JP2004206520A - Document image distribution system and device, terminal device, and document image distribution program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a document image distribution system and device, a terminal device, and a document image distribution program, for searching and distributing formatted documents via a network while reducing the amount of data sent and received between devices. <P>SOLUTION: A keyword search is performed over the search area of a document image and an image of the search area on a page where there is a string of characters as the result of a character recognition process that match the keyword is distributed. The user checks whether or not the keyword is described in the document distributed. If the keyword is described there, the user requests the document image distribution device to distribute the entire image of the document. The document image distribution device distributes the entire image of the document. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a document image delivery system, a document image delivery device, a terminal device, a document image delivery program, and an image creation program, and more particularly to a document image delivery system, a document image delivery device, and a terminal that output a document image of a page including a search keyword. The present invention relates to an apparatus, a document image distribution program, and an image creation program.
[0002]
[Prior art]
One method of extracting necessary information from information written on paper or the like represented by a book or a document is a method using an OCR (optical character reading device). In this method, an image written on paper or the like is read, and character data extraction processing of the read image and character recognition processing of the extracted character part are performed to convert the data into text data. Also, the image data of the read image is held. Then, a character string is searched using the text data, and image data is output using the search result.
[0003]
As an apparatus used for such a search, an apparatus that complements the accuracy of OCR using a search keyword input by a user and a similar keyword that is similar to the search keyword has been proposed (for example, Patent Document 1). ).
[0004]
In addition, in addition to performing matching between the character recognition processing result of the document and the keyword used in the search, which is performed at the time of the search, with respect to the character of the character recognition processing result and the keyword character, the character of the character recognition processing result and the keyword are compared. There is also a document search method and apparatus for extracting characters that are likely to be erroneously recognized as characters and extracting a character resulting from the character recognition processing as a search result if any of the characters is found (for example, Patent Document 2).
[0005]
[Patent Document 1]
JP 2000-305935 A (Page 4-9, FIG. 1)
[Patent Document 2]
JP-A-7-152774 (page 4-9, FIG. 1)
[0006]
[Problems to be solved by the invention]
There is a case where a server holds image data of a document, a user searches for the document via a network which is a communication line, and a client terminal connected to the network, and there is a case where the user wants to retrieve the image data of the search result. At this time, the server searches for a document and transmits image data in response to a request from the client terminal.
[0007]
However, the amount of image data is large. Even if a document search process is performed by the client terminal and only documents that match the search conditions among all the documents are received from the server, a load is applied to the network.
[0008]
For example, when a document image of 100 pages is extracted as a result of the search, the server transmits the image data of 100 pages to the client terminal. Assuming that the data amount of one page is 200 KB, the total amount of data transmitted and received on the network is 20 MB, which increases the load on the network.
[0009]
The document filing apparatus described in Patent Document 1 discloses a method for complementing the accuracy of OCR. However, there is no viewpoint of connecting the document filing device to a network as a document image search server, and no method for reducing the output data amount of the document image is disclosed.
[0010]
The method and apparatus described in Patent Literature 2 disclose a method for preventing search omission due to an error in character recognition processing. However, there is no viewpoint for connecting to a network, and a method for reducing the amount of output data of a document image. Does not disclose anything.
[0011]
When the devices described in Patent Literature 1 and Patent Literature 2 are connected to a network and used as a server, there is a problem that the load on the network is large. In particular, in the case of image data in a fixed format, the area required for the search may be limited to a part of the entire format, and even in such a case, the entire image data is stored in the server every time the search is performed. If the data is transmitted from the client to the client terminal, the load on the network increases accordingly.
[0012]
Therefore, the present invention provides a document image distribution system, a document image distribution device, a terminal device, and a device that can reduce the amount of data transmitted and received between devices when searching and distributing documents in a fixed format via a network. A document image distribution program and an image creation program are provided.
[0013]
[Means for Solving the Problems]
A document image distribution system according to the present invention includes a document image distribution device that distributes image data according to a search keyword, and image data that transmits a search keyword to a document image distribution device via a communication network and receives from the document image distribution device. And a terminal device for displaying a document image. The document image distribution device reads a document in a fixed format, and searches for image data of the entire document and a search target area of the entire document. A document reading unit that generates image data of the area and performs character recognition processing on the search area to create character candidate data of characters described in the search area; and a terminal device based on the character candidate data. An area image data specifying unit for specifying image data of a search area corresponding to the received search keyword; Area image data transmitting means for transmitting image data of the search area specified by the means to the terminal device; and overall image data transmitting means for transmitting image data of the entire document designated by the terminal device to the terminal device. And
[0014]
The area image data transmitting means may, when the area image data specifying means specifies the image data of the plurality of search areas, sequentially deliver the image data of each search area to the terminal device. The image data of the search area received from the means may be sequentially displayed. According to such a configuration, the user can sequentially browse image data corresponding to the search keyword.
[0015]
A document image distribution device according to the present invention is a document image distribution device that receives a search keyword from a terminal device via a communication network, and distributes image data according to the search keyword. Generates the entire image data and the image data of the search area, which is the area to be searched in the entire document, and performs character recognition processing on the search area, thereby performing character recognition processing on the search area. Document image reading means for creating a search area, area image data specifying means for specifying image data of a search area corresponding to a search keyword received from a terminal device based on character candidate data, and a search area specified by the area image data specifying means. Area image data transmission means for transmitting the image data of the document to the terminal device, and image data of the entire document designated by the terminal device Characterized in that an entire image data transmitting means for transmitting to the terminal device.
