JPS6087572A - Picture signal binary coding system - Google Patents

Abstract

PURPOSE:To improve the resolution and also to decrease generation of noise on a picture image by keeping a threshold value signal level when an absolute value of a difference between a picture signal level to a preceding picture element and the threshold signal level is a fixed value or below. CONSTITUTION:A digitized picture signal (a) and an output of a counter 4 [the output becomes a threshold signal (d)] are given to a decoder 3, and outputs of the decoder 3 are inputted to the counter 4 as an up-down control signal (g) and an enable signal (h). The decoder 3 is switched properly depending on the relation of the picture signal (a) and the threshold value signal (d) being two input signals in this case, and when the absolute value of the difference between the level of the picture signal to the preceding picture element and the level of the threshold value signal is a fixed value or below, the level of the threshold signal to the present picture element is kept to the same level as that to the preceding picture element and the threshold value signal is changed only as required.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to an image signal binarization method for binarizing an image signal obtained by reading an original in a facsimile machine, an IJ device, etc. into black and white binary values. .

Configuration of a conventional example and its problems Figure 1 shows a block diagram of a device that performs a conventional image signal binarization method. 01 is a comparator, 2 is a threshold generation circuit, and one input of the comparator 1 is The image signal d is inputted to the other input of the comparator 1, and the threshold value of a constant level outputted from the threshold value generation circuit 2 is inputted to the other input of the comparator 1. Then, the comparator 1 outputs the comparison result between the image signal a and the threshold value S as a binarization signal C, thereby performing binarization processing.

By the way, corresponding to the pixel pitch of the image sensor, 1
When an image sensor reads the butter 7 of a document whose black and white color changes for each pixel, a phenomenon occurs in which the contrast of the photoelectric conversion output of the image sensor decreases due to a phase shift between the document and the sensor pixels and the resolution limit. .

However, in the conventional binarization method, depending on the size of the threshold value, it is difficult to faithfully decompose the document pattern into white and black binary code C, as shown in FIG. Met.

That is, in FIG. 2, a' is an analog representation of the digitized image signal ai, and b.

b', b'' indicate three types of threshold values, c, a', c''Fi, and z) binarization codes for the aforementioned threshold values, l, /, b//. As is clear from the figure, the binarization code C for the threshold value does not faithfully reproduce black and white in both area A of thin black lines on white ground and area B of thin white lines on black ground in the manuscript. .

Further, the binarization code C' for the threshold value b' is reproduced for the area A, but not for the area BK.

Furthermore, the binary symbol C'' for the threshold value b'' is
However, region A is not reproduced.

In this way, for image signals with reduced contrast, there is an optimal threshold value depending on the pattern of the image signal.
The conventional binarization method using -C with a fixed threshold value has the disadvantage that it is not possible to obtain high-resolution black and white images for all patterns.

To solve the above-mentioned problem, refer to the binarization code for the previous pixel, and if it is white, the threshold value is raised by a predetermined level -, ΔV, while if it is black, it is lowered by the threshold value tΔV. can be dealt with.

However, in that case, the threshold value always changes by either increasing or decreasing, so when the image signal and the threshold value are close to each other, black and white appear alternately in the binarization signal, and the image When viewed as an image, a new problem arises in that the image has noticeable noise.

OBJECT OF THE INVENTION The present invention solves the above-mentioned conventional problems, and even for image signals with reduced contrast, a binary code that converts the same image signal into a binary signal has a sufficiently high resolution and can be converted into black and white. An object of the present invention is to provide an image signal binarization method that can suppress the occurrence of noise on an image by reproducing n and L signals.

Structure of the Invention The image signal binarization method according to the present invention usually refers to the level of the image signal and the level of the threshold signal for the previous pixel, and changes the level of the threshold signal for the current pixel depending on the magnitude relationship between them. When the absolute value of the difference between the level of the image signal for the previous pixel and the level of the threshold signal is less than a certain value, the above purpose is achieved by making the level of the threshold signal for the current pixel the same as that for the previous pixel. This is what we aim to achieve.

Description of examples Hereinafter, the present invention will be explained based on embodiments shown in the drawings.

FIG. 3 shows a block 1 of an apparatus for realizing an image signal binarization method according to an embodiment of the present invention, where 3 is a decoder and 4 is a decoder.
is an up/down counter, and 5 is a comparator.

The decoder 3 receives the digitized image signal a and the output of the counter 4 (this output becomes the threshold signal d). A clock signal f is input. counter 4
The up/down control input and enable input of
The output of the decoder 3 is input as the up/down control signal q1 enable signal. As a result, the following three modes are available for the decoder 3 to control the counter 4. However, in the following, it is assumed that the enable is on at h=1, up at q-1, and down at q-0.

