JP3171626B2 - Character recognition processing area / processing condition specification method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION This invention is, read the image of the original document
In particular, a method of designating a document processing area and its processing conditions for such a system.
About the law .

[0002]

2. Description of the Related Art Conventionally, when a processing area of a document is designated,
Using a marker (such as a color felt-tip pen), a line surrounding the processing area is directly written on the original and recognized (for example, JP-A-2-253377), or using a tablet or displaying the original image. (For example, Japanese Patent Application Laid-Open No. 63-86015) in which a region is designated by coordinates on a display screen using a mouse or the like. To specify what kind of processing (processing condition or content) is to be performed on a specified area, a method is generally used in which a key for processing selection is used after the area is specified.

[0003]

However, the method of separately designating the area and designating the processing conditions or contents for the area is not only troublesome in the designation operation, but also requires processing of a large number of originals for each original. Since the processing area and the processing conditions are usually different, the process of interrupting the processing each time one sheet is processed, resuming the processing by designating the area and the processing conditions is repeated, and the continuous processing (batch processing) is performed. There is a big problem that you can not.

An object of the present invention is to provide a character recognition processing system.
In a simple operation, the character recognition processing area and its processing conditions
The present invention provides a method for specifying a processing area and a processing condition which can specify a processing area and which is not inconvenient for batch processing of a large number of documents.

[0005]

The gist of the present invention is as follows. When specifying the character recognition processing area and processing conditions of the manuscript, use a marker such as a color felt-tip pen,
A line figure surrounding the processing target area is directly written on the document. At this time, a figure having a shape characteristic defined in advance corresponding to the processing condition is selected as the line figure to be written. In the character recognition processing system, the original is read, and from the read original image, a line figure written on the original with a marker is recognized,
The range of the line figure is determined as the character recognition processing area, and the processing conditions (character direction, line direction) are determined from the shape characteristics of the line figure. Here, the shape features of the line figure include the sides of the line figure.
Add an arrow to the line segment to be formed, the number, continuity, and length of the line segment
Use the differences.

[0006]

As described above, by directly writing a line figure having a shape characteristic corresponding to the character recognition processing condition on the document, the character recognition processing area and the processing condition can be specified at once, and the specification work is extremely simple. It is.

Further, when a large number of originals are to be subjected to character recognition processing, only by writing necessary line figures on all the originals in advance, the character recognition processing system determines a processing area and processing conditions for each original. By successively performing the necessary processing, continuous character recognition processing (batch processing) for a large number of documents is possible.

[0008]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows an optical character according to each embodiment described later.
1 shows a schematic configuration of a recognition processing system. The outline of this system is as follows. An original is read by a scanner 1 and an image thereof is input to an area / condition determining device 2. And processing conditions. The determined contents and the original image are sent to the image processing device 3, where processing is performed on the processing area according to the determined processing conditions, and the processed image is output by the output device 4.

More specifically, the image processing device 3
In other words, the recognition process is performed on the specified area under the specified conditions, and the recognition result is printed out by the output device 4, displayed on a screen, or stored in a storage medium.

FIG. 2 is a schematic block diagram of the area / condition determining device 2. The original image input from the scanner 1 is sent to the image processing device 3 and the image
Buffer 11. The marker line segment detection unit 12 includes:
The original image is read from the image buffer 11, the line segment of the line figure directly written on the original is extracted by the marker, and the line figure of the marker is recognized by the line figure recognition unit 13 based on the information of the extracted line segment. The condition determining unit 14 compares the recognized shape feature of the line figure with the shape feature of the line figure defined in advance for each processing condition, and notifies the image processing apparatus 3 of the processing condition whose shape feature matches. Also, the area determination unit 1
5 detects a rectangle of the upper left corner and lower right corner coordinates which circumscribes the recognized line drawing, image processing apparatus 3 as the area information
Notify to

The processing for recognizing a line figure from a binary image can be easily realized by a known technique. In addition, techniques such as marker line segment detection
For example, it is described in detail in JP-A-2-25377. Hereinafter, each embodiment will be specifically described.

Embodiment 1 In this embodiment, in an optical character recognition system, a rectangle having one side as an arrow is used as a line figure for specifying an area, and character recognition is performed according to the direction of the arrow on one side of the line figure.
Specifies the character direction, which is one of the processing conditions for knowledge .

That is, as shown in FIG.
4b is defined corresponding to the character direction. When the character direction is specified as upward, the figure 51a or 51b is selected, and this is written on the manuscript with the marker. The same applies to other character directions. The system condition determination unit 14
The direction of the arrow of the line figure on the document can be used as a shape feature for determining the character direction.

