JPH0326878B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Technical Field) The present invention relates to a character recognition device, and particularly to a stroke extraction method in a character recognition device for handwritten characters.

(Background technology) As a conventional method for extracting strokes, (1) the curvature is calculated for a series of contour points detected by tracing the contour of a character figure, and the points with large values of curvature are used as dividing points to divide the contour series. Either the strokes are extracted by dividing and combining the divided series, or (2) the character figure pattern is thinned to create a skeleton, and the connectivity of the skeleton pattern and the skeleton pattern are traced and the strokes are extracted. There is a method of extracting strokes by detecting points of change in angle.

However, method (1) loses track of the stroke direction when strokes are concentrated and extracts strokes in the wrong direction.Furthermore, as the character/figure pattern becomes larger and more complex, the amount of processing increases, resulting in processing speed. (2)
In this method, since it is necessary to thin the character/figure pattern, distortion of the pattern, whiskers, etc. occur due to the thinning, and if the strokes are concentrated, the connectivity of the pattern may be incorrect, and the strokes cannot be extracted stably. There were flaws.

(Problems to be solved by the invention) The present invention provides a stroke extraction method that improves the above-mentioned drawbacks of the conventional technology and enables stable and high-speed stroke extraction. The purpose of this method is to delete unnecessary segments, extend only valid segments, combine segments connected at end points, and extract strokes based on the connection point coordinates or intersection coordinates of the segments.

(Structure and operation of the invention) FIG. 1 is a block diagram showing an embodiment of the invention. 1 is a binary pattern input signal, 2 is a pattern register, 3 is a line width calculation section, 4a is a vertical sub-pattern extraction section, 4b is a horizontal sub-pattern extraction section,
4c is a left diagonal sub-pattern extraction section, 4d is a right diagonal sub-pattern extraction section, 5a is a vertical segment extraction section, 5b is a horizontal segment extraction section, 5c is a left diagonal segment extraction section, 5d is a right diagonal segment extraction section, and 6 is a right diagonal segment extraction section. , a primitive segment table memory that stores the primitive segments detected by each segment extracting unit; 7 a connecting point intersection detection unit that detects the connecting point coordinates and intersection coordinates of each of the detected primitive segments; 8 a connecting point coordinate; 9 is a connection point/intersection table memory that stores intersection coordinates; 9 is a valid segment detection unit that detects valid segments using detected connection points and intersection table memory; 10 is a valid segment table memory that stores detected valid segments; , a stroke extractor which combines the detected effective segments and extracts strokes, and 12 a stroke table memory which stores the extracted strokes.

Binary pattern input signal 1 is stored in pattern register 2 . At the same time, the line width calculation section 3 calculates the line width of the input pattern. The vertical sub-pattern extraction section 4a vertically scans the entire pattern register 2, extracts the continuous black bits whose length is sufficiently longer than the line width calculated by the line width calculation section 3, and extracts the continuous black bits. Extract subpatterns (VSP). Similarly, the horizontal sub-pattern extraction section 4b generates a horizontal sub-pattern (HSP) by horizontal scanning, and the left diagonal sub-pattern extraction section 4c generates a left diagonal sub-pattern (LSP) by scanning diagonally 45 degrees to the left.
, the right diagonal sub-pattern extraction section 4d extracts the right diagonal
Extract the right diagonal subpattern (RSP) by 45° scanning.

FIG. 2 is a diagram showing an example of extracted sub-patterns of the letter "L". Figure 2a is the original pattern,
Figure 2b is a horizontal sub-pattern, Figure 2c is a vertical sub-pattern, Figure 2d is a left diagonal sub-pattern,
FIG. 2e shows a right diagonal subpattern, but the right diagonal subpattern is not extracted in this example.

