JPS61157158A - Picture processing device - Google Patents

Abstract

PURPOSE:To obtain a picture processing device provided with the identifying function of a picture tone where the high speed processing can be executed by identifying the picture tone of attention picture element data from the characteristic of a space frequency of the attention picture element data and the linearity which the attention picture element data have by the identifying means, in the digital copying machine, facsimile, etc. CONSTITUTION:At an output edge of static RAMs 1-1-1-7 and delayed flip-flop (D-F/F) circuits 2-1-2-8, 3-1-3-8, the binary picture corresponding to an original can be two-dimensionally and simultaneously detected for eight picture elements in the sub-scanning direction and three picture elements in the main scanning direction. 4-1-4-23 are 'or else' circuits (hereinafter called 'EX-OR gate'), and these EX-OR gates calculate the 'or else' between adjoining picture elements with six picture elements equivalent to the output of D-F/F circuits 2-2-2-7. 5-1 and 5-2 are ROMs, input the calculating result of 23 EX-OR gates, judges whether or not the level of two picture elements adjoining by sandwiching the attention picture element concerning six picture elements is inverted for the attention picture element, and executes the counting of the inverting arrangement.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field] The present invention relates to an image processing apparatus equipped with an image identification function suitable for use in digital copying machines, facsimile machines, and the like.

[Prior Art] Conventionally, in this type of device, when reproducing or encoding and transmitting a document containing a mixture of text and halftone images, the image tone (nature or characteristic of the image) is different, so that the entire document is The same playback process or the same code or decoding could not be applied.

Therefore, there is a need to distinguish between different image tones in the manuscript,
Various proposals have been made. However, it has been difficult to provide a processing device that is suitable for high-speed processing, has a small hardware scale, and has an image tone identification function. Furthermore, it is also difficult to accurately distinguish between halftone dot images and line images such as characters.

[Objective] The present invention was made in view of the above points, and an object of the present invention is to provide an image processing apparatus having a small hardware scale and an image tone identification function capable of high-speed processing.

Another object of the present invention is to provide an image processing apparatus having an accurate image tone identification function.

[Example] Embodiments of the present invention will be described in detail below with reference to the drawings.

FIG. 1 is a block diagram of the image processing apparatus in this embodiment. In the figure, reference numeral 16 has a sensor such as a CCD, and this sensor reads a document and generates 6-bint image data IS.
This is an image reading section that outputs. This sensor is a line sensor, which electrically scans the document in the line direction (main scanning direction) and mechanically scans it in a direction perpendicular to the line direction (sub-scanning direction) to read the entire document. . 10 is a two-dimensional smoothing circuit (smoothing unit) that performs two-dimensional smoothing processing on 6-bit image data IS, and 11 is image data (smoothed signal SS) smoothed by the two-dimensional smoothing circuit 10. ) is a dither processing circuit (dither binarization unit) that performs dither processing (halftone processing) using a predetermined dither matrix and binarizes it. With the above configuration, a binarized signal DS with moiré suppressed can be obtained from the dither processing circuit.

Further, 12 is the other binarization circuit, which inputs the 6-bit image data IS and also inputs the smoothed signal SS.
The 6-bit image data IS is binarized using the threshold value.

In this embodiment, by using the smoothed image data SS as a threshold value, it is possible to extract and reproduce even character parts that are likely to be buried in the background image (the background of the original). 13 is a 6-discrimination circuit which inputs the z-valued signal (high resolution signal) HS outputted from the z-valued circuit 12 and identifies the image tone. The identification circuit 13 counts feature amounts (described later) from the array in the vicinity of the pixel of interest, performs threshold processing on the feature amounts, and identifies the image tone. Then, according to the image tone identification result RE, the switch 15 is changed to select either the binary code H5 or the binary code DS and output it to the recording device 14 such as a laser beam printer. That is, when the identification circuit 13 determines that the pixel of interest is a halftone dot area, it selects the binarization code DS in which moiré is suppressed, and when it determines that the pixel is a line drawing area such as a character, it selects the binarization code DS. Select the No. HS.

The two-dimensional smoothing circuit shown in FIG. 1 and the like are explained in detail in Japanese Patent Application No. 59-248131 filed by the applicant of the present invention, so the explanation thereof will be omitted.

Next, the identification circuit 13 shown in FIG. 1 will be explained in detail. First, the identification algorithm in this embodiment will be explained in detail. In this example, in order to distinguish between characters and halftone dot images, we focus on the following two features.

