43 derives the relative orientation RD, and the reference orientation identification unit 44 identifies the reference orientation SD based on the derived relative orientation RD and the traveling orientation MD. As explained. However, the embodiment of the present invention is not limited to this, and a configuration in which the reference orientation SD is specified directly from the image recognition result without directly deriving the relative orientation RD can be adopted. In such a configuration, for example, the information of the traveling direction MD is input to the image recognition unit 43, and the image recognition unit 43 directly specifies the reference direction SD from the information of the traveling direction MD and the image recognition result. can do. That is, in the present invention, “the reference direction SD of the target feature is specified based on the relative direction RD between the target feature recognized by the image recognition unit 43 and the host vehicle 100 and the traveling direction MD of the host vehicle 100”. Includes not only a configuration in which the relative orientation RD is directly derived but also a configuration in which the relative orientation RD is conceptually used in the process of specifying the reference orientation SD without directly deriving the relative orientation RD.

(3) In the above embodiment, when the estimated vehicle position correction unit 40 is in the roundabout 5, the estimated vehicle position EP is corrected by the feature-use vehicle position correction unit 42. In other cases, the configuration in which the normal vehicle position correction unit 41 corrects the estimated vehicle position EP has been described as an example. However, the embodiment of the present invention is not limited to this, and the estimated vehicle position correction unit 40 also uses the feature-based vehicle position correction unit even when the estimated vehicle position EP is not within the roundabout 5. It is also possible to adopt a configuration in which the estimated own vehicle position EP by 42 is corrected. For example, the feature-use vehicle position correction unit 42 is also used on roads (including intersections) having a predetermined radius of curvature or less and roads having a predetermined number (for example, 2, 3, 4, etc.) or more lanes (lanes). The estimated vehicle position EP can be corrected by the above.

(4) In the above embodiment, the estimated vehicle position correction unit 40 corrects the estimated vehicle position EP by the feature-use vehicle position correction unit 42 when the estimated vehicle position EP is within the roundabout 5. The configuration for performing the above has been described as an example. However, the embodiment of the present invention is not limited to this, and when the estimated own vehicle position correcting unit 40 has the estimated own vehicle position EP within the roundabout 5, estimation by the normal own vehicle position correcting unit 41 is performed. It is also possible to adopt a configuration for correcting the own vehicle position EP. Further, the circular intersection 5 is determined by the difference between the outer diameter and the inner diameter of the circular road 51 (that is, the width of the circular road 51), the curvature of the circular road 51, the number of connecting roads 52 connected to the circular road 51, and the like. 5 is classified into two groups, and when the estimated own vehicle position EP is within the roundabout 5 belonging to one group, the estimated own vehicle position EP is corrected by the feature-use own vehicle position correcting unit 42; When the estimated own vehicle position EP is within the roundabout 5 belonging to the other group, the normal own vehicle position correcting unit 41 may correct the estimated own vehicle position EP.

(5) In the above-described embodiment, the configuration in which the feature on the road surface that is the target feature is the road marking M has been described as an example. However, the embodiment of the present invention is not limited to this, and the features on the road surface other than the road marking M provided with the azimuth parallel to the extended azimuth ED of the road as the reference azimuth SD are the target features. It is also possible to do. For example, a structure provided on a road surface such as a curb or a gutter can be used as the target feature.

(6) In the above embodiment, the configuration in which the camera 16 is a back camera that captures the road surface behind the vehicle 100 has been described as an example. However, the embodiment of the present invention is not limited to this, and the camera 16 may be a front camera that photographs a road surface in front of the vehicle 100. In addition, the camera 16 may be a camera that captures an oblique direction on the front side, an oblique direction on the rear side with respect to the vehicle 100, or a road surface on the side of the vehicle 100.

(7) In the above embodiment, the configuration in which the link K information (link attribute) includes information on whether or not the link K is a circular intersection link forming the circular intersection 5 has been described as an example. . However, the embodiment of the present invention is not limited to this, and information on whether or not the node N is a circular intersection node forming the circular intersection 5 is included as information (node attribute) of the node N. It can be set as the structure which has. In such a configuration, it is determined whether or not the estimated vehicle position EP is within the roundabout 5 based on whether or not the estimated vehicle position EP has reached the roundabout node or the coordinate range near the roundabout node. It can be set as the structure determined.

(8) The assignment of the function units described in the above embodiment is merely an example, and a plurality of function units can be combined or one function unit can be further divided.

(9) In the above embodiment, the case where the vehicle position recognition system according to the present invention is applied to the navigation device 1 has been described as an example. However, the embodiment of the present invention is not limited to this, and the vehicle position recognition system according to the present invention can also be applied to devices other than the navigation device provided in the vehicle 100 (such as a travel control device). It is.

(10) In the above embodiment, the configuration in which the vehicle position recognition unit 3 and the road map database 12 are provided in the vehicle 100 has been described as an example. However, at least a part of the functional units and road map of the vehicle position recognition unit 3 are described. The database 12 may be provided outside the vehicle 100 so that information and signals are transmitted and received via a communication network such as the Internet.

(11) Regarding other configurations as well, the embodiments disclosed herein are illustrative in all respects, and embodiments of the present invention are not limited thereto. That is, as long as the configuration described in the claims of the present application and a configuration equivalent thereto are provided, a configuration obtained by appropriately modifying a part of the configuration not described in the claims is naturally also included in the present invention. Belongs to the technical scope.