[0016]
A document image distribution program according to the present invention is a document image distribution program installed in a document image distribution device that receives a search keyword from a terminal device via a communication network and distributes image data according to the search keyword. Then, image data of the entire document and image data of a search area, which is a search target area in the entire document, are generated from the information obtained by reading the document in the fixed format, and a character recognition process is performed on the search area. Processing to create character candidate data of characters described in the search area, processing to specify image data of the search area corresponding to the search keyword received from the terminal device based on the character candidate data, A process of transmitting image data to the terminal device, and a process of transmitting image data of the entire document designated by the terminal device to the terminal device. Characterized in that to execute a process of transmitting to.
[0017]
A document image distribution system according to the present invention includes a document image distribution device that distributes image data according to a search keyword, and image data that transmits a search keyword to a document image distribution device via a communication network and receives from the document image distribution device. And a terminal device for displaying a document image. The document image distribution device reads a document in a fixed format and generates image data of a search area that is a search target area in the entire document. Document reading means for creating character candidate data of a character described in a search area by performing a character recognition process on the search area, and a search corresponding to a search keyword received from a terminal device based on the character candidate data An area image data specifying unit that specifies the image data of the area, and a search area that is specified by the area image data specifying unit. A region image data transmitting unit that transmits image data to the terminal device, the terminal device includes a basic image data storage unit that stores image data of at least a region other than the search region in the entire region of the document in the fixed format, Image data combining means for combining the image data of the search area received from the area image data transmission means with the image data stored in the basic image data storage means to create image data of the entire document. I do.
[0018]
When the area image data specifying means specifies the image data of the plurality of search areas, the area image data transmitting means may sequentially transmit the image data of each search area to the terminal device. The image data of the search area received from the means may be sequentially displayed. According to such a configuration, the user can sequentially browse image data corresponding to the search keyword.
[0019]
The image data synthesizing unit of the terminal device synthesizes the image data specified by the user of the terminal device and the image data stored in the basic image data storage unit out of the image data of the search area received from the area image data transmitting unit. May be. According to such a configuration, when image data of only the search area is received, the entire image data can be displayed without receiving image data of an area other than the search area.
[0020]
A document image distribution device according to the present invention is a document image distribution device that receives a search keyword from a terminal device via a communication network, and distributes image data according to the search keyword. Document reading means for generating image data of a search area that is a search target area of the whole, and performing character recognition processing on the search area to create character candidate data of characters described in the search area, Area image data specifying means for specifying image data of a search area corresponding to a search keyword received from the terminal apparatus based on the character candidate data, and transmitting the image data of the search area specified by the area image data specifying means to the terminal apparatus And an area image data transmitting unit that performs the operation.
[0021]
A terminal device according to the present invention receives image data of a search area according to a search keyword from a document image distribution device that distributes image data of a search area that is a search target area in a whole fixed format document, A terminal device for displaying data, wherein a basic image data storage means for storing image data of at least an area other than a search area in a whole area of a document in a fixed format, and a search area received from an area image data transmission means. An image data synthesizing unit for synthesizing image data and image data stored in the basic image data storage unit to create image data of the entire document is provided.
[0022]
A document image distribution program according to the present invention is a document image distribution program installed in a document image distribution device that receives a search keyword from a terminal device via a communication network and distributes image data according to the search keyword. Then, from information obtained by reading a document in a fixed format, image data of a search area, which is a search target area in the entire document, is generated, and the search area is described in the search area by performing a character recognition process. A process of creating character candidate data of a character, a process of specifying image data of a search area corresponding to a search keyword received from a terminal device based on the character candidate data, and transmitting the image data of the specified search region to the terminal device Is performed.
[0023]
An image creation program according to the present invention receives image data of a search area according to a search keyword from a document image distribution device that distributes image data of a search area which is an area to be searched out of a whole document in a fixed format, An image creation program installed in a terminal device for displaying image data, wherein a process of causing a computer to store, in a storage device, image data of at least an area other than a search area in a whole area of a fixed format document, and receiving the image data. The image data of the search area to be searched is combined with the image data stored in the storage device to execute a process of creating image data of the entire document.
[0024]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0025]
Embodiment 1 FIG.
FIG. 1 is a block diagram showing a first embodiment according to the present invention. The document image distribution device 8 in this embodiment is an information processing device such as a server, and is used for searching and distributing a document in a fixed format.
[0026]
The document image distribution device 8 is connected to a client terminal 7 via a communication line 6 such as the Internet. The document image distribution device 8 sets an area to be searched (hereinafter, referred to as a search area) in the area of the fixed format. Then, the data storage unit 5 in the document image distribution device 8 stores the entire image data of each document and the text data of the search area of each document so that the document data can be identified. . When receiving the search keyword from the client terminal 7, the document image distribution device 8 searches the text data for a character string that matches the search keyword. Then, image data obtained by extracting a search area from the entire image data of the text data document including a character string that matches the search keyword is transmitted from the data storage unit 5 to the client terminal 7. When the client terminal 7 requests the entire image data of a specific document, the document image distribution device 8 transmits the entire image data of the document requested by the client terminal 7.
[0027]
FIG. 2 shows an example of a fixed format document. The document 30 shown in FIG. 2 is a fixed format document. FIG. 2 shows an example in which a description field of a management code is used as a search area. When storing the document data of the “management code RS”, the document image distribution device 8 stores the entire image data of the document 30 shown in FIG. 2 and the data of the character string included in the search area 31. The same data is stored for a document in which another management code is described. When the search keyword is input, the document image distribution device 8 specifies text data including a character string that matches the search keyword, and transmits the image data of the search area 31 corresponding to the text data to the client terminal 7. Therefore, the client terminal 7 that has transmitted the search keyword displays the image of the search area 31. When the user browses not only the search area 31 but also the entire document 30, the client terminal 7 requests the entire image data of the document 30 from the document image distribution device 8 according to the operation of the user, and the entire document 30 is Is displayed.