■Generally, the difference between character features and halftone image features lies in their spatial frequencies. In other words, a halftone image has a high spatial frequency in two orthogonal directions, but a line drawing such as a character has a high spatial frequency in two orthogonal directions.
Frequency often increases only in one direction.

■While halftone dots are a collection of round dots, characters (especially kanji) are comprised of a collection of orthogonal straight lines.

Therefore, the identification algorithm used in this example focuses on the difference in frequency components mentioned above, and at the same time takes advantage of the fact that characters are composed of thin lines. This makes it possible to identify images from several (fine) halftone dots.

In this embodiment, a feature quantity Pf (X, V) is calculated for each pixel in order to identify the tone of each pixel.

Neighborhood (2n+1)X(
2m+1) (where n and m are natural numbers), the feature quantity Pf(x, y) is defined by the following equation only when all of the following conditions are satisfied, and otherwise pf(x, y)=0.

KokoteP (u, S) A P(x+i, y+j)■
P (x + i ten.

y◆j+s) ■ is exclusive OR, and * is logical product.

P(r,y)IP(x-1,y)IP(4+1.y)'
I.O.

Condition P(x,y)IP(x-1,y)I
P(xtl,y)iP(1!,ri"P(x,y-1)
IP(x,y+1)”tO.

P(K,y)IP(x,y-1)IP(x,y+1)'
The y3 binarized image P (x, y) and the neighboring binarized image sequence are shown in FIG. 2, and the physical meaning of the feature quantity pf (x, y) will be explained. In addition, in Fig. 2, the binarized image is z
The binarized signal H5 outputted from the digitization circuit 12 is used.

Figure 2 shows the binarized state of 255 pixels in the vicinity including the pixel of interest P(x,y) (center in the figure), where black pixels are represented by ゛・pa and white pixels are represented by "0". . Also, when the pixel of interest is black, it is represented by "@", and when it is white, it is represented by 'O°'. As shown in Figure 9, the pixel of interest p (x, y
) adjacent 8 pixels (p(x-t, y), P(xtl,
y), P(x-1, y-1), P(xtl,
y+1), P (xtl, y-1), P(x-
1, y+1), P(x.

y-1) and P (x, y+1)), the isolation and linearity of the pixel of interest P (x, y) with respect to neighboring pixels can be recognized. In other words, in Figure 2 (a), P (-1, 0) = IP (x, y) fP (x-1,
7) = 0P(1,0) = P(x,y)■P(x
tl,y) = 1P(0,-1)=P(x,y)■
IP(xy&,y,,) = 11P(0,1)
= F(x, y) ■ IP(x7J, y,,)
= 0fP(-1,-1) = 1)(x, y
)■FP(!-1.7-1) =1P(1,1) =
PCx,y) (31P(!+1.!$1) 211P
(1, -1) = IP (x, y) ■ F (!'+
1. ! -1) = 0F(-1,1)=IP(
x, y) @ IP (x-1, y+1) = O, so P (-1, -1) ・ IP (1, 1) = 11P
(0,-1)-fP(0,1) = 0IP(1,-1
)・IP(-1,1)=0IP(-1,0)・I
P(1,0) = 0.

Therefore, in FIG. 2(a), two adjacent pixels sandwiching the pixel of interest P(x, y) (for example, p(x+y-1), P(x,
Array P (x-1, y-1) in which the levels of y+1)) are both inverted with respect to the pixel of interest P (x, y) −
IP (xtl, y+1) exists once. Note that this array will hereinafter be referred to as an inverted array, and the value obtained by performing similar counting (counting of the inverted array) over each pixel around the pixel of interest becomes the feature quantity Pf(x, y).

Here, the feature amount Pf(x) for halftone dots and characters.

y) will be explained using FIG. 2(b) and FIG. 2(C). The image in FIG. 2(b) shows a case where halftone dots are assumed, and Pf(x,y) becomes Pf(x,y)=20 if n=m=1 in the formula 0. Similarly, Figure 2 (
c) shows the case assuming a thin line, and Pf (x,
y)=O. Therefore, if the value of the feature amount Pf(x,y) is subjected to threshold processing, it is possible in principle to distinguish between a halftone dot area and a line drawing area such as a character.

That is, when Pf (x, y) > K(7), the pixel of interest P
(x.

y) belongs to the halftone area.

When Pf(x, y)≦K(7), the pixel of interest P(x.

y) belongs to the line drawing area.

