Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
1999-02-23
2003-04-15
Robinson, Greta (Department: 2177)
Data processing: database and file management or data structures
Database design
Data structure types
C382S116000, C382S118000
Reexamination Certificate
active
06549913
ABSTRACT:
This application is based on application No. 10-45609 filed in Japan, the contents of which is hereby incorporated by reference.
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to a method for compiling an image database from which an image of a specific person can be retrieved, and relates more specifically to a method for compiling and automatically indexing a database of facial images by finding a specific person in a dictionary of human facial images and generating an index of personal attributes such as the name, employee number, and/or address of the person in the facial image stored to the image database. The present invention also relates to an image database system, and to an image information storage medium.
2. Description of the Related Art
Rapid advances in information processing technology, combined with improvement in performance of personal computers and other information products, have helped expand the use of image database systems and electronic filing systems from graphic designers, publishers, printers, and other users of large-volumes of images and graphics to even casual users. These image database systems and electronic filing systems typically enable the user to capture images and text using a scanner or other input device, store and manage the captured information in a database together with keywords and other ancillary information for searching and retrieval, and find a specific image in the database for output to hard copy using a printer, or transmission to another user.
A desirable feature of such image storage and retrieval systems is the ability to search the facial area of stored images to confirm the identity of a person. One type of such an image searching and identification system is a recognition system whereby, for example, images of the objects that will be searched for are segmented into an image mosaic, and images are recognized by applying an aspect method using image features detected from the mosaics.
Japanese Patent Laid-Open Publication No. 5-159063, for example, teaches an image searching system whereby images of reference objects are segmented into image mosaics and stored in a reference dictionary of search objects. An image of an unidentified object is then similarly segmented into an image mosaic for comparison with the reference object mosaics in the reference dictionary. The desired image is then selected from among a plurality of candidate images displayed on screen.
Another image searching system as taught in Japanese Patent Laid-Open Publication 5-159064 similarly compares a mosaic of an unidentified image against the content of a dictionary containing mosaics of reference images using an approximate-match searching means for comparison using a coarse image mosaic, and a detail searching means for comparison using a dense image mosaic.
A paper in Proceedings of the electronic information communication association of Japan D-II No. 11 (1995 November issue) also describes a method of optimizing an image dictionary based on image-to-image distances detected from differences in facial orientation, and statistical analysis of differing intrafacial distances in individual faces.
The known systems described above, however, make no mention of assigning or creating an index for efficiently searching a database for specific image data. When optimized keywords or other identifying information are not added to the images stored in a large-volume image database so that stored images can be retrieved as and when required, the above-described image database systems and electronic filing systems are unable to quickly retrieve a needed image. As a result, these systems are generally difficult to deploy for practical purposes.
The present invention is therefore directed to a method for resolving the aforementioned problems of conventional image database systems and electronic filing systems. More specifically, an object of the present invention is to provide a method for compiling an image database storing facial image information for a large number of people so that desired image information can be easily retrieved as required.
A further object of the present invention is to provide an image database system for automating the compilation of an image database storing facial image information for a large number of people.
A yet further object of the present invention is to provide an image database system for capturing and compiling external image information in an image database.
Yet a further object of the present invention is to provide an image database system for capturing facial image information and related attributes from a business card[, for example,] on which a person's face is printed, and storing the captured facial information together with the captured attribute values in an image database.
Another object of the present invention is to provide an image database system for compiling an image database using a face dictionary stored to, for example, CD-ROM, a hard disk, or other computer-readable medium.*
Yet another object of the present invention is to provide an image database system for finding desired image information in an image database, and then outputting the necessary image information to an external device, including to a printer for printing a desired number of images to hard copy.
SUMMARY OF THE INVENTION
To achieve the above-noted objects, an image database compiling method according to the present invention relates to an image database in which a facial image of a person is stored with related attribute information to enable searching and retrieving required image information based on the related attribute information. This compiling method comprises a registration step for storing human facial information together with related attribute information in a face dictionary; and a comparison step for comparing a facial image contained in image information stored or to be stored in the image database with dictionary image information stored in the face dictionary. When dictionary image information matching registered image information in the image database is detected, the attribute information stored with the dictionary image information is stored with the registered image information in the image database.
Facial images for a large number of people are previously stored in a face dictionary together with related attribute information. When new image information is stored in the image database, the facial image information to be registered in the image database is compared with facial images in the face dictionary. If a match is detected, the attribute information stored with the facial image in the face dictionary is retrieved and stored with the facial image in the image database.
Alternatively, the present invention achieves the above-noted objects by means of an image database system for storing facial images for a plurality of people together with related attribute information to enable searching and retrieving required image information based on the related attribute information. This image database system comprises a face dictionary for storing human facial information together with related attribute information; a means for comparing a facial image contained in image information stored or to be stored in the image database with dictionary image information stored in the face dictionary; and a means for storing attribute information stored with dictionary image information with image information registered in the image database when dictionary image information matching image information registered in the image database is detected.
Facial images for a large number of people are previously stored in a face dictionary together with related attribute information. When the means compares image information stored in the image database the facial images in the face dictionary and detects a match, the storing means stores the attribute information stored with the facial image in the face dictionary with the facial image in the image database.
The image dat
Black Linh
Minolta Co. , Ltd.
Robinson Greta
Sidley Austin Brown & Wood LLP
LandOfFree
Method for compiling an image database, an image database... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for compiling an image database, an image database..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for compiling an image database, an image database... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3111900