JPH0782522B2 - Document reader - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a document reading device for recognizing and reading characters from a document containing characters, and more particularly to a mechanism for registering a document reading format.

[Conventional technology]

As a conventional document reading device of this type, a so-called optical character reading device is shown in FIG. FIG. 6 is a block diagram showing a conventional document reading device. As shown in the figure, this device scans a document, that is, characters written on a sheet 1 by a scanning means 2 and photoelectrically converts them, and stores an image for one line, that is, image data in a row storage means 3. Further, a character pattern for one character is cut out from the image data for one line by the character cutting means 4,
A feature is extracted from the cut-out character pattern, the extracted feature value is compared with the feature value of the recognition reference pattern stored in the character recognizing means 5 to be recognized, and a character code is output as a recognition result. Was. In the case of performing the above processing, the information necessary for reading indicating the position to be cut out and the character type to be recognized, that is, the reading format is input from the input unit 6 and stored in the reading format table of the control unit 7. Processing was performed based on this. That is, the reading format required for recognition indicating the position of the characters to be written on the paper 1 and the character type to be recognized is set in advance in correspondence with the document, and the characters are cut out according to this reading format for recognition. It was

[Problems to be solved by the invention]

By the way, with the spread of this type of device, the applications for reading a document immediately at the time of creating a document have increased, and the types of documents have become diversified. Therefore, in extreme cases, the reading format of each document has become different. . However,
According to the above-mentioned conventional apparatus that specifies the reading area and the character type to be recognized in advance, it takes a huge amount of time to input each reading format corresponding to diversifying documents, and it is possible to cope with diversifying documents. It had a problem that it could not be done.

The present invention has been made to solve the above problems, and an object of the present invention is to provide a document reading device that can easily input a reading format on the spot and can cope with diversified documents.

[Means for solving problems]

A document reading apparatus according to the present invention includes storage means for storing an image of an input document, display means for displaying the image stored in the storage means, and characters to be recognized in the image displayed by the display means. Input means for selecting an area and an attribute such as a character type of a character displayed in the character area, storage means for storing the character area and the attribute selected by the input means, and the storage means Character cutout means for cutting out the stored character in the selected character area, and recognition means for recognizing the character cut out by the character cutout means based on the selected attribute stored in the storage means. It is equipped with and.

[Action]

In the present invention, the image of the entire document is temporarily stored in the storage means, and the content is displayed on the display means. The character area to be recognized in the image displayed by the display means and the attribute such as the character type of the character displayed in this character area are selected by the input means, and the storage means stores the selected character area and attribute. Then, the character cut-out means cuts out the character in the stored selected character area, and the recognition means recognizes the character based on the stored selected attribute.

〔Example〕

 Hereinafter, embodiments of the present invention will be illustrated and described.

FIG. 1 is a block diagram showing an embodiment of the document reading apparatus of the present invention. The same reference numerals as those in the conventional example of FIG. 6 indicate the same components, and the description thereof will be omitted.
In the figure, 8 is an image of the entire document, that is, page storage means for storing image data for one page, 9 is display means for displaying all image data stored in the page storage means 8, and 10 is the display means. This is an input means for instructing and inputting the information necessary for recognizing the position of the reading area and the characters in the reading area, that is, the reading format while looking at the display screen of 1. On the other hand, 11 is control means for storing the information input by the input means in the read format table and controlling the read format of the document, and 12 is first controlled by the control means from the image data stored in the page storage means. It is a character cutting means for cutting out a line and cutting out a character pattern from the line.

Next, the schematic operation of the present embodiment will be described with reference to the same drawing, and the main parts thereof will be described in detail with reference to FIGS.
As shown in FIG. 1, the characters written on the sheet 1 are scanned by the scanning means 2 and photoelectrically converted, and the image data for one page is stored in the page storage means 8. Then, the image for one page is displayed on the display means 9. Here, the operator observes the display screen of the displayed document image, and inputs an input character area on the display screen from the input means 10 which can perform cursor scanning with the keyboard. Furthermore, the input means
The keyboard 10 is scanned to input the numbers corresponding to the character types to be recognized, and these are sent to the control means 11. The entered character area and the character set to be recognized are stored in the reading format table in the control means 11. The character cutting means 12 automatically cuts out a line in which characters are written in the written character area, and further cuts out a character pattern one by one from the line and sends the character recognition means 5. The character recognition means 5 extracts a feature using the cut-out character pattern, compares the extracted feature value with the feature value of the recognition reference pattern stored in the character recognition means 5, and recognizes the feature value. As a result, the character code is output.

