JPH10207985A - Method and device for segmenting character - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the calculation amount for word generation processing and word collation processing and to more efficiently segment characters by controlling the number of characters constituting a word to generate the word. SOLUTION: A segment extraction part 17 extracts each primary segment being a cluster area of black bits from original picture data stored in a memory of a picture input part 13 in such state that its coordinates can be recognized. Extracted primary segments are integrated in accordance with a prescribed rule by a segment integrating part 19 to generate a secondary segment. A character recognition part 21 performs character recognition on the assumption that each of primary segments and secondary segments is one character, and a candidate character code as the result is stored. A word generation part 23 combines respective candidate characters of primary and secondary segments obtained by character recognition to generate a word. In this case, the number of characters constituting a word is controlled to generate the word. Thus, unnecessary candidate character codes and their combinations are removed from the calculation object at the time of word generation.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character extracting method in character recognition technology and a character extracting apparatus suitable for implementing the method.

[0002]

2. Description of the Related Art A handwritten character string has a greater change in character spacing and character shape than a printed character string. Therefore, if characters are cut out at regular intervals when extracting characters from image data of a handwritten character string, characters cannot be extracted accurately. This causes misrecognition. As a conventional technique for solving this, there is a character cutout method disclosed in, for example, Japanese Patent Application Laid-Open No. 5-35917.

In this prior art, a character block is first cut out from a line image. Here, the character block is a block region of black bits. Note that the character block cut out from the line image is also referred to as a primary segment in this specification. This character block may constitute a character pattern by itself or may be a part of the character pattern. Next, this character block (primary segment)
Are integrated to generate a character pattern (also referred to as a secondary segment). Next, character recognition is performed on the character pattern (primary segment, secondary segment). If the character pattern does not include the same character block in relation to another character pattern (when there is no overlap of character blocks), the character pattern is cut out as it is. On the other hand, if there are duplicate characters,
A word is generated by combining the recognition result of each of these character patterns with the recognition result of several characters before or after the character pattern. The generated words are checked against a word dictionary.

[0004] For a generated word that matches a registered word in the word dictionary with the generated word, its character pattern is cut out. If they do not match, the evaluation value of each character pattern before word generation and the evaluation value of the word are obtained, and the character pattern with the highest evaluation value is cut out.

[0005] According to this conventional character extraction method, an evaluation value of the certainty of the recognition result is obtained only for a character pattern that has been determined to be impossible to extract by word collation due to overlapping character blocks. Therefore, there is no waste in the process of obtaining the evaluation value, and the efficiency of the character segmentation process can be increased.

[0006]

However, in the prior art, when a word is created by combining the recognition result (candidate character code) for the primary segment and the recognition result (candidate character code) for the secondary segment, It merely states that each recognition result is combined with several characters before and after it, and the length of the word to be created is not clear. Therefore, in some cases,
Unnecessarily long words are created, and the amount of calculation may be wasted.

[0007] Depending on the field of character recognition, the number of characters in a word may be limited. If the number of characters in a word can be limited in advance, the amount of calculation in word creation processing and word collation processing can be reduced, so that characters can be cut out more efficiently.

[0008]

Therefore, according to the invention of the character extracting method of this application, the original image data including the image data of the character string stored in the memory is converted into a primary area which is a block area of black bits. A process of extracting each segment in a state where its coordinates are known, a process of integrating each extracted primary segment according to a predetermined rule, and creating a secondary segment in a state where its coordinates are known, a process of extracting each primary segment and each of the two segments Character recognition processing for each of the next segments, and the primary segments and 2 obtained by the character recognition
A process of creating a word by combining candidate character codes of each of the next segments; a process of matching the created word with a word dictionary; and the primary segment and the secondary segment based on a result of the word matching. Deciding one of the characters as a character cutout area for one character, wherein the word is created by limiting the number of characters constituting the word.

According to the invention of the character extracting method, the number of characters of a word to be created is restricted, so that the amount of calculation in the word creating process and the word matching process is reduced.

[0010] Here, it has been described that the primary segment is a lump area of black bits. The word “black bit” here means that the character constituent pixels are considered as black pixels, and is not intended to limit the black bits. Note that in the case of black and white inversion, the lump area of white bits becomes the primary segment.

In practicing the invention of the character extracting method, it is preferable that a reference value for regulating the number of characters is determined in accordance with the field of character recognition. Here, the field to be subjected to character recognition is a field in which the number of characters of each piece of data can be expected to be approximately within the number of characters, such as address data and first and last name data. When the reference value is determined according to the field in this way, words can be created within an appropriate number of characters.

Further, in implementing the invention of the character segmentation method, the secondary segment has a direction in which characters are arranged in X.
The direction is created by integrating m primary segments that are continuous in the X direction according to a predetermined rule, and
The word is a candidate path represented by the concatenation of the primary segment and / or the secondary segment, and includes the following (a) to
It is preferable to create it based on the candidate path created by the process including the process (d). Here, m is an integer of 2 or more.

(A) Coordinates for dividing each of the m primary segments in the X direction are determined as candidate clipping positions Ci
When (i = 0 to m), a process of extracting all the segments whose extraction candidate position C0 is the extraction start point from the m primary segments and the created secondary segments.

(B) With respect to each of the segments extracted in the process (a), the segmentation candidate positions Cj (j = 1 to m) on the end point side of the segment are connected since the segmentation start positions are set. Other segments that can be connected, and other segments that have the same relationship as the above-described segmentation candidate positions to the other segments, have another segment that can be further connected to the segment, and the segmentation candidate position on the end point side is Cm. A process of extracting from the m primary segments and the created secondary segments until another segment appears.

(C) Every time the other segment is extracted in the process of (b), a process of determining whether or not the number of segments of the candidate path including the other segment is within a specified number.

(D) A candidate path having a number of segments within the specified number and a candidate position for extracting the end point of the last segment in the candidate path being Cm is set as a candidate path for word creation. Processing.