It can be determined that

Now, using the above-mentioned algorithm, a halftone image of 100 lines/inch or more is distinguished from a character image, and the former is subjected to dither processing to suppress moiré by a dither processing circuit 12 as shown in FIG. If the latter is to be binarized by the binarization circuit 12, it is desirable to set m=n=3 and select the value of K from 1'6 to 20.

Next, the circuit configuration of the identification circuit 13 will be explained using FIG. 3. The circuit shown in FIG. 3 executes the above-mentioned algorithm, and is configured to count the feature amount Pf(x,y) over 6×6 pixels in the vicinity including the pixel of interest.

In the figure, 1-1-1-7 is a study R of 4 x 1 bit.
AM, the input binary images H5 that are input sequentially are 1
Delay in the sub-scanning direction line by line. 2-1 to 2-8 are delayed flip-flop circuits (hereinafter referred to as D-F/F circuits), which receive input from terminal T and static RAM.
The outputs of l -1 to 1-7 are each delayed by one pixel in the main scanning direction. Also 3-1 to 3-8 are D-F/F
This circuit further delays the outputs of the D-F/F circuits 2-1 to 2-8 by one pixel. Therefore, static RAM l-1 to 1-7 and D-F/F circuits 2-1 to 2-
8. At the output ends of 3-1 to 3-8, z-value images corresponding to the document can be detected two-dimensionally simultaneously for 8 pixels in the sub-scanning direction and 3 pixels in the main-scanning direction. 4-1 to 4-23 are exclusive OR circuits (hereinafter referred to as EX-OR gates);
- Calculate the exclusive OR between the six pixels corresponding to the outputs of the F/F circuits 2-2 to 2-7 and the adjacent pixels. 5-1.5-
2 t* ROM with TE, EX-OR gate 23 (1
1 (7) Input the calculation results and judge whether or not the levels of two adjacent pixels sandwiching the pixel of interest are inverted with respect to the pixel of interest, and count the inversion array. is 2 (=4K) word c7)
Using two ROM5-1.5-2, EX-OR gate 4
-12 output is input to both chips. Therefore ROM5
-1.5-2 performs the above-described judgment and inversion array coefficients for each three pixels. The outputs of ROM5-1.5-2 are added by adder 6,
The calculation result (integrated value of the inverted array for 6 pixels) is D-F
/F circuits 7-1 to 7-5 delay and hold the adder 8.
-1 to 8-3 are input. The adders 8-1 to 8-3 each add the above calculation results for two columns in the main scanning direction, and the addition results of each adder are input to the ROM 9.

Therefore, the address input terminal of the ROM 9 has the feature amount Pf(x,
A signal corresponding to y) is added. The ROM 9 has a table whose addresses are the outputs of the adders 8-1 to 8-3, and this table stores a 1-bit identification result RE indicating whether it is a halftone image or a line image. There is. That is, the ROM 9 inputs the feature Pf (x, y) and outputs the comparison result (1-bit identification signal RE) as to whether or not this feature amount Pf (X, y) is larger than the value. be.

As explained above in Figure 3, memories such as RAM and ROM,
Image identification processing can be performed in real time with a simple configuration using gate circuits and the like.

[Effect J] As explained above, according to the present invention, image tone can be identified in real time with a simple configuration.

Furthermore, according to the present invention, since the image tone is determined based on the spatial frequency characteristics of the image and the linearity of the image, the image can be identified with high precision. Images and halftone images can be accurately identified.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of the image processing device in this embodiment,
FIG. 2 is a diagram for explaining the image tone discrimination algorithm in this embodiment, and FIG. 3 is a detailed circuit diagram of the discrimination circuit 13. Here, 10 is a two-dimensional smoothing circuit, 11 is a dither processing circuit, 12 is a binarization circuit, 13 is an identification circuit, 14 is a printer, 15 is a changeover switch, 16 is an image reading section, 1-1-1
-7 is static RAM, 2-1~2-8.3-1~
3-6. 7-1 to 7-5 are delayed flip-flop circuits, 4-1 to 4-19 are exclusive OR circuits, 5-1.
5-2.9 is a ROM, and 6.8-1 to 8-3 are adders.

Claims (1)

[Claims] means for inputting binarized pixel data; and identification means for identifying the tone of the pixel data of interest from the pixel data of interest input from the input means and adjacent pixel data sandwiching the pixel data of interest; An image processing apparatus characterized in that the identifying means identifies the tone of the pixel data of interest based on the spatial frequency characteristics of the pixel data of interest and the linearity of the pixel data of interest.