JP2000311216A - Character data connecting device and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a character data connecting device/method which deletes automatically one of 1st and 2nd character data groups which are partly overlapping each other and secures the connection between two character data which are acquired from the image data by the recognition of characters. SOLUTION: This character data connecting device includes a character recognition means 90 which generates the character data by analyzing the image data and recognizing the characters, the 1st and 2nd character data storage parts 10 and 20 which can access the character data generated by the means 90 in every optional character data unit, a detection means which compares a 1st character data group stored in the part 10 with a 2nd character data group stored in the part 20 and detects the partial overlapping data detected by the detection means and a connection means 40 which deletes one of partial overlapping data and connects together the 1st and 2nd character data groups.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character data linking apparatus and method.

[0002]

2. Description of the Related Art The connection of character data simply connects the end of a first character data group and the head of a second character data group.

[0003]

However, a simple connection may be inconvenient. For example, when a long sentence such as an article in a magazine is read in a plurality of times using a handy scanner, and this image data is converted into character data by character recognition means, the character data obtained by one reading and the next reading In some cases, some of the character data obtained by the method may be duplicated. In such a case, if the character data is simply connected, the overlapping portion has to be manually checked and deleted later.

An object of the present invention is to automatically search a character data obtained by analyzing image data even if a part of the first character data group and a part of the second character data group overlap. It is another object of the present invention to provide a character data linking apparatus and method for linking two character data groups by removing one of the overlapping portions.

[0005]

SUMMARY OF THE INVENTION A character data linking device according to the present invention analyzes character data by analyzing image data to generate character data, and character data generated by the character recognition device. A first and second character data storage unit capable of accessing data in arbitrary character data units, a first character data group stored in the first character data storage unit, and a first character data group stored in the second character data storage unit. Detecting means for comparing the two character data groups to detect overlapping partial data; and removing one of the overlapping partial data detected by the detecting means to connect the first character data group and the second character data group. Connection means for performing the connection. Further, the character data linking method of the present invention is a character recognition step of generating character data by analyzing image data and performing character recognition,
Storing the character data generated in the character recognition step in the first character data storage unit or the second character data storage unit so that data can be accessed in arbitrary character data units; and storing the character data in the first character data storage unit A detecting step of comparing the extracted first character data group and the second character data group stored in the second character data storage unit to detect overlapping partial data, and an overlapping partial data detected in the detecting step. And connecting the first character data group and the second character data group.

[0006]

According to the character data linking apparatus and method of the present invention, in the character recognition stage, conversion from image data to character data is performed. Then, with respect to the obtained character data, in a detection stage, an overlapping portion between the first character data group stored in the first character data storage unit and the second character data group stored in the second character data storage unit is determined. Is detected.
In the concatenation stage, one of the overlaps is removed and the first
The character data group and the second character data group are linked as one character data.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the first to third inventions will be described below with reference to FIGS. FIG. 1 is a configuration block diagram of the present embodiment. The character data linking device 1 according to the present embodiment includes a first character data storage unit 10 storing a first character data group, a second character data storage unit 20 storing a second character data group, and a first character data storage unit 20. Data storage unit 1
0 and a reading means 30 for reading partial data of a row, a column, or a character string from the second character data storage unit 20, and a collating means 40 for collating the partial data read by the reading means 30
And a control means 50 for controlling the reading means 30 and a connecting means 60 for connecting the first character data group and the second character data group.

[0008] The first character data storage unit 10 and the second character data storage unit 20 have a data structure that allows data access in units of arbitrary partial data. The arbitrary partial data includes, for example, partial data in units of rows, partial data in units of columns, partial data in units of clauses, and partial data in units of one character. The character data connected by the linking means is stored in the character data 3 storage unit 70.

Next, the contents of the processing of this embodiment will be described with reference to the flowchart of FIG. First, the upper half of the type data 80 such as a letter is read by a handy scanner to obtain image data 81 (step 100). The image data 81 is image-analyzed by the character recognition unit 90 to generate a first character data group 120 (step 101).
Further, the lower half of the print data 80 is read by a handy scanner to obtain image data 82 (step 10).
2). The image data 82 is image-analyzed by the character recognition unit 90 to generate a second character data group 121 (step 1).
03).

The first character data group 120 is stored in the first character data storage unit 10, and the second character data group 121 is stored in the second character data storage unit 10.
Each is stored in the character data storage unit 20. The third and fourth lines of the first character data group 120 thus stored and the second
Since the first and second lines of the character data group 121 are duplicated,
One of these is removed in the subsequent processing.

To remove duplicate rows, first
The character data for each line, which is partial data, is read from the character data storage unit 10 and the second character data storage unit 20 for each line.
Read using. Specifically, first, "1" (indicating the first line) is set in a search counter indicating the number of lines of the second character data group 121 (step 104).

Next, the last line of the first character data group 120 is read (step 105). And step 10
Using the search counter set in step 4, an arbitrary line of the second character data group 121 is read (step 106).
After reading, 1 is added to the search counter (step 10).
7). Steps 106 to 111 have a loop structure, and the search counter reads character data sequentially from the first line of the second character data group 121. If character data cannot be correctly read in step 106 because the last line of the second character data group 121 has been read in this process (step 10).
8), the search counter is set to “1” (step 1).
09), the processing of the connecting means is performed.

