JPS60206371A - Picture signal processing system - Google Patents

Abstract

PURPOSE:To enable a titled system to scarcely annihilate a thin line of black or white, and also to emphasize the contrast at the post-stage of a correcting circuit, by indicating the density of an object picture element by the value of a picture signal derived by using a specified expression from a digital signal for indicating the density of the object picture element and a picture element being adjacent to said picture element, respectively. CONSTITUTION:A contrast emphasizing circuit of this invention has a flip-flop 11 and a memory 12. In a waveform A, for instance, when a digital picture signal g0 and a digital picture signal (f) denote a density of a picture element P2, and the density of a picture element P1 which is one picture element before in the main scanning direction seen from the digital picture signal g0, respectively, the value g1 shown by an expression I is written in the memory 12 by using a value of the digital signal g0 and (f) as an address. The value of g1 of a waveform B is read out of the memory 12 by applying as an address signal, the digital picture signal (f) inputted to the flip-flop 11 and delayed by one picture element portion by a picture signal clock C, and the digital signal g0. That is to say, the picture signal waveform A is converted consequently to the picture signal waveform B which has emphasized the contrast of a thin line.

Description

[Detailed Description of the Invention] [Technical Field of the Invention] The present invention processes a group of digital image signals that determine the density of each pixel in an image to obtain a group of digital image signals with enhanced contrast. The present invention relates to an image signal processing method for emphasizing the contrast of thin lines, and particularly to an image signal processing method for emphasizing the contrast of thin lines.

[Technical background of the invention]

Conventionally, when converting an analog image signal that determines the density of each pixel in an image into a digital image signal and obtaining a digital halftone signal and a binary signal from this, a block diagram is shown in FIG. 1 or 2. A signal processing circuit is used.

That is, what is shown in FIG. 1 converts the analog image signal obtained by converting the density information of each pixel into an electrical signal by the photoelectric conversion element 1,
After the level converter 2 converts the level to match the range of the A/D converter 3, the A/D converter 3
Then, a correction circuit 4 removes distortion of the image signal caused by the optical system to obtain a digital halftone output signal, and a binarization circuit 5 obtains a binarized signal. .

Regarding the second cause, a thin line contrast enhancement circuit 6 was inserted between the photoelectric conversion element 1 and the level converter 2 in FIG. 1, and this enhanced the contrast of the a-line before digitizing it. .

[Problems with background technology]

However, the image signal processing circuit shown in FIG. 1 has a drawback that thin black or white lines tend to disappear. For example, in FIG. 3, three pixels P1 consecutive in the main scanning direction
, Pt, and Ps, and read in a thin line whose thickness is almost the same as the width of one pixel.If the position of the pixel P and the position of the thin line match as shown in Figure 3A, then As shown in FIG. 4A, an output voltage with a large contrast is obtained; however, as shown in FIG. 3B, when the center of the thin line is on the boundary between pixels P1-Px, As shown in Fig. 4 BK, a small contrast output voltage is obtained, and therefore, if this is binarized, there is a risk that the thin line will disappear.In addition, in the circuit shown in Fig. 2, the negative contrast enhancement circuit 6, the above-mentioned drawbacks of the circuit shown in FIG. Because of this, an image signal with enhanced contrast of thin lines is input to the correction circuit 4, which has the disadvantage that accurate correction is impossible.

[Purpose of the invention]

The present invention has been made in order to eliminate the drawbacks of the conventional image signal processing circuits described above, and is an image signal processing method that prevents thin black or white lines from disappearing and that can enhance contrast at a later stage of the correction circuit. The purpose is to provide

[Summary of the invention]

In the first invention of the present application, when the digital image signals that represent the densities of the target pixel and the pixels adjacent thereto are respectively go and f, the values of these two digital image signals go and f are used as addresses, and gt= Write the value g1 calculated by a((1+b)go bf) (where a and b are positive integers) into the memory, and calculate the density of the target pixel using the value of the output image signal g1 read from this memory. This achieves the above objective.

