JPH0697471B2 - How to cut out contact characters - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a contact character cutting method, and more particularly to a contact character cutting method suitable for reading a contact character written in a non-dropout color entry frame. Regarding

[Prior Art] Most of the conventional character reading methods read characters in an entry frame printed in dropout color.

However, in recent years, there has been a demand for a copying machine that can easily create a form and a technique for reading a form with a black print entry frame output by a laser beam printer.

In this type of character reading technique, it is necessary to automatically distinguish between a character and an entry frame, and also to cut out a character in contact with the entry frame so that the character can be read separately from the entry frame. As a conventional technique of this kind, Japanese Patent Laid-Open No. 60-
There is a technique described in Japanese Patent No. 126777.

[Problems to be Solved by the Invention] In the above-mentioned conventional technique, the character that is in contact with the entry frame is separated by cutting the entry frame and the character at the contact portion.

Therefore, there is a problem that some of the characters existing in the entry frame image are missing, separated as the original normal characters, and cannot be read.

An object of the present invention is to solve the above-mentioned problems of the prior art and to separate a character in contact with the entry frame from the entry frame without missing a part of the characters in the entry frame image, and to provide a non-dropout color. An object of the present invention is to provide a method of cutting out a contact character that can be read with high reliability even if the character is in contact with the entry frame.

[Means for Solving Problems] The purpose is to cut a character image by cutting at the first boundary set inside the outline of the entry frame image, assuming that the character image is not included in the entry frame image. When a plurality of cutting points are generated by taking out and cutting at the first boundary,
Assuming that a character image is included between the plurality of cutting points, the character image is extracted by cutting from the cutting point at a second boundary including a portion outside the contour of the entry image,
This is achieved by selecting which of the first and second boundaries the character image can be read from.

[Operation] In the present invention, when separating the entry frame and the characters displayed on the form, it is first assumed that the character image is not included in the entry frame image, and the first boundary set inside the entry frame image is used. Cut and take out the character image.

Then, when a plurality of cutting points are generated by cutting at the first boundary, it is assumed that a character image is included between the plurality of cutting points. Based on this assumption, the character image is cut out from the cutting point at the second boundary including the portion outside the outline of the entry frame image and taken out.

The character image cut out at the first boundary and the second image
What kind of character is read by the character image cut out at the boundary of.

As a result, the characters that are in contact with the entry frame will be separated from the entry frame without losing some of the characters in the entry frame image, and the characters that are in contact with the non-dropout color printing entry frame will be displayed. However, it is possible to read with high reliability.

[Embodiment] An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a flow chart showing an embodiment of the method of the present invention,
2 (a) to (j) are diagrams showing specific examples of character cutout,
FIG. 3 is a block diagram showing an example of an apparatus for carrying out the method of the present invention.

The apparatus of the embodiment shown in FIG. 3 includes a form transport mechanism 1 for transporting a form 2, a light emitting source 3, a light receiving element 4, a binarization circuit 5, a storage unit 6, and a cutout / recognition unit 7. It has and.

Then, in the apparatus shown in FIG. 3, the form 2 is conveyed by the form conveying mechanism 1 to a reading section including a light emitting source 3 and a light receiving element 4.

The form 2 is irradiated with light by the light emitting source 3, and the light is reflected on the form 2 and enters the light receiving element 4. The light receiving element 4 converts the sequentially incident patterns into electric signals and outputs the electric signals. A one-dimensional semiconductor CCD sensor or the like is used for the light receiving element 4.

The binarization circuit 5 outputs the output signal from the light receiving element 4 to the storage unit 6 as a binary signal of “0” and “1” according to a predetermined threshold value.

The storage unit 6 stores the pattern on the form 2 as a binary image of “0” and “1”, and a semiconductor memory or the like is used for the storage unit 6.

The cutout / recognition unit 7 is composed of a microprocessor. The cutout / recognition unit 7 cuts out a character pattern for each character from the binary image in the storage unit 6, compares the cut out character image with the standard pattern, and outputs the result as a recognition result.

Next, a method of cutting out a contact character by the cutout / recognition unit 7 will be specifically described with reference to FIGS. 1 and 2A to 2J.

When a character pattern including an input frame as shown in FIG. 2 (a) is input, first, in order to separate the input frame and the character pattern, the block of the image is outlined by the contour C in step 100 shown in FIG.
Extract as ₁ , C ₂ , C ₃ .

Next, in step 200 shown in FIG. 1, the extracted contour C ₁ ,
C ₂ and C ₃ are classified into the outline of the input frame and the outline of the character, and the outline of the input frame is selected. Although various classification methods can be considered, in this embodiment, classification is performed according to whether the size of the rectangle circumscribing the contour is larger or smaller than a specified value. In the example of FIG. 2B, C ₂ and C ₃ are selected as entry boxes. At this time, the character contour which is in contact with the outline of the entry frame is included in the outline of the entry frame as shown in FIG. 2 (c).

The character outline C _{1 that} is not in contact with the entry frame outlines C ₂ and C ₃ is extracted and used in combination with another character outline during character recognition.

Subsequently, in step 300 shown in FIG. 1, a boundary between the character contour and the entry frame contour is set in order to separate the character contour within the entry frame contour. As shown in FIG. 2D, it is assumed that a part of the character outline is not included in the entry frame line segment, and the first boundary B _1- setting the B _4. Here, inside the rectangle formed by the first boundaries B _{1 to} B ₄ , the entry frame contours C ₂ and C ₃ and the first boundaries B _{1 to} B 4
The crossing points of ₄ are defined as cutting points P ₁₀ and P ₁₁ . The cutting direction at the cutting points P ₁₀ and P ₁₁ is the same as the boundary B ₁ .