For example, in the case of the original 56 shown in FIG. 4 , the line graphic 57 written by the marker is the graphic 53a or 5 in FIG.
It has the same shape characteristics as 3b. Also, a line figure 58
Is figure 51a or 51b and the shape characteristics of FIG. 3 are the same.

In the case of the original 56 , in the system,
The area / condition determination device 2 recognizes the line figures 57 and 58 on the document, determines the range of each circumscribed rectangle as a processing area, determines the character direction of the area of the line figure 57 to the left, and The character direction of the area 58 is determined to be upward.
Then, the image processing apparatus 3 performs recognition processing on the area of the line graphic 57 under the condition that the character direction is leftward, and recognizes that the character direction is upward on the area of the line graphic 58. Perform

Note that, instead of the arrow, another symbol may be written at the tip of the side line segment. This is the implementation
The same applies to Examples 2 and 3 .

Embodiment 2 In this embodiment, in an optical character recognition system , a rectangle having one side as an arrow is used as a line figure for designating an area, and character recognition is performed according to the direction of the arrow on one side of this line figure.
Specifies the line direction, which is one of the processing conditions of the knowledge .

FIG . 5 shows the shape of a line figure defined corresponding to the row direction. When the row direction is designated as the vertical direction , the figure 61a or 61b is selected, and this is written on the manuscript with the marker. The same applies when the row direction is horizontal.
It is like. The condition determining unit 14 of the system can use the direction of the arrow of the line graphic as a shape feature for determining the row direction.

For example, in the case of the original 65 shown in FIG. 6 , the line graphic 66 written with the marker is the graphic 61a or 6a in FIG.
It has the same shape characteristics as 1b. Also, the line graphic 67
Is the same shape feature a figure 62a or 62b in FIG. 5 are the same.

In the case of the original 65 , the system
The area / condition determination device 2 recognizes the line figures 66 and 67 on the document, determines the range of each circumscribed rectangle as a processing area, determines the row direction of the area of the line figure 66 as the vertical direction , The row direction of the region of FIG. 67 is determined as the horizontal direction .
Then, the image processing device 3, the recognition for a region of the line drawing 66, as a recognition condition row direction performs a recognition process as a vertical direction, also with respect to areas of the line drawing 67 the row direction is a horizontal direction Perform processing.

Embodiment 3 This embodiment is directed to an optical character recognition system in which the upper side or right side of a rectangle to be written on a document to specify an area is designated.
The character direction is specified by the direction of the single arrow (first arrow) attached to the line segment (first line segment), and the double arrow ( second line segment) is added to the lower or left side line segment (second line segment) of the rectangle . (2 arrow) to specify the row direction. Specifically, a line figure obtained by combining the figures 71 to 74 and the figures 75 and 76 shown in FIG. 7 is used as a processing condition for character recognition.

The condition determining unit 14 of the system can use the direction of the arrow on the upper side or right side of the line figure on the document as a shape feature for determining the character direction, and can use the direction of the arrow on the lower side or left side of the line figure. Presence or absence can be used as a shape feature for determining the row direction.

[0024] For example, in the case of the document 77 shown in FIG. 8, line figure 78 filled with a marker has the shape features of shape features and figure 76 figure 73 in FIG. Also, the line graphic 79
Has the shape features of both figures 71 and 75 in FIG.

In the case of the original 77, the system
The area / condition determination device 2 recognizes the line figures 78 and 79 on the document, determines the range of each circumscribed rectangle as a processing area, and sets the line direction of the line figure 78 to the vertical direction and the character direction to the left. It determines that the row direction in the horizontal direction of the area of the line drawing 79, determines the character direction and upward. Then, the image processing device 3 performs a character recognition process on each of the designated areas under the determined recognition conditions.

Embodiment 4 In this embodiment, figures 81 and 82 shown in FIG. 9 are used as line figures in an optical character recognition system. In this line figure, two sides corresponding to the row direction are two lines. The area / condition determination device 2 of the system determines the number of lines on the side of the line figure.
It can be used as a shape feature for determining processing conditions for character recognition .

In the case of the document 83 shown in FIG. 10, for example, a line drawing 84 filled with a marker has a shape characteristic of figure 81 in FIG. 9, line drawing 85 has a shape characteristic of figure 82 in FIG. Therefore, in the system, the line figure 84
For the area of, the line direction is recognized as the vertical direction ,
For the region of the line graphic 85, the recognition process is performed with the row direction as the horizontal direction .