If the last line has not been read, the processing is shifted to the collating means 40, and the first character data group 1 is read.
20 and the character data read from the second character data group 121 are collated (step 110). If the character data matches by this collation, the linking means 6
The process is shifted to 0 (step 111). If they do not match, the process returns to step 106 to read the character data of the next line of the second character data group 121. The linking means 60 links all the lines of the first character data group with the lines of the second character data group that follow the line indicated by the search counter (step 112).

In order to perform the processing of the matching means more accurately, the comparison processing may be performed using a plurality of rows. In this case, first, in step 105, the number of lines indicated by the search counter is sequentially read from the last line of the first character data group. Similarly, in step 106, the number of lines indicated by the search counter is read from the first line of the second character data group.
Then, in step 110, a plurality of lines of the first character data group and the second character data group are collated.

Next, an example of character data connected in this embodiment will be described with reference to the character data connection diagrams of FIGS.

FIG. 3A shows an example of connection in the case where the third and fourth lines of the first character data group and the first and second lines of the second character data group overlap. In this example, the third character data is created by connecting the third line of the second character data group to the end of the first character data group. The same applies to FIG.

FIG. 4A shows an example of connection in the case where the third to fifth columns of the first character data group and the first to third columns of the second character data group overlap. In this example, the third character data is created by connecting the fourth column of the second character data group to the end of the first character data group. FIG. 4B shows a connection example in the case where the 17th to 19th columns of the first character data group and the 1st to 3rd columns of the second character data group overlap. In this example, the third character data is created by connecting the fourth column of the second character data group to the end of the first character data group.

FIG. 5A shows an example of a concatenation in a case where the character string "Early Summer" of the first character data group does not exist in the second character data group. In this example, at the end of the first character data group,
The third character data is created by connecting the head of the second character data group. FIG. 5B shows that the character string “Epson” of the sixth word of the first character data group is the fourth character string of the second character data group.
It is a connection example in the case of overlapping with the word. In this example, at the end of the first character data group, 4
The third character data is created by connecting the words.

Next, a hardware configuration when the present embodiment is realized by a personal computer will be described with reference to FIG. The personal computer 210 includes a CPU 211 operating as the control unit 50,
2 in which a program such as a memory and a collating means 40 is stored
12 and a RAM 213 storing a system program such as a BIOS. The personal computer 210 includes a CRT 220 for displaying a first character data group, a second character data group, and the like; a keyboard 221 for inputting a command from a user; And a magnetic disk 223 for storing character data input by the handy scanner 222 or the like.

This embodiment is particularly applicable to the handy scanner 222.
This is effective when concatenating a plurality of character data input using. That is, since the width of the input unit of the handy scanner 222 is small, it is usually necessary to scan the character data several times. In this scanning, input is generally performed while overlapping end portions so that input omission does not occur. By using this embodiment, it is possible to automatically delete the duplicated portion generated at this time.

It should be noted that whether partial data of a row, a column, or a character string is used in this embodiment may be determined by a command from a user, or two or more partial data may be used together. As an example of using them together, there is a means for first collating overlapping portions on a row-by-row basis, then collating overlapping portions on a column-by-column basis, and linking.

[0022]

According to the character data linking apparatus and method of the present invention, when a part of a plurality of character data overlaps,
It is possible to automatically search for an overlapping portion, remove one of the overlapping portions, and concatenate character data.

Therefore, when linking a plurality of character data, the user does not need to manually delete the overlapped portion. For this reason, it is possible to prevent occurrence of a correction error, and to save labor and time.

In particular, as an application of the character data linking apparatus of the present invention, a long sentence such as a magazine article is read a plurality of times by a handy scanner, and this image data is converted into character data by a character recognition means. This is effective when the original sentence is reproduced by linking a plurality of character data obtained as described above.

[Brief description of the drawings]

FIG. 1 is a configuration block diagram of a present embodiment.

FIG. 2 is a flowchart illustrating the contents of a process according to the embodiment.

FIG. 3 is a conceptual diagram illustrating a character data linking process according to the embodiment.

FIG. 4 is a conceptual diagram illustrating a process of connecting character data according to the present embodiment.

FIG. 5 is a conceptual diagram illustrating a process of linking character data according to the embodiment.

FIG. 6 is a hardware configuration diagram of the present embodiment.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Character data coupling device, 10 ... First character data storage part, 20 ... Second character data storage part, 30 ... Reading means, 4
0: verification means, 50: control means, 60: coupling means, 90
… Character recognition unit

Claims (2)

[Claims] 1. Character recognition means for generating character data by analyzing image data and performing character recognition; and accessing the character data generated by the character recognition means in arbitrary character data units. First and second character data storages, and a first character data group stored in the first character data storage and a second character data group stored in the second character data storage are compared and overlapped Detecting means for detecting partial data to be copied, and connecting means for removing one of the overlapping partial data detected by the detecting means and connecting the first character data group and the second character data group. Character data concatenation device. 2. A character recognition step for generating character data by analyzing image data and performing character recognition, and allowing the character data generated in the character recognition step to be accessed for each character data unit. Storing in the first character data storage unit or the second character data storage unit; and storing the first character data group stored in the first character data storage unit and the second character data group stored in the second character data storage unit. A detecting step of comparing the character data group to detect overlapping partial data; and removing one of the overlapping partial data detected in the detecting step to remove the first character data group and the second character data group. Concatenating the character data.