JPH11232296A - Document filing system and document filing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a document filing system and a document filing method to eliminate omission in retrieval, erroneous extraction and capable of retrieving the document with high accuracy in retrieval of a document using character recognition. SOLUTION: A character whose character code is uniquely decided and the one whose character code is not uniquely decided are sorted from results of the character recognition and a post-processing and are stored together with a flag capable of discriminating each character. In the case of the retrieval, collation with the character to be uniquely decided is regarded to be successfully performed when the character codes coincide and regarding the characters not to be uniquely decided, a matching degree between a key word and a text in the document is calculated by using similarity to be outputted by a similarity output part 7.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document filing system and a document filing method for electronically filing images such as documents and drawings, and more particularly to a document filing system and a document for retrieving filed documents using keywords. It concerns the filing method.

[0002]

2. Description of the Related Art Conventionally, document images are stored electronically,
In order to retrieve and display the keyword, a method of manually adding keyword information to a document image and storing the keyword image is used. In addition, in order to save the trouble of manually inputting a keyword, a method is used in which a system having a character recognition function is used, whereby a character in a document image is recognized, and a related keyword or a whole sentence is stored together with the document image. Have been.

In the latter case, erroneous recognition occurs due to imperfect character recognition performance of the system. For this reason, a so-called “erroneous extraction” occurs in which a character string different from the keyword input for the search is displayed as a search result. In addition, even though the characters in the document image are the same as the input keyword, "search omission" occurs in which the characters are erroneously recognized and are not displayed as a search result.

Therefore, in order to improve search accuracy,
It is necessary to minimize the above erroneous extraction and search omission. As a method of reducing erroneous extraction and search omission at the time of the search, (1) improving the character recognition performance (improving the correct character holding ratio) (2) partial mismatch between the input keyword and the search target character string And assist the imperfect character recognition performance.

[0005] As an example of the above (1), there is a method of increasing the probability of retaining correct characters by retaining a plurality of character recognition results for one character image. For example, "Experimental Study on Fusion Technology of Document Recognition and Full Text Search" (Information Processing Society of Japan, Information Science Basics 39-9, September 1995,
Marukawa et al.) Makes the number of candidate characters to be held variable according to the degree of similarity of individual recognition with respect to the recognition candidate characters, and holds a plurality of them, whereby a more accurate search can be performed as compared to holding one character at a time. Is possible.

[0006] For example, in a filing apparatus disclosed in Japanese Patent Application Laid-Open No. Hei 8-272713, the recognition result for each character image is fixed and held up to the fourth candidate, and matching with a search keyword is performed from a document code including candidate characters. Is going.

As an example of the method according to the above (2), as described in JP-A-8-272813, the degree of coincidence m between a keyword and a recognition result is calculated as follows: m = (number of matched characters / number of characters of keyword) ) × 100 (%) (1) Even if all the search characters are not included in the candidate characters as a result of character recognition, some of them are output as search results.

[0008] The operation of the apparatus according to Japanese Patent Application Laid-Open No. Hei 8-272713 will be briefly described below. <Description of Data Storage Method> FIG.
FIG. 28 is a block configuration diagram of a filing device according to No. 2813. In FIG. 1, a document image read by a scanner 101 is a scanner interface (I / F) circuit 102.
Is converted into a digital signal. If the original is a character image, the CPU 105 receives a signal from the I / F circuit 102.
Performs a character recognition process and outputs 4 characters to one character of a character image.
A candidate character as a recognition result up to the character is output to the external storage device 110 as a document storage unit.

The RAM 107 is a work area for developing character images and performing character recognition processing. The external storage device 110 is a device for storing registered data, such as a bird disk, and stores not only data but also a dictionary for character recognition.

In this filing apparatus, when filing an input document image, instead of registering only the character code of the first candidate of the character obtained in the character recognition processing as text data, the character code of up to the fourth candidate is registered. sign up. That is, for each character image, four character images and candidate characters of the recognition result are stored.

<Description of Search Method> FIG.
The text data candidates in the device according to 272813 are shown. In the figure, the arrow indicates the character recognition result and the matching part (matching part) of the keyword when the search keyword is “internal processing integrated type”. CPU 1 described above
In step 05, the keyword is matched with all the candidate characters up to the fourth place as a document search unit.

When the value of m in the above equation (1) is equal to or greater than a certain threshold (for example, 60 (%)), and this is considered as a search result candidate, FIG.
Since there is a match of six characters, m = (6/7) × 100 = 85.7 (%), and these are search result candidates.

[0013]

However, in the conventional method of holding a plurality of character recognition results for one character image, if the number of candidate characters to be stored is reduced with respect to the accuracy of character recognition, correct character characters are not included in the candidate characters. Is less likely to be included, and search leakage is likely to occur. In addition, if the number of candidate characters is large, the possibility of including correct characters is high, and search omissions are reduced. However, since many characters other than correct characters are also stored, erroneous extraction often occurs. There is. Another problem is that holding a large number of candidate characters increases the memory capacity for document storage.