[0028]
In the document image distribution device 8, the scanner 1 is an input device for reading information described on a medium such as paper. The program storage unit 3 is a storage device that stores a document image distribution program according to the present invention. The communication interface 4 transmits and receives information to and from the client terminal 7 via the communication line 6. The data storage unit 5 is a storage device that stores data created by the control unit 2. The data storage unit 5 stores a character recognition dictionary used in the character recognition processing.
[0029]
The control unit 2 executes processing according to the document image distribution program stored in the program storage unit 3. The control unit 2 causes the scanner 1 to read a document in a predetermined format. Then, the whole image data which is the image data of the whole document of the information read by the scanner 1 is created. The control unit 2 is designated by the user with a search area. In each document of the same fixed format, the search area is located in the same area.
[0030]
Further, the control unit 2 performs a character recognition process on the information in the search area specified by the user, and creates text data. At this time, the control unit 2 determines whether the search area of the information read by the scanner 1 is vertical writing or horizontal writing, and uses a number indicating the number of the character on the page for each character in the text data. Set a certain character number. Further, the control unit 2 specifies one or more character candidates that are candidates for each character in the text data. In addition, at the time of identification, character recognition evaluation information indicating the degree of certainty that each character candidate is the same character as each character in the document is generated. Further, the control unit 2 generates character position information indicating the position of each character candidate in the entire image data.
[0031]
The control unit 2 outputs the whole image data, the character number, the character candidates, the character recognition evaluation information of each character candidate, and the character position information to the data storage unit 5 and stores them. At this time, based on the character recognition evaluation information, the control unit 2 stores the character candidates in the data storage unit 5 with the order of the degree of certainty that the characters are the same as the characters in the document.
[0032]
When receiving the search keyword from the client terminal 7, the control unit 2 checks whether there is a character string candidate that matches the search keyword. Here, the character string candidate is a character string of a character candidate determined for each character, and is a character string that matches the search keyword. If there is a character string candidate that is a character string that matches the search keyword, the entire image data of the document including the character string that is the basis of the character string candidate is extracted from the data storage unit 5. The document including the character string that is the source of the character string candidate can be determined using information indicating which document the character string candidate is for. The control unit 2 cuts out the image data of the search area from the entire image data and creates image data of the search area (hereinafter, referred to as area image data). The control unit 2 outputs the area image data to the client terminal 7 via the communication line 6. Further, the control unit 2 outputs the entire image data of the designated document in response to a request from the client terminal 7.
[0033]
The client terminal 7 is, for example, an information processing terminal such as a personal computer. The client terminal 7 transmits the search keyword input by the user to the document image distribution device 8, and displays the area image data received from the document image distribution device 8. When the region image data corresponding to the search keyword is displayed, the client terminal 7 requests the entire image data of the page of the displayed region image data from the document image distribution device 8, and The received whole image data is displayed.
[0034]
The document reading unit, the area image data specifying unit, the area image data transmitting unit, and the entire image data transmitting unit are realized by the control unit 2. The control unit 2 functions as each unit by the document image distribution program stored in the program storage unit 3.
[0035]
The document image distribution program is a document image distribution program installed in a document image distribution device that receives a search keyword from a terminal device via a communication network and distributes image data according to the search keyword. Generates image data of the entire document and image data of a search area, which is a search target area of the entire document, from the information obtained by reading the document in the format, and performs character recognition processing on the characters described in the search area. Performing a process of generating character candidate data of a character, a process of specifying image data of a search area corresponding to a search keyword received from the terminal device based on the character candidate data, and a step of transmitting the image data of the specified search area to the terminal. Processing for transmitting to the device and processing for transmitting image data of the entire document designated by the terminal device to the terminal device It is intended to execute.
[0036]
Next, the operation of the present embodiment will be described. First, a description will be given of how information written on a medium such as paper is stored in the data storage unit 5. Here, a management code is described in a specific area of each page in a document of a fixed format composed of a document of 10 pages, and as shown in FIG. The case where the number is 31 will be described as an example. It is assumed that the management code “RS” is described in the search area 31 of the first page. The character string described in the search area 31 of each page and used for the search may be, for example, a character string such as a person's name or a number such as a management number, as long as it can be input as a search keyword. Good.
[0037]
The scanner 1 first reads information written in the first page of the document, and outputs the read information to the control unit 2. The control unit 2 creates overall image data that is image data of the entire document of the first page read by the scanner 1. Then, the entire image data is output to and stored in the data storage unit 5 after it is determined that the image data is the image data of the first page. Further, the control unit 2 sets a search area 31 which is a search target area on the image data according to a user's instruction. Here, the area in which the management number “RS” is written is set as the search area 31.
[0038]
The control unit 2 creates position information of the search area 31 in the entire image data. As shown in FIG. 3, for example, the position information of the search area 31 includes an X coordinate (for example, an X coordinate of an upper left position of the search area 31), a width, a Y coordinate (for example, a Y coordinate of an upper left position of the search area 31), Although indicated by the vertical width, any information may be used as long as the position of the search area 31 in the entire image data is known. The control unit 2 outputs and stores the position information of the search area 31 to the data storage unit 5.
[0039]
The control unit 2 performs a character recognition process on the information in the search area 31 among the information read by the scanner 1. At this time, the control unit 2 divides the information in the search area read by the scanner 1 into units of one character and separates the information into individual characters. Then, character recognition processing is performed on each character, and character candidates that can be candidates for the result of the character recognition processing are specified. Thus, character candidate data is created. Then, it is determined whether the information in the search area is in vertical writing or horizontal writing, and a character number is set for each character to indicate the number of the character on the page.