Embodiment 5 In this embodiment, figures 91 and 92 shown in FIG. 11 are used as line figures in an optical character recognition system. In this line figure, two sides coinciding with the row direction are dotted lines (broken lines). The area / condition determination device 2 of the system can use the continuity of the sides of the line graphic as a shape feature for determining the processing conditions for character recognition .

In the case of the document 93 shown in FIG. 12, for example, a line drawing 94 filled with a marker has a shape characteristic of figure 91 in FIG. 11, a line drawing 95 has a shape characteristic of figure 92 in Figure 11. Accordingly, in the system, recognition processing is performed on the area of the line graphic 94 with the row direction as the vertical direction , and on the area of the line graphic 95, the recognition processing is performed with the row direction as the horizontal direction .

Embodiment 6 In this embodiment, an optical character recognition system is used to specify a processing area on a manuscript as shown in FIG.
A line figure such as 1,102 is used. This line figure specifies the line direction, which is one of the processing conditions for character recognition, by reducing the length of two sides that do not match the line direction to about half. In the area / condition determining unit 2 of the system, the position of the short side of the line figure on the document can be used as a shape feature for determining the processing conditions for character recognition .

For example, in the case of the original 103 shown in FIG. 14 , the line figure 104 written with the marker is the figure 101 shown in FIG.
The line figure 105 is the figure 102 in FIG.
It has the shape characteristics. Therefore, in the system, the recognition process is performed on the region of the line graphic 104 with the row direction as the vertical direction, and the recognition process is performed on the region of the line graphic 105 with the row direction as the horizontal direction .

Embodiment 7 In this embodiment, in the optical character recognition system , a graphic 11 shown in FIG.
A line figure such as 1,112 is used. This line figure is subjected to character recognition depending on whether the side is a single line or a double line.
Specify the character type (Kanji / English) that is one of the logical conditions .
Here, “kanji” means a general Japanese sentence, and “English character” means an English sentence. When three or more character types are specified, 3
A figure using multiple lines such as a multiple line and a multiple line may be additionally defined. In the region-condition determining unit 2 of the system, sentence number of lines constituting each side of the line figure on a document
It can be used as a shape feature for determining processing conditions for character recognition .

For example, in the case of the original 113 shown in FIG. 16 , the line figure 114 written with the marker is replaced with the figure 111 in FIG.
15 , and the line figure 115 is the figure 112 in FIG.
It has the shape characteristics. Therefore, in the system, the character type is recognized as alphabetic for the region of the line graphic 114, and the character type is recognized as kanji for the region of the line graphic 115.

In this embodiment, the processing condition to be specified is a character type. However, it is also possible to specify a character direction or a line direction depending on the difference in the number of lines on the same side.

Embodiment 8 In this embodiment, in the optical character recognition system, the character type is a kanji (a general Japanese sentence) or an English character (an English sentence) as a line figure used for designating a processing area on a document. Is used, a solid rectangle such as the graphic 121 shown in FIG. 17 or a dotted rectangle (the broken line) such as the graphic 122 is used. To specify three or more character types, 1
A figure using a line type having different continuity, such as a chain line or a two-point chain line, may be defined. In the area / condition determination unit 2 of the system, a difference in the continuity (line type) of lines constituting each side of the line figure on the document can be used as a shape feature for determining processing conditions for character recognition. .

For example, in the case of the original 123 shown in FIG. 18 , the line graphic 124 written with the marker is replaced with the graphic 122 shown in FIG.
The line figure 125 is a figure 121 in FIG.
It has the shape characteristics. Accordingly, in the system, the character type is recognized as an alphabet for the area of the line graphic 124, and the character type is recognized as a kanji for the area of the line graphic 125.

In this embodiment, the character type is specified as the processing condition. However, it is also possible to specify the character direction and the line direction by the similar continuity of the line.

[0038]

As described in detail above, according to the present invention, a line image having a shape characteristic corresponding to a processing condition is directly written on an original, so that an original image is read and a character is read.
The processing area and processing conditions can be specified at a time for the character recognition system that performs the recognition processing, and the specification work becomes extremely simple . In particular, in a character recognition system, it is necessary to input many parameters for designating processing conditions for each area, so that the designation operation can be greatly simplified, and a line figure required in advance for a plurality of originals can be formed. By simply writing, continuous processing (batch processing) of a plurality of documents becomes possible.

[Brief description of the drawings]

FIG. 1 is a diagram showing a schematic configuration of a character recognition processing system according to each embodiment of the present invention.

FIG. 2 is a block diagram illustrating a configuration example of a region / condition determination device.