The second invention of the present application is to calculate the density of a target pixel and a plurality of pixels surrounding the target pixel. When the digital image signals are go +fl, ft...fn-1, fn, first two Digital image signal go, fl
Use the value of gs”a+4 (1+b,)go as the address
bl fl ) (where al and bl are positive constants) The value g! is written into the first memory, the output image signal g1 is read out from this first memory, and then gt=a*((1+bt)gt bmfg
) (where am and bn are positive constants) Write the value g to be filtered in the second memory, read out the output image signal gt from this second memory, and repeat the same operation sequentially. Finally, using the output image signal gn-1 read from the n-1th memory and the digital image signal fnQ value as an address, i gn=arl((]+bn)gn-1''n'n) (however, a n r bn is a positive constant)
was written into the n-th memory, and the density of the target pixel was determined using the value of the digital image signal go using the value of the output image signal gnO read out from the n-th memory, thereby successfully achieving the above object.

The third invention of the present application calculates the density of one pixel and a plurality of pixels around it by go I fl *
fn...fn 1. When fn, the digital image signal go = 1-ft -f, the value of f'n-In is used as the address, and gn = "n4 (1+b, + bt+...+bn
I+b11)g.

btf, -btfm-"'-bn, fn, -bnfn)
(However, all + 1) t + 1) t ”-bn is a positive constant] Write the value gn to the memory,
The above object was achieved by adjusting the density of the target pixel using the output image signal gnO value read from this memory.

[Embodiments of the invention]

FIG. 5 is a circuit diagram of a contrast enhancement circuit applied to carry out the first invention of the present application, and includes a flip-flop 11 and a memory 12.

The flip-flop 11F'i is a set of flip-flops whose number corresponds to the number of bits of the digital image signal.

In addition, FIG. 6 shows the waveform of a digital image signal that shows the density of pixels P1 to P4, and pixel P! has a density of X″t′, but at the pixel P*′f and at the pixel P1
It is assumed that the subsequent concentration is zero. Furthermore, when the digital image signal go changes the density of the pixel P8, the digital image signal f is the image signal of the pixel one pixel before the main scanning direction from the digital image signal g0, that is, the pixel P adjacent to the pixel P.
The concentration of , shall be considered.

The digital image signal g is stored in the memo 1J12, which is a ROM.

Using the value of and f as an address, the value g1 determined by gs=a4 (1+b) go bf ) - (1) is committed.

a and b are appropriate positive constants, for example am1.

becomes.

Therefore, in FIG. 5, by adding the digital image signal f input to the flip-flop 11 and delayed by one pixel by both signal clocks C and the digital image signal g6 as address signals to the memory 12, The value of g is read. FIG. 6 shows the waveform of the BFi output image signal g+. That is, (a) Up to pixel P1, go-0, f=o, so double -〇 (b) In pixel P, go-X because g6wx, fxO (c) In pixel P1, go 1,000, gl-X of f This means that the waveform has been converted to the image signal waveform shown in $6 B, which emphasizes the contrast of the thin lines. Similarly, from the image signal waveforms shown in Figures 7 and 8, Figure 7B is used.
An image signal waveform with enhanced contrast shown in FIG. 8B is obtained.

Table 1 go (lower bit) Table 1 shows the data to be written into the ROMK according to equation (2), where go is the lower bit of the ROM and f is the upper bit of the ROM, and the numerical values are expressed in hexadecimal. However, in this case, in equation (2), when the value of g is less than θ, all 0 is written, and when it is more than F, all F is written.

The first invention of the present application has been clarified above, but in the embodiment described above, the pixel signal g of the target pixel, and the pixel signal f of the pixel one pixel before the target pixel in the main scanning direction. When inputting and processing, the pixel signal g of the target pixel
o and the pixel signal of a pixel one pixel later in the main scanning direction with respect to the target pixel may be input and processed, and the pixel signal g of the target pixel and the pixel signal g of the target pixel with respect to the target pixel in the sub-scanning direction may be input and processed. A pixel signal of a pixel before or after the pixel may be input. Furthermore, the memory 12 may be a RAM""c instead of a ROM, the data written to the memory 12 may be of negative logic, or the bits of the data written to the ROM&C-1F may be replaced.

FIG. 9 shows a block diagram of a circuit for converting an analog image signal into a digital halftone signal and a binary signal, which includes a contrast enhancement circuit 10 to which the image signal processing method according to the first invention of the present application is applied. As is clear from the figure, since the contrast enhancement circuit 10 is a circuit that performs digital processing, it can be inserted between the correction circuit 4 and the binarization circuit 5, allowing accurate correction.

Next, the second invention of the present application will be explained with reference to FIG. 10 and subsequent figures. The second invention of the present application is an image signal processing method that sequentially inputs and processes a pixel signal g0 of a target pixel and pixel signals of n (plural) pixels surrounding the target pixel.

