JPH02162967A - Detection system for black character - Google Patents

Abstract

PURPOSE:To detect a black character with high accuracy by tracing a character area in a local area and detecting the black character after extracting a character core from a 3 primary color. CONSTITUTION:A character extraction section 11 discriminates a picture element as an achromatic picture element depending whether a maximum value of differences of 2 primary color signals of different combination in an inputted 3 primary color signal is smaller than a prescribed 1st threshold level or not. Moreover, an achromatic picture element with high density is selected as a character core and a picture element not achromatic with high density is discriminated as a dark element and other picture elements are discriminated as other elements. The data are inputted to a trace section 12, in which a character is discriminated whether it is a black character or a non black character depending whether a noted picture element is a character core or not in a local area of a prescribed size.

Description

[Detailed Description of the Invention] [Industrial Field of Application] The present invention relates to a method for detecting black characters, and in particular to a method for detecting a character portion from a color document and detecting black characters among the character portions with high precision. .

In a color copying machine or a color facsimile, a color image read from a color original is subjected to various types of image processing to improve color image quality, and then the color image is transferred onto transfer paper using toner or ink. In such color copying machines and color facsimiles, it is desired to improve the reproduction quality of black characters as one method for improving color image quality.

[Conventional technology]

In conventional color copying machines, the three primary color signals of red (R), green (G) and blue (B) obtained by scanning a color original with a scanner and photoelectric conversion are converted into complementary colors. The signal is converted into three subtractive primary color signals of cyan (C), magenta (M) and yellow (Y). At this time, since the C, M, and Y color materials (ink and toner) do not have ideal spectral reflection characteristics, faithful color reproduction cannot be achieved even if the color materials are simply used as complementary colors for R, G, and B. , after color correction is performed on the three subtractive primary color signals, a black (Bk) signal is generated based on these signals.

The three primary color signals (ink control data) Y, M, C and the black signal (ink control data) of the above color-corrected subtractive color method are:
After being subjected to known image quality improvement processing for halftone reproduction and image sharpening, Y, M.

Single-color images of Y, M, and C are generated on transfer paper using each color material of C, and these images are superimposed to obtain a full-color image.

Here, black characters can be printed by overlapping equal amounts of the three primary color materials Y, M, and C, but in that case, the density in high-density areas is mainly due to the effect of reflection on the color material surface. The shortage is noticeable. For this reason, under color removal (UCR) has conventionally been used to compensate for this lack of density.
o1or Removal) and inking.

The above UCR is obtained by subtracting, for example, the minimum value of the three primary color ink control data y, m, and c that represents the character color (i.e., removing the gray component), and then using black ink that is the same as the removed gray component. It is a color process that obtains control data. Also, inking is printing using this black ink control data in addition to the ink control data of the remaining primary colors from which the gray component has been removed. This LICR can save the amount of color ink or color toner.

[Problem to be solved by the invention]

However, conventionally, black characters have not been detected, and in the case of black characters, three primary colors of ink IIJIII representing the character color are used.
Printing was performed by indicating the same amount of data y, m, and c, but in reality, these ink control data were not the same even for black characters, so even if LICR was executed, black ink or Black characters were printed (reproduced) with some Y, M, and C ink or toner added to the toner. For this reason, conventional methods have had the disadvantage that the reproduction quality of black characters is not good.

SUMMARY OF THE INVENTION It is therefore a general object of the present invention to provide a new and useful black character detection method that eliminates the above-mentioned drawbacks.

Another object of the present invention is to provide a black character detection method that detects black characters by extracting the core of black characters from three primary color signals read from a color original and then tracing the character area within a local area. The purpose is to provide.

[Means for Solving the Problems] FIG. 1 shows a block diagram of the principle of the present invention. In the same figure, 11
is a character kernel extraction unit that separately detects achromatic pixels and high-density pixels from the three primary color signals of additive or subtractive methods read from a color original, and extracts achromatic and high-density pixels as characters. It is detected as a nucleus, and pixels that are not achromatic but have high density are detected as dark.

