JPH0467674B2 - - Google Patents

Description

[Detailed description of the invention]

[Industrial Field of Application] The present invention relates to a document reading device, and particularly to an improvement in means for cutting out a character pattern from a character string in a read document. [Prior art] To recognize characters, an image pattern is created by photoelectrically converting the characters written on paper into binary signals of "1" and "0" corresponding to the character part and its background part. The character pattern must be cut out one character at a time. Fig. 7 is a diagram showing the configuration of a conventional device of this kind shown in Japanese Patent Publication No. 55-42434. Scanning means to be replaced; 3 is a pattern memory that stores image information of one line of characters that have been photoelectrically exchanged; 4 is a peripheral distribution value determined in advance by scanning the pattern memory 3 up and down for each line of characters; The image pattern in the pattern memory 3 (hereinafter referred to as (referred to as "basic pattern")
This is left end/right end detection means for detecting the left end coordinates and the right end coordinates of. Reference numeral 5 denotes frame information detection in which the coordinates of a character frame are detected from mark information, etc., given in advance as document format information or written on the sheet 1, and this information is stored. 6 is a character cutting means for cutting out a pattern of one character, and 7 is a cutting pattern buffer for storing the cut out character pattern. The conventional character pattern cutting device is configured as described above, and compares the coordinates of the left end and right end of the basic pattern detected by the left end/right end detection means 4 with the coordinates of the character frame detected by the frame information detection means 5. By doing this, a pattern of one character can be cut out. [Problems to be Solved by the Invention] However, in the conventional device as described above, character patterns are cut out character by character based on the coordinates of the character frame and the size of the character frame. For characters such as ``word'' and ``go'' that can be read as independent characters even if separated to the left and right, if the characters are written with the character frame in mind, the two character elements ``go'' and ``go'' can be written correctly without separating them. It is possible to cut out characters as characters, but if the characters are handwritten strings written in a free format without being aware of the character frame, there is a problem that the character pattern cannot be extracted correctly because the character frame information cannot be used. Ta. The present invention was made in order to solve such problems, and an object of the present invention is to provide a document reading device that can correctly cut out character strings written in a document without a character entry frame one character at a time. It is said that [Means for solving the problem] The document reading device according to the present invention binarizes the marginal distribution value obtained by scanning a character string image of one line in a direction orthogonal to the character string using a threshold value, Basic pattern area detection means detects the left, right, top, and bottom coordinates of the basic pattern cut out from the character string image for each basic pattern based on the continuity of the result, and a plurality of adjacent basic patterns are combined. The character evaluation value indicating the character-likeness of each shape of the combined pattern and basic pattern obtained by A character evaluation value calculation means calculates each character evaluation value based on the character string image, and calculates all possible combinations of character pattern cutout candidate positions in the character string image from the position information of the basic pattern, and calculates the cutout candidate positions for each set of cutout candidate positions. A cutout evaluation value for one line of character strings is calculated based on the character evaluation values of the basic pattern and combined pattern cut out at each position, and a set of character cutout positions based on the basic pattern and combined pattern that has the maximum cutout evaluation value is determined. a character cutting position determining means for determining an optimal combination of character cutting positions, and a character cutting means for cutting out character patterns one character at a time from a character string image based on positional information of a combination of a basic pattern and a combined pattern determined by the character cutting position determining means. It is equipped with the following. [Function] In the present invention, after the basic pattern detection means detects the positional information of the basic pattern of character elements constituting a character, the character evaluation value calculation means uses the positional information of the basic pattern to detect the basic pattern and the basic pattern. A character evaluation value that quantitatively represents how much the combined pattern resembles a character pattern when the basic patterns are combined is calculated, and the character cutting position determining means cuts out the character as a character based on the size of this character evaluation value. Determine the position of the basic pattern. Therefore, characters can be correctly extracted even without information such as character frames. [Embodiments of the Invention] The present invention will be described in detail below with reference to the drawings. FIG. 1 is a diagram showing the configuration of an embodiment of the present invention. In the figure, 1, 2, 3, and 7 are the same as the conventional device. Reference numeral 8 denotes basic pattern area detection means for detecting the coordinates of the upper and lower ends of the basic pattern in addition to the coordinates of the left and right ends of the basic pattern obtained using the same method as in the conventional apparatus. 9 is a value that quantifies how much each basic pattern and a combined pattern formed by combining a plurality of consecutive basic patterns are likely to be patterns of one character, based on the basic pattern position information obtained by the basic pattern area detection means 8. This is a character evaluation value calculation means for calculating a character evaluation value (hereinafter referred to as a "character evaluation value"). Reference numeral 10 denotes a character cutting position determining means for determining the position of a character pattern to be cut out based on the size of the character evaluation value calculated by the character evaluation value calculating means 9. Reference numeral 11 denotes a character cutting means for reading character patterns one character at a time from the pattern memory 3 based on the cutting position information determined by the character cutting position determining means 10 and transferring the read character patterns to the cutting pattern buffer 7. FIG. 2 is a diagram showing an example of an input character string as a one-line character string image written on a form. In the figure, 1 is the paper, 12 is the input character string "read" as an example, and 13 is the input character string. It is a character frame. FIG. 3 is a diagram for explaining the processing of the basic pattern area detection means 8, in which 14 is the image information stored in the pattern memory 3, 15 is a character projection diagram of the image information 14, and 16 ,17,1
8 and 19 are the coordinates of the left end, right end, top end, and bottom end of the basic pattern. The operation of the basic pattern area detecting means 8 will be explained below with reference to FIGS. 2 and 3. First, the character string "read" 12 written on the paper 1 in FIG. 2 is photoelectrically converted by the scanning means 2, and
This becomes image information for one line and is stored in the pattern memory 3. Note that the character frame 13 shown in FIG.
Although it is not particularly related to the processing of the present invention, the input character string "read" 12 is presented for convenience to make it clear that it is written at irregular pitches, and the input character string "read" 12 is presented for the sake of convenience to make it clear that it is written at an irregular pitch. The image information 14 does not include image information corresponding to character frames. Next, the image information 14 in the pattern memory 3 is
The data is transferred to the basic pattern area detection means 8. The basic pattern area detection means 8 extracts basic patterns that are constituent elements of characters from the image information 14, and detects the coordinates of the left end, right end, top end, and bottom end of a rectangle circumscribing each basic pattern as position information of the basic pattern. do. Specifically, each pixel in the pattern memory 3 is scanned in the vertical direction perpendicular to the input character string "read" 12 and the number of pixels representing the character part of the image information 14 is counted, thereby reading the image information 14. The marginal distribution value is calculated, and if this marginal distribution value is greater than or equal to a predetermined threshold value K, it is converted into binary image information of "1", otherwise it is "0", and the characters corresponding to the original image information are converted. A projection diagram 15 is created. Next, in the character projection diagram 15, the width of the part where the image information with the value "1" is continuous is regarded as the width of the basic pattern,
The coordinates 16 of the edge and the coordinate 17 of the right edge of each basic pattern are detected from the image information 14. Further, after detecting the coordinates 16 of the left end and the coordinates 17 of the right end of each basic pattern, the coordinates 18 of the top end and the coordinates 19 of the bottom end of each basic pattern are detected. For example, the coordinate 18 of the upper end of the basic pattern "word" in Figure 3 and the coordinate 1 of the lower end
9 scans each pixel of the image information 14 in the range surrounded by the leftmost coordinate 16 and the rightmost coordinate 17 of the basic pattern "word" in the horizontal direction from the top row to the bottom row. At this time, the coordinates when the first pixel representing the character part is encountered are determined as the coordinates 18 of the upper end of the basic pattern "word", and the coordinates of the row of pixels representing the last character part encountered are determined as the coordinates 19 of the lower end of the basic pattern. do. Table 1 is a table showing the position information of the basic pattern, including "word", "sale", and "sale" extracted from the image information 14.
Position information is shown for a total of six basic patterns: "mi", "mimi", "mata", and "ru".