In the method of changing the number of candidate characters to be held according to the similarity, for example, when the density of the document image is not appropriate and the document image is crushed or blurred, the standard pattern and the character image are used for recognition. Since the difference between the feature amounts is large, the correct character is not included in the candidate character, and the recognition rate is reduced, which causes a problem that a search miss occurs.

Further, in this case, since the similarity of the recognition result is reduced (the possibility of correct answer is reduced), it is necessary to store more candidate characters in order to satisfy a certain correct character content rate. As a result, there is a problem that erroneous extraction becomes large at the time of retrieval.

On the other hand, in a search method that allows a certain degree of mismatch as disclosed in Japanese Patent Application Laid-Open No. Hei 8-272713, a character that does not match a character to be matched with a keyword is matched regardless of the character. As long as the portions are common, there is a problem that the same degree of coincidence is calculated.

Thus, for example, when the search keyword is "Japanese", the character string "Japanese", "Japanese", "Japan", "Japan", "Japan is", etc.
The same degree of coincidence of (2/3) × 100 = 67% is obtained, which is output and displayed as a search result.
Here, "Japanese" is misrecognized as "Japanese", and the character string you actually want to search for is "Japanese", but the above "Japanese", "Japanese country", "Japanese", "Japan is" Since the degree of coincidence is the same as the degree of coincidence, if the items are displayed in descending order of the degree of coincidence, "Japan" will be buried in these and displayed.

Then, the user selects one of the erroneous extractions displayed on the display 108 as the display means.
You also need to look for the desired result. Then, as the threshold value for permitting the mismatch is smaller, a large amount of erroneous extraction is also output, so that a document that the user really wants to search is buried in the erroneous extraction, and as a result, the device becomes difficult for the user to use. There is a problem. In addition, there is also a problem that search omission increases when the threshold value is increased.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems, and has as its object to provide a document filing system in which search omission of a document search hardly occurs and erroneous extraction does not occur even if the number of candidate characters is increased. And a document filing method. Another object of the present invention is to provide a document filing system and a document filing method that can calculate a proper degree of matching even if a character partially mismatched with a keyword exists in a recognition result, and can execute a highly accurate search. It is to be.

[0020]

According to a first aspect of the present invention, there is provided a document filing system for retrieving a predetermined document from a document image in accordance with a keyword by recognizing a character in the document image. A matching unit that matches the keyword with the recognized character; and an output unit that outputs information on the document image as a document search result based on the result of the matching. The similarity between the character in the keyword and the recognized character is calculated for each character, and the matching is performed based on the result of calculating the matching degree between the character in the keyword and the recognized character using the similarity. And outputting the document corresponding to the keyword as a document search result based on the value indicating the degree of coincidence. To provide.

According to a second aspect of the present invention, in the first aspect, a predetermined grammatical analysis is performed on the recognized character with reference to a word dictionary stored in advance, and based on a result of the grammatical analysis, Means for assigning an evaluation value indicating a matching state between the character and the word dictionary to each character, wherein the collation means performs the collation based on an overall evaluation value for each character obtained from the similarity and the evaluation value. Provide a document filing system that performs

In a third aspect based on the second aspect, the collating means uniquely assigns a character code to the recognized character based on whether or not the comprehensive evaluation value is equal to or greater than a predetermined value. Provided is a document filing system that distinguishes a character that can be determined from a character that cannot be uniquely determined and stores the character that can be uniquely determined by associating the character code with a predetermined flag.

According to a fourth aspect of the present invention, in the third aspect of the invention, the collating means may determine whether the character which can be uniquely determined and the character in the keyword are based on whether or not the character codes are the same. And the similarity is calculated using the similarity. For the characters that cannot be uniquely determined, the similarity between the character in the keyword and the recognized character is used to calculate the similarity. Provide a document filing system that calculates the degree of coincidence.

According to a fifth aspect of the present invention, in the third aspect, there is further provided a means for extracting a characteristic amount of each of the character in the keyword and the recognized character, and the collating means is provided for the character which cannot be uniquely determined. And a document filing system that calculates the similarity of the feature amount and calculates the coincidence using the similarity.

According to a sixth aspect of the present invention, in the third aspect, there is further provided means for calculating a ratio between the number of characters of the characters that cannot be uniquely determined and the number of characters of the recognized characters.
The output means provides a document filing system for displaying that the predetermined document cannot be properly searched when the ratio is equal to or more than a certain value.

According to a seventh aspect, in a document filing method for retrieving a predetermined document from a document image according to a keyword, a step of recognizing a character in the document image, and a step of recognizing the character in the keyword and the recognized character are performed. Calculating the degree of similarity for each character; calculating the degree of coincidence between the character in the keyword and the recognized character using the degree of similarity; and calculating the degree of similarity in the keyword based on the degree of coincidence. A document filing method comprising: a collation step of collating a character with the recognized character; and an output step of outputting a document corresponding to the keyword as a document search result based on a result of the collation.