When the number of cutting points corresponding to each of the first boundaries B _{1 to} B ₄ is plural, it is assumed that a part of the character contour is included between the cutting points. Then, based on this assumption, a second boundary including a portion outside the entry frame contour is set from the plurality of cutting points. In the case of this example,
As shown in FIG. 2 (e), it is assumed that a part of the character contour is included between the cutting points P ₁₀ and P _{11 with} respect to the boundary B ₁ , and the character is lowered at a right angle from the cutting points P ₁₀ and P _11. The second boundaries B ′ _{1 to} B ₄ that step on a part of the outline C ₂ of the entry frame are set. If the outline of the character is not in contact with the outline of the entry frame, the cutting point is not detected even if this process is executed, so there is no problem.

Next, in step 400 shown in FIG. 1, the first boundaries B _{1 to} B _{4 are}
And the second boundaries B ′ _{1 to} B ₄ separate the character contour from the entry frame contour. The character boundaries C ₄ and C _{5 of} FIG. 2 (f) are separated by the first boundaries B _{1 to} B ₄ , and the second boundaries B ′ ₁
Character outline C ₆ of FIG. 2 (g) is separated by .about.B _4.

Then, in step 500 shown in FIG. 1, character recognition is performed. Character recognition is performed using a combination pattern of character contours C ₄ , C ₅ and C ₆ separated from the previously extracted character contour C ₁ and entry frame contours C ₂ and C ₃ . That is, in the case of character outline separated in the first boundary B ₁ .about.B ₄ as shown in FIG. 2 (h), a combination of character outline C _1, C _4, C _5, the character outline
Character recognition is performed using the combined character pattern of C ₁ , C ₄ , and C ₅ , and the recognition result (unread) is output. On the other hand, in the case of character outlines separated by the second boundaries B ′ _{1 to} B ₄ ,
As shown in FIG. (I), a combination of character outline C _1, C _6, the character outline C _1, character recognition in combination with the character pattern of C ₆ is performed, the recognition result "5" is output .

After the character recognition is completed, in step 600, the character contours separated by the most appropriate boundaries from the recognition results shown in FIGS. 2 (h) and (i) are selected. That is, in this embodiment, the character contours separated by the second boundaries B ′ _{1 to} B ₄ which are to be correctly read are selected and processed as the correctly read character “5” as shown in FIG. 2 (j). . If there are no boundaries that deserve to be read correctly, or if there are multiple boundaries that deserve to be read correctly,
Treat as unreadable characters.

In this embodiment, the boundary is selected under the condition of whether the recognition result is correct reading or non-reading, but it goes without saying that the judgment may be made based on other recognition result information such as similarity.

Further, the separation and recognition process may be interrupted when it is determined from the previously recognized result that the separation and recognition of the remaining boundary need not be executed. Thereby, the processing time can be shortened.

As described above, according to this embodiment, even if a part of the written character comes into contact with the entry frame and a part of the character outline is buried in the outline of the entry frame, the entry frame is non-dropout color printing. Also, the character contour is separated from the entry frame contour by the first and second boundaries to be recognized, and the first and second boundaries are recognized based on the recognition result information.
It is possible to accurately determine which of the boundaries of the character can be used to read the character with the separated character outline.

Note that the present invention is not limited to the method of separating the character image from the entry frame image based on the entry frame contour and the character contour, in short, the character image is separated from the entry frame image at the first and second boundaries,
Any method can be used as long as it is possible to select which of the boundaries the character image can be read from.

[Effect of the Invention] According to the present invention described above, assuming that a character image is not included in the entry frame image, the character image is cut and taken out at the first boundary set inside the outline of the entry frame image. When a plurality of cutting points are generated by cutting at the first boundary, it is assumed that a character image is included between the plurality of cutting points, and a portion outside the outline of the entry frame image is cut from the cutting points. It is cut out at the second boundary that includes it and is taken out, and which of the first and second boundaries separates the character image from which the character can be read is selected. Characters touching the entry frame can be separated from the entry frame without missing a part of the character,
Therefore, even a character in contact with the non-dropout color entry frame can be read with high reliability.

[Brief description of drawings]

FIG. 1 is a flow chart showing an embodiment of the method of the present invention,
2 (a) to (j) are diagrams showing a specific example of character extraction, and FIG. 3 is a block diagram showing an example of an apparatus for carrying out the method of the present invention. 1 ... Form transport mechanism, 2 ... Form, 3 ... Light emitting source, 4 ...
… Light receiving element, 5 …… Binarization circuit, 6 …… Storage unit, 7 ……
Cutout / recognition unit, C ₁ , C ₂ , C ₃ …… Entry frame and character outline, B _{1 to} B ₄ …… First boundary, B ′ _{1 to} B ₄ …… Second boundary, P ₁₀ , P ₁₁ …… Cutting point, C ₁ , C ₄ , C ₅ , C ₆ …… Separated character outline.

Claims (1)

[Claims] 1. A method of cutting out a contact character from a binary image including a writing frame and characters displayed on a form by separating the writing frame image and the character image, the character image being included in the writing frame image. If not, cut at the first boundary set inside the outline of the entry frame image, take out the character image,
And, when a plurality of cutting points are generated by cutting at the first boundary, it is assumed that a character image is included between the plurality of cutting points, and the outside of the outline of the entry frame image is cut from the cutting points. A contact characterized in that a character image is taken out by cutting at a second boundary including a part, and which of the first and second boundaries separates the character image to read the character. How to cut out characters.