[Table] Next, the processing of the character evaluation value calculation means 9 will be explained. The character evaluation value calculation means 9 uses the position information of each basic pattern shown in Table 1 to determine how much a pattern composed of one basic pattern is like a character, and also to combine multiple consecutive basic patterns. A character evaluation value indicating how much the combined pattern formed by the above characters resembles a character is calculated. The character evaluation value is calculated by the following method. Characters are generally small, independent figures, and the following properties have been experimentally confirmed. (1) Each letter has a well-balanced shape.
A rectangle circumscribing a character has a shape close to a square. (2) There is a blank space between characters that separates them. (3) Even if a character element such as "word" can be separated to the left and right as a single character, the blank space within this character is the same as the width of the blank space between characters as shown in item 2 above. narrow in comparison. Therefore, in the present invention, a character evaluation value indicating the degree to which the combined pattern obtained by combining a plurality of adjacent basic patterns and the basic pattern resembles a character is provided between the shape (aspect ratio) of the basic pattern and the basic pattern. Calculated based on the width of the blank area. Specifically, P ₁₁ , P ₂₂ ... PMM (total M pieces) in order from the left basic pattern in the image information 14
Then, the i-th basic pattern Pii from the left
The character evaluation value V(i,
The degree to which the pattern Pij resembles a character pattern is evaluated based on the magnitude of the value of j). V(i,j)=C ₁ −C ₂ × |Wij−Hij|/MAX(Wij, Hi
j) −C ₃ × IWij / Hij + C ₄ × Owij / Hij ... (1) Here, Wij : Width of pattern Pij Hij : Height of pattern Pij IWij : Blank area existing in pattern Pij OWij : Width of pattern Pij Blank portions C ₁ to C ₄ at both ends are constants. The character evaluation value V(i, j) determined by Equation 1 is determined by Takes a large value. In this case, Wij to OWij can be determined from the coordinates of the left end, right end, upper end, and lower end of each basic pattern determined by the basic pattern area detecting means 8. For example, when the coordinates of the left and right ends of basic pattern Pii are SXi, EXi (SXi≦EXi), and the coordinates of the top and bottom ends are SYi, EYi (SYi≦EYi), Wij, Hij,
IWij, OWij, can be determined using equations 2 to 7 shown below. Wij=EXj−SXi+1 …(2) Hij=MAX(EYk)−MIN(SYk)+1……(3) k=i,j k=i,j IWij=O (i=j) Σ(SX _{k+ 1} −EXk−1(i<j) k=i, j−1 …(4) OWij{(SXi−EX _il −1)+(SX _j+1 −EXj−1)
}/2 (1<i, j<M) SXi−EX _il −1 (≠1, j=M) SX _j+1 −EXj−1 (i=1, j≠M) O (i=1, and j=M)...(5) Table 2 is a table showing the processing results in the character evaluation value calculation means 9, and is a table that shows the processing results of the character evaluation value calculation means 9, and is a table that shows the processing results of the character evaluation value calculation means 9. Basic patterns ``goto'', ``sale'', ``mi'', ``mimi'', ``mata'', ``ru'' and combined patterns that combine multiple consecutive basic patterns ``yomi'', ``sale'', ``mimi'' , "tori", "mataru",
The character evaluation value of ... is calculated. for example,
The character evaluation value for the basic pattern “Word” is “68”
This shows that the character evaluation value for the pattern combining the two basic patterns "goto" and "sale" is "146". Note that this example shows an example in which character evaluation values are obtained when up to three consecutive basic patterns are combined.