[0040]
In this example, since "RS" is written in the search area 31, these two characters are separated from each other, and a character candidate is specified for each of these two characters. In FIG. 2, a character “management code” is also described in the search area 31. However, since this character is common to all pages, in the following description, only the character following the character is used for character recognition. Target. Then, the character number of the first character “R” in the search area is 1 and the character number of the second character “S” is 2. Here, the number of character candidates for each character in the document is not limited to one. The likelihood that each character candidate matches a character actually described in the document changes depending on the description state of the character in the document. Accordingly, character candidates for the character "R" in the document include "R", "P", and the like, and the probability of the character candidate "P" may be determined to be higher. In the following description, a certain character X actually described in a document is referred to as a “character“ X ”in a document”.
[0041]
The control unit 2 checks the characters read one character at a time against characters in a character recognition dictionary stored in the data storage unit 5 in advance. Then, a character having a large matching element is selected from the character recognition dictionary and set as a character candidate. The number of character candidates is not limited, and the number of character candidates may be different for each character of each image data. For example, it is assumed that there are four character candidates for the character “R” in the document, “P”, “R”, “8”, and “B”, and the character candidates for the character “S” in the document are “8”, “ It is assumed that there are five pieces of "S", "3", "R", and "u".
[0042]
Further, the control unit 2 generates character recognition evaluation information indicating the degree of certainty that each character candidate is the same character as each character in the document. The character recognition evaluation information is determined based on the degree of matching by comparing characters in a character recognition dictionary stored in the data storage unit 5 in advance with characters read by the scanner 1 and cut out by the control unit 2. Here, as an example of the character recognition evaluation information, FIG. 4 shows an example of evaluation using character recognition evaluation values expressed by numerical values. In FIG. 4, the first line shows an example of the character recognition evaluation value of each character candidate of the character "R" in the document, and the second line shows the character recognition evaluation value of each character candidate of the character "S" in the document. Is shown. When the character recognition evaluation value is used, each character candidate can be given a candidate ranking. For example, the character candidate with the first candidate rank of the character “R” in the document is “P”, and the character candidate with the second highest rank is “R”.
[0043]
Further, the control unit 2 generates character position information indicating the position of each character candidate in the entire image data. In this example, the character position information is indicated by an X coordinate (for example, an X coordinate of the upper left position of the character area), a character width, a Y coordinate (for example, a Y coordinate of the upper left position of the character area), and a character height. Any image may be used as long as the position of the character in the image data is known.
[0044]
Then, the control unit 2 outputs each character number, each character candidate, each character recognition evaluation value, and each character position information to the data storage unit 5, and the data storage unit 5 outputs the data for the first page. Is stored in such a form that it can be understood that
[0045]
Next, the reading operation for the second and subsequent pages will be described. The scanner 1 reads information written on each of the second and subsequent pages of the document, and outputs the read information to the control unit 2. The control unit 2 creates overall image data of each page read by the scanner 1. Then, the entire image data of each page is output to the data storage unit 5 and stored in such a manner that the image data of which page is known. Note that the position information set when reading the first page can be used as the position information of the search area, and there is no need to newly create the position information here.
[0046]
Hereinafter, similarly to the first page, the control unit 2 performs a character recognition process on characters in the search area of each page, and determines a character number, a character candidate, a character recognition evaluation value, and character position information. Then, the control unit 2 outputs each character number, each character candidate, each character recognition evaluation value, and each character position information to the data storage unit 5, and the data storage unit 5 stores the The data is stored after being understood.
[0047]
FIG. 5 is an explanatory diagram showing a specific example when the present invention is applied to a document in a fixed format. In the example shown in FIG. 5, the area described as “2002-11-29 back number” in the back number document as the fixed format document 31 as shown in FIG. Is specified. Accordingly, when the first page of the back number document is read, the control unit 2 detects the X coordinate, the horizontal width, the Y coordinate, and the vertical width of the search area 31 and determines the position as shown in FIG. Create information. In the example shown in FIG. 5B, “200”, “300”, “30”, and “50” are detected as the X coordinate, the horizontal width, the Y coordinate, and the vertical width of the search area 31. Have been.
[0048]
Further, the control unit 2 divides the search area 31 into individual character areas, and detects coordinates and sizes (character width and character height) for each character area. Further, a character candidate is specified by performing a character recognition process. In the example shown in FIG. 5C, the first character candidate in the candidate order of the first character is “2”, and the second character candidate in the candidate order is “Otsu”. The first character candidate in the candidate order of the second character is “0”, and the second character candidate in the candidate order is “o”.
[0049]
Next, a search operation in which a user inputs a search keyword to be searched and image data including the search keyword is output will be described with reference to FIG. FIG. 6 is a flowchart for explaining the operation at this time. Here, it is assumed that the user has requested information written on the page of the management code “RS”.
[0050]
The client terminal 7 receives an input of a search keyword from a user (step S101). As an example, assume that the user inputs “RS”. The client terminal 7 transmits the search keyword “RS”, which is a character string input by the user, to the document image distribution device 8 via the communication line 6 (Step S102). In the document image distribution device 8, the communication interface 4 receives and outputs it to the control unit 2.
[0051]
The control unit 2 specifies a character string candidate that is a character string that matches the search keyword by combining the character candidates stored in the data storage unit 5 for all pages (step S103). At this time, the control unit 2 extracts character candidates one by one for each character in the search area stored in the data storage unit 5 and combines them in the order of the character numbers to generate a character string in order to eliminate omission in the search. It is determined whether or not it matches the search keyword. As a result, since the combination of the character candidate “R” and the character candidate “S” on the first page matches the search keyword, the combination becomes the character string candidate “RS”.