Reference numeral 12 denotes a tracking unit, which determines that when the pixel of interest is a character nucleus, it is a black character, and when the pixel of interest is dark, it determines whether or not it is a black character depending on whether or not there are character nuclei in the surrounding pixels. If it is neither a character core nor a dark character, it is determined that the character is a non-black character.

(Operation) The character kernel extraction unit 11 determines that a pixel is an achromatic pixel based on whether the maximum value of the difference between two primary color signals of different combinations among the three input primary color signals is smaller than a predetermined first threshold value. This is because although the levels of the three primary color signals of achromatic pixels (monochrome characters) are theoretically equal, the maximum difference value described above should be zero in principle.

In addition, since the input three primary color signal levels correspond to the density, the character kernel extraction unit 11 detects the minimum level for each pixel, and when the minimum level is greater than the second threshold, the pixel is raised. It is determined that the pixel is a density pixel.

Furthermore, the character kernel extraction unit 11 uses the above-mentioned achromatic pixels, and
Pixels with high density are considered to be text parts, pixels that are not achromatic but have high density are determined to be dark, and other pixels are determined to be other.

These data are input to the tracking unit 12, where it is determined that the pixel of interest is a black character or a non-black character within a local area of a predetermined size depending on whether or not it is a character part. That is,
For example, as shown in FIG. 2, this local area is a 3×
In the local area 3, of the remaining 8 pixels in this area, 4 pixels indicated by squares are pixels for which tracking processing has been completed, and 4 pixels indicated by circles
A pixel is a pixel before tracking processing.

Here, when the pixel of interest is a character portion, the tracking unit 12 determines that the pixel of interest is a black character. Further, when the pixel of interest is dark, if there is a pixel detected as a character part among the surrounding eight pixels, it is determined to be a black character, and if there is no pixel, it is determined to be a non-black character.

Further, when the pixel of interest is neither a character portion nor a dark pixel, it is determined to be a non-black character, and a detection signal is output.

Thereby, it is possible to detect whether or not the pixel of interest is a black character depending on three cases: a character part, dark, and other cases.

〔Example〕

FIG. 3 shows a block diagram of one embodiment of the method of the present invention. In the figure, the same components as in FIG. 1 are denoted by the same reference numerals, and their explanations will be omitted. In FIG. 5, the three primary color signals of R, G, and B obtained by reading a color original with the scanner 15 (these are digital signals obtained by sampling and stringing the analog three primary color signals) are as follows: It is supplied to the color correction section 16 and also to the literature section extraction section 11 .

The color correction unit 16 converts the input three primary color signals B, G and R into the three primary color signal Y of the subtractive color method, which is their complementary color.

The three primary color signals Y, M, and C are converted into M and C, and color correction is performed to reproduce colors faithful to the original image, and the three primary color signals Y, M, and C obtained thereby are input to the black signal generation section 18 and one literature detection section, respectively. do.

The black signal generation section 18 generates and outputs a black signal Bk based on the three primary color signals Y, M, and C, and also outputs the three primary color signals Y, M, and C described above.

On the other hand, the literature section detection unit 19 separately detects whether or not each of the three primary color signals Y, M, and C is a text section, and when the Y signal is detected as a text section, the output signal y char is set to "1".
Similarly, when a character part is detected for M signal and C signal, output signal @M char.

(:, char is largely set to "1". As a detection method by the literature detection unit 19, for example, the degree of edge is determined for each pixel, and the pixel with the highest degree of edge is detected as a local area consisting of a plurality of pixels. If there are more than a set number of pixels within the pixel, there is a method of determining that the pixel is a character part.

In addition, the three primary color signals Y extracted from the black signal generation section 18
, M, C and the black signal Bk are supplied to a gradation processing section 20 and a resolution processing section 21, respectively. The gradation processing unit 20 is a known circuit that smoothes and dithers the input signal, and outputs the four signals Y, M, C, and Bk subjected to gradation processing.

The resolution processing unit 21 is a known circuit that emphasizes the edges of the human input signal and then performs dither processing, and outputs four signals Y, M, C, and Bk subjected to resolution improvement processing.

The signals Y, M, C, and Bk respectively extracted from the gradation processing section 20 and the resolution processing section 21 are supplied to the selection circuit 22, where the signal Ychar from one literature detection section is
, Mchar and CChar, and the Faculty of Literature extraction unit 11. which is the main part of this embodiment. It is selectively outputted based on the black character detection signal 3kchar generated by the circuit section consisting of the tracking section 12 and the expansion section 17.

The selection circuit 22 receives detection signals Ychar, Mchar,
When C-char is logic "1", the corresponding primary color signals Y, M, and C taken out from the resolution processing unit 21 are selected and output, and Ychar, Mchar, and Cchar are logic "1".
When it is 0'', the corresponding primary color signals Y, M, and C taken out from the gradation processing section 20 are selectively output.

However, this selection operation is based on the detection signal Ychar, M −
char, Cchar and black character detection signal 3kchar
This is the case when all of r are not logic "1" at the same time, and when all of these four signals become logic "1" at the same time, the primary color signals Y, M, and C are not output, and for example, W
Only the black signal 13 from the image intensity processing section 21 is output.

That is, according to this embodiment, when a black character is detected, only the black signal Bk is output from the selection circuit 20.

As mentioned above, when the selection circuit 22 does not detect a black character, it selects and outputs the three primary color signals of either the gradation processing section 20 or the resolution processing section 21 depending on whether it is a character part or not. As a result, pixels with text areas are processed with high resolution, and pixels without text areas are processed with 111g4 processing, which allows conflicting color reproduction and gradation to be achieved without reducing the resolution of text areas. Maintained.

Next, the main part of this embodiment is the character extraction section 11゜tracking section 1.
The configuration and operation of the circuit section consisting of the expansion section 2 and the expansion section 17 will be explained in more detail with reference to the block diagram of FIG. In FIG. 4, the same components as those in FIG. 1 are designated by the same reference numerals, and their explanations will be omitted. In the M4 diagram, additive color method 3
Two different combinations of the primary color signals R1G and B are supplied to computing units 26, 27 and 28, respectively, where they are subtracted and their absolute values are taken.

The comparator 29 selects and outputs the larger value of the output signal IRG-1 of the arithmetic unit 26 and the output signal IG-81 of the arithmetic unit 27 and supplies it to the comparator 30. A comparison is made in magnitude with the signal IB-Rl. The comparator 30 selects and outputs the higher level signal among these input signals. Therefore, the computing units 26 to 28. A circuit consisting of comparators 29 and 30 allows IR-G to be output from comparator 30.
1. The maximum level signal of IG-81 and IB-RI, that is, the maximum difference value described above, is output.

The maximum difference value for each pixel taken out from the comparator 30 is
A logic "1" is supplied to the comparator 31 where it is compared with a first threshold TH1, indicating that the pixel is an achromatic pixel when the maximum difference value is less than the first threshold TH1. In other cases, the signal is a logic "0" signal.

On the other hand, the three primary color signals R, G and B are also subjected to modulation transfer? - After passing through function (MTF) correction circuits 32, 33 and 34 separately for edge emphasis, the signal is supplied to a minimum level detection circuit 35, where a minimum level primary color signal is detected for each pixel. This minimum level is fed to a comparator 36 where a second threshold T
H2 is compared, and the lowest level is the second threshold value.
When it is greater than H2, it is a logic "1" signal indicating that it is a high density pixel, and in other cases it is a logic "0" signal.

Based on the logic of the Okayama power signals of the comparators 31 and 36, the code synthesis circuit 37 determines that the input pixel has a high density
One of three codes in total (this is 2 bits): an achromatic color indicates a character part, a high density and chromatic color indicates a dark color, and otherwise indicates something else. Generate and output.