According to an eighth aspect, in the seventh aspect, a predetermined grammatical analysis is performed on the recognized character with reference to a word dictionary stored in advance, and based on a result of the grammatical analysis, A step of assigning an evaluation value indicating a matching state between the character and the word dictionary to each character, wherein the matching step includes the step of performing the matching based on the overall evaluation value for each character obtained from the similarity and the evaluation value. Provide a document filing method that performs

In a ninth aspect based on the eighth aspect, the collating step uniquely identifies a character code of the recognized character based on whether the comprehensive evaluation value is equal to or greater than a predetermined value. Provided is a document filing method for distinguishing between a character that can be determined and a character that cannot be uniquely determined, and storing the uniquely determined character in association with its character code and a predetermined flag.

[0029] In a tenth aspect based on the ninth aspect, in the collation step, the uniquely determinable character and the character in the keyword are determined based on whether or not the character codes are the same. And the similarity is calculated using the similarity. For the characters that cannot be uniquely determined, the similarity between the character in the keyword and the recognized character is used to calculate the similarity. A document filing method for calculating the degree of coincidence is provided.

An eleventh invention according to the ninth invention further comprises a step of extracting a characteristic amount of each of the character in the keyword and the recognized character, and the collating step is a step of uniquely determining the character. A document filing method is provided for calculating the similarity of the feature amount for a character that cannot be obtained, and calculating the coincidence using the similarity.

In a twelfth aspect based on the ninth aspect, the method further comprises the step of calculating a ratio of the number of characters that cannot be uniquely determined to the number of characters of the recognized character. A document filing method for displaying a message indicating that the predetermined document cannot be properly searched if the value is equal to or more than a predetermined value.

[0032]

Embodiments of the present invention will be described below with reference to the accompanying drawings. Embodiment 1 FIG. FIG. 1 is a block diagram showing a configuration of a filing system according to Embodiment 1 of the present invention. In the system shown in FIG. 1, a document image is input from the input unit 1, and the input document image and the result recognized by the character recognition unit 3 are stored in the document storage unit 2. The character recognizing unit 3 obtains a character area from the input document image and recognizes characters in the character area. Further, the dictionary 8 holds the features of the standard pattern for each character which the character recognition unit 3 holds for character recognition.

The document image database 4 electronically stores the input image, and the character recognition result database 5 stores a character code string as a result of recognition by the character recognition unit 3.
The document search unit 6 searches the character recognition result database 5 for a document including a character string that matches the keyword 20 input by the user. The similarity output unit 7 calculates the similarity between the two character codes with reference to the dictionary 8, and the display unit 9 displays the search result by the document search unit 6.

FIG. 2 is a flowchart showing a document storage processing procedure in the system according to the present embodiment. First, a method for storing a character recognition result according to the present embodiment will be described with reference to the flowchart shown in FIG.
In step S100 of FIG. 2, the input unit 1 inputs an image and transfers it to the document storage unit 2. The input unit 1 may, for example, perform photoelectric conversion of a document image using a scanner, or may input a photoelectrically converted image via a network or the like.

In step S110, the document storage unit 2 transfers the image input from the input unit 1 to the character recognition unit 3, and receives a character code as a result of character recognition in the image from the character recognition unit 3. The character recognizing unit 3 performs known character string cutting, character cutting, and character recognition, and returns a character recognition result obtained thereby to the document storage unit 2. In addition, the character string cutout uses, for example, a method of connecting portions where black pixels are continuous in a document image, and determining whether or not it is a character string from the width and height of the connected components of the black pixels. .

In the character extraction method, for example, a character string image determined by character string extraction is scanned in the vertical direction and the horizontal direction, a peripheral distribution of the number of black pixels is obtained, and a portion having a small number of black pixels is extracted. Is divided into images for each character. In the character recognition process, for example, an 8 × 8-dimensional density feature is extracted from an image divided into single characters by character segmentation, and a sum of a difference for each dimension from a standard pattern is obtained. A few characters are output as a recognition result from the standard pattern having the smallest sum of.

In the next step S120, the document storage unit 2
Saves a character code as a character recognition result in the character recognition result database 5, but here, only one character recognition result is stored for each one character image. Note that FIG.
5 shows an example of storing a character recognition result in the present embodiment. The document storage unit 2 stores the input binary document image in the character image database 4. Thus, the document storage processing is completed.

Next, a keyword search method according to the present embodiment will be described. FIG. 4 is a flowchart showing a search processing procedure in the system according to the present embodiment. That is, FIG. 4 is a flowchart showing a procedure for collating the keyword with the character code of the recognition result.

In step S 200 of FIG. 4, the keyword 20 input by the user is input to the document search unit 6. In the following step S210, an array including the recognition result (hereinafter, te
xt) and an array including a keyword (hereinafter, key)
) Is initialized. The document search unit 6 extracts the recognition result of one document from the character recognition result database 5 and sets it in text.

In the next step S220, text and k
Check the characters in ey. This collation is based on the character text
The character code of [k] and the character key [j] are input to the similarity output unit 7, and the similarity output unit 7 refers to the dictionary 8 and
This is performed by outputting the similarity s between the characters xt [k] and key [j]. here, Is returned as the similarity. In the above equation (2), Ntext [k] n is the n-dimensional density value of the character text [k] in the dictionary 8, and Ntext [k]
key [k] n is an n-dimensional density value of the character key [k] in the dictionary 8. Further, n takes a value of 1 to 64, and | | represents an absolute value.