【table】

〔Effect of the invention〕

As explained above, this invention cuts out characters one by one based on the shape of the patterns and the positional relationship between the patterns, so it is possible to cut out characters one by one based on the shape of the patterns and the positional relationship between the patterns. It is possible to accurately cut out characters one by one.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention, FIG. 2 is a diagram showing an example of an input character string written on a sheet, and FIG. 3 is a diagram for explaining the processing of the basic pattern area detection means. , FIG. 4 is a diagram showing an example of a combination of a basic pattern and a combined pattern, FIG. 5 is a diagram for explaining a method for obtaining a cutout evaluation value, and FIG. 6 is a diagram showing an example of a cutout evaluation value.
FIG. 7 is a block diagram showing the configuration of a conventional device. DESCRIPTION OF SYMBOLS 1...Paper, 2...Scanning means, 3...Pattern memory, 8...Basic pattern area detection means, 9...
...Character evaluation value calculation means, 10...Character cutting position determining means, 11...Character cutting means, 12...
...Cut out pattern buffer.

Claims (1)

[Claims] 1 In a document reading device that reads an image of characters written on a document and recognizes the characters based on the read image, the marginal distribution value obtained by scanning one line of character string image is binarized using a threshold value. , basic pattern area detection means for detecting the left, right, top, and bottom coordinates of the basic pattern cut out from the character string image for each basic pattern based on the continuity of the result, and combining a plurality of adjacent basic patterns. The character evaluation value indicating the character-likeness of each shape of the combined pattern and basic pattern obtained by A character evaluation value calculation means that calculates each character evaluation value based on A cutout evaluation value for one line of character strings is calculated based on the character evaluation values of the basic patterns and combined patterns cut out at candidate positions, and a set of character cutout positions based on the basic pattern and combined pattern that has the maximum cutout evaluation value is determined. a character cutting position determining means for determining an optimal combination of character cutting positions, and a character cutting means for cutting out character patterns one character at a time from a character string image based on positional information of a combination of a basic pattern and a combined pattern determined by the character cutting position determining means. A document reading device comprising: