JP2000148735A - Image processing device - Google Patents

Abstract

(57) [Summary] [PROBLEMS] To suppress deterioration of the quality of small characters when copying an original image. A document image is read by a scanner, and a recognition unit detects a character size and a character position and performs character recognition. The CPU 3 reads a font from the dictionary 6 according to the character recognized by the recognition unit 7 and generates an image based on the character size and character position detected by the recognition unit 7 and the copy magnification set by the MMI 1.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for recognizing characters in a read image and generating a reproduced image based on the recognized characters.

[0002]

2. Description of the Related Art In recent years, the character recognition technology has been improved at a high speed of CPUs and semiconductors.
Near real-time processing is possible. In recent years, since most copied originals are created by a word processor or the like, the recognition accuracy has been improved, and has reached a practical level.

[0003] Regarding the recognition of the character size in a document and the technology of recognizing characters, the present applicant has disclosed, for example, Japanese Patent Laid-Open No. 61-107876.
No. has already been disclosed.

[0004]

By the way, in this type of copying machine, in order to treat and process an image as dots separated at a fixed resolution, when a document once copied is copied again as a manuscript, especially when the document is reduced in size, In principle, the quality of reproduced images is inevitably reduced. In view of the above conventional example, an object of the present invention is to provide a copying apparatus that recognizes a character from an image, develops the character into a dot font according to the recognition result, and reproduces a copied image. In particular, an object of the present invention is to generate a high-quality scaled reproduction image.

[0005]

In order to solve the above-mentioned problems, the present invention provides a reading means for reading an image of a document, a character recognizing means for recognizing characters in the image read by the reading means, and a character font. Storage means for storing; read means for reading a character font from the storage means in accordance with the recognition result of the character recognition means; detection means for detecting information about characters in an image read by the read means; And a generating means for generating a reproduced image based on the information on the character font read by the method and the character detected by the detecting means.

[0006]

FIG. 1 is a block diagram of an image processing apparatus according to an embodiment of the present invention. The CPU 3 controls the scanner 4 via the CPU bus 9 to read the image of the original, and stores the image data read by the scanner 4 in the image buffer 8 via the image data bus 10. The recognition unit 7 detects the size and position of the character in the document from the image data stored in the image buffer 8, performs character recognition in the image data with reference to the dictionary 6, and stores the character recognition result in the image buffer 8. Size information and position information are temporarily stored.

After determining the character size and font type to be used for the reproduced image in accordance with the instruction input from the MMI unit 1, the character code which has already been stored is sequentially developed into dot font information from the stored character code, and the image buffer 8 is again stored. Is stored as recording dot information. Thereafter, the dot information is supplied to the printer unit 5 to perform recording. As described above, it is possible to record and reproduce a small character which cannot be obtained by the conventional digital copying machine as a clear font character.

[0008] A method of creating reproduced image data according to the present invention will be described in detail with reference to the flow chart of FIG. The control based on this flow is performed by the CPU 3 based on a program stored in the ROM 11 or the RAM 12. The program stored in the RAM 12 may be read from a hard disk (not shown),
It may be read from a CD-ROM or floppy disk. Further, the functions of the blocks other than the scanner 4 and the printer 5 in FIG. 1 may be executed by a computer. At this time, you can read the program from a CD-ROM or floppy disk, or download the program from a network.

Now, when reading the width of the short side of 200 mm of an A4 document at a resolution of 600 DPI, the total number of pixels read in the buffer (step 100) is 4724 pixels, and there are 40 characters on this document. Is detected from a histogram or the like on the image. In this case, the width of one character is 4724/40 =
This is 118.1 pixels (step 101). This character width is defined as the width of the actual character plus space.

The character size within each character width is also detected by a known method such as a histogram (step 102). If the 12 points are 100 dots, 118.1-100 = 18.1 pixels correspond to the characters of the document and the space between the characters. This is the spacing (character spacing). Based on these information, the whole area of the document is recognized as characters and the code data is stored in the image buffer (step 103).

Assuming that the setting of the copying magnification of the operator is set to 70% (step 104), 40 characters must be recorded in a width of 4724 × 0.7 = 3307 pixels. That is, the character width assigned to one character is 3306.8 / 40 = 82.67 pixels, and the character size is 12 × 0.7 = 8.4 points (70 dots) (steps 105 and 106).

In this case, since the character width is not an integer, it is preferable to set the character width to an integer so as to be suitable for recording. However, if the character width is all 82 pixels or all 83 pixels, the character size will be the size according to the set copy magnification, but the character width (character spacing) will be slightly different from that according to the copy magnification. Therefore, reproduction cannot be performed faithfully at the set copy magnification. Therefore, in this embodiment, 1
The character width in the line can be used for both 82 pixels and 83 pixels.
The distribution is such that 3307 pixels are seen when viewing the entire row.

That is, 83 pixels are 40 × 0.67 = 26.8 ≒ 27 characters, and 82 pixels are 40-27 = 13 characters. In this way, if the number of characters of the document is arranged in a predetermined ratio using two types of character width N pixels and N + 1 pixels according to the set copying magnification, the number of characters of the original can be almost within the total number of pixels at the predetermined magnification. Can be arranged. It goes without saying that the two types of character widths are evenly arranged according to the character number ratio. In this example, 83, 83, 82, 83,
83, 82,...