According to this preferred embodiment, the extraction candidate position C is selected from the extraction area defined by the m primary segments.
All the candidate paths consisting of the segment sequence in which 0 is the starting point, the cutout candidate position Cm is the end point, and the number of segments is within the specified number are extracted. Since the recognition result (candidate character code) of each of the segments constituting the extracted candidate path has already been found in the recognition processing, a word having a specified number of characters or less is generated from the extracted candidate path. You.

Specifically, the candidate path creation process in the preferred embodiment described above is preferably created by a process including the following processes (1) to (9).

(1) Coordinates for dividing each of the m primary segments in the X direction are designated as cutout candidate positions C0
To Cm, the cutout candidate position Ci (i
= 0 to m) is Cm.

(2) A second process which is executed when it is determined that Ci = Cm in the first process, and records the current candidate path in the candidate path memory.

(3) This is executed when it is determined that Ci 判定 Cm in the first processing, and the segment S sandwiched between the extraction candidate position Ci and the extraction candidate position Cj (j = i + 1)
Third processing for determining whether or not k + 1 exists.

(4) Executed when it is determined in the third processing that a segment exists, and the segment Sk +
A fourth process of determining whether the number of segments of the candidate path does not exceed a specified number when 1 is added to the candidate path.

(5) A fifth process which is executed when it is determined in the fourth process that the number is within the specified number, and adds the segment Sk + 1 to the candidate path.

(6) The sixth processing which is executed subsequent to the fifth processing and is executed again from the first processing by regarding the cut candidate position Cj as the focused cut candidate position Ci.
Processing.

(7) For the candidate path created by executing the fifth and sixth processes, the seventh segment for deleting the segment most recently added to the candidate path from the candidate path. processing.

(8) When the determination is negative in the third processing, when the determination is negative in the fourth processing, or when the seventh processing is performed, Eighth processing for changing j defining the candidate position to j = j + 1 and determining whether or not the changed j satisfies j> m in relation to m.

(9) A ninth process that is executed when j ≦ m is determined in the eighth process, and is re-executed from the third process.

According to the processes (1) to (9), from the cutout area defined by the m primary segments, the cutout candidate position C0 is the starting point and the cutout candidate position Cm is the end point. All the candidate paths composed of the segment sequence in which the number of segments is within the specified number are extracted.

Further, in implementing the invention of the character segmentation method, when the direction in which the characters are arranged is set to the X direction as the predetermined rule when the primary segments are integrated, m continuous m characters in the X direction are set. The height H of the segment having the highest height among the primary segments is obtained, and another primary segment existing in the H × N coordinate range in the X direction with respect to the primary segment of interest is focused on. It is preferred to use the rule of merging into primary segments. Here, N is a predetermined value.

Generally, the size of a handwritten character, the space between characters, and the like change depending on the writer and the circumstances surrounding the writer. Therefore, it is common that the row height also changes. Therefore, by considering the height of the segment, the above variation in the row height is taken into account. As a result, the secondary segment can be created under conditions that take into account the size of the character that changes depending on the writer and the circumstances surrounding the writer, and so an appropriate secondary segment can be created. The value of N is preferably determined empirically (statistically). According to the inventor's research on this application,
It has been found that preferable integration is realized by setting N to 1.2.

In order to carry out the above-described character extracting method, it is preferable to configure a character extracting device as follows.

A segment extraction unit for extracting, from input image data including image data of a character string stored in the memory, a primary segment, which is a block area of black bits, in a state where its coordinates are known; A segment integration unit for integrating each primary segment according to a predetermined rule and creating a secondary segment with its coordinates known;
A character recognizing unit for recognizing each of the next segment and each of the secondary segments, a word generating unit for generating a word by combining candidate character codes of the primary and / or secondary segments obtained by the character recognition,
A word matching unit that matches the created word with a word dictionary; and determines one of the primary segments and the secondary segments as a character cutout area for one character based on a result of the word matching. And a character selecting unit for controlling the number of characters constituting the word in the word creating unit.

In implementing the invention of the character extracting apparatus, the segment integrating unit integrates m primary segments continuous in the X direction according to a predetermined rule when the direction in which characters are arranged is the X direction. A segment integrating unit, wherein the word creating unit (A) includes a means (1) to (8) below and creates a candidate path represented by a concatenation of the primary segment and / or the secondary segment. (B) operates when the third means included in the candidate path generating unit determines that a segment exists, and when the segment Sk + 1 is added to the candidate path, the number of segments of the candidate path is defined. It is preferable that the number of characters is controlled by determining whether the number does not exceed the number of characters. Here, m is an integer of 2 or more.

(1) The coordinates for dividing each of the m primary segments in the X direction are defined as a candidate cutout position C0.
To Cm, the cutout candidate position Ci (i
= 0 to m) is Cm.

(2) A second means which operates when the first means determines that Ci = Cm, and records the current candidate path in the candidate path memory.

(3) The operation is performed when Ci と Cm is determined by the first means, and the segment Sk sandwiched between the extraction candidate position Ci and the extraction candidate position Cj (j = i + 1)
Third means for determining whether or not +1 exists.

(4) A fourth means which operates when the number-of-characters rule checking unit determines that the number is within the specified number, and adds the segment Sk + 1 to the candidate path.

(5) Fifth means which operates following the fourth means, and regards the cut candidate position Cj as the focused cut candidate position Ci and starts the operation of the first means.

(6) For a candidate path created as a result of the operation of the fourth means and the fifth means, a sixth means for deleting from the candidate path the segment most recently added to the candidate path. .

(7) When the third means determines that no,
Or, if the character number rule checking unit determines that the number exceeds a specified number, or operates after the sixth means is operated,
A seventh means for changing j defining the cutout candidate position to j = j + 1 and determining whether the changed j satisfies j> m in relation to the m.

(8) An eighth means for operating when the seventh means determines that j ≦ m, and for operating the third means.

In implementing the invention of the character segmenting apparatus, when the direction in which the characters are arranged is the X direction, the segment integration section sets the height of the m primary segments continuous in the X direction. Means for obtaining the height H of the segment having the highest segment, and means for integrating another primary segment present in the H × N coordinate range in the X direction with respect to the primary segment of interest into the primary segment of interest. It is preferable to adopt a configuration including Here, N is a predetermined value.