The above code is fed to a density filter 38 to remove small text parts such as dust, where a code indicating a text part when the number of text parts in a local area is greater than a third threshold TH3. Otherwise, if the pixel of interest is a character part, a code indicating dark is used, and if the pixel is different, the code of the pixel of interest is output as is.

The output code (data) of the density filter 38 is supplied to the tracking processing circuit 39, where it is determined whether the pixel of interest within the 3×3 local area is a temporary black character or a non-black character according to the processing algorithm of the tracking unit 12 described above. .

Data retrieved from tracking processing circuit 39 is provided to density filter 40 . This density filter 40 constitutes an expansion unit 17 that expands the tracked data of the temporary black character to determine whether it is a black character or a non-black character, and there is at least one temporary black character within a local area of a predetermined size. If so, the pixel of interest is detected as a black character, and if it is not, a black character detection signal of 3kchar is output indicating that the pixel is a non-black character. This results in
Black text can be detected from color originals.

Note that the tracking processing circuit 39 described above can be realized by a circuit with a simple configuration as shown in FIG.

In the figure, the code (pixel data) input to the terminal 42 is transferred to the shift register 43. While the output data of the FF 44 is supplied to the judgment circuit 46 through the flip 70 tubes (hereinafter referred to as FF) 44 and 45, the output data of the FF 44 is sent to the shift register 47.
The signal is further supplied to the determination circuit 46 via the FF 48 and the FF 48.

A delay of one line is applied to the input data by the shift register 43 and FF 44, and similarly, the shift register 47 and FF
It is delayed by one line by F48. Therefore, FF48
Assuming that the output data of is the pixel data of interest at the center of the 3×3 local area indicated by the X mark in FIG. This is the next pixel data to be processed, and also FF44
The output data of the shift register 43. is the pixel data indicated by ■ in FIG.
Each output data of the FF 45 is also pixel data one line in the future, indicated by ■ and ■ in FIG.

The output data of FF48 is FF49. shift register 50
The signal is further delayed by one line by the FF 51 and supplied to the determination circuit 46, and further delayed by one pixel by the FF 52 and supplied to the determination circuit 46. Therefore, FF51.
Each output data of FF 49.52 is pixel data of one line past the pixel of interest shown in FIG. The data supplied from the shift register 50 to the determination circuit 46 is the pixel data shown by and in FIG.

As a result, the pixel data of the 3×3 local area is input to the determination circuit 46, and based on these, the determination circuit 46 uses the tracking algorithm to determine whether the pixel of interest is a temporary black character if it is a character liquid, or a temporary black character if it is not. When the pixel of interest is dark and there are temporary black characters in the surrounding eight pixels, it is determined to be a temporary black character, otherwise it is determined to be a non-black character, and the determination result is output to the output terminal 53.

The primary color signals Y, M, C or black signal @Bk extracted from the selection circuit 22 in the embodiment shown in FIG. 3 described above are as follows:
The data is output to the printer as ink control data to control the printer's color materials. Although this printer is not directly related to the gist of the present invention, for example, there is a laser printer having a configuration as shown in FIGS. 6 and 7. In FIG. 6 and FIG. 7, the same components are denoted by the same reference numerals, and suffixes Y, M, C, and [3 are attached corresponding to the input signals.

The Y, M, C, and Bk ink control data described above are laser beams 1157Y and 57M shown in FIG.

57G and 578k to modulate the laser beam. The modulated laser beam irradiated by the laser beam +IQ578 is incident on the rotating polygon 111588k shown in FIGS. 6 and 7, respectively, and is deflected there and passes through the imaging lens 598k (shown only in FIG. 6) to the photosensitive drum. Irradiated onto 60Bk.

Since the rotating polygon ff1588 is rotated in a fixed direction by the motor 61Bk, the laser beam deflected here repeatedly linearly scans the surface of the photosensitive drum 608 along the direction of the drum rotation axis. (called scanning). Further, by rotating the photosensitive drum 608k in the direction of the arrow, the laser beam is also scanned in a direction perpendicular to the main scanning direction (this scanning is called sub-scanning).

