JPS5829910B2 - Image signal correlation processing method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for converting line-sequentially scanned analog image signals such as facsimile signals into binary digital signals using a threshold value.

Conventionally, this type of system generally converts an input signal into a binary signal using a fixed threshold value.

FIG. 1 shows the state of signal conversion in this case.

The first scanning signal A is compared with a threshold value l□ and binarized as shown in C, and the second scanning signal B is compared with a threshold value l□.
It is compared with , and is binarized as D.

The problems in this case are the following two points.

The first is that when the input signal portions AO and BO are sampled by the clock E and binarized, the output waveforms Co and DO are shown. Even if the temporal positions of the signals AO and BO vary slightly, that variation is magnified to a variation of one sampling interval, resulting in jitter that fluctuates at both ends of the shaded areas of CO and Do in the figure for each scanning line. .

The second problem with the conventional method is the input signal portion Bl.

Dl when the maximum or minimum value of the signal amplitude is at a level near the threshold value as shown by B2.

The problem is that the output waveform shown in B2 cannot be stably reproduced.

That is, as shown by the hatched portions of DI and D2 in the figure, Dl changes from 1" to O", and D2 changes from 0" to 1".

B1. The signal B2 is a signal that can always be generated if line sequential scanning is performed with a constant aperture, and it is desirable that the above-mentioned fluctuation does not occur in this portion.

From this point of view, the present invention refers to the binarized output value of the pre-sampling point immediately before the target sampling point of interest and the binarized output value of the corresponding sampling point one scanning line before. Judging from the correlation of The present invention provides an image signal correlation processing method that avoids

FIG. 2 is an explanatory diagram of the image signal correlation processing method according to the present invention.

A indicates the relative position of nearby sampling points a*E)sCsd with respect to any target sampling point X of interest, and when binarizing at the sampling point of X, a, b, c, d At each sampling point, an output binary signal has already been determined.

Therefore, by controlling the binary determination at the X point from these already binarized signals in consideration of the signal correlation, more accurate binarization can be achieved.

Figure 2 B, C, D, and E are sampling points a=d, respectively.
It shows various states of binary output of , and the shaded area is 1
'', the blank area is assumed to output nOn, and the * area may be either 1'' or ``0''.

In FIG. 2B, sampling points a, b of FIG. 2A,
In this case, the previous scan line of X changes from 1'' to O'' at the sampling point C.

Therefore, in order to reduce the jitter at the X point, it is sufficient to similarly make it easier to change from 1'' to 0''.

For this purpose, the threshold value for binarization may be slightly increased from the standard state.

Similarly, in C of the same figure, at the sampling point C,
Since the threshold value is changing from "0" to 1", the threshold value may be slightly decreased from the standard state in order to facilitate the change from "0" to 1".

Next, consider the same case.

In this case, the sampling point b a Cs in FIG. 2A
d, all previous scanning lines for X are 1'', and it is assumed that X is a signal that should be O as shown in FIG. 1 B2.

In this case, if it is made easier to generate 0'' by increasing the threshold value at point X, it becomes easier to detect a signal with an attenuated amplitude response.

That is, it becomes easier to set the signal as shown in FIG. 1B2 to an output value different from the previous scanning line output value, and the loss of information due to binarization is reduced.

Similarly, in FIG. 2E, the threshold value may be decreased at point X.

Expressing this relationship in a logical formula, the threshold value Vt for binarizing the X point is: b c d + a b c = 1 and vt = vto + △Vt;
V to - ΔVt1 Otherwise, V t =V to may be used.

(2) to is a threshold value under standard conditions, and is usually selected at the center of the maximum and minimum signal amplitude values.

ΔVt is the amount of change in the threshold value, and is a parameter that controls the strength of correlation.

For the purpose described above, ΔVt preferably has a value of about o, i or 0.2 times the maximum signal change width.

That is, the threshold value is 0.4-0.6 or 0.3-0.
It changes to 7.

a, b, c, d are sampling points a, b, c, respectively.
The binary outputs et and y5 of d, a, b, τ, and i each indicate "O".

FIG. 3 is a block diagram showing one embodiment of the present invention.

The analog image signal inputted from the input terminal 11 is compared with a threshold value Vt by the comparator 12, binarized, and outputted to the output terminal 13.

This output is further connected in cascade to shift registers 14 to 14.
I am led to 17.

The shift registers 14, 16, and 17 each provide a 1-bit delay, and the shift register 15 provides an N-2 bit delay for the number of samples N for one scanning line. Driven by a clock.

Each output of shift register 14, 15° 16.17 a s
b s Cs d corresponds to the aforementioned sampling point,
bcd+a at AND circuit 21, 22 and OR circuit 23
bc is calculated, AND circuits 24, 25 and OR circuit 2
6, bcd+abc is calculated.

These calculation outputs are supplied to an analog subtraction circuit 27, where they are subtracted by setting the human power "1" to voltage V and "Ojt" to voltage 0, thereby generating the threshold control voltage ΔVt.

This threshold control voltage is the threshold value V of the comparator 12.
t is varied, and the correlation processing operation described with reference to FIG. 2 is performed.

Note that in the above, the threshold value is controlled by referring to the binary outputs of sampling points a to d, but it is also possible to refer to the binary outputs of more sampling points, or only the four points and point C. Furthermore, addition or subtraction of ΔVt is not limited to the above logical formula, but may be other logical formulas.

As explained above, according to the present invention, by controlling the threshold value, analog image signals can be binarized with good jitter characteristics and detection characteristics of signals with attenuated amplitude responses. It is effective as a binarization method in a transmitter case or a receiver.

[Brief explanation of drawings]

Fig. 1 is a waveform diagram for explaining the signal conversion of the conventional method of binarizing an analog signal, Fig. 2 is a diagram for explaining the principle of the image signal correlation processing method of the present invention, and Fig. 3 FIG. 1 is a block diagram showing an embodiment of the inventive system. 11: Input terminal, 12: Comparator, 13: Output terminal, 14, 16.17: 1-bit shift register, 15:
N-2 bit shift register, 18: clock input terminal, 27: subtraction circuit.

Claims (1)

[Claims] 1. In a method in which an analog image signal scanned line-sequentially is compared with a threshold value in a comparator and the analog image signal is converted into a digital signal using this threshold value, for one target sampling point of interest, a first binary output value of the preceding pre-sampling point, a second binary output value of the sampling point one scanning line previous corresponding to the pre-sampling point, - the target sampling of the previous scanning line; With at least reference to the third binary output value of the sampling point corresponding to the point, the logical value "1" and the logical value "Ou of the first, second, and third binary output values are respectively a s b s
For C, a, b, and c, a threshold value changing means is provided for increasing the threshold value from the standard state at abc"1 and decreasing the threshold value from the standard state for abc"'1. Image signal correlation/processing method.