Feature information collecting apparatuses, methods, and programs

• US 8,213,682 B2
• Filed: 03/11/2008
• Issued: 07/03/2012
• Est. Priority Date: 03/30/2007
• Status: Expired due to Fees

• Current Assignee: Aisin AW Company Limited (Aisin Corp.)
• Original Assignee: Aisin AW Company Limited (Aisin Corp.)
• Inventors: Nakamura, Masaki, Nakao, Koichi, Kanba, Motoki
• Primary Examiner(s): Koziol, Stephen R
• Application Number:
  US12/073,867
• Publication Number:
  US 20080240505A1
• Time in Patent Office: 1,575 Days
• Field of Search: 382/100, 382/103, 382/104
• US Class Current: 382/104
• CPC Class Codes: G01C 21/26 specially adapted for navigation in a road network
• CPC Class Codes: G01C 21/3815 Road data
• CPC Class Codes: G01C 21/3837 Data obtained from a single source
• CPC Class Codes: G01C 21/3848 Data obtained from both position sensors and additional sensors

Abstract

Apparatuses, methods, and programs acquire vehicle position information that represents a current position of a vehicle, acquire image information of a vicinity of the vehicle, and carry out image recognition processing of a target feature that is included in the image information to determine a position of the target feature. The apparatuses, methods, and programs store recognition position information that is based on the acquired vehicle position information and that represents the determined recognition position of the target feature. The apparatuses, methods, and programs determine an estimated position of the target feature based on a set of a plurality of stored recognition position information for the target feature, the plurality of stored recognition position information for the target feature being stored due to the target feature being subject to image recognition processing a plurality of times.

Claims

1. A feature information collecting apparatus, comprising:

a memory that stores road information that represents a road network by connection relationships between nodes that correspond to intersections and links that correspond to roads linking the intersections; and

a controller that;

acquires vehicle position information that represents a current position of a vehicle;

acquires image information of a vicinity of the vehicle;

carries out image recognition processing of a target feature that is included in the image information to determine a recognition position of the target feature;

stores recognition position information in the memory that is based on the acquired vehicle position information and that represents the determined recognition position of the target feature;

determines an estimated position of the target feature based on set of a plurality of stored recognition position information for the target feature, the plurality of stored recognition position information for the target feature being stored due to the target feature being subject to image recognition processing a plurality of times; and

corrects a position of a stored node based on the estimated position of the target feature and an estimated position of another target feature, the target feature and the other target feature surrounding the node and being disposed at substantially equal distances from the stored node.

2. The feature information collecting apparatus according to claim 1, wherein the controller:

determines a representative value of a distribution of the plurality of stored position information for the target feature to be the estimated position.

3. The feature information collecting apparatus according to claim 1, wherein the controller:

uses the plurality of stored recognition position information for the target feature as learning values for a predetermined position range, the predetermined position range including the determined recognition positions of the plurality of stored recognition position information for the target feature;

adds the learning values to each of the position range each time the target feature is recognized and stores the results; and

determines a position that represents the position ranges for which the learning value for the target feature has become equal to or greater than a predetermined learning threshold value to be the estimated position of the target feature.

4. The feature information collecting apparatus according to claim 1, wherein the controller:

determines a feature distance, which represents the distance between two target features, based on sets of the recognition position information for the two target features.

5. The feature information collecting apparatus according to claim 1, wherein the controller:

generates feature information that associates the estimated position of the target feature and the attribute information, the attribute information based on the recognition position off the target feature.

6. The feature information collecting apparatus according to claim 1, wherein:

the memory stores feature information including at least one of;

learned feature information that associates the estimated position of the target feature and the attribute information, the attribute information based on the recognition position of the target feature; and

initial feature information that includes position information and attribute information that have been prepared in advance for a plurality of target features.

7. The feature information collecting apparatus according to claim 6, wherein:

the controller;

carries out image recognition processing on the image information to identify target features that that have been extracted from the feature information based on the vehicle position information;

stores recognition failure information, which represents that the image recognition of the target features that are shown by the feature information by the image recognition device has failed, such that the target features can be distinguished; and

deletes feature information for a target feature in the feature information based on a set of recognition failure information for the target feature, the set of recognition failure information stored due to the image recognition processing having been carried out a plurality of times on the same target feature.

8. The feature information collecting apparatus according to claim 7, wherein the controller:

uses the recognition failure information for each of the target features as a learning value for the target features;

decrements a learning value each time the image recognition of a corresponding target features fails; and

deletes the feature information for a target feature when the corresponding learning value becomes equal to or less than a predetermined threshold value.

9. The feature information collecting apparatus according to claim 1, wherein the controller:

is configured to communicate with a plurality of vehicles; and

stores the recognition position information obtained by the plurality of vehicles.

10. A vehicle position recognition apparatus, comprising:

the feature information collecting apparatus according to claim 5;

wherein the controller revises the vehicle position information based on the image recognition results for the target features and the feature information for the target features.

11. A navigation apparatus, comprising:

the vehicle position recognition apparatus according to claim 10; and

a memory that stores map information;

wherein the controller outputs guidance information based on the map information.

12. A feature information collecting method, comprising:

accessing a memory that stores road information that represents a road network by connection relationships between nodes that correspond to intersections and links that correspond to roads linking the intersections;

acquiring vehicle position information that represents a current position of a vehicle;

acquiring image information of a vicinity of the vehicle;

carrying out image recognition processing of a target feature that is included in the image information to determine a recognition position of the target feature;

storing recognition position information in the memory that is based on the acquired vehicle position information and that represents the determined recognition position of the target feature;

determining an estimated position of the target feature based on a set of a plurality of stored recognition position information for the target feature, the plurality of stored recognition position information for the target feature being stored due to the target feature being subject to image recognition processing a plurality of times; and

correcting a position of a stored node based on the estimated position of the target feature and an estimated position of another target feature, the target feature and the other target feature surrounding the node and being disposed at substantially equal distances from the stored node.

13. The feature information collecting method according to claim 12, further comprising:

using the plurality of stored recognition position information for the target feature as learning values for a predetermined position range, the predetermined position range including the determined recognition positions of the plurality of stored recognition position information for the target feature;

adding the learning values to each of the position range each time the target feature is recognized and stores the results; and

determining a position that represents the position ranges for which the learning value for the target feature has become equal to or greater than a predetermined learning threshold value to be the estimated position of the target feature.

14. The feature information collecting method according to claim 12, further comprising:

carrying out image recognition processing on the image information to identify target features that that have been extracted from the feature information based on the vehicle position information;

storing recognition failure information, which represents that the image recognition of the target features that are shown by the feature information by the image recognition device has failed, such that the target features can be distinguished; and

deleting feature information for a target feature in the feature information based on a set of recognition failure information for the target feature, the set of recognition failure information stored due to the image recognition processing having been carried out a plurality of times on the same target feature.

15. The feature information collecting method according to claim 14, further comprising:

using the recognition failure information for each of the target features as a learning value for the target features;

decrementing a learning value each time the image recognition of a corresponding target features fails; and

deleting the feature information for a target feature when the corresponding learning value becomes equal to or less than a predetermined threshold value.