[0052]
The control unit 2 determines on which page the character string candidate is located based on the information stored in the data storage unit 5 (step S104). Taking the document 31 shown in FIG. 2 as an example, it can be seen that the character string candidate “RS” exists on the first page, so the control unit 2 reads the entire image data of the first page from the data storage unit 5. The extracted image data of the search area 31 (hereinafter, referred to as area image data) extracted from the entire image data based on the position information of the search area 31 is transmitted to the client via the communication interface 4 and the communication line 6. Output to the terminal 7 (step S105). As a result, “RS” is written in the output area image data. Thus, the area image data is specified and output.
[0053]
If another character string candidate exists, the control unit 2 also outputs, to the client terminal 7, area image data cut out from the entire image data of the page where the other character string candidate exists. Here, the order of outputting the region image data when outputting a plurality of region image data will be described.
[0054]
When there are a plurality of character string candidates, the control unit 2 pays attention to the candidate ranks of the character candidates constituting the character string candidates (that is, the character candidates that match the characters in the search keyword), and averages the candidate ranks. The image data is output in order of decreasing value. At this time, the control unit 2 reads the character recognition evaluation value of each character candidate stored in the data storage unit 5 and analyzes each character string candidate using the candidate ranking based on the character recognition evaluation value.
[0055]
Specifically, the candidate rank of the character candidate “R” on the first page is second, the candidate rank of the character candidate “S” is second, and the average of the candidate rank is 2.0. For example, if there is a character "B" in the document having the character candidate "R" in the first candidate rank on the seventh page and a character "8" in the document having the character candidate "S" in the second candidate rank, the character The average value of the candidate rank of column “B8” is 1.5. In addition, when a document character “P” having a character candidate “R” in the third place and a document character “G” having a character candidate “S” in the second place are present on the ninth page, The average value of the candidate rank of the character string “PG” is 2.5. Among these three character candidates “R” and “S”, “B8” which is the area image data of the seventh page having the smallest candidate ranking average value is output first. Next, “RS”, which is the area image data of the first page, and “PG”, which is the area image data of the ninth page, are finally output. The client terminal 7 receives and displays the area image data in the order output by the control unit 2.
[0056]
When there are a plurality of character string candidates, the control unit 2 continuously outputs the area image data to the client terminal 7. Here, the control unit 2 outputs the image data in ascending order of the average value of the candidate rank. However, the present invention is not limited to this, and the degree of certainty of the character recognition process of each character candidate is high using the character recognition evaluation information and the like. Output them in order. Then, on the client terminal 7, the characters are continuously displayed in the descending order of the likelihood of the character recognition processing. Note that the output order may be output in the order of smaller page numbers without performing any special processing.
[0057]
The user checks whether or not the search keyword “RS” is written in the area image data displayed by the client terminal 7. Here, the area image data of the seventh page is displayed first, the area image data of the first page is displayed second, and the area image data of the ninth page is displayed third.
[0058]
In this example, since “RS” is written in the area image data of the first page, the user knows that the first page has desired information. Then, the client terminal 7 requests the document image distribution device 8 to output the entire image data of the first page (step S106). At this point, when the user has displayed the area image data on the first page, which is the area image data displayed second, the user has found that the desired information is on the first page. The display of the first region image data may be interrupted, and an output of the entire image data of the first page may be requested.
[0059]
In the document image distribution device 8, the control unit 2 receives a request from the client terminal 7 via the communication interface 4. In response to a request from the client terminal 7, the control unit 2 retrieves the entire image data of the first page from the data storage unit 5 and transmits it to the client terminal 7 via the communication interface 4. The entire image data of the page is displayed (step S107). Thus, the entire image data is output.
[0060]
In this example, the data transmitted from the document image distribution device 8 to the client terminal 7 includes three area image data and one entire image data. When the conventional technique is used, three pieces of whole image data are transmitted because the whole image data of all documents meeting the search condition is transmitted without using the area image data. Here, for example, assuming that one area image data is 10 KB and one entire image data is 200 KB, when the conventional technique is used, 600 KB of data must be transmitted. The document image distribution device 8 only transmits 230 KB data, and can reduce the load on the network and shorten the data transmission time.
[0061]
In addition, if transmission of the same amount of data as that of the related art is permitted, it is possible to transmit higher-resolution area image data. For example, if image data having a data amount 10 times larger per unit area is transmitted for only the search area, one area image data becomes 100 KB. Then, a data amount of 500 KB is required to transmit three area image data and one entire image data. Therefore, area image data with higher resolution can be transmitted with a smaller amount of data than in the related art.
[0062]
Embodiment 2 FIG.
FIG. 7 is a block diagram showing a second embodiment according to the present invention. The document image distribution device 20 in this embodiment is an information processing device such as a server, and has a fixed format in which an area of information common to each document and an area of difference exist at fixed positions. Used for searching and delivering documents. In this embodiment, the form of connection between the document image distribution device 20, the client terminal 10, and the communication line 6 is the same as in the first embodiment. Here, the same blocks as those in the first embodiment are denoted by the same reference numerals as in FIG. 1, and description thereof will be omitted. The difference between the client terminal 10 and the first embodiment is that the client terminal 10 has a basic image storage unit 11 that stores image data, and that the area image data and the image data that the basic image storage unit 11 stores are combined. And an image data synthesizing unit 12 that performs the following. The image data synthesizing unit 12 is realized by, for example, a CPU that executes processing according to a document image distribution program.