Next, a description will be given of a character size developed in both character widths. Generally, when a character font is expanded from the dictionary into dot data of an arbitrary size, the size is set to an integer. In the above case, if you try to reproduce an image with the correct size at a recording density of 600 DPI, the character size will be 600 × 8.4
/ 72 = 70 pixels, which is an integer.
If it were 14 points, the size would be 0.7 × 600 × 14 /
72 = 81.67 pixels, which is not an integer.

Therefore, in such a case, if the same processing as the above-described character width is performed, two types of character sizes of 81 pixels and 82 pixels are uniformly set at a ratio of 1: 2, respectively, and each character is set according to the setting. What is necessary is just to develop with a different size (step 107).

From the above, the image reduced by 70% has a character width of 82, respectively.
By developing and arranging a character font of 8.4 points with a size of 70 pixels in 83 pixels, image data for one page of recording can be created. Although the description has been made in only one direction, the same processing is performed in the orthogonal direction. FIG. 3 shows the result of the above processing. The expanded character dots in one character area divided by a character width of 82 pixels or 83 pixels may be arranged at an arbitrary position, but FIG. 3 is based on the upper left of the drawing.

[Determination of Character Size] In order to avoid a reduction in the speed at which the character size is changed by one pixel for each character,
The fraction of the character size obtained by the operation is rounded down to an integer. In the above example, it is set to 81 pixels. That is, since the character width is controlled to the substantially specified magnification, if the character size developed within one character width is equal to or smaller than the character width, it can be reproduced as a document.

[Decision of Character Width] The magnification is not set as a number as in the previous example. For example, in the case of a fixed-size scaling instruction from A4 to A5, the character width in the previous example is strictly 82.67. Becomes an integer of 82. In other words, the reduction ratio is exactly 82 ×
40/4724 = 69.43%, but it is only necessary to be able to record the information of the A4 document in the A5 size paper. Also, since the number of pixels that can be recorded on paper of the same size differs depending on the recording device actually used, the maximum character width may be the same as the maximum character width in which no information is lost in the entire recordable area.

[Determination of Font] The above example is based on the premise that a font used for a reproduced image should be selected closer to the original for copying the original. An image can be reproduced by selecting an arbitrary font according to the instruction.

[Image reading resolution] In the above example, the original is read at the same magnification, stored in the image buffer 8, and the original information is recognized or detected, regardless of the type of the subsequent scaling processing. If the original is considerably smaller than the A4 size, such as business cards and postcards, if it is expected that the original text will be small in advance, read the original by enlarging it and perform character recognition with this image signal. It is possible to increase recognition accuracy.

[0021]

As described above, according to the present invention,
Recognizes characters in the scanned document image, detects information about the characters in the read image, reads out character fonts according to the character recognition results, and generates a reproduced image based on the information about these character fonts and characters. So
It is possible to suppress the deterioration of the quality of small characters during copying. Further, when the image reproduced in the above configuration is further copied using the original as a document, the image quality does not deteriorate.

[Brief description of the drawings]

FIG. 1 is a block diagram of an image processing apparatus according to an embodiment of the present invention.

FIG. 2 is a flowchart illustrating a flow of processing in the image processing apparatus.

FIG. 3 is a diagram showing reproduced character sizes and character intervals.

[Explanation of symbols]

 1 MMI 3 CPU 4 Scanner 5 Printer 6 Dictionary 7 Recognition unit

Claims (9)

[Claims] A reading means for reading an image of a document; a character recognizing means for recognizing characters in the image read by the reading means; a storing means for storing a character font; Reading means for reading out a character font from the storage means, detecting means for detecting information on characters in an image read by the reading means, and a character font read by the reading means and detected by the detecting means. An image processing apparatus comprising: a generation unit configured to generate a reproduced image based on information on characters. 2. The character font used for a reproduced image is a font closest to a character on a document.
The image processing apparatus according to any one of the preceding claims. 3. The size of a character used for a reproduced image with respect to the same character size on a document is at least different in size.
2. The image processing apparatus according to claim 1, wherein there are two types. 4. The image processing apparatus according to claim 1, wherein the character size used for the reproduced image is a maximum size that allows all characters in the document to be reproduced as the reproduced image. 5. The image processing apparatus according to claim 1, wherein the generation unit reproduces a character having a size corresponding to the information detected by the detection unit and a set copy magnification. 6. The image processing apparatus according to claim 5, wherein said generating means reproduces characters at character intervals according to a set copy magnification. 7. The character reproducing apparatus according to claim 6, wherein said generating means reproduces the character by combining a plurality of character intervals when the number of pixels of the character interval calculated according to the set copy magnification is not an integer. Image processing device. 8. An image of a document is read, information on characters in the read image is detected, characters in the read image are recognized, a character font is read from a storage unit according to a result of the character recognition, and the read character is read. An image processing method for generating a reproduced image based on information on a font and a detected character. 9. An image of a document is read, information on characters in the read image is detected, characters in the read image are recognized, and a character font is read from a storage unit in accordance with a result of the character recognition. A computer-readable storage medium storing a program for executing a process of generating a reproduced image based on information on fonts and detected characters.