As the photosensitive drum 608 rotates, its surface is uniformly charged by a charger 628k shown in FIG. An electrostatic latent image is formed corresponding to the image. This electrostatic latent image is transferred to the developing device 63'.
It is developed and visualized with Bk.

On the other hand, tray 64. Transfer sheets 63Y and 63M are transferred to the photosensitive drum 638 while being sandwiched between the rows 566 after passing through the sheet distribution device 65 and the rollers 66, respectively.

It is sent onto the conveyor belt 67 in synchronization with the rotation of 63C. After a visible image on the surface of the transfer paper on the conveyance belt 67 is electrostatically transferred to the photosensitive drum 608 by the action of a transfer discharger 688, the transfer paper is conveyed to the next photosensitive drum 60Y direction.

Thereafter, in the same manner as above, a yellow-white image is formed on the surface of the transfer paper by the photosensitive drum 60Y and the transfer discharger 68Y, a magenta image is formed by the photosensitive drum 60M and the transfer discharger 68M, and a cyan image is formed by the photosensitive drum C and the transfer discharger 68C. After being electrostatically transferred in order of overlapping, the fixing unit 69 performs fixing, and the paper is discharged to the tray 71 via the paper discharge roller 70.

By applying this embodiment to a laser printer with such a structure, the character portion determined to be black is electrostatically transferred onto the transfer paper using only black toner, and images of the other three primary colors are not electrostatically transferred. , the reproduction quality of black characters can be improved compared to the conventional method.

It should be noted that the present invention is not limited to the above-mentioned embodiments, and it goes without saying that, for example, the detection of the maximum difference value and the detection of the minimum level may be performed based on the three primary color signals of the subtractive color method.

〔Effect of the invention〕

As described above, according to the present invention, after extracting a character part from the three primary colors, black characters are detected by tracing the character area within a local area. The nucleus extraction section and tracking section can be configured easily, so
It is possible to detect black characters with small-scale hardware, and it is also possible to print text parts that are determined to be black on transfer paper using only black toner or black ink. It has features such as being able to faithfully reproduce the color of black text without color.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of the principle of the present invention, FIG. 2 is a diagram showing a local area for explaining tracking processing, FIG. 3 is a block diagram of an embodiment of the present invention, and FIG. 4 is a main part of the present invention. FIG. 5 is a block diagram of an embodiment of the main part of FIG. 4, and FIGS. 6 and 7 are diagrams each explaining the structure of a laser printer to which the present invention can be applied. be. 11... Character kernel extraction section, 12... Tracking section, 15...
- Scanner, 17... Expansion unit, 26-28... Arithmetic unit, 29-31. 36... Comparator, 35... Minimum level detection circuit, 37... Code synthesis circuit, 38.40
...Density filter, three...Tracking processing circuit. Kazu Diagram

Claims (1)

[Claims] Three primary color signals obtained by scanning a color original with a scanner are color corrected and converted into three primary color signals of a subtractive color method, and image quality is improved with respect to the three primary color signals of the subtractive color method. A method for detecting black characters used in a device that prints a color image of the color document on transfer paper by controlling color materials in a subtractive color method based on control data of three primary colors obtained by processing, the method comprising: Three primary color signals of the color method or subtractive color method are input, achromatic pixels are extracted based on the maximum value of the difference between the two primary color signals of different combinations, and high density pixels are detected. , and a character kernel extraction unit that detects the high density pixel as a character nucleus, and detects and outputs the high density pixel as dark instead of the achromatic pixel, and an output signal of the character kernel extraction unit. Based on this, when the pixel of interest is in the text area, it is determined to be a black character, and when the pixel of interest is dark, if the character area exists in the surrounding pixels, it is determined to be a black character, and when it does not exist, it is determined to be a black character. A method for detecting a black character, comprising: a tracking unit that determines that the character is a black character, and determines that the pixel of interest is a non-black character when the pixel of interest is neither the character portion nor the dark character.