[0043]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of a character extracting apparatus according to an embodiment of the present invention; However, the drawings used in the description are schematically shown to the extent that the present invention can be understood.

FIG. 1 is a diagram showing a configuration of a character cutout device according to the embodiment. The character segmenting apparatus 10 includes a control unit 11, an image input unit 13, a character number rule input unit 15, a segment extraction unit 17, a segment integration unit 19, a character recognition unit 21, a word creation unit 23, a word comparison unit 25, and a result selection unit. 27 and a result output unit 29.

These components 11 to 29 can be constituted by a computer and its peripheral devices, respectively. Hereinafter, the configuration and operation of each component will be sequentially described.

The control section 11 controls the operation of each of the components 13 to 29.

The image input section 13 includes a memory (not shown), inputs original image data to be recognized, and stores the input image data in the memory. Specifically, original image data represented by black and white binary is input.

The image input section 13 can have any suitable configuration. For example, the image input unit may have a configuration in which a scanner is provided and photoelectrically converts an optical signal from a document and takes in the original image data into a memory, or another database that stores the original image data in the first place. .
Of course, a binary image may be obtained from a multivalued image.

The number-of-characters rule input unit 15 includes a word creating unit 23.
Enter the reference value for regulating the number of characters in the word created in. The character number rule input unit 15 can be constituted by, for example, a computer and its keyboard.

For example, the operator can input a reference value for regulating the number of characters of a word created by the word creating section 23 from a keyboard. Here, an upper limit value of the number of segments constituting a candidate path (described later) is input as a reference value. The reference value can be input, for example, according to the field of the character recognition target. For example, in the case of recognizing data related to surnames, the surname is at most four characters or less, for example, so 4 is input as a reference value.

The segment extraction unit 17 is provided with the image input unit 13
The primary segment, which is a block area of black bits, is extracted from the original image data stored in the memory of FIG.

The segment extraction processing in the segment extraction unit 17 can be easily performed by a method using the horizontal and vertical projection distributions of black bits (also referred to as black points) which are well known in the art.

More specifically, the memory in which the original image data is stored is scanned in the horizontal direction, and a histogram of black points is obtained. Each minimum point in the histogram is set as a candidate cutting position Ci in the horizontal direction. The same processing is performed by changing the scanning direction to the vertical direction, and the extraction candidate position in the vertical direction is extracted. The primary segment is included in a rectangular area surrounded by the extraction candidate positions.

In order to deepen the understanding of the operation of the segment extracting section 17, FIG. 2 shows primary segments S0 to S5 extracted from the original image data 31 and each primary segment in a direction in which characters are arranged (in this example, (X direction), that is, the cutout candidate positions C0 to C6 are shown.

The X-direction start coordinates Xs, X-direction end coordinates Xe, Y-direction start coordinates Ys, and Y-direction end coordinates Ye of each of the primary segments S0 to S5 are stored in an internal segment coordinate table. (Not shown). FIG. 3 shows primary segments S0 to S
5 schematically shows the segment coordinate table. For example, Xs to Ye for the primary segment S0
Xs = 10, Xe = 26, Ys = 61, Ye = 10
It turns out that it is 5.

Further, the cutout candidate positions (coordinates) C0 to C6 in the X direction necessary for creating a candidate path later.
Is stored in an internal predetermined memory (not shown).

The segment integrating section 19 integrates the primary segments extracted by the segment extracting section 17 according to a predetermined rule to create a secondary segment. Specifically, in consideration of the shape characteristics of a plurality of adjacent primary segments, if there is a possibility that they will be one character even if they are integrated, the primary segments are integrated to create a secondary segment. Here, a secondary segment is created by the following procedure.

First, the height H of the segment having the highest height among the m primary segments continuous in the X direction is obtained. The height H is obtained by calculating the difference between the Ys coordinate and the Ye coordinate for each primary segment. Considering the example of the six primary segments in the example of FIG. 2, the Y coordinate difference of the segment S5 is 112-5 = 107, which is higher than any of the other segments S0 to S4. Therefore, in the example of FIG. 2, the segment having the highest height is the segment S5.

Next, a secondary segment is created by integrating another primary segment present in the H × N coordinate range in the X direction with respect to the primary segment of interest into the primary segment of interest.

Regarding the secondary segment creation processing, the illustration of the primary segment shown in FIG. 2, the segment coordinate table shown in FIG. 3, the flow chart of the segment integration processing shown in FIG. 4, and FIG. This will be described more specifically with reference to the explanatory diagram of the secondary segment creation procedure.

First, the processing of loop 1 starting from each input segment is started. Therefore, as a target segment (segment A in FIG. 4), the process of loop 1 starting from the primary segment S0 is first started (steps 41 to 47 in FIG. 4).

That is, the primary segment S0 and the primary segment S1 as the segment B arranged to the right thereof
The distance D in the direction in which the characters are arranged (here, the X direction) is determined (steps 42 and 43 in FIG. 4). This distance D is obtained from, for example, the difference between the Xs coordinates of each of the segments S0 and S1. Then, in this example, the distance D = 52−10
= 42.

Next, it is determined whether or not the distance D is in the range of H × N (step 44 in FIG. 4). Here, N is a predetermined value. Here, N = 1.2. H is 107 here as described above. Therefore, in this case, it is determined whether or not D ≦ 1.2 × 107 = 128.4 is satisfied.

In the case of the example of FIG. 2, the primary segments S0 and S1
Since the distance D from the next segment S1 satisfies the condition of D ≦ 1.2 × 107, the primary segment S1 is integrated into the primary segment S0 to create the secondary segment S6 (step 45 in FIG. 4, (See FIG. 5).

Next, the loop 2 is executed again (FIG. 4
Steps 46 and 42), this time the primary segment S0
The distance D between the first segment S2 is obtained (step 43 in FIG. 4). This distance D is 126-10 = 116.