Next, the vertical segment detection unit 5a scans the vertical sub-pattern (VSP) horizontally from left to right, and detects the black point coordinates when the white point (character background part) changes to a black point (character line part), and the black point coordinates when the white point (character background part) changes to a black point (character line part). The coordinate position of the continuous area of the black point is calculated from both coordinates of the white point coordinate when it changes to a point, and the number of points of change is detected. This operation is performed from top to bottom, and the center coordinate series of the continuous area of black points is detected as a vertical primitive segment from the coordinate positions of the continuous area of black points in the previous row and the current line and the number of changing points. Similarly, horizontal segment detection section 5b
scans the horizontal sub-pattern (HSP) from top to bottom to detect the horizontal primitive segment, and the left diagonal segment detection unit 5c scans the horizontal sub-pattern (HSP) from top to bottom, and the left diagonal segment detection unit 5c scans the horizontal sub-pattern (HSP) from top to bottom to detect the horizontal primitive segment.
The right diagonal segment detection section 5d scans the right diagonal subpattern (RSP) from left to right to detect a left diagonal primitive segment. Regarding the detection of this segment, a patent application filed by the applicant in 1983
-31228 (Japanese Unexamined Patent Publication No. 57-146379), the method will be explained using an example of detecting vertical primitive segments by processing the stroke components of a vertical sub-pattern (VSP) into core lines. .

Now, suppose that a pattern of segment components A and B as shown in FIG. 3 is obtained in the vertical sub-pattern (VSP). When starting from the upper left point and scanning from left to right to detect change points, no change points are detected until the line before the Y1 line, and it is only after scanning the Y1 line that the two change points of stroke component A are detected. In other words, a point of change from a white point to a black point X _BY1 and a point of change from a black point to a white point X _WY1 are obtained. At this time, calculate (X _BY1 +X _WY1 )/2,
These coordinates are registered in the primitive segment table memory 6 as the starting point of the primitive segment 1.

Furthermore, the following equation (1), where the two change points obtained by scanning the next row are X _BY1+1 and X _WY1+1 ,
When both equations (2) are satisfied, it is defined that the coordinate position of the continuous region of the currently detected black point overlaps with that of the black point detected in the previous line, and it is assumed that the segment is a continuation of the previous line.

X _BY1 ≦X _WY1+1 ...(1) X _WY1 ≧X _BY1+1 ...( ₂ ) ₍ However, X _{BY1 <} In this case, the center coordinate position of the continuous area of black dots detected in the current row is registered in the primitive segment table memory 6 as belonging to a primitive segment different from the previous row.

In this way, the line before the Y2 line is scanned and the coordinates of the center point of the original segment 1 are registered. When row Y2 is scanned, the number of change points becomes four, indicating that a new segment component B has appeared. By repeating the above-mentioned operations, the coordinate position of the center point, which is registered as a continuation of primitive segment 1, and the starting point of new primitive segment 2, separate from primitive segment 1, are determined and registered in the primitive segment table memory 6. Divided into things. In this way, scan to the previous line of Y3 line, segment component A,
The coordinates of the center point of B are respectively registered in the primitive segment table memory 6. When the Y3 column is scanned, there are two changing points, and it can be seen from the coordinate positions of the continuous area of the black dots that it is a continuation of the primitive segment 2, and that the detection of the primitive segment 1 has been completed.
There are two changing points from now to the previous line of Y4, and they are registered as a continuation of primitive segment 2. When you scan line Y4, there are no change points, and primitive segment 2
This means that the detection of the vertical subpattern (VSP) has been completed, and the segment extraction of the vertical subpattern (VSP) has been completed. A similar function is provided by the horizontal segment detection section 5.
b, the left diagonal segment detector 5c and the right diagonal segment detector 5d, and the detected primitive segments are stored in the primitive segment table memory 6, respectively.

FIG. 4 shows the original segment extracted by converting the subpattern of the letter "L" shown in FIG. 1 into core lines as described above. Figure 4a is
A horizontal primitive segment, FIG. 4b shows a vertical primitive segment, and FIG. 4c shows a left diagonal primitive segment.

