Image analysis – Image transformation or preprocessing – Image storage or retrieval
Reexamination Certificate
2000-08-29
2004-08-17
Johnson, Timothy M. (Department: 2625)
Image analysis
Image transformation or preprocessing
Image storage or retrieval
C382S173000, C382S175000, C382S199000, C382S203000, C382S229000
Reexamination Certificate
active
06778712
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a data sheet identification device that can be suitably used for a data sheet processing in a financial institution and others.
BACKGROUND IF THE INVENTION
In recent years, there has been developed a data sheet identification device as a device for identifying a data sheet (a medium exclusively used for a recognition processing) based on a process of reading information on the data sheet as optical image information, processing the read image and then identifying the data sheet. This data sheet identification device has now been widely used by various industries to improve their operation efficiency.
In a financial institution and a like industry, operators at windows are processing data sheets by using data sheet identification devices. In order to improve the work efficiency of data sheet processing, it has been required that one data sheet identification device can automatically process data sheets having various kinds of formats, not only that the data sheet identification device can process a large volume of data sheets of the same kind. As a data sheet identification device that meets this requirement, there has been a data sheet identification device that catches ruled lines printed on a data sheet as a feature for identifying the data sheet. The data sheet identification device that catches the ruled lines as the feature of the data sheet has been disclosed in PCT International Patent Publication No. WO97/05561.
According to the data sheet identification device disclosed in the above publication, a data sheet X shown in
FIG. 35A
is discriminated from a data sheet X′ shown in
FIG. 35C
based on a difference between ruled lines printed on both data sheets. In this example, an oval Ka portion is different from an oval Kb portion between the data sheet X and the data sheet X′. In other words, the data sheet X is different from the data sheet X′ in that while a ruled line does not exist at the oval Ka portion in the data sheet X, a ruled line exists at the oval Kb portion in the data sheet X′.
The operation of identifying the data sheet X shown in
FIG. 35A
will be explained next. First, the data sheet identification device optically reads an image (ruled lines, characters, graphics) printed on the data sheet X, and obtains image information. Then, the data sheet identification device processes the image information to extract only ruled-line information Xk shown in FIG.
35
B. Next, the data sheet identification device collates the ruled-line information Xk with a database relating to ruled-line information of various data sheets, and identifies the data sheet X from among these various data sheets.
Similarly, for identifying the data sheet X′ shown in
FIG. 35C
, the data sheet identification device optically reads an image (ruled lines, characters, graphics) printed on the data sheet X′, and obtains image information. Then, the data sheet identification device processes the image information to extract only ruled-line information Xk′ shown in FIG.
35
D. Next, the data sheet identification device collates the ruled-line information Xk′ with the database relating to ruled-line information of various data sheets, and identifies the data sheet X′ from among these various data sheets.
In this case, the ruled-line information Xk is different from the ruled-line information Xk′ in that an oval Ka′ portion is different from an oval Kb′ portion. In other words, while a ruled line does not exist at the oval Ka′ portion in the data sheet Xk, a ruled line exists at the oval Kb′ portion in the data sheet Xk′. Therefore, the data sheet identification device recognizes that the data sheet X and the data sheet X′ are different kinds of data sheets.
As explained above, according to the conventional data sheet identification device (PCT International Patent Publication No. WO97/05561), the data sheet identification device identifies data sheets based on ruled lines. Therefore, when the printing precision is poor on a certain data sheet, there has been a problem that this data sheet is identified by error as the same kind of data sheet as the other data sheet although they are actually different kinds of data sheets.
As a specific example, when the data sheet X′ shown in
FIG. 35C
has been printed in a state that the ruled line of the oval Kb portion has been blurred and dropped, the data sheet identification device obtains the ruled-line information Xk′ shown in
FIG. 35D
in a state that the ruled-line information at the oval Kb′ portion has been dropped. In other words, the data sheet identification device recognizes the ruled-line information Xk′ as the ruled-line information Xk (
FIG. 35B
) that is actually different from the ruled-line information Xk′. As aresult, the data sheet identification device recognizes by error that the data sheet X and the data sheet X′ are of the same of data sheets.
Further, in financial institutions and others, data sheets are also identified based on a difference between data sheet identification codes printed on data sheets, instead of based on a difference between formats like ruled lines printed on data sheets. The operation of identifying data sheets
1000
A to
1000
C shown in
FIG. 36A
to
FIG. 36C
based on data sheet identification codes will be explained next. In this case, a data sheet identification code is a 10-digit code of “customer code”.
A data sheet identification code of the data sheet
1000
A is “1234567890”, and a data sheet identification code of the data sheet
1000
B is “1234567890” which is the same as the data sheet identification code of the data sheet
1000
A. On the other hand, a data sheet identification code of the data sheet
1000
C is “9876543210” which is different from the data sheet identification codes of the data sheet
1000
A and the data sheet
1000
B. Therefore, in the financial institutions, the data sheet
1000
A and the data sheet
1000
B are handled as the same data sheets because of the same data sheet identification code.
However, among the data sheet
1000
A to the data sheet
1000
C, the ruled lines of the data sheet
1000
A are the same as the ruled lies of the data sheet
1000
C, and the ruled lines of the data sheet
1000
A and the data sheet
1000
C are different from the ruled lines of the data sheet
1000
B. Therefore, according to the conventional data sheet identification device, there has been a problem that the data sheet
1000
A and the data sheet
1000
C are identified as the same data sheets by error because of their same ruled lines although the data sheet
1000
A and the data sheet
1000
B should actually be handled as the same data sheets.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a data sheet identification device having improved identification precision.
In order to achieve the above object, according to a first aspect of the present invention, there is provided a data sheet identification device comprising: character/graphics extracting unit (corresponding to a character/graphics extracting section
50
in a first embodiment to be described later) for extracting characters (including character strings) and graphics from image information of a data sheet that has been read by image reading unit; identical shape deciding unit (corresponding to a identical shape deciding section
60
in the first embodiment to be described later) for deciding whether or not there exist a plurality of characters and graphics having the same shape among a plurality of characters and graphics that have been extracted by the character/graphics extracting unit; graphic collating unit (corresponding to a graphics collating section
80
in the first embodiment to be described later) for collating graphics that have been decided to have the same shape with a graphic database in which a plurality of graphics showing features of a plurality of data sheets respectively have been registered;
Chiba Koichi
Eguchi Shinichi
Kanamoto Kouichi
Katsumata Yutaka
Kobara Katsutoshi
Hung Yubin
Johnson Timothy M.
LandOfFree
Data sheet identification device does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Data sheet identification device, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Data sheet identification device will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3354309