Next, it is determined whether or not the distance D is in the range of H × N (step 44 in FIG. 4). In this case, the primary segment S2 is D ≦ 1.2 × with respect to the primary segment S0.
Since the condition of 107 is satisfied, they are integrated and the secondary segment S7 is created (see step 45 in FIG. 4 and FIG. 5).

Next, since the loop 2 is executed again (FIG. 4)
Steps 46 and 42), this time the primary segment S0
The distance D between the first segment S3 and the primary segment S3 is determined (step 43 in FIG. 4). This distance D is 180-10 = 170.

Next, it is determined whether or not the distance D is in the range of H × N (step 44 in FIG. 4). In this case, the primary segment S3 is D ≦ 1.2 × with respect to the primary segment S0.
Since the condition of 107 is not satisfied, it is not integrated.

Next, the primary segment of interest is changed to primary segment S1, and the processing of loop 1 is started (FIG. 4).
Steps 41 and 42). Therefore, the primary segment S1
The distance D between the first segment S2 is obtained (step 43 in FIG. 4). This distance D is 126-52 = 74.

Next, it is determined whether or not the distance D is in the range of H × N (step 44 in FIG. 4). In this case, the primary segment S2 is D ≦ 1.2 × with respect to the primary segment S1.
Since the condition of 107 is satisfied, they are integrated and a secondary segment S8 is created (see step 45 in FIG. 4 and FIG. 5).

Next, the distance D between the primary segment S1 and the primary segment S3 is determined (step 43 in FIG. 4). This distance D is 180-52 = 128.

Next, it is determined whether or not the distance D is in the range of H × N (step 44 in FIG. 4). In this case, the primary segment S3 is D ≦ 1.2 × with respect to the primary segment S1.
Since the condition of 107 is satisfied, they are integrated and a secondary segment S9 is created (see step 45 in FIG. 4 and FIG. 5).

Next, the distance D between the primary segment S1 and the primary segment S4 is determined (step 43 in FIG. 4). This distance D is 231-52 = 179.

Next, it is determined whether or not the distance D is in the range of H × N (step 44 in FIG. 4). In this case, the primary segment S4 is D ≦ 1.2 × with respect to the primary segment S1.
Since the condition of 107 is not satisfied, it is not integrated.

Hereinafter, similarly, the primary segments S2 to S
5 by performing the processing of Loop 1 with each of them as a starting point (as a target segment) and integrating the primary segments,
Create a secondary segment. As a result, in this example, S6 ~
Nine in S14 (however, in FIG. 5, S6 to S9, S13,
A secondary segment (only S14 is shown) is created.

The segment integrating unit 19 stores the X direction start coordinate Xs, X direction end coordinate Xe, Y direction start coordinate Ys, and Y direction end coordinate Ye of each of the created secondary segments in the segment coordinate table. It is additionally stored (step 48 in FIG. 4).

FIG. 6 schematically shows a segment coordinate table for the primary segments S0 to S5 and the secondary segments S6 to S14.

Further, with respect to the above-mentioned primary segments S0 to S5 and secondary segments S6 to S14, a table in which what segments are sandwiched between a certain extraction candidate position and another extraction candidate position (this Is called a “segment table”) as shown in FIG.

This segment table is an adjacency matrix in graph theory. That is, considering an adjacent matrix in which the start cutout point is a row and the end cutout point is a column, the table can be created by giving a segment number (in this case, any of S0 to S14) to the element.
However, NULL in FIG. 7 indicates a blank, that is, no segment to be sandwiched.

As shown in FIG. 7, the primary segment S0 is a segment sandwiched between the clipping candidate position C0 and the clipping candidate position C1, and the primary segment S6 is sandwiched between the clipping candidate position C0 and the clipping candidate position C2. It can be seen that the secondary segment S7 is a segment sandwiched between the cutout candidate position C0 and the cutout candidate position C3.

The character recognition section 21 performs character recognition on the assumption that each primary segment and each secondary segment is one character, and stores a candidate character code as a result. The character recognition processing itself can be performed by any conventionally known method.

FIGS. 8A and 8B show an example in which the primary segment S0 shown in FIG. 2 is recognized as one character and is recognized as a character. FIG. 8A shows a primary segment S0, and FIG. 8B shows a set 51 of candidate characters as a result of recognition. Note that the candidate characters are actually given by candidate character codes. Although the number of candidates is not particularly limited, here, up to the K-th position is considered. In this case, an example considering K = 10 is considered. However, the case of FIG. 8B shows a case where only one candidate character appears.

The recognition results of the other primary segments S1 to S5 are shown in FIGS. 9A to 9E, respectively. Some examples (S6, S7, S11 to S1) of the recognition results of the secondary segments S6 to S14, respectively.
4A) to 10 (A) to 10 (C) and FIG.
1 (A) to (C).

The word creating section 23 creates a word by combining the candidate character codes of the primary segment and the secondary segment obtained by the character recognition.

The word creating section 23 of this embodiment is configured to include a candidate path creating section 23a and a character number rule checking section 23b. The configuration and operation of the candidate path creation unit 23a and the character number rule checking unit 23b will be described below with reference to FIG.

The candidate path creation unit 23a performs the following (1) to
A candidate path represented by the concatenation of the primary segment and / or the secondary segment including the means of (8) is created.

(1) When the coordinates for dividing each of the m primary segments in the X direction are set as the extraction candidate positions C0 to Cm, the extracted extraction candidate positions Ci (i = 0 to 0)
First means for determining whether or not m) is Cm (step 61 in FIG. 12).

(2) The second means for operating when the first means determines that Ci = Cm and recording the current candidate path in a candidate path memory (not shown) (step 62 in FIG. 12). ).

(3) The operation is performed when Ci ≠ Cm is determined by the first means, and the segment Sk sandwiched between the extraction candidate position Ci and the extraction candidate position Cj (j = i + 1)
Third means for determining whether or not +1 exists (steps 63 to 65 in FIG. 12).

(4) The fourth means for operating when the number-of-characters rule checking unit determines that the number is within the specified number and adding the segment Sk + 1 to the candidate path (step 6 in FIG. 12)
6).

(5) Fifth means (FIG. 5) which operates following the fourth means and starts the operation of the first means by regarding the cut candidate position Cj as the focused cut candidate position Ci. Twelve steps 67).