Then, the document search unit 6 outputs the similarity output unit 7
Is calculated by using the similarity output by..., (3).

In step S230, the document search unit 6
It is determined whether or not the degree of coincidence is equal to or greater than a certain value. If it is equal to or larger than the predetermined value, the process proceeds to step S240, where each of the pointers k and j is incremented by one, and in subsequent step S250, it is determined whether or not matching with all keywords has been performed.

If it is determined in step S250 that all keywords have been collated, the process proceeds to step S250.
At 270, the output result is saved, and the process ends. In addition,
The output result means a matched document name, a matching degree, a page number, coordinates of a matched character on an image, and the like.
In the present embodiment, they are output to the display unit 9 and displayed.

On the other hand, if it is determined in step S230 that the coincidence is equal to or smaller than the predetermined value, the process proceeds to step S280 to check whether the end of the document is reached. If it is the end of the document, this process ends. If it is not the end of the document, the process proceeds to step S290, where i
Is incremented by 1 and substituted into k, the pointer j of the keyword is set to 0, and the process returns to step S220.

If it is determined in step S250 that matching with all keywords has not been performed,
In a succeeding step S260, it is checked whether or not the end of the document is reached. If it is the end of the document, the process ends. If it is not the end of the document, the process returns to step S220. In the same manner, the above processing is repeated for all the documents in the character recognition result database 5.

Here, a method of matching the character recognition result with the keyword “internal processing integrated type” in the present embodiment will be described by way of example. Step S220 in FIG.
The search process proceeds until pointers k = 0 and j = 0,
That is, the character recognition result is compared with the first character of each of the keywords. That is, text [0] = “in”,
Key [0] = “inside” is collated, and their similarity is determined. FIG. 5 shows the standard pattern “de” in the dictionary 8.
FIG. 6 also shows the feature of “inside”.
Numerical values in these patterns indicate respective feature amounts.

The similarity output unit 7 calculates the similarity s from the patterns of FIGS. 5 and 6 according to the above equation (2), and s = s
1− {91 / (94 + 95)} = 0.52, and outputs it to the document search unit 6. Then, step S230
Then, the document search unit 6 calculates the coincidence based on the similarity s, and obtains the coincidence = 0.52 / 1 = 0.52. here,
If the threshold value of the degree of coincidence s is 0.6, the determination in step S230 is “No”, so the processing is performed in step S280.
Proceed to.

At this stage, since it is not the end of the document, the determination in step S280 is "No", and in step S290, i is incremented, that is, k is incremented by 1, and the process returns to step S220. As a result, in step S220, text [1] = “line”
And key [0] = “inside”.

When the similarity output unit 7 performs the same calculation as above and outputs, for example, the similarity 0.4 to the document search unit 6, the determination in step S230 is also "No" in this case. The process proceeds to step S280. Hereinafter, similarly, these processes are executed, and when the result of the comparison indicates that the keyword and the “internal processing vertical type” in the text show a degree of coincidence equal to or more than a certain value, this is set as an output candidate.

As described above, according to the present embodiment, since only one character is stored for each character image in the recognition result, the memory capacity required for storing the document can be reduced. Also, the characters to be collated are not limited to several characters that are candidates for the recognition result, all characters are to be collated, and
Since all the degrees of matching for each character are used at the time of collation, search omission is reduced, and even if a character partially mismatching with the keyword exists in the recognition result, it is possible to calculate an appropriate degree of matching. Therefore, there is a remarkable effect that the user does not have to check the search result.

Note that the character recognition method is not limited to the above method, and for example, a structural analysis method described in "Pattern Recognition" (Noboru Funakubo, Kyoritsu Shuppan) may be used. Further, in the above embodiment, the similarity output unit 7 calculates from the dictionary used by the character recognizing unit 3, but the present invention is not limited to this, and a similarity table between each character is stored in advance. You may make it. FIG. 7 is an example in which the similarity between each character is tabulated. further,
It goes without saying that the threshold value of the degree of coincidence and the method of calculating the degree of coincidence are not limited to the above numerical values and expressions.

Embodiment 2 FIG. 8 is a block diagram showing a configuration of a filing system according to Embodiment 2 of the present invention. In the figure, the same components as those of the system according to the first embodiment shown in FIG. 1 are denoted by the same reference numerals, and the description thereof is omitted here. The post-processing unit 10 in FIG. 8 grammatically verifies the character recognition result of the character recognition unit 3 stored in the document storage unit 2 and outputs the result. The word dictionary 11 is a dictionary used by the post-processing unit 10.

First, a method for registering a document in the system according to the present embodiment will be described. FIG. 9 is a flowchart showing a character recognition result storing process executed by the filing system according to the present embodiment. In step S300 in the figure, the character recognition unit 3 performs a character recognition process. Since the method of character recognition here is the same as the method in the first embodiment, the description is omitted. It should be noted that a plurality of recognition results, including candidate characters, are output for one character image.