FIG. 2 is an example in which an image of a document is displayed on the screen of the display means 9. Further, FIG. 3 is an example showing a rectangular area cut out by specifying an area. First, the above document example includes the first character type including kanji, hiragana, katakana, symbols, and numbers.
Can be divided into three areas, that is, a second area 14 including the figure, and a third area 15 including the character types of alphanumeric characters and symbols. The cursor movement key of the input means 10 is used to indicate the area of the image of the document at the cutout indication point corresponding to the displayed cursor position. That is, the first region 13 is the first
The first rectangular area 18 shown in FIG. 3 is designated by instructing with the cutout instruction point 16 and the second cutout instruction point 17.
In addition, the second rectangular area 21 is designated in the second area 14 by instructing with the third cutout instruction point 19 and the fourth cutout instruction point 20. Similarly, the third region 15 has the fifth cutout point 22.
And the sixth cut-out instruction point 23 to specify the third rectangular area 24.

FIG. 4 is a block diagram showing the contents of the read format table stored in the control means 11. The value "1" indicating that the recognition target character type of the first rectangular area 18 is kanji, hiragana, katakana, symbol, or numeral is in the character type column 25, and the second rectangular area 21 is a figure area. The value “0” indicating that there is no recognition target character type is also in the character type column 26, and the value “2” indicating that the recognition target character type of the third rectangular area 24 is alphanumeric or symbol is the same. The character type is stored in the column 27. In addition, in correspondence with the position of the document image in the first rectangular area 18,
Coordinate values 1, 1 and 20, 16 in the X and Y directions of the first and second cutout instruction points 16 and 17 are stored in the cutout area columns 28 and 29. Similarly, the coordinate values in the X and Y directions of the third cut-out designating point to the sixth cut-out designating point 19, 20, 22, 23 are stored in the cut-out area columns 30, 31, 32, 33, respectively.

FIG. 5 is a diagram showing an example of a method of cutting out a line from the document image data in the first rectangular area 18 and further cutting out a character pattern in the line. This is a method of cutting out lines and character patterns by using the marginal distribution of the black pattern of the character part which is already widely used. First, as shown in FIG. Direction to obtain a black pattern histogram 33, and a first portion 34 in which the histogram value is 1 or more and which is continuous is cut out as a first row 35,
The second portion 36 is cut out as the second row 37. Further, as shown in FIG. 6B, the first row 35 is scanned in the column direction to obtain a histogram 38 of a black pattern, and a continuous portion where the value of this histogram is 1 or more is cut out, The eleventh character portions 39 to 49 are cut out from the first character portion.

The character pattern cut out as described above is recognized by the character recognition means 5 with the character type designated by the reading format shown in FIG. 4 as the recognition target. The second rectangular area 21 is a drawing area, and character cutting and character recognition are not performed.

Therefore, according to the present embodiment, the operator immediately inputs and inputs the cutout area and the recognition target character type while displaying the image of the read document, and the characters in the specified area are automatically cut out and recognized. As a result, there is an advantage that the working time for creating the reading format can be significantly shortened and can be easily input, and even a document having a different reading format can be efficiently input. Further, since the read document image is observed on the screen and the format is input, there is an advantage that the cutout area can be accurately set.

In the above description, as the recognition information registered in the reading format table, the recognition target character type and the position of the cutout area have been described, but the present invention is not limited to this, and whether the recognition target character is a handwritten character or a printed character is recognized. The information shown may be registered in the read format table for use. Although the keyboard has been described as the input means, the input means is not limited to this, and any input means capable of inputting point coordinates and numerical values may be used.

〔The invention's effect〕

As described above, the document reading device according to the present invention is
Since the character area of the image displayed by the display means is selected and the character cutting means cuts out the character from the selected character area,
The recognition process can be performed at high speed by selecting the recognition area in advance. Further, by selecting an attribute such as a character type of a character displayed in the character area, highly accurate recognition processing can be performed.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of a document reading apparatus according to the present invention, FIG. 2 is a view showing an example of a document image display screen, and FIG. 3 is a rectangular area cut out by the above embodiment. FIG. 4, FIG. 4 is a configuration diagram of a reading format table, FIG. 5 is a diagram showing an example of a character cutting method, and FIG.
FIG. 1 is a block diagram showing a conventional example. 1 ... document, 8 ... storage means, 9 ... display means, 10 ...
Input means. The same reference numerals are used for the same or corresponding parts in the drawings.

Claims (1)

[Claims] 1. A storage means for storing an image of an input document, a display means for displaying the image stored in the storage means, a character area to be recognized by the image displayed by the display means, and this character. Input means for selecting an attribute such as a character type of a character displayed in the area, storage means for storing the character area and the attribute selected by the input means, and storage means for storing the character area And a recognition unit for recognizing the character cut out by the character cutting unit based on the selected attribute stored in the storage unit. A document reading device characterized by the above.