(6) For a candidate path created as a result of the operation of the fourth means and the fifth means, a sixth means for deleting, from the candidate path, the segment most recently added to the candidate path. (Step 68 in FIG. 12).

(7) If the third means determines no,
Alternatively, when the character number rule checking unit determines that the number exceeds the specified number, or when the sixth means is operated, it is operated, and j defining the cutout candidate position is changed to j = j + 1 and changed. Seventh means for determining whether j satisfies j> m in relation to m (step 6 in FIG. 12)
9, 70).

(8) An eighth means (steps 70 and 64 in FIG. 12) that operates when the seventh means determines that j ≦ m and operates the third means.

On the other hand, the character number rule checking unit 23b uses the segment Sk by the third means included in the candidate path creating unit 23a.
The operation is performed when it is determined that +1 exists, and when the segment Sk + 1 is added to the candidate path, the number of characters is regulated by determining whether the number of segments of the candidate path does not exceed a specified number (FIG. 12). Step 71).

In order to deepen the understanding of the candidate path creating unit 23a and the character number rule checking unit 23b, a specific example of the candidate path creating process will be described. However, since the principles of the candidate path creation process and the character number rule checking process need only be explained, an example in which the number of segments and the number of extraction candidate positions are reduced will be described. That is, there are three extraction candidate positions C0 to C2, and the segment is S0
In the case where the relationship between each of the cutout candidate positions C0 to C2 and each of the segments S0 to S2 has a relationship represented by a segment table as shown in FIG. The creation processing and the character number rule inspection processing will be described below.

Note that the cutout candidate position C0 is the first cutout candidate position in the direction in which the characters are arranged, and the cutout candidate position C2 is the last (rightmost) cutout candidate position in the direction in which the characters are arranged.

First, the path P in the candidate path memory (not shown) is cleared, and the process for the function Funct (Ci, P), here (C0, P), is first started (step 60 in FIG. 12).

That is, first, the candidate path creation unit 23a determines whether or not the cut candidate position Ci, here C0, is the rightmost cut candidate position (ie, whether or not C0 = C2) (step 61 in FIG. 12). Note that the clipping candidate position Ci
Here, the control unit 11 transfers from the segment extraction unit 17 to the candidate path creation unit 23b.

Here, since C0 is not the rightmost cut-out candidate position, the process proceeds to step 63. That is, j =
The processing of i + 1 = 0 + 1 = 1 is performed. As a result, Cj
Becomes C1.

The segment Sk + 1 sandwiched between the extraction candidate position C0 and the extraction candidate position C1 is sent from the segment extraction unit 17 or the segment integration unit 19 to the control unit 11
Is transferred to the candidate path creation unit 23b. In the case of the example of FIG. 13, the segment S0 is transferred. If there is no corresponding segment, the control unit 11 sends a signal to that effect (NU
LL) is transferred to the candidate path creation unit 23a.

The candidate path creation section 23a sets the segment Sk
It is determined whether or not +1 exists (step 6 in FIG. 12).
5). In this case, since the segment S0 exists, the process proceeds to step 71.

In step 71, it is determined whether or not the number of segments in a column (segment column) formed by adding the segment Sk + 1 to the candidate path P is within a specified number. Here, the prescribed number is considered to be four. In this case, since the segment of the segment row is only S0, the number of segments is 1, so that
Since the specified number is satisfied, the process proceeds to step 66.

At step 66, processing for adding the segment S0 to the candidate path P is performed. As a result, the candidate path P = {S0}. Thereafter, the process proceeds to step 67.

In step 67, since Cj is set as the target extraction candidate position this time, the extraction candidate position C1 is regarded as the target extraction candidate position Ci. That is, the function is set to Funct (C1, P). Then, the processing is restarted from the processing of step 61.

Therefore, the candidate path creation unit 23a determines whether the cutout candidate position C1 is the rightmost cutout candidate position (ie, whether or not C1 = C2) (FIG. 12).
Step 61).

Since C1 is not the rightmost cut-out candidate position, the process proceeds to step 63. That is, j =
The processing of i + 1 = 1 + 1 = 2 is performed. As a result, C
j becomes C2.

The segment Sk + 1 sandwiched between the extraction candidate position C1 and the extraction candidate position C2 is sent from the segment extraction unit 17 or the segment integration unit 19 to the control unit 11
Is transferred to the candidate path creation unit 23b. In the case of the example of FIG. 13, the segment S1 is transferred.

[0109] The candidate path creation unit 23a calculates the segment Sk
It is determined whether or not +1 exists (step 6 in FIG. 12).
5). In this case, since the segment S1 exists, the process proceeds to step 71.

In step 71, it is determined whether or not the number of segments of a column (segment column) formed by adding the segment Sk + 1 to the candidate path P is within a specified number. In this case, the number of segments in the segment row is two, S0 and S1, so that the specified number is satisfied.

At step 66, processing for adding the segment S1 to the candidate path P is performed. As a result, the candidate path P = {S0, S1}. Thereafter, the process proceeds to step 67.

In step 67, the cut candidate position C2 is regarded as the target cut candidate position Ci this time. That is, the function is set to Funct (C2, P). Then, the processing is restarted from the processing of step 61.

Therefore, the candidate path creation unit 23a determines whether or not the cutout candidate position C2 is the rightmost cutout candidate position (ie, whether or not C2 = C2) (FIG. 12).
Step 61).

Since C2 is the rightmost cut-out candidate position, the process proceeds to step 62. Therefore, the candidate path P = {S0, S1} is recorded in the candidate path memory (not shown). As a result, a candidate path P = {S0, S1} is created as one of the candidate paths whose start point is C0, whose end point is C2, and whose number of segments is equal to or less than the specified number. Also, by the processing up to this point, the function Funct
The processing of (C2, P) ends.

This candidate path P = {S0, S1} is the fourth
And candidate paths created as a result of the operation of the fifth means and the fifth means. That is, it is a candidate path created as a result of the processing of steps 66 and 67. Therefore, the process proceeds to step 68. In this step 68, the candidate path P
= S0, S1}, the segment S1 newly added to this is deleted. As a result, the candidate path P
= {S0}.

