Image analysis – Image transformation or preprocessing – Image storage or retrieval
Reexamination Certificate
1998-12-23
2004-01-13
Mehta, Bhavesh M. (Department: 2621)
Image analysis
Image transformation or preprocessing
Image storage or retrieval
C382S305000
Reexamination Certificate
active
06678427
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to document identification registration systems that are used for identification (or discrimination) of document types and registration (or entry) of information with respect to a variety of documents such as slips and notes of books.
This application is based on Patent Application No. Hei 9-355631 filed in Japan, the content of which is incorporated herein by reference.
2. Description of the Related Art
Conventionally, document identification registration systems are used to identify document types when collectively reading many kinds of documents together. ID characters are written on prescribed locations of the documents to represent their document types. Based on results of reading of the ID characters written on the documents, the document identification registration system determines formats suited to the documents. Then, the content of the document is read with respect to each field of the document in accordance with the format.
The paper of Japanese Patent Application, Publication No. Hei 7-152856 discloses a device that a human operator enters an identification area of a document, which is used for identification of a document type, in advance by operating a mouse.
By the way, banks and post offices issue a variety of notes in response to many kinds of businesses and duties. They sometimes deal with notes on which the aforementioned ID characters are not written.
Therefore, in the case of the businesses and duties of the banks and post offices, the conventional document identification registration system cannot identify the notes on which the ID characters are not written.
In addition the aforementioned device has a drawback that it takes an enormous amount of human work for registration of a plenty of documents. At a registration mode, it cannot perform verification whether the identification areas are effective or not. Therefore, the aforementioned device has a problem that an identification precision for identification of documents is relatively low.
SUMMARY OF THE INVENTION
It is an object of the invention to provide a document identification registration system that is capable of identifying documents on which ID characters are not written with a high identification precision.
A document identification registration system of this invention is provided for identification and registration of (constituent elements of) documents such as slips and notes of books. For example, the document identification registration system inputs digital picture data representing a picture of a slip, which is picked up by an image pickup device. From the digital picture data, the system extracts unregistered constituent elements such as characters and/or lines of rectangular frames, which are printed on the slip. Herein, extraction of the constituent elements is performed using character identification areas, which are located on areas of characters printed on the slip. From the constituent elements, the system selects a first constituent element to meet a prescribed condition. Then, the first constituent element is compared with the registered constituent elements, which are stored in advance.
Next, the system further selects a second constituent element, which is clearly identified from the registered constituent elements, from the first constituent element on the basis of comparison result. Thus, the system registers the second constituent element therein.
To improve an identification precision, binary processing using different threshold values are performed on the digital picture data to produce first and second pictures derived from an original picture of the slip, for example. A dispersion value is calculated between data of corresponding areas of the first and second pictures. Thus, the system selects the first constituent element whose dispersion value is relatively small and which meets the prescribed condition.
Thus, it is possible to perform identification of the slips with ease and with a high identification precision.
REFERENCES:
patent: 3905045 (1975-09-01), Nickle
patent: 4933984 (1990-06-01), Nakano et al.
patent: 5583950 (1996-12-01), Prokoski
patent: 5628008 (1997-05-01), Hayashi
patent: 61 59568 (1986-03-01), None
patent: 7-152856 (1995-06-01), None
patent: 9-138837 (1997-05-01), None
Azarian Seyed
Mehta Bhavesh M.
NEC Corporation
LandOfFree
Document identification registration system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Document identification registration system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Document identification registration system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3248671