Figure 10 shows an image signal processing circuit when n=2, and is a circuit that inputs and processes the image signal go of the target pixel and the image signals f, , f, of the two surrounding pixels. . In this case, as shown in FIG. 11, the image signal f1 is similar to the image signal f described above, and the target pixel P! The image signal of the pixel P1, which is one pixel before in the main scanning direction when viewed from the image signal g0, is the image signal f.
tFi image signal g.

The image signal is the image signal of the pixel Pa' that is one pixel before in the sub-scanning direction.

In Fig. 10, first, in the ROM memo 1J12,
Using the values of the image signals go and fl as addresses, gl-a+((], +bl)go bIL) -(3)
I'm going to write a direct gt that will be hailed. If a, , bI are appropriate positive constants, for example, al -1x b1 = d, then the data in Table 1 is read in which memory 2VctI'if is replaced with f. Therefore, the image signal f1 delayed by one pixel by the flip-flop 11
, and the image signal go are input to the memory 2 as an address signal, g,
The value of is read.

After this output image signal g1 is delayed by the flip-flop 13, the second memo! It is input as the J14vc address signal.

The image signal clock C is input to the counter 5, and its counter value becomes the address of the line memories 16 and 17.

The counter value Fil is reset for each line. FIG. 12 shows a timing chart of the image signal clock C, the gate signal G1 or 2, and the write signal WEI or 1 yoke 2. In addition, FIG. 18 shows a timing chart of 18 selections.

The image signal is output to the line 19 when the gate signal 101 or υ2 is at a low level, and is written to the line memory 16 across 17 at the timing of the rise of the write signal WEI or WF2. Commitment signals WEI, wE2 and gate 1g G1. G2 is output or not output alternately for each line. For example, line memory 1
When writing an image signal to the line memory 17, a fill-in signal WEI and a gate signal Gl are output, and during this time, the data of the previous line is outputted from the line memory 17. In the next line, the roles of line memories 16 and 17 change. In addition, 21.
22 is a driver. ) The selector 18 selects and outputs the line 19 or 20 of the sword where the data l lines before was output. Therefore, the flip-flop n
An image signal f of one pixel before the image signal go in the sub-scanning direction is outputted from the image signal go, and is inputted together with the image signal g1 as a second memo IJ14K address signal.

Memory 14 K t;j, image signal g1 and f, value as address g*=a,=4 (1+b, Igx b*ft)
...The value g2 obtained in (4) is stored.

a.

b, Fi are suitable positive constants, for example, Bx-1°, then the values shown in Table 2 are written into the second memory 14, and the image signals g1 and fl are used as address signals to write into the memory. 14, the value of g is read out from the memory 14, and this output image signal g becomes an image signal in which the contrast of thin lines is emphasized in both the main scanning direction and the sub-scanning direction.

Table 2 gl (lower bit) The above is a case where the pixel signals of the target pixel and two surrounding pixels are input and processed, but by further increasing the number of memories and repeating the same processing, The above image signals may be input.

In general, if the image signals of n (plural) pixels around the target pixel are f, , f, ... fn-1"n, then 3, jt[
Using the values of the output image signal gn-1 and image signal fn read from the memory of n-1 as addresses, gn=a n ((1+b n) gn 1
b nf nF...(5) (However, L/% an, bn#
-1 positive constant) is finally written into the n-th memory, and the value of the image signal go is converted by the value of the output image signal gn read out from this n-th memory. Furthermore, an image signal with enhanced contrast of thin lines can be obtained.

Next, the third invention of the present application will be explained.

From equations (3) and (4) above, gy=at[(1+b*)at((1+b,)go b
, f, ) btft]=a+at((1+b++bt+
b+b*) go (b1+'Jb2)L btfx) Tonalnote, if you replace atat with a2, bI+ Jbt with su, you get gn-at((1+b1+bt) go-btL klt
ft) −(63L7 Therefore, the image signal go of the target pixel
The image signals f+ and ft of two pixels around it are simultaneously input as the same ROMK address, and the image signal g2 generated by equation (6) is output using the stored ROM. can be obtained.

Similarly, the image signal go of the target pixel and the image signals of n (plural) pixels around it f'l+fl+...f, f
At the same time, input as the same ROMK address -In, gyl=an((1+b,+bt+---+bn)go
l)+L bsfs −・= bnfJ...(7) (However, LA an, b, ~bn are positive constants) gn
An output signal gn can be obtained using a memory that has been assigned as an output.