Next, the flow advances to step S310 to perform post-processing. Here, the post-processing unit 10 grammatically analyzes the recognition result including the candidate character and determines a combination considered to be correct as a sentence. Hereinafter, a post-processing method using a known morphological verification will be described.

FIG. 10 is a flowchart showing the procedure of post-processing according to the present embodiment. Here, this post-processing is performed on the recognition result candidate characters shown in FIG. FIG. 11 is a diagram showing the correspondence between words in the word dictionary 11 and morpheme numbers. Note that the morpheme number is a number uniquely assigned to a morpheme (a word such as a common noun, proper noun, or savari noun). FIG. 12 illustrates the connection relationship between these morpheme numbers.

In step S410 of FIG.
Performs a comparison between the character recognition result and the word dictionary 11. Here, when the word shown in FIG. 11 is compared with the recognition result shown in FIG. 18, the “processing”, “integration”, and “type” in FIG. 11 are successfully matched with the recognition result. Then, the process proceeds to step S420, and a morphological connection test is performed according to the description shown in FIG.

As shown in FIG. 11, the morpheme numbers of “processing” are “55”, and the morpheme numbers of “integration” are “55” and “5”.
"9" and "46" for "type". Therefore, when the morphological connection test is performed using the description contents shown in FIG. 12, the connections of 55 and 55 and the connections of 55 and 46 are all established (indicated by a circle in the figure). Next, the process proceeds to step S430, in which the number of matching morpheme characters is assigned to each candidate character as an evaluation value.

FIG. 13 shows a collation result output from the post-processing unit 10. As shown in FIG.
The evaluation value is “2” for “go”, the evaluation value is “1” for “type”, and all other characters are assigned “0” because they do not match the word dictionary.

When the above post-processing is completed, step S320 in FIG. 9 is executed. That is, the document storage unit 2 calculates a comprehensive evaluation value for each character from the character recognition result and the post-processing result. This comprehensive evaluation value is calculated by the following equation (4). Total evaluation value = α × (similarity of character recognition) + (1−α) × (evaluation value of post-processing) (4) Here, α is a value satisfying 0 ≦ α ≦ 1, and In the embodiment, α = 0.8.

Now, of the candidate characters shown in FIG. 18, the similarity of “vertical” is set to 0.7 and the similarity of “to” is set to 0.5, and the matching result shown in FIG. When applied to the equation (4), the overall evaluation value of “vertical” is 0.8 ×
0.7 + 0.2 × 0 = 0.56, the overall evaluation value of “To” is 0.8 × 0.5 + 0.2 × 2 = 0.8, and “To” is more than “Vertical”. Evaluation value is high.

In the following step S330, the document storage unit 2
Distinguishes between characters that can be uniquely determined and characters that cannot be uniquely determined. The method of this distinction is that, for example, when the total evaluation value is equal to or more than a certain value, it can be uniquely determined, and when it is less than that, it cannot be determined. Then, step S34
In the case of 0, the character that can be uniquely determined by the document storage unit 2 is only the character code having the highest overall evaluation value.
Save with flag "0". However, for a character determined not to be uniquely determined, the character code having the highest overall evaluation value is stored together with the flag “1”. In other words, character recognition and post-processing are performed, and
Characters are stored as recognition results regardless of whether they can be uniquely determined.

FIG. 14 shows an example of storing a character recognition result in this embodiment. As shown in the figure, as for the character recognition result, in the present embodiment, the character "" is correctly stored as compared with the case of FIG.

Hereinafter, a search method in the system according to the present embodiment will be described. Note that the search method here is basically the same as the search method according to Embodiment 1 shown in FIG. 4, but the processing in step S220 is different. That is, in the first embodiment, the degree of coincidence of all characters with the keyword is calculated using the similarity from the similarity output unit 7, but in the second embodiment, te
For a character whose xt flag is 1, the degree of coincidence is calculated using the value from the similarity output unit 7 as the similarity. For a character whose flag is 0, the character codes of text and key are completely changed. If they match, it is determined that the matching is successful, and the similarity 1 is given. In addition, the character whose flag is 0, t
If the character codes of ext and key do not match, the similarity 0 is returned.

Therefore, the search method in the present embodiment will be described with reference to the flowchart shown in FIG. 4, taking as an example a case where the text shown in FIG. 14 is collated with the keyword “internal processing integrated type”. The document search unit 6 receives the keyword “internal processing integrated type” (step S20 in FIG. 4).
0), the text and key pointers are initialized (step S210), and in the following step S220, the text and the key are collated. Here, as in the first embodiment, initially, text [0] = “in” and key [0].
= Check with “inside”.

As shown in FIG. 14, since the flag of text [0] is “0”, the document search unit 6 sets the text [0]
And whether or not key [0] has the same character code. In this case, since they are not the same, it is determined that the matching has failed, and the document search unit 6 returns the similarity 0. Then, the process proceeds to step S230, and the degree of coincidence is calculated. Since the degree of coincidence is 0/1 = 0, the process proceeds to step S28.
0, and again in step S220, this time with tex
Matching is performed between t [1] = “line” and key [0] = “inside”.