M-1: Holding mode h = O, g = don't'
careM=2: addition mode h==1. g=0M-
3: Subtraction mode h=1. g=0 The decoder 3 switches the following modes as appropriate based on the relationship between the image signal a and the threshold signal d, which are input to the decoder 302, and the algorithm is as follows. .

That is, the upper limit level and lower limit level of the threshold signal d are
If suppressed to Vmax and Vmin, respectively, (i) d
<When Vmin, M=2(!t) d>V
When max, M=3(iil) d=Vmin
When 01a-dl≧L (LiJ constant value) and a
) In the case of d, M = 2 ■ In other cases, M = 1 (iv) When d = Vmax, ■ If 1a-dl≧L and a < d, M23 ■ In other cases , M=1 (V) When Vmax > d > Vmin, ■1
If a-dl≦L(1), 1ld=1■1a-dl≧L
Then, if a) d, then M=2. If 1a-di≧L, and if a<d, then M=3.

A change in the output of the counter 4 (ie, the dark value signal d) due to such a difference in mode occurs at the rise of the clock signal amount. Therefore, with reference to the pixel signal a and the threshold signal dv for the previous pixel, the threshold signal d for the next pixel (current pixel) is determined as needed as follows.

That is, the level of the image signal a for the previous pixel VC is set as ao,
When the level of the threshold signal d for the previous pixel is do, and the level of the threshold signal d for the current pixel is dl, (1) When do<Vml n, d1 is raised from d0 by a predetermined level ΔV.

(II) When do>Vmax, d1'i5d. The voltage is lowered by a predetermined level ΔV.

Qll) When d0=Vmin, ■l When ao-d01≧L and ao>do, d
l id. It is increased by a predetermined level ΔV.

■Otherwise, keep dl at the same level as do.

(IV) do=Vmax (When D, ■l ao-
If do l≧L and ao<d o, then d1'
id. The voltage is lowered by a predetermined level ΔV.

■Otherwise, keep dl at the same level as do.

(V) When Vmax > do > Vmin O, ■
1ao-do If 1≦L, dlid. be maintained at the same level.

■l ao-do When l≧L and ao>d o, dl is increased by a predetermined level ΔV above do.

■l ”o-do 1.: qL, and a o< d
If o, dl is d. The voltage is lowered by a predetermined level ΔV.

The threshold signal d and image signal a obtained as described above are
Comparator 5 is manually operated. Then, the comparator 6 outputs the comparison result as a binary code e.

As a result, if the binarization code is e, a) if d, white, a〈
If it is d, it is black.

Figure 4 is a waveform timing diagram showing an example of binarization using this method, in which al (solid line) and d' (broken line)
are the image signal a, which is a digital signal, and the threshold signal d ((shown in analog form). In the same figure, the binarization signal e is a process in which the low level is black, and the high level is white. As is clear from the figure, with this method t''L, areas A and B, which could not be reproduced with the conventional image signal binarization method using a fixed threshold value, can be reproduced.
Together they can reproduce black and white.

In addition, in experiments in which actual reading and recording were performed, it was confirmed that according to this method, images with high resolution and low noise generation can be obtained.

Effects of the Invention According to the present invention, when the absolute value of the difference between the level of the image signal for the previous pixel and the level of the threshold signal is less than a certain value, the level of the threshold signal for the current pixel is set to be the same as that for the previous pixel. Since the level is maintained at the same level and the -value code is changed only when necessary, even for an image signal with reduced contrast, the binarization code that binarizes the same image signal,
If the resolution is sufficiently high and black and white is reproduced, an excellent effect can be obtained in that the occurrence of noise on the image can be suppressed to a minimum.

[Brief explanation of the drawing]

Figure 1 is a block diagram of a device that realizes the conventional image signal binarization method, and Figure 2 shows the signal 2 by the conventional image signal binarization method.
A waveform Tabeminck diagram showing an example of digitization, FIG. 3 is a block diagram of an apparatus for realizing an image signal binarization method according to an embodiment of the present invention, and FIG. 4 is an example of signal binarization using the actual thickness example. FIG. a...Both signals, d...Threshold signal.

Claims (1)

[Claims] The level of the image signal for the previous pixel is ka□, and the level of the threshold signal for binarizing the image signal for the previous pixel is d. , the level of the threshold signal for the current pixel d1
In principle, la. -d01≦L (L is a constant value), hold dl at the same level as do, l ao-clol > L, and a. >do, raise dl by a predetermined level from do, and ao<d o, d, 1 id. An image signal binarization method that lowers the image signal by a predetermined level.