[0063]
The document image distribution program according to the present embodiment is a document image distribution program installed in a document image distribution device that receives a search keyword from a terminal device via a communication network and distributes image data according to the search keyword. A computer that generates image data of a search area that is a search target area of the entire document from information obtained by reading a document in a fixed format, and performs a character recognition process on characters described in the search area. Processing for creating character candidate data of a character, processing for identifying image data of a search area corresponding to a search keyword received from the terminal device based on the character candidate data, and image data of the identified search area to the terminal device. This is for executing the transmission process. The control unit 21 of the document image distribution device 20 performs processing according to the document distribution program.
[0064]
The client terminal 10 performs processing according to an image creation program stored in a storage device (not shown). The image creation program receives image data of a search area corresponding to a search keyword from a document image distribution device that distributes image data of a search area, which is an area to be searched, in an entire document in a fixed format, and Processing for causing a computer to store, in a storage device, image data of at least an area other than a search area in a whole area of a fixed format document, and a search to be received. The image data of the area is combined with the image data stored in the storage device to execute a process of creating image data of the entire document.
[0065]
The image data stored in the basic image storage unit 11 may be the entire image data of the document of a specific page, or may be the common image data of the area indicating information common to the documents. Examples of a document having an area indicating common information between documents include a certificate of diploma, a graduation certificate, a blank travel expense report, and the like. All of the above-mentioned examples differ only in the area where the serial number such as the student registration number and the name are described, and the information of the other areas is common.
[0066]
The basic image storage unit 11 stores the entire image data or the common image data in advance. The image data combining unit 12 combines the area image data received from the document image distribution device 20 with the image data stored in the basic image storage unit 11. The document image distribution device 20 differs from the configuration of the first embodiment in that the control unit 21 does not create the entire image data and that the image display position information is stored in the data storage unit 5. .
[0067]
The operation of the present embodiment will be described. Here, the document shown in FIG. 2 will be described as an example. First, a description will be given of how information written on a medium such as paper is stored in the data storage unit 5. As an example, a management code is described in a specific area of each page of a 10-page document, and a search is performed using this management code. It is also assumed that the basic image storage unit 11 of the client terminal 7 has previously stored common image data which is image data of an area having no difference between pages used in this example. That is, image data of an area other than the search area 31 in the document shown in FIG. 2 is stored. However, this area shall be left blank. It is also assumed that a management code “RS” is described in the search area of the document on the first page. The character string used for the search is described in the search area of each page, and may be a character string such as a person's name or a number such as a management number as long as it can be input as a search keyword.
[0068]
The scanner 1 first reads information written in the first page of the document, and outputs the read information to the control unit 21. At this time, the scanner 1 may read the information of the entire document on the first page, or may read only the area specified by the user. When the scanner 1 reads the entire first page of the document, the control unit 21 creates a region image data by setting a search region, which is a search target region, on the entire image data in accordance with a user's instruction. When only the area indicated by the above is read, the read area is set as a search area and area image data is created. Here, it is assumed that the area in which the management number “RS” is written is set as the search area, and the area image data is created. The control unit 21 outputs the created area image data to the data storage unit 5 after storing it so as to be recognized as the image data of the first page. . Here, in this embodiment, unlike the first embodiment, only the area image data is stored without storing the entire image data.
[0069]
Further, the control unit 21 creates image display position information which is information on where the search area is located in the document of the first page. The control unit 21 outputs the created region image data of the first page and the image display position information to the data storage unit 5 and stores them. Note that the image display position information is represented by an X coordinate, a horizontal width, a Y coordinate, and a vertical width as shown in FIG. 3, for example, as in the first embodiment. Any method that can be understood may be used.
[0070]
The control unit 21 performs a character recognition process on information in the area set as the search area. At this time, the control unit 21 divides the information in the search area read by the scanner 1 into units of one character, and separates the information into individual characters. Then, character recognition processing is performed on each separated character, character candidates that can be candidates for the character recognition processing result are specified, and character recognition evaluation values of the specified character candidates and character position information are set and data storage is performed. It is stored in the unit 5. Thus, character candidate data is created. Here, the method of setting the character number, the method of specifying the character candidate, and the method of setting the character recognition evaluation value of the specified character candidate and the character position information are the same as in the first embodiment. .
[0071]
The control unit 21 outputs each character number, each character candidate, each character recognition evaluation value, and each character position information to the data storage unit 5, similarly to the control unit 2 of the first embodiment. The data storage unit 5 stores each of them.
[0072]
Next, the reading operation for the second and subsequent pages will be described. The scanner 1 reads information written on each page of the second and subsequent pages of the document, and outputs the read information to the control unit 21. The control unit 21 creates area image data of each page from information of each page read by the scanner 1. As in the method of creating the region image data of the first page, the image data of the search region may be cut out from the entire image data, or only the search region may be read by the Image data may be created. The position of the search area at this time is the same position as the search area of the first page. The control unit 21 outputs the data to the data storage unit 5 after storing the created region image data of each page as the image data of which page. Note that the image display position information set when reading the first page can be used, and thus it is not necessary to newly create the image display position information.
[0073]
Hereinafter, similarly to the control unit 2 of the first embodiment, the control unit 21 performs a character recognition process on the characters in the search area of each page, and stores a character number, a character candidate, a character recognition evaluation value, and character position information. decide. Then, the control unit 21 outputs each character number, each character candidate, each character recognition evaluation value, and each character position information to the data storage unit 5, and the data storage unit 5 stores the data for each page. Is stored after being understood.
[0074]
Further, the client terminal 10 stores the common image data of an area common to each document (an area other than the search area 31) in the basic image storage unit 11. The client terminal 10 may acquire the common image data from the document image distribution device 20, for example. That is, the document image distribution device 20 may create the common image data, transmit it to the client terminal 10, and store it in the basic image storage unit 11.