Hereinafter, the process proceeds similarly, and text [5]
Is compared with key [0], a matching degree equal to or greater than a predetermined value is obtained, and the determination result in step S230 is “Y
es ", and the process proceeds to step S240. In this step S240, the pointer of the text and the key is set to 1
Increment, and then steps S250 and S26
After 0, the process returns to step S220.

Text [6] = “city” and key [1] =
In the case of collation with “part”, since the flag of text [6] is “1”, the document search part 6 sets the result of the similarity output part 7 as the similarity. For example, assuming that the similarity (city, part) returned by the similarity output unit 7 is 0.6, the coincidence obtained in step S230 is (1 + 0.6) /2=0.8.
As a result, since the degree of coincidence exceeds the fixed value 0.6, the process proceeds to step S240. Hereinafter, the same processing is performed, and if the collation with the text “internal processing vertical type” is successful, it is output as a search result.

As described above, according to the present embodiment, for a character that can be uniquely determined, only when the character code is completely the same, the collation succeeds. If the following candidate characters are excluded, their matching becomes unnecessary, and the probability of detecting erroneous extraction can be further reduced.

Also, since the degree of coincidence with a keyword is always evaluated using the degree of similarity for each character, no matter what character comes to the mismatched part, they are not displayed with the same degree of coincidence. Based on the obtained coincidence, it is possible to more clearly separate erroneous extraction from correct search results.

In the second embodiment, the character code and the flag 1 are stored for characters whose character codes cannot be uniquely determined. However, the present invention is not limited to this. The reason why it cannot be determined is, for example, that the density value of the input image is not appropriate and the input image is blurred or crushed. In this case, for a character that cannot be uniquely determined, storing a feature used for character recognition enables more accurate collation and a more accurate search than storing a character code.

FIG. 15 shows an example of a dictionary relating to the above feature storage. As shown in FIG.
The characteristic values (characters indicated by 21 in the figures) are stored for the characters of 8 × 8-dimensional (64-dimensional) as shown in FIGS. 5 and 6, for example. It is displayed in a line.

As a process at the time of retrieval, the similarity output unit 7 compares text [k] with key [j] when the flag of text [k] is 0 in step S220 in FIG. If the flag of text [k] is 1, te
The feature value shown in FIG. 15 is used as the feature of the character of xt [k], and the similarity is calculated according to the above expression (2) using the feature in the dictionary 8 as the feature of the character of key [j]. You may make it calculate.

On the other hand, if the number of characters that cannot be uniquely determined is larger than the number of characters in the document image, the document has a wider allowable range for various keywords, and the probability of being searched as erroneous extraction increases. Therefore, in such a case, for example, a message to the effect that “this document may not be correctly retrieved” may be displayed on the display unit 9 together with the image input from the input unit 1. .

Therefore, this operation will be described with reference to FIGS. In this case, in step S340 of FIG.
The document storage unit 2 executes a document storage procedure according to the flowchart shown in FIG. That is, in step S510 of FIG. 16, the document storage unit 2 calculates the number of characters A output by the character recognition unit 3. Then, in step S520,
The number of characters that cannot be uniquely determined (the number of characters whose flag is 1), calculated from the result of the character recognition unit 3 and the result of the post-processing unit 10
Calculate F.

Next, in step S530, it is determined whether F / A is equal to or more than a predetermined value. If it is above a certain value,
Proceeding to step S540, a warning message stating that "this document may not be correctly searched" is displayed on the display unit 9, and this processing ends without saving the document. Also,
If F / A is smaller than the predetermined value, the above-mentioned warning message is not displayed, the document is saved, and the process is terminated.

In this manner, the user can guess the cause of the incorrect retrieval of the document by looking at the message and the image. This is an effective means for storing correct data. Become.

[0077]

As described above, according to the first aspect, in a document filing system for retrieving a predetermined document from a document image in accordance with a keyword, means for recognizing characters in the document image, Matching means for comparing the character with the recognized character, and output means for outputting information on the document image as a document search result based on the result of the matching. And the similarity with the recognized character is calculated for each character, and using the similarity, the matching is performed based on the result of calculating the degree of coincidence between the character in these keywords and the recognized character, and The output unit outputs a document corresponding to the keyword as a result of the document search based on the value indicating the degree of matching, so that a document necessary for storing the document is output. It is possible to reduce the re-capacity, reduce search omissions, and even if a character that does not match the keyword partially exists in the recognition result, it is possible to calculate an appropriate matching degree. You don't have to check the search results.

According to the second invention, a predetermined grammatical analysis is performed on the recognized character with reference to a word dictionary stored in advance, and the character and the word dictionary are determined based on the result of the grammatical analysis. Means for assigning an evaluation value indicating a matching state to each character, and the collating means performs the collation based on an overall evaluation value for each character obtained from the similarity and the evaluation value. Thus, a more appropriate degree of coincidence can be calculated.