The connection point/intersection coordinate detection unit 7 detects connection/intersection relationships for all the primitive segments stored in the primitive segment table 6, and stores the results in the connection point/intersection table memory 8. The chart is shown in Figure 5. First, one primitive segment (referred to as primitive segment 1) is taken out from the primitive segment table memory 6, and the distance between each point of another primitive segment (referred to as primitive segment 2) and each point of the retrieved primitive segment 1 is is calculated, and the coordinates of each point in primitive segment 1 and primitive segment 2 (hereinafter referred to as connection/intersection candidate points) whose distance is less than or equal to C ₁ (C ₁ is a constant and is 1 in this example) are detected, and the connection/intersection candidate points are calculated. When an intersection candidate point exists, primitive segment 1 and primitive segment 2 are considered candidates for connection/intersection. Whether it is a connection or an intersection is determined by the following process.

Normally, if there are connection/intersection candidate points, there are multiple consecutive connection/intersection candidate points, so for each of primitive segment 1 and primitive segment 2, the period in which connection/intersection candidate points are consecutive within the own segment is detected, and the start point and end point between these points are detected. Calculate the distance from each to the end points of the own segment, and when this distance is less than or equal to C ₂ W (C ₂ is a constant and is 1 in this example, and W is the line width), the distance between primitive segment 1 and the primitive segment is calculated. Segment 2 is a connection (the connection/intersection candidate point is determined to be a connection point), and at the same time, for each of primitive segment 1 and primitive segment 2, the midpoint coordinates of consecutive connection points of the segment, and the Information on whether the segment is connected to the start point or the end point is stored in the connection point/intersection table memory 8. Other than this, it is primitive segment 1,
It is determined that primitive segment 2 is an intersection (the connection/intersection candidate point is determined to be an intersection), and at the same time, for each of primitive segment 1 and primitive segment 2, the coordinates of the midpoint of the consecutive intersections of the segments are determined as the connection point.・Intersection table memory 8
Store in.

The valid segment detection unit 9 deletes unnecessary segments by the following two types of processing.

Unnecessary segment deletion processing 1: Unnecessary segments are deleted from a plurality of segments detected in strokes having intermediate angles in four directions.

Unnecessary segment deletion process 2: Unnecessary segments detected when strokes are concentrated or at connection points/intersections are deleted.

An operational flowchart of unnecessary segment deletion processing 1 is shown in FIG. 6, and an operational flowchart of unnecessary segment deletion processing 2 is shown in FIG. 7. Each process will be explained below in order.

The operation of unnecessary segment deletion processing 1 will be explained using FIG. From the connection point/intersection table 8, one valid segment (valid segment 1, H segment in Fig. 8) is extracted from among the segments for which the flag indicating an unnecessary segment is not set (hereinafter referred to as valid segment), and its Valid segment intersecting valid segment 1 (valid segment 2, R segment in Figure 8)
, and draw perpendicular lines from the start point (start point H in Figure 8) and end point (start point H in Figure 8) of the effective segment 1 to the effective segment 2, respectively, and when the perpendicular lines intersect with the effective segment 2, respectively. When the distances are both less than or equal to C ₄ W (C ₄ is a constant and is 2 in this example, and W is the line width), a flag indicating that the valid segment 1 is an unnecessary segment is set. is set in the connection point/intersection table memory 8.

The operation of unnecessary segment deletion processing 2 will be explained using FIG. 9. Connection point/intersection table memory 8
One valid segment (valid segment 1 and V segment in Fig. 9) is extracted from among them, all valid segments that intersect or connect to that valid segment are detected, and each of the detected valid segments is at least Only two valid segments that are connected to or intersect with one other detected valid segment are extracted. In FIG. 9, the _H1 segment and _H2 segment connected to the V segment, and the intersecting L segment are extracted.

Next, a combination in which a triangle is formed by effective segment 1 and the two extracted effective segments is detected (△ABC, △ in Fig. 9).
Points where the other two valid segments connect or intersect in valid segment 1 (where AB'C' is detected) (points A, B, or A in FIG. 9)
(point D or point D' in Figure 9)
Calculate. Furthermore, the coordinates of the point where the two effective segments connect or intersect (point C or point C' in Figure 9) are calculated, and the distance from the calculated midpoint (point C or C' in Figure 9) is calculated. If the distance is less than C ₃ W,
The area in valid segment 1 up to the point where each of the two valid segments connects or intersects is set as a deletion candidate. After performing the above processing for all of the detected triangles, the areas within distance C ₂ W from both end points of effective segment 1 are candidates for deletion. If all valid segments 1 are deletion candidates, valid segments 1 are deleted as unnecessary segments. Other valid segments are registered in the valid segment table memory 10.