[0075]
Next, a search operation in which a user inputs a search keyword to be searched and output image data including the search keyword will be described with reference to FIG. FIG. 8 is a flowchart for explaining the operation at this time. Here, it is assumed that the user has requested area image data of a page in which the management code “RS” is written.
[0076]
The client terminal 10 receives an input of a search keyword from a user (step S201). As an example, assume that the user inputs “RS”. The client terminal 10 transmits the search keyword “RS”, which is a character string input by the user, to the document image distribution device 20 via the communication line 6 (Step S202). In the document image distribution device 20, the communication interface 4 receives and outputs the received data to the control unit 21.
[0077]
The control unit 21 specifies a character string candidate that is a character string that matches the search keyword by combining the character candidates stored in the data storage unit 5 for all pages (step S203). At this time, the control unit 21 extracts character candidates one by one for each character in the search area stored in the data storage unit 5 and combines them in the order of the character numbers to generate a character string in order to eliminate search omission. It is determined whether or not it matches the search keyword. As a result, since the combination of the character candidate “R” and the character candidate “S” on the first page matches the search keyword, the combination becomes the character string candidate “RS”.
[0078]
The control unit 21 determines on which page the character string candidate exists based on the information stored in the data storage unit 5. (Step S204). Since it is known that the character string candidate “RS” exists on the first page, the control unit 21 extracts the area image data of the first page from the data storage unit 5 and, together with the image display position information, communicates with the communication interface 4. The data is output to the client terminal 10 via the line 6 (step S205). As a result, “RS” is written in the output area image data. Thus, the area image data is specified and output.
[0079]
If another character string candidate exists, the control unit 21 also outputs the area image data based on the character position information of the character candidate of the other character string candidate to the client terminal 10. Here, when outputting a plurality of area image data, the order of outputting the area image data is the same as that of the first embodiment.
[0080]
In this example, since “RS” is written in the area image data of the first page, the user can understand that the first page is a desired document. Therefore, the user instructs the image data synthesizing unit 12 to synthesize the area image data of the first page with the image data stored in the basic image storage unit 11. The image data combining unit 12 combines the area image data with the common image data, which is the image data stored in the basic image storage unit 11, based on the image display position information and displays the combined image data (Step S206). At this time, the image data previously stored in the basic image storage unit 11 may be common image data or entire image data. When the image data previously stored in the basic image storage unit 11 is the entire image data, the area image data is displayed in a superimposed manner at a predetermined position of the entire image data based on the image display position information. Thus, the image data of the entire document is created.
[0081]
Here, when the area image data of the first page is displayed second, and the user knows that the desired document is on the first page at the time when the second area image data is displayed. The display of the third and subsequent area image data may be interrupted to instruct the image data synthesizing unit 12 to synthesize the area image data of the first page with the image data stored in the basic image storage unit 11.
[0082]
In this example, the data transmitted from the document image distribution device 20 to the client terminal 7 is three region image data. When the conventional technique is used, three pieces of whole image data are transmitted because the whole image data of all documents meeting the search condition is transmitted without using the area image data. Here, for example, assuming that one area image data is 10 KB and one entire image data is 200 KB, when the conventional technique is used, 600 KB of data must be transmitted. The document image distribution device 8 only transmits 30 KB data, and can reduce the load on the network and shorten the data transmission time.
[0083]
In addition, if transmission of the same amount of data as that of the related art is permitted, it is possible to transmit higher-resolution area image data. For example, when image data having a data amount 20 times larger per unit area is transmitted for only the search area, one area image data becomes 200 KB. Then, the transmission of three area image data requires a data amount of 600 KB. Therefore, it is possible to transmit the region image data with higher resolution with the same data amount as the conventional technology.
[0084]
This embodiment is significant, for example, when displaying a blank form with only an individual's name or employee number (student registration number) added. A user searches for area image data indicating his / her name or the like at a client terminal that stores image data of an unfilled portion of a form as common image data. By printing the result of synthesizing the area image data and the common image data, it is possible to obtain a form in which the user's name and the like are already described. When creating this form, the load on the network can be reduced.
[0085]
In addition, the present embodiment is also significant when making copies of various award certificates and the like. The user searches for area image data indicating a name or the like on the client terminal 10 that stores image data of a portion other than the name or the like described in the award certificate (a place where common text is described) as common image data. By printing the result of synthesizing the area image data and the common image data, a copy such as a certificate can be obtained. Also in this case, the load on the network can be suppressed.
[0086]
Here, an example has been described in which an unfilled part of a form or a place where a common text of a prize is described is used as the common image data, but the application example of the present embodiment is not limited to the above example.
[0087]
【The invention's effect】
As described above, according to the present invention, when the document to be distributed is not determined, when the document is being searched, only the image data of the search target area is distributed, and when the document to be distributed is determined. Since the entire image data is distributed or combined with the stored image data, the load on the network can be reduced as compared with the configuration in which the image data of the entire document is distributed for each search.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating an example of a first embodiment of the present invention.
FIG. 2 is an explanatory diagram showing an example of a document in a fixed format.
FIG. 3 is an explanatory diagram showing an example of position information of a search area according to the present invention.
FIG. 4 is an explanatory diagram showing an example of each information in the present invention.
FIG. 5 is an explanatory diagram showing a specific example when the present invention is applied to a fixed format.
FIG. 6 is a flowchart illustrating an operation of the first exemplary embodiment of the present invention.
FIG. 7 is a block diagram illustrating an example of a second embodiment of the present invention.
FIG. 8 is a flowchart illustrating an operation of the second exemplary embodiment of the present invention.