FIG. 3 is an explanatory diagram of a line figure used in the first embodiment.
is there.

FIG. 4 shows an example of writing a line figure on an original in the first embodiment .
FIG.

FIG. 5 is an explanatory diagram of a line figure used in a second embodiment.
is there.

FIG. 6 shows an example of writing a line figure on an original in the second embodiment .
FIG.

FIG. 7 is an explanatory diagram of a line figure used in a third embodiment.
is there.

FIG. 8 shows an example of writing a line figure on a document in a third embodiment .
FIG.

FIG. 9 is an explanatory diagram of a line figure used in a fourth embodiment.
is there.

FIG. 10 is a diagram illustrating an example of writing a line figure on a document according to the fourth embodiment .
FIG.

FIG. 11 is an explanatory diagram of a line figure used in a fifth embodiment.
It is.

FIG. 12 is a diagram illustrating an example of writing a line figure on a document in a fifth embodiment .
FIG.

FIG. 13 is an explanatory diagram of a line figure used in a sixth embodiment.
It is.

FIG. 14 is a diagram illustrating an example of writing a line figure on a document according to the sixth embodiment .
FIG.

FIG. 15 is an explanatory diagram of a line figure used in a seventh embodiment.
It is.

FIG. 16 shows an example of writing a line figure on a document in the seventh embodiment .
FIG.

FIG. 17 is an explanatory diagram of a line figure used in the eighth embodiment.
It is.

FIG. 18 is a diagram illustrating an example of writing a line figure on a document according to the eighth embodiment .
FIG.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 scanner 2 region / condition determining device 3 image processing device 4 output device 11 image buffer 12 marker line segment detecting unit 13 line figure recognizing unit 14 condition determining unit 15 region determining unit

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) G06K 9/00-9/82 H04N 1/387 G06T 11/60

Claims (5)

(57) [Claims] 1. A method for designating a character recognition processing area and processing condition in a system for reading an image of a document and performing character recognition processing, wherein the character recognition processing area on the document is surrounded by a line figure, and the processing condition is defined.
A document indicated by adding an arrow to a line segment constituting a side of the line figure is read, a line figure written on the document is recognized, and an area surrounded by the line figure is determined as a character recognition processing area. A character recognition processing region / processing condition specification method, wherein the character direction or the line direction of the character recognition processing is determined based on the direction of an arrow of a line segment constituting a side of the line figure . 2. A method for designating a character recognition processing area and processing condition in a system for reading an image of a document and performing character recognition processing, wherein the character recognition processing area on the document is surrounded by a line figure, and the processing condition is defined.
A first arrow is placed on a first line segment forming a side of the line graphic,
Read the original indicated by adding a second arrow to line 2
Recognize the line figure entered in the manuscript, and
Is determined as a character recognition processing area, and
Depending on the direction of the first arrow of the first line segment forming the side of the shape
The character direction of the character recognition process is determined, and the second
A sentence characterized by determining the line direction according to the direction of the arrow
How to specify the processing area and processing conditions for character recognition. 3. A method for specifying a character recognition processing area and processing condition in a system for reading an image of a document and performing character recognition processing, wherein the character recognition processing area on the document is surrounded by a line figure, and the processing condition is defined.
Original indicated by the difference in the number of line segments constituting the sides of the line figure
Is read, and the line figure written on the original is recognized.
To determine the area surrounded by the figure as the character recognition processing area
In the first place, the character is determined by the direction of the line segment indicating the specific number of the line figure.
Specially, it determines the character direction or line direction of the recognition process.
How to specify the character recognition processing area and processing conditions. 4. A method for specifying a character recognition processing area and processing condition in a system for reading an image of a document and performing character recognition processing, wherein the character recognition processing area on the document is surrounded by a line figure, and the processing condition is defined.
The original represented by the difference in the continuity of the line segments constituting the sides of the line figure
Read the manuscript, recognize the line figure written on the manuscript,
If the area surrounded by the line figure is determined as the character recognition processing area
Both the direction of the line segment indicating a certain continuity of 該線diagram form
Character direction or line direction for character recognition
Character recognition processing area and processing condition specification method
Law. 5. An image of a document is read and subjected to character recognition processing.
Of processing area and processing conditions for character recognition in different systems
A method for enclosing a character recognition processing area on a document with a line figure and setting processing conditions
Original indicated by the difference in the length of the line segment constituting the side of the line figure
Is read, and the line figure written on the original is recognized.
Character recognition processing according to the direction of the line segment indicating the specific length of the figure
Character direction or line direction is determined
How to specify the processing area and processing conditions for character recognition.