According to the third aspect of the present invention, the collating means determines whether or not the character code of the recognized character can be uniquely determined based on whether or not the comprehensive evaluation value is equal to or greater than a predetermined value. Characters that cannot be uniquely determined are distinguished from each other, and the characters that can be uniquely determined are stored by associating the character codes with predetermined flags, so that the probability of detection of erroneous extraction can be reduced. Since parts are not displayed with the same degree of matching, erroneous extraction and correct search results can be more clearly separated based on the obtained degree of matching.

According to the fourth aspect, the collating means calculates the similarity between the uniquely determinable character and the character in the keyword based on whether or not the character codes are the same. At the same time, the degree of matching is calculated using this degree of similarity, and for characters that cannot be uniquely determined,
By calculating the matching degree using the similarity between the character in the keyword and the recognized character, it is possible to calculate a proper character matching degree for document search.

According to the fifth aspect, there is further provided a means for extracting a characteristic amount of each of the character in the keyword and the recognized character, and the collating means determines the characteristic amount of the character which cannot be uniquely determined. By calculating the degree of similarity and calculating the degree of coincidence using the degree of similarity, more accurate collation and search can be performed than in the case where the character code is stored.

According to the sixth aspect of the present invention, there is further provided means for calculating a ratio of the number of characters of the characters which cannot be uniquely determined to the number of characters of the recognized characters. In the case of (1), by displaying an indication that the predetermined document cannot be searched properly, it is possible to guess the cause of the incorrect search of the document by looking at the display.

According to a seventh aspect of the present invention, in a document filing method for retrieving a predetermined document from a document image according to a keyword, a step of recognizing a character in the document image, and a step of recognizing a character in the keyword and the recognized character Calculating the degree of similarity for each character; calculating the degree of coincidence between the character in the keyword and the recognized character using the degree of similarity; and calculating the degree of similarity in the keyword based on the degree of coincidence. And a collation step of collating the recognized character with the recognized character, and an output step of outputting a document corresponding to the keyword as a document search result based on the result of the collation. Calculate the correct matching degree even if the recognition result requires less memory capacity, less search omissions, and characters that do not match the keyword partially. Can, the user would Habukeru also troublesome to confirm the search result, there is an effect that.

According to the eighth invention, a predetermined grammatical analysis is performed on the recognized character with reference to a word dictionary stored in advance, and the character and the word dictionary are analyzed based on the result of the grammatical analysis. And a step of assigning an evaluation value indicating a matching state to each character, wherein the collation step performs the collation based on an overall evaluation value for each character obtained from the similarity and the evaluation value. , A more appropriate degree of coincidence can be calculated.

[0085] According to the ninth aspect, the collating step determines whether or not the character code of the recognized character can be uniquely determined based on whether or not the comprehensive evaluation value is equal to or greater than a predetermined value. Characters that cannot be uniquely determined are distinguished from each other, and the characters that can be uniquely determined are stored by associating the character codes with predetermined flags, so that the probability of detection of erroneous extraction can be reduced. Since parts are not displayed with the same degree of matching, erroneous extraction and correct search results can be clearly separated based on the obtained degree of matching.

According to the tenth aspect, the collating step comprises:
Based on whether or not the character codes are the same, calculate the similarity between the uniquely determinable character and the character in the keyword, calculate the degree of coincidence using this similarity, For the characters that cannot be uniquely determined, the degree of similarity is calculated using the similarity between the character in the keyword and the recognized character, thereby calculating an appropriate degree of character matching for document search. be able to.

Further, according to the eleventh aspect, there is further provided a step of extracting a characteristic amount of each of the character in the keyword and the recognized character. By calculating the degree of similarity of the feature amount and calculating the degree of coincidence using the degree of similarity, it is possible to perform more accurate collation and search as compared with the case where a character code is stored.

According to the twelfth invention,
A step of calculating a ratio between the number of characters of the characters that cannot be uniquely determined and the number of characters of the recognized characters, wherein the output step indicates that the predetermined document cannot be properly searched if the ratio is equal to or more than a certain value. By doing so, it is possible for the user to look at the display and guess the reason why the document is not correctly searched.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a filing system according to Embodiment 1 of the present invention.

FIG. 2 is a flowchart showing a document storage processing procedure in the system according to the first embodiment of the present invention.

FIG. 3 is a diagram illustrating an example of storing a character recognition result according to the first embodiment of the present invention.

FIG. 4 is a flowchart showing a search processing procedure in the systems according to the first and second embodiments of the present invention.

FIG. 5 is a diagram showing characteristics of a standard pattern (character “de”) in the dictionary according to the first embodiment.

FIG. 6 is a diagram showing characteristics of a standard pattern (character “in”) in the dictionary according to the first embodiment.

FIG. 7 is a diagram showing a similarity between characters in a table format.

FIG. 8 is a block diagram showing a configuration of a filing system according to Embodiment 2 of the present invention.

FIG. 9 is a flowchart showing a character recognition result saving process executed by the filing system according to the second embodiment.

FIG. 10 is a flowchart illustrating a procedure of post-processing according to the second embodiment.

FIG. 11 is a diagram showing correspondence between words in a word dictionary and morpheme numbers according to the second embodiment;

FIG. 12 is a diagram showing a description of a connection rule between morphemes in the word dictionary according to the second embodiment.