The stroke extraction unit 11 sequentially extracts valid segments from the valid segment table memory 10, and when the valid segments can be extended to other valid segments, extends the process of combining them with other valid segments to form one valid segment. This process is performed until there are no more possible valid segments, and the remaining valid segments are stored as strokes in the stroke table memory 12. The operation flow is shown in FIG. 11. The operation will be explained in detail below.

A valid segment whose used flag is not set (referred to as valid segment 1) is retrieved from the valid segment table memory 10, and other valid segments whose used flags are not set are retrieved from the valid segment table memory 10 and are connected to valid segment 1 from end to end. A segment (defined as valid segment 2) is detected, and valid segment 1 and valid segment 2 are combined to form valid segment 1. At the same time, a flag indicating that valid segment 2 has been used is set in valid segment table memory 10. The above processing is executed until there are no valid segments whose end points are connected to valid segment 1 and whose used flags are not set, and valid segment 1 is registered in the stroke table memory 12 as a stroke.

FIG. 10a shows an example in which an effective segment of the hiragana "u" is obtained, and FIG. 10b shows an example in which two strokes are extracted from the obtained target segment.

(Effects of the Invention) As explained above, in the present invention, unnecessary segments detected at connection points and intersections, unnecessary segments detected at strokes having intermediate angles in four directions, Furthermore, unnecessary segments that occur when strokes are concentrated are deleted, remaining valid segments are extended, and strokes are extracted, so the connections and intersecting relationships between strokes become clear.

Furthermore, when extending segments, it does not require complicated calculations such as angles between segments, and instead focuses only on the end points of valid segments, resulting in a stable and high-speed character recognition device that has the advantage of fast processing speed. It can be used for.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIGS. 2 a to e are diagrams showing examples of sub-pattern extraction, FIG. 3 is a diagram explaining a stroke extraction method, and FIG. 4 a to c are diagrams showing segments extracted from the sub-patterns in FIG. 2, FIG. 5 is an operation flowchart in the connection point/intersection detection section 7, and FIGS. 6 and 7 are diagrams showing the segments extracted from the sub-pattern in FIG. The operation flowcharts, Figures 8 and 9 are as follows:
FIGS. 10A and 10B are diagrams explaining the effective segment detection method, FIGS. 10A and 10B are diagrams explaining the stroke extraction method, and FIG. 11 is an operation flowchart of the stroke extraction section 11. 1...Binary pattern input signal, 2...Pattern register, 3...Line width calculation section, 4a...Horizontal sub-pattern extraction section, 4b...Vertical sub-pattern extraction section, 4c...Left diagonal sub-pattern extraction section , 4d...
...Right diagonal sub-pattern extraction section, 5a...Horizontal segment extraction section, 5b...Vertical segment extraction section,
5c...Left diagonal segment extraction unit, 5d...Right diagonal segment extraction unit, 6...Primitive segment table memory, 7...Connection point/intersection detection unit, 8...
...Connection point/intersection table memory, 9... Valid segment detection section, 10... Valid segment table memory, 11... Stroke extraction section, 12...
Stroke table memory.

Claims (1)

[Claims] 1. A process of converting the brightness of cells constituting a character/figure pattern into electrical signals and storing them in a memory, and a sub-pattern that scans the stored pattern in a plurality of predetermined directions to represent stroke components corresponding to each scanning direction. , a process of extracting primitive segments obtained by skeletonizing the stroke components of each sub-pattern, and detecting and removing unnecessary segments using the skeleton coordinate series of the skeleton segments, and creating effective segments other than unnecessary segments. a step of obtaining segments; a step of combining two valid segments when a condition that two valid segments are connected at each end point is met; and a step of repeating the step of combining the two segments until the condition no longer holds true. A method for extracting strokes in character recognition, the method comprising the following steps: 1. A method for extracting strokes in character recognition, comprising the following steps: 1. A method for extracting strokes in character recognition.