[Explanation of symbols]
1 Scanner
2 Control unit
3 Program storage
4 Communication interface
5 Data storage unit
6 communication lines
7 Client terminal
8 Document image distribution device
10 Client terminal
11 Basic image storage
12 Image data synthesis unit
20 Document image distribution device
21 Control unit
30 documents
31 Search area

Claims (11)

A document image distribution device that distributes image data according to the search keyword; and a terminal device that transmits the search keyword to the document image distribution device via the communication network and displays image data received from the document image distribution device. A document image distribution system,
The document image distribution device reads a document in a fixed format, and generates image data of the entire document and image data of a search area that is a search target area in the entire document. Document reading means for creating character candidate data of characters described in the search area by performing a character recognition process,
Area image data specifying means for specifying image data of a search area corresponding to a search keyword received from the terminal device based on the character candidate data;
Area image data transmitting means for transmitting image data of the search area specified by the area image data specifying means to the terminal device;
A document image distribution system comprising: an entire image data transmitting unit that transmits image data of an entire document designated by a terminal device to the terminal device. The area image data transmitting means, when the area image data specifying means specifies the image data of the plurality of search areas, sequentially delivers the image data of each search area to the terminal device,
2. The document image distribution system according to claim 1, wherein the terminal device sequentially displays the image data of the search area received from the area image data transmitting means. A document image distribution device that receives a search keyword from a terminal device via a communication network and distributes image data according to the search keyword,
Reading a document in a fixed format, generating image data of the entire document, and image data of a search area that is a search target area in the entire document, and performing a character recognition process on the search area; Document reading means for creating character candidate data of characters described in the search area by
Area image data specifying means for specifying image data of a search area corresponding to a search keyword received from the terminal device based on the character candidate data;
Area image data transmitting means for transmitting image data of the search area specified by the area image data specifying means to the terminal device;
A document image distribution apparatus comprising: an entire image data transmission unit configured to transmit image data of an entire document designated by the terminal device to the terminal device. A document image distribution program installed in a document image distribution device that receives a search keyword from a terminal device via a communication network and distributes image data according to the search keyword,
On the computer,
From the information obtained by reading the document in the fixed format, image data of the entire document and image data of a search area that is a search target area in the entire document are generated, and a character recognition process is performed on the search area. Processing to create character candidate data of characters described in the search area by performing
A process of specifying image data of a search area corresponding to a search keyword received from the terminal device based on the character candidate data;
A document image distribution program for executing a process of transmitting image data of a specified search area to a terminal device and a process of transmitting image data of an entire document specified by the terminal device to the terminal device. A document image distribution device that distributes image data according to the search keyword; and a terminal device that transmits the search keyword to the document image distribution device via the communication network and displays image data received from the document image distribution device. A document image distribution system,
The document image distribution device reads a document in a fixed format, generates image data of a search area that is a search target area in the entire document, and performs a character recognition process on the search area to perform the search. Document reading means for creating character candidate data of characters described in the area,
Area image data specifying means for specifying image data of a search area corresponding to a search keyword received from the terminal device based on the character candidate data;
Area image data transmitting means for transmitting the image data of the search area specified by the area image data specifying means to the terminal device,
A terminal device, a basic image data storage unit that stores image data of at least an area other than a search area in an area of the entire document in the fixed format;
Image data combining means for combining the image data of the search area received from the area image data transmission means with the image data stored in the basic image data storage means to create image data of the entire document. Document image distribution system. The area image data transmitting means, when the area image data specifying means specifies the image data of the plurality of search areas, sequentially transmits the image data of each search area to the terminal device,
6. The document image distribution system according to claim 5, wherein the terminal device sequentially displays the image data of the search area received from the area image data transmitting means. The image data synthesizing unit of the terminal device synthesizes the image data specified by the user of the terminal device and the image data stored in the basic image data storage unit out of the image data of the search area received from the area image data transmitting unit. The document image distribution system according to claim 5 or 6, wherein A document image distribution device that receives a search keyword from a terminal device via a communication network and distributes image data according to the search keyword,
By reading a document in a fixed format, generating image data of a search area that is a search target area in the entire document, and performing a character recognition process on the characters described in the search area, thereby obtaining the characters of the characters. Document reading means for creating candidate data;
Area image data specifying means for specifying image data of a search area corresponding to a search keyword received from the terminal device based on the character candidate data;
A document image distribution device comprising: a region image data transmitting unit that transmits image data of a search region specified by a region image data specifying unit to a terminal device. A terminal device that receives image data of a search area corresponding to a search keyword from a document image distribution apparatus that distributes image data of a search area that is a search target area in the entire document in a fixed format, and displays the image data And
Basic image data storage means for storing image data of at least an area other than the search area in the entire area of the document in the fixed format;
Image data combining means for combining the image data of the search area received from the area image data transmission means with the image data stored in the basic image data storage means to create image data of the entire document. Terminal device to do. A document image distribution program installed in a document image distribution device that receives a search keyword from a terminal device via a communication network and distributes image data according to the search keyword,
On the computer,
From the information obtained by reading the document in the fixed format, image data of a search area that is a search target area in the entire document is generated, and the image data described in the search area is obtained by performing a character recognition process on the search area. Processing to create character candidate data for
Document image distribution for executing a process of specifying image data of a search area corresponding to a search keyword received from a terminal device based on the character candidate data and a process of transmitting the image data of the specified search region to the terminal device program. A terminal device that receives image data of a search area corresponding to a search keyword from a document image distribution apparatus that distributes image data of a search area that is a search target area in the entire document in a fixed format, and displays the image data An image creation program installed in the
On the computer,
A process of storing image data of at least an area other than the search area in the entire area of the document in the fixed format in the storage device, and combining the received search area image data with the image data stored in the storage apparatus to obtain the document. An image creation program for executing a process of creating entire image data.

Images