Next, j = j + 1 is set (step 69 in FIG. 12). Here, the current j is 2, so j = 2 + 1
= 3.

Next, it is determined whether or not j is the maximum clipping candidate position (whether or not j> m) (step 7 in FIG. 12).
0).

Since j = 3 in this case exceeds the maximum cutout candidate position 2, the processing of the function Funct (C1, P) is completed. Therefore, this time the original function Fun
The processing from step 68 is performed for ct (C0, P).

Accordingly, from the candidate path P = {S0},
Then, a process of deleting the segment S0 added most recently is performed. As a result, the candidate path P = {｛} = 0.

Next, j = j + 1 is set (step 69 in FIG. 12). Here, the current j is 1, so j = 1 + 1
= 2.

Next, it is determined whether or not j is the maximum clipping candidate position (whether or not j> m) (step 7 in FIG. 12).
0).

Since j = 2 in this case does not exceed the maximum clipping candidate position 2, the processing from step 64 is performed. Therefore, it is determined whether or not there is a segment Sk + 1 sandwiched between the extraction candidate position C0 and the extraction candidate position C2. Segment Sk in this case
Since the segment S2 exists as +1 (see FIG. 13), it is determined whether or not the number of segments in the segment string obtained by adding the segment S2 to the candidate path P is within a specified number.

Since the segment of this segment row is only S2, it satisfies the specified number 3 or less. Therefore, the segment S2 is added to the candidate path P. As a result, the candidate path P = {S2}. Thereafter, the process proceeds to step 67.

Since Cj in this case is 2, in this step 67, the extraction candidate position C2 is regarded as the extraction candidate position Ci of interest. That is, the function is set to Funct (C2, P). And step 61
Processing is restarted from the processing of.

Therefore, the candidate path creating section 23a determines whether or not the cutout candidate position C2 is the rightmost end cutout candidate position (ie, whether or not C2 = C2) (FIG. 12).
Step 61).

Since C2 is the rightmost cut-out candidate position, the process proceeds to step 62. Therefore, the candidate path P = {S2} is recorded in the candidate path memory (not shown). Thus, the starting point is C0, the ending point is C2,
In addition, a candidate path P = {S2} is created as one of the candidate paths whose number of segments is equal to or less than the specified number. Further, the processing for the function Funct (C2, P) is completed by the processing up to this point.

This candidate path P = {S2} is a candidate path created as a result of the operation of the fourth means and the fifth means. That is, it is a candidate path created as a result of the processing of steps 66 and 67. Therefore, the process proceeds to step 68. In this step 68, the candidate path P = ｛S
From 2｝, processing is performed to delete the segment S2 that has been added to this most recently. As a result, the candidate path P = {｛} = 0
become.

Next, j = j + 1 is set (step 69 in FIG. 12). Here, the current j is 2, so j = 2 + 1
= 3.

Next, it is determined whether or not j is the maximum clipping candidate position (whether or not j> m) (step 7 in FIG. 12).
0).

Since j = 3 in this case exceeds the maximum clipping candidate position 2, the processing of the function Funct (C0, P) ends.

The process described with reference to FIG. 12 is a process called a recursive algorithm. This is a process capable of recursively tracing from the leftmost segment of the character string to the rightmost segment.

According to this process, all segments starting from C0 are extracted. In addition, other connected segments with the end point of the extracted segment as a starting point are further sequentially extracted. In addition, a candidate path composed of a sequence of segments (including one segment) whose start point is C0 and whose end point is C2 and whose number of segments is equal to or less than a specified number is easily created.

The processing described with reference to FIG.
The primary segment S of the original image data 31 shown in FIG.
When candidate paths are created by applying the candidate paths to 0 to S5 and the secondary segments S6 to S14, as shown in FIG.
To [17], a total of 17 candidate paths are created. However, this is the case where the specified number of segments constituting the candidate path is 4 and the candidate path is created.

As can be seen from FIG. 14, the segment S where the extraction candidate position C0 is the extraction start point
With 0, S6, and S7 as vertices, candidate paths composed of segments connected to these segments in a tree shape are created. However, in each candidate path, the segment at the right end is a segment (specifically, one of S13, S14, and S15) ending at the cutout candidate point C6. Moreover, the number of segments constituting each candidate path is equal to or less than a specified number (here, equal to or less than 4).

FIG. 15 shows the principle of creating a word from the 17 candidate paths shown in FIG.

However, FIG. 15 shows an example in which a word is created from each of candidate path [1], candidate path [2], candidate path [3] and candidate path [17] among the 17 candidate paths. Is shown.

In FIG. 15, the candidate path [1] includes segments S0, S1, S2 as described with reference to FIG.
And S13. Also, here, segment S
The recognition results of each of 0, S1, and S2 are shown in FIGS.
As shown in (B), there is only one and each is “1”. Therefore, “1 (1)” is entered as a candidate character in each of the positions of the segments S0, S1, and S2 in the column of the candidate path [1] in FIG. On the other hand, segment S1
As shown in FIG. 11B, there are a total of ten recognition results for the first to tenth places. Therefore, at the position of the segment S13 in the column of the candidate path [1] in FIG. 15, each candidate character such as “kawa (1)”... Therefore, from this candidate path [1],
"111 river" ... ten words up to "111 le" are created. Similarly, a candidate word can be created from the candidate path [2], the candidate path [3],..., The candidate path [17].

The word matching section 25 checks the word created by the word creating section 23 against a word dictionary (not shown). The word matching unit 25 stores “0” as an evaluation value of the candidate path if the word indicated by the candidate path does not exist in the word dictionary as a result of the word matching. For example, "1" is stored as the evaluation value of the candidate path.

For example, in the case of the example shown in FIG. 15, from the word “111 river” in the column of the candidate path [1] to the word “11” in the column of the candidate path [3], and from the candidate path [1]. 17] does not exist in the word dictionary. Therefore, “0” is stored as the evaluation value of each of these candidate paths. On the other hand, the word "Ogawa" in the column of the candidate path [17] exists in the word dictionary. Therefore, “1” is stored as the evaluation value of this candidate path.