FIG. 13 is a diagram illustrating a collation result output from a post-processing unit.

FIG. 14 is a diagram showing an example of storing a character recognition result according to the second embodiment.

FIG. 15 is a diagram illustrating an example of a dictionary related to feature storage used for character recognition.

FIG. 16 is a flowchart showing another procedure for storing a document according to the second embodiment.

FIG. 17 is a diagram showing a block configuration of a conventional filing apparatus.

FIG. 18 is a diagram showing text data candidates in a conventional device.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Input part, 2 ... Document storage part, 3 ... Character recognition part, 4 ... Document image database, 5 ... Character recognition result database,
6 Document search unit, 7 Similarity output unit, 8 Dictionary, 9 Display unit, 10 Post-processing unit, 11 Word dictionary, 20 Keywords

Claims (12)

[Claims] 1. A document filing system for retrieving a predetermined document from a document image according to a keyword, means for recognizing a character in the document image, collation means for collating the keyword with the recognized character, Output means for outputting information on the document image as a document search result based on a result of the matching, wherein the matching means calculates a similarity between the character in the keyword and the recognized character for each character Then, using the similarity, the matching is performed based on the result of calculating the degree of coincidence between the character in the keyword and the recognized character, and the output unit determines the value based on the value indicated by the degree of coincidence. And outputting a document corresponding to the keyword as the document search result. 2. A predetermined grammatical analysis is performed on the recognized character with reference to a word dictionary stored in advance,
Based on the result of the grammatical analysis, a means for assigning an evaluation value indicating a matching state between the character and the word dictionary to each character, wherein the matching means determines each of the similarity and the evaluation value 2. The document filing system according to claim 1, wherein the collation is performed based on a comprehensive evaluation value for each character. 3. The method according to claim 1, wherein the collating unit determines, based on whether or not the comprehensive evaluation value is equal to or more than a predetermined value, a character whose character code can be uniquely determined and a character that cannot be uniquely determined for the recognized character. 3. The document filing system according to claim 2, wherein the character that can be uniquely determined is stored by associating the character code with a predetermined flag. 4. 4. The collation unit calculates a similarity between the uniquely determinable character and a character in the keyword based on whether or not the character codes are the same,
The similarity is calculated using the similarity, and for the character that cannot be uniquely determined, the coincidence is calculated using the similarity between the character in the keyword and the recognized character. The document filing system according to claim 3, wherein 5. The apparatus according to claim 1, further comprising a unit configured to extract a feature amount of each of the character in the keyword and the recognized character, wherein the matching unit calculates a similarity of the feature amount for the character that cannot be uniquely determined. 4. The document filing system according to claim 3, wherein the similarity is calculated using the similarity. 6. The apparatus according to claim 1, further comprising: means for calculating a ratio between the number of characters of the characters that cannot be uniquely determined and the number of characters of the recognized characters. If the ratio is equal to or more than a predetermined value, the output means outputs the predetermined document. 4. The document filing system according to claim 3, wherein an indication that the search cannot be properly performed is displayed. 7. A document filing method for retrieving a predetermined document from a document image according to a keyword, wherein a step of recognizing a character in the document image, and a step of recognizing a similarity between the character in the keyword and the recognized character by a character Calculating for each character; using the similarity to calculate the degree of coincidence between the character in the keyword and the recognized character; and, based on the degree of coincidence, the character in the keyword and the recognized character. A document filing method, comprising: a collation step of collating a character; and an output step of outputting a document corresponding to the keyword as a document search result based on the result of the collation. 8. A predetermined grammatical analysis is performed on the recognized character with reference to a word dictionary stored in advance,
Based on the result of the grammatical analysis, a step of assigning an evaluation value indicating a matching state between the character and the word dictionary to each character, wherein the matching step includes determining each of the similarities and the evaluation value. 8. The document filing method according to claim 7, wherein the collation is performed based on a comprehensive evaluation value for each character. 9. The collating step includes, for the recognized character, a character for which a character code can be uniquely determined and a character for which the character code cannot be uniquely determined based on whether or not the comprehensive evaluation value is equal to or more than a predetermined value. 9. The document filing method according to claim 8, wherein the distinction is made, and the character which can be uniquely determined is stored in association with the character code and a predetermined flag. 10. The collation step calculates a similarity between the uniquely determinable character and a character in the keyword based on whether or not the character codes are the same. And calculating the degree of coincidence using the similarity between the character in the keyword and the recognized character for the character that cannot be uniquely determined. 9. The document filing method according to item 9. 11. The method according to claim 11, further comprising the step of extracting a characteristic amount of each of the character in the keyword and the recognized character, wherein the collating step calculates a similarity of the characteristic amount for the character that cannot be uniquely determined. 10. The document filing method according to claim 9, wherein the similarity is calculated using the similarity. 12. The method according to claim 12, further comprising: calculating a ratio between the number of characters of the characters that cannot be uniquely determined and the number of characters of the recognized characters. 10. The document filing method according to claim 9, wherein an indication that the search cannot be properly performed is displayed.