1. In equation (7), for example, bl fl + bllf
Use another ROM in advance, set this as Bl, and then go to gn==an((1+b*'+ba+・=+bn)
It is also possible to perform specific calculations in advance in another ROM, such as calculating the bt'f*' bsfs -bHf meter, and then calculate gn.

〔Effect of the invention〕

As is clear from the above explanation, according to the present invention, a signal that emphasizes the contrast of thin lines can be obtained using extremely simple means. Furthermore, the contrast enhancement circuit to which the method of the present invention is applied can be inserted after the correction circuit if necessary, so that the image signal can be accurately corrected. This has the advantage of allowing correction.

[Brief explanation of drawings]

1 and 2 are block diagrams of conventional image signal processing circuits, and FIGS. 3A and 3B are diagrams showing the positional relationship between pixels and thin lines.
Figures 4A and B are Figure 3A. Condition B? FIG. 5 is a circuit diagram of a contrast enhancement circuit to which the first invention of the present application is applied; FIG. 6 A, B; FIGS. 7 A, B; and FIG.
Figures A and B are diagrams showing input image signal waveforms and output image signal waveforms in the circuit of Figure 5, and Figure 9 is a block diagram of a signal processing circuit equipped with a contrast enhancement circuit to which the image signal processing method of the present invention is applied. 10th factor is a circuit diagram of a contrast enhancement circuit to which the second invention of the present application is applied. FIG. 11 is a diagram showing the relationship between pixels and image signals. FIGS. 12A to 12C are image signal clocks and gate signals. FIGS. 13A to 13E are timing charts showing the relationships among write signals, gate signals, and selector selections. 11, 13, 23...Flip-flop, 12
, 14...Memory, 15-...Counter, 16.17
-・Line mesori, 18...-Selector, 21.22・
··driver. Mockery η [1''' (Ku Sushikan Hiroshi 1q Ori Kazumachi PCK Spear/('/ @ 5011-) Q〆J-/l-/ 1 Kuno0 Kuno00LLl

Claims (1)

[Claims] α) An image processing method for processing a group of digital image signals that calculates the density of each pixel in an image to obtain a group of digital image signals with enhanced contrast, comprising: When the digital image signal that changes the density is go, and the digital image signal that changes the density of the pixel adjacent to the target pixel is f, the values of the digital image signals go and f are used as addresses, -g1=a4 (] +b) go bf ) (However, a,
b is a positive constant) is written into a memory, and the density of the target pixel is set to 1 using the value gM of the output image signal read from the memory. (2) Calculate the density of each pixel in an image In an image processing method that processes a digital image signal group to obtain a digital image signal group with enhanced contrast, calculate the density of the target pixel. Let go be the digital image signal to calculate the density of a plurality of pixels around the target pixel, and let nfl#f and ``'frl be the digital image signals to calculate the density of a plurality of pixels around the target pixel, respectively.
−1, fn, first the digital image signal g
Using the values of o and f as addresses, gI = am ((1+b+)go b+L) (however, [1,
al, bl are positive constants) is written into the first memory, the output image signal g1 is read out from the first memory, and then the output image signal g1 and the digital image signal f, Using the value of as an address, fold the value gtaki given by go = &鵞4 (1+b鵞)go bmft) (where am l b is a positive constant) into the second memory, and Read the output image signal gt from the memory, repeat the same operation one after another, and finally, using the output image signal gn-1 read from the n-1th memory and the digital image signal fnO value as an address, gH=an ((1+b n ) g II 1 b
n f n ) (where an and bn are positive constants) Write the value gn to the n-th memory VC, and calculate the density of the target pixel using the output image signal gn read from the n-th memory. An image signal processing method characterized by hail. (3) In an image signal processing method that processes digital image signal groups that calculate the density of each pixel in an image and obtains a digital image signal group with enhanced contrast, 1. Calculate the density of the target pixel. Let the digital image signal be go, and divide the digital image signal to calculate the density of multiple pixels around the target pixel, respectively f 1 + ' 2 +
``'' nl 1 f n, the value of the digital image signal gO+fl*fl・''fn,, fn is used as an address, and gn=an((1+b++b,+・'・bn,+bn
)g(1b+f11)zf*-...-bn-1'n-
1"nfn) (where an, b, , b, -b are positive constants) is written in the memory, and the value gn of the output image signal read from the memory is used as the above value. An image signal processing method characterized by determining the density of a target pixel.