The result selecting section 27 determines one of the primary segment and the secondary segment constituting the candidate path as a character cutout area for one character based on the result of the word collation. Here, each segment of the candidate path having the best evaluation value in word matching is set as a cutout area. In the example of FIG. 15, the segments S7 and S13 constituting the candidate path "Small (1) river (1)" in the candidate path [17] are each character extraction regions.

The result output unit 29 outputs a cutout candidate position corresponding to the character cutout area determined by the result selection unit 27 to, for example, the control unit 11. In the example of FIG. 15, C
0, C3, and C6 are output to the control unit 11 as character extraction candidate positions. The control unit 11 can instruct character cutout based on the result.

When a plurality of candidate paths having an evaluation value of “1” appear in the result of word matching, for example,
According to the second evaluation method, the third evaluation method, or the like, a word to be a criterion for character segmentation is determined, such as selecting a word composed of a recognition result having a high candidate rank by focusing on the candidate rank in the recognition result. good.

The embodiment of the present invention has been described above. However, the present invention is not limited to the above-described embodiment, and many modifications or changes can be made.

For example, in the above-described embodiment, when defining the number of characters, the number of characters is defined to be equal to or less than the reference value. However, two reference values such as T1 and T2 (T1 <T2) are provided as reference values. The number of characters may be restricted such that T1 <number of characters (in the embodiment, the number of segments) <T2. In this case, the number of candidate paths on the lower limit side can be restricted, so that the number of candidate paths can be further reduced. Therefore, the amount of processing such as word creation processing can be further reduced.

[0145]

As is apparent from the above description, according to the invention of the character extracting method of this application, a black bit block area is obtained from the original image data including the image data of the character string stored in the memory. , A process of integrating the extracted primary segments according to a predetermined rule to create a secondary segment, and a process of recognizing each primary segment and each secondary segment with a character. A process of creating a word by combining candidate character codes of the primary segment and the secondary segment obtained by the character recognition; a process of matching the created word with a word dictionary; Determining a character cut-out area based on the number of characters constituting the word by controlling the number of characters constituting the word. To create a word. Therefore, useless candidate character codes and combinations thereof can be omitted from the calculation target when creating words. Therefore, it is possible to eliminate the waste of the calculation amount, and it is possible to perform more efficient character segmentation.

Further, according to the invention of the character extracting device of this application, the invention of the above-described character extracting method can be easily implemented.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of a character cutout device according to an embodiment.

FIG. 2 is a diagram for explaining a primary segment and a clipping candidate position.

FIGS. 3A and 3B are diagrams showing examples of a segment coordinate table before integration.

FIG. 4 is an explanatory diagram of a segment integration unit.

FIG. 5 is an explanatory diagram of a secondary segment creation procedure.

FIG. 6 is a diagram showing an example of a segment coordinate table after integration.

FIG. 7 is a diagram showing an example of a segment table after integration.

FIG. 8 is a diagram showing an example (part 1) of a recognition result;
FIG. 9 is a diagram showing a recognition result of a segment S0.

FIGS. 9A to 9E are examples of recognition results (part 2); FIGS.
And is a diagram showing recognition results of the segments S1 to S5, respectively.

FIGS. 10A to 10C are diagrams illustrating an example (part 3) of a recognition result, and include segments S6, S7, and S11.
It is the figure which showed each recognition result.

FIGS. 11A to 11C are diagrams illustrating an example (part 4) of a recognition result, and include segments S12, S13, and S.
It is the figure which showed each of 14 recognition results.

FIG. 12 is an explanatory diagram of a candidate path creation unit and a character number rule checking unit.

FIG. 13 is an explanatory diagram of a specific example of a candidate path creation process.

14 is an explanatory diagram of a candidate path, and is an explanatory diagram of a candidate path created from the original image data shown in FIG. 2 with a specified number of 4; FIG.

FIG. 15 is an explanatory diagram of the principle of creating a word from a candidate path.

[Explanation of symbols]

 10: Character extraction device of embodiment 11: Control unit 13: Image input unit 15: Character number rule input unit 17: Segment extraction unit 19: Segment integration unit 21: Character recognition unit 23: Word creation unit 23a: Candidate path creation unit 23b: Character number rule checking unit 25: Word matching unit 27: Result selecting unit 29: Result output unit S0 to S5: Segment (primary segment) S6 to S14: Segment (secondary segment) C0 to C6: Cutout candidate position 51: Set of candidate characters

Claims (9)