  The present invention includes an estimated own vehicle position acquisition unit that acquires an estimated own vehicle position that is an estimated position of the own vehicle, a traveling direction acquisition unit that acquires a traveling direction of the own vehicle, and a road that includes information related to a road extending direction A vehicle position recognition system including a road information acquisition unit that acquires information and a road surface image acquisition unit that acquires a captured image of the road surface, an estimated vehicle position that acquires an estimated vehicle position that is an estimated position of the vehicle An acquisition function, a traveling direction acquisition function for acquiring the traveling direction of the host vehicle, a road information acquisition function for acquiring road information including information related to the extension direction of the road, and a road surface image acquisition function for acquiring a shot image of the road surface , A vehicle position recognition program for realizing the above, an estimated vehicle position acquisition step for acquiring an estimated vehicle position that is an estimated position of the vehicle, and a traveling direction acquisition step for acquiring a traveling direction of the vehicle, Road A road information obtaining step of obtaining road information including information on orientation, and the road surface image acquiring step of acquiring the captured image of the road surface, can be suitably used in the vehicle position recognition method comprising a.

5: Roundabout 31: Estimated own vehicle position acquisition unit 32: Traveling direction acquisition unit 33: Road surface image acquisition unit 34: Road information acquisition unit 41: Normal own vehicle position correction unit 42: Feature-use own vehicle position correction unit 43: Image recognition unit 44: reference direction specifying unit 51: ring road 100: vehicle (own vehicle)
ED: Extension direction EP: Estimated vehicle position M: Road marking (target feature)
MD: traveling direction P0: feature direction matching position RD: relative direction RI: road information SD: reference direction

Claims (7)

An estimated own vehicle position acquisition unit that acquires an estimated own vehicle position that is an estimated position of the own vehicle, a traveling direction acquisition unit that acquires a traveling direction of the own vehicle, and road information including information related to a road extending direction is acquired. A vehicle position recognition system comprising a road information acquisition unit and a road surface image acquisition unit that acquires a captured image of a road surface,
An image recognition unit that performs image recognition processing of the target feature included in the photographed image, using a feature on a road surface provided with a direction parallel to the extending direction of the road as a reference direction as a target feature;
A reference azimuth identifying unit that identifies the reference azimuth of the target feature based on the relative azimuth between the target feature and the host vehicle recognized by the image recognition unit, and the traveling azimuth of the host vehicle;
Based on the road information around the estimated vehicle position, a feature direction coincidence position that is a position on the road where the extended direction coincides with the reference direction specified by the reference direction specifying unit is detected. A feature-use own vehicle position correcting unit that corrects the estimated own vehicle position based on an object orientation matching position;
The vehicle position recognition system equipped with. A normal vehicle position correction unit that corrects the estimated vehicle position based on the extension direction of the road and the traveling direction of the vehicle included in the road information around the estimated vehicle position;
The own vehicle position according to claim 1, wherein the estimated own vehicle position is corrected by the feature-use own vehicle position correcting unit when at least the estimated own vehicle position is in a roundabout that is an intersection having a ring road. Recognition system.   The vehicle position recognition system according to claim 1 or 2, wherein a feature on a road surface having a rotational symmetry of less than 180 degrees with an axis orthogonal to the road surface as a center is the target feature.   The feature on the road surface having a shape having at least one of a linear component parallel to the extending direction of the road and a linear component orthogonal to the extending direction of the road is defined as the target feature. 4. The vehicle position recognition system according to any one of 3 above.   The feature use own vehicle position correcting unit corrects the estimated own vehicle position based on a distance between the target feature recognized by the image recognition unit and the own vehicle. The vehicle position recognition system according to the item. Obtain road information including information on the estimated vehicle position acquisition function that acquires the estimated vehicle position, which is the estimated position of the vehicle, a travel direction acquisition function that acquires the travel direction of the host vehicle, and information on the extension direction of the road A vehicle position recognition program for realizing a road information acquisition function and a road surface image acquisition function for acquiring a captured image of a road surface,
An image recognition function for performing image recognition processing of the target feature included in the photographed image, with a feature on the road surface provided with a direction parallel to the extending direction of the road as a reference direction as a target feature;
A reference azimuth specifying function for specifying the reference azimuth of the target feature based on a relative azimuth between the target feature and the host vehicle recognized by the image recognition function, and a traveling azimuth of the host vehicle;
Based on the road information around the estimated vehicle position, a feature direction matching position that is a position on the road where the reference direction specified by the reference direction specifying function matches the extended direction is detected. A feature-use own vehicle position correcting function for correcting the estimated own vehicle position based on an object orientation matching position;
A vehicle position recognition program that enables a computer to realize a vehicle. An estimated own vehicle position obtaining step for obtaining an estimated own vehicle position that is an estimated position of the own vehicle, a traveling direction obtaining step for obtaining a traveling direction of the own vehicle, and road information including information relating to a road extending direction are obtained. A vehicle position recognition method comprising a road information acquisition step and a road surface image acquisition step of acquiring a captured image of the road surface,
An image recognition step for performing image recognition processing of the target feature included in the photographed image, with a feature on the road surface provided with a direction parallel to the extending direction of the road as a reference direction as a target feature;
A reference azimuth specifying step for specifying the reference azimuth of the target feature based on a relative azimuth between the target feature and the host vehicle recognized in the image recognition step, and a traveling azimuth of the host vehicle;
Based on the road information around the estimated vehicle position, a feature for detecting a feature orientation coincidence position that is a location on the road whose extended orientation coincides with the reference orientation identified in the reference orientation identifying step. A direction matching position detection step;
A feature using own vehicle position correcting step for correcting the estimated own vehicle position based on the feature direction matching position detected in the feature direction matching position detecting step;
A vehicle position recognition method comprising:

Images