[Claims] 1. A process for extracting primary segments each of which is a black bit lump area from original image data including image data of a character string stored in a memory, and extracting each of the extracted primary segments by a predetermined process. Integrate according to the rules 2
A word is created by combining a process of creating a next segment, a process of character recognition of each primary segment and each secondary segment, and a candidate character code of each of the primary segment and secondary segment obtained by the character recognition. Processing, processing of matching the created word with a word dictionary, and extracting one of the primary segments and the secondary segments based on the result of the word matching into a character cutout area for one character. Determining the number of characters constituting the word, and creating the word. 2. The character extracting method according to claim 1, wherein a reference value for regulating the number of characters is determined according to a field to be recognized. 3. The character segmenting method according to claim 1, wherein the secondary segment is defined as follows: when the direction in which characters are arranged is the X direction, m primary segments that are continuous in the X direction are determined according to a predetermined rule. The word is a candidate path represented by the concatenation of the primary segment and / or the secondary segment, and the following words (a) to
A character segmentation method (where m is an integer of 2 or more), which is created based on a candidate path created by a process including the process (d). (a) Coordinates for dividing each of the m primary segments in the X direction are determined as candidate clipping positions Ci (i = 0 to 0).
m), from among the m primary segments and the created secondary segment, a cutout candidate position C
A process of extracting all the segments for which 0 is the extraction start point. (b) With respect to each of the segments extracted in the process (a), the segmentation candidate position Cj (j = 1 to m) on the end point side of the segment is the segmentation start position, so that the segments can be connected. Other segments which have the same relationship with the other segment and have the same candidate clipping position as described above, and which can be further connected to each other, are replaced by other segments whose candidate candidate positions on the end point side are Cm. A process of extracting from the m primary segments and the created secondary segments until appears. (c) A process of determining whether or not the number of segments of a candidate path including the other segment is within a specified number every time the other segment is extracted in the process of (b). (d) A process in which a candidate path having a number of segments within the specified number and a candidate position for extracting the end point of the last segment in the candidate path being Cm is set as a candidate path for word creation. . 4. The character segmentation method according to claim 1, wherein, when the direction in which characters are arranged is the X direction, m secondary segments that are continuous in the X direction are determined according to a predetermined rule. The word is a candidate path represented by the concatenation of the primary segment and / or the secondary segment, and is expressed by the following (1) to
A character segmentation method characterized by being created based on a candidate path created by a process including the process (9) (where m is an integer of 2 or more). (1) When coordinates for partitioning each of the m primary segments in the X direction are set as candidate clipping positions C0 to Cm, it is determined whether or not the focused candidate clipping position Ci (i = 0 to m) is Cm. First processing for determination. (2) A second process that is executed when it is determined that Ci = Cm in the first process, and records the current candidate path in the candidate path memory. (3) This is executed when it is determined that Ci ≠ Cm in the first processing, and the extraction candidate position Ci and the extraction candidate position Cj
Third processing for determining whether or not a segment Sk + 1 sandwiched between (j = i + 1) exists. (4) This is executed when it is determined that a segment exists in the third processing, and when the segment Sk + 1 is added to a candidate path, it is determined whether or not the number of segments of the candidate path does not exceed a specified number. Fourth processing. (5) A fifth process which is executed when it is determined in the fourth process that the number is within the specified number, and adds the segment Sk + 1 to the candidate path. (6) A sixth process which is executed subsequent to the fifth process, and regards the clipping candidate position Cj as the focused clipping candidate position Ci and re-executes the first process. (7) A seventh process of deleting, from the candidate path, a segment that is most recently added to the candidate path for a candidate path created by executing the fifth process and the sixth process. (8) This is executed when it is determined to be no in the third process, or when it is determined to be no in the fourth process, or when the seventh process is executed, and the cutout candidate position is determined. An eighth process of changing the specified j to j = j + 1 and determining whether the changed j satisfies j> m in relation to the m. (9) A ninth process that is executed when j ≦ m is determined in the eighth process, and is executed again from the third process. 5. The character segmenting method according to claim 1, wherein the predetermined rule for integrating the primary segments is that, when the direction in which the characters are arranged is the X direction, The height H of the segment having the highest height among the m primary segments is determined, and the other primary segments existing in the H × N coordinate range in the X direction with respect to the primary segment of interest are obtained. A character segmentation method using a rule of integrating segments into the primary segment of interest (where N is a predetermined value). 6. The character extracting method according to claim 1, wherein the character string is a character string including a handwritten character. 7. A segment extraction unit for extracting a primary segment, which is a block area of black bits, from input image data including image data of a character string stored in a memory, and each extracted primary segment. A segment integrating unit that creates a secondary segment by integrating the primary segment and the secondary segment according to predetermined rules, a character recognizing unit that recognizes each primary segment and each secondary segment, and a primary segment and / or a secondary segment obtained by the character recognition. A word creating unit that creates a word by combining the candidate character codes; a word matching unit that matches the created word with a word dictionary; and the primary segment and the 2nd segment based on the result of the word matching. A result selection unit for determining any one of the next segments as a character cutout area for one character;
A character segmentation device comprising: a character number regulation inspection unit for regulating the number of characters constituting a word in the word creation unit. 8. The character segmenting device according to claim 7, wherein, when the direction in which characters are arranged is the X direction, m segment segments that are continuous in the X direction are defined as the segment integrating unit according to a predetermined rule. A segment integrating unit for integrating, the word generating unit includes: (A) generating a candidate path represented by a concatenation of the primary segment and / or the secondary segment including the following means (1) to (8): (B) operates when the third means included in the candidate path creation unit determines that a segment exists, and when the segment Sk + 1 is added to the candidate path, the number of segments of the candidate path A character cutout device (where m is an integer of 2 or more), comprising a character number rule checker that controls the number of characters by determining whether the number does not exceed a prescribed number. (1) When coordinates for partitioning each of the m primary segments in the X direction are set as candidate clipping positions C0 to Cm, it is determined whether or not the focused candidate clipping position Ci (i = 0 to m) is Cm. First means for determining. (2) The second operation is performed when the first means determines that Ci = Cm, and records the current candidate path in the candidate path memory.
Means. (3) The first means operates when it is determined that Ci ≠ Cm, and the cut-out candidate position Ci and the cut-out candidate position Cj
Third means for determining whether or not a segment Sk + 1 sandwiched between (j = i + 1) exists. (4) A fourth means which operates when the number-of-characters rule checking unit determines that the number is within the specified number, and adds the segment Sk + 1 to the candidate path. (5) A fifth means which operates following the fourth means, and regards the cutout candidate position Cj as the focused cutout candidate position Ci, and starts the operation of the first means. (6) A sixth means for deleting, from the candidate path, a segment most recently added to the candidate path for the candidate path created as a result of the operation of the fourth means and the fifth means. (7) When the third means determines no, or when the character number rule checking unit determines that the number exceeds a specified number, or operates after the sixth means operates, and specifies the cutout candidate position. A seventh means for changing j being performed to j = j + 1 and determining whether or not the changed j satisfies j> m in relation to m. (8) operates when the seventh means determines that j ≦ m,
Eighth means for operating the third means. 9. The character segmenting device according to claim 7, wherein the segment integrating unit is configured to calculate m primary segments that are continuous in the X direction when a direction in which characters are arranged is the X direction. A means for calculating the height H of the segment having the highest height, and the other primary segments existing in the H × N coordinate range in the X direction with respect to the primary segment of interest as the primary segment of interest. A character extracting device (where N is a predetermined value).