JPS6226631B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides two
Concerning value determination methods and devices.

An analog image signal obtained by a reading device in a facsimile machine or the like always contains a DC component, and furthermore, this DC component includes a component caused by the density of the background of the original. However, in conventional facsimile devices, the analog image signal containing the DC component is compared with a threshold value fixed at a certain level to perform binary discrimination, and then converted into a binary digital image signal. was.

Therefore, if the background is not white, such as a diazo (blueprint) original, you will not be able to obtain a good received image unless you perform density correction for the background. The disadvantage is that a separate circuit must be provided for this purpose.

In addition, with only a circuit that performs density correction for the background of the original, there is a problem with the drift of the fluorescent lamp used as the light source of the reading device. The problem is that the white level of the analog image signal output from the reading device corresponds to the halftone level after stabilization, and if left as it is, the background of the received image will become black. However, there was also the drawback that a density correction function for this purpose had to be provided separately.

The present invention has been made in order to eliminate the above-mentioned conventional drawbacks, and an object of the present invention is to provide a binary discrimination method and device that do not require separate provision of the various density correction circuits. That is, the present invention creates a discrimination signal by removing the direct current component originally contained in the analog image signal to be discriminated into two values, and compares this discrimination signal with an appropriate threshold. , the analog image signal to be determined into two values is determined into two values.

The present invention will be described below based on embodiments shown in the drawings.

FIG. 1 shows an embodiment in which the present invention is applied to a binary discrimination device for a facsimile machine, and FIG. 2 shows waveforms of the main parts. An analog image signal input terminal 1 to which an analog image signal a is input is connected to one terminal of a capacitor 2, and the other terminal of this capacitor 2 is connected to an in-phase input terminal (+) of an analog comparator 3.
One terminal of resistor 4, the cathode of diode 5,
It is connected to the cathode of diode 6 and the anode of diode 7.

The other terminal of the resistor 4 is grounded. The anode of the diode 5 is supplied with a voltage of an appropriate level from a power supply connection terminal 8 via a volume 9. The anode of the diode 6 and the cathode of the diode 7 are both grounded via a capacitor 10, and the common connection point of these diodes 6, 7 and the capacitor 10 is connected to the negative phase input terminal (-) of the analog comparator 3 via a resistor 11.
Cathode of diode 12 and diode 13
connected to the anode of the

The anode of the diode 12 is supplied with an appropriate level of voltage from the power supply connection terminal 14 through the volume 15, and the cathode of the diode 13 is supplied with an appropriate level of voltage from the power supply connection terminal 16 through the volume 17. be done. 18 is an output terminal of the analog comparator 3.

Next, the operation of this embodiment will be explained with reference to the operation waveform diagram shown in FIG.

In FIG. 2, the portions , , , of the analog image signal a are read by a reading device (not shown) in the original document (which has a background density at the level indicated by Va [V] in column a of FIG. 2). ) shows the waveform generated when reading the following part.

Part...A part where multiple dark and thick lines exist at wide intervals...A part where multiple dark and thin lines or multiple thin and thin lines exist at narrow intervals...A part where multiple dark and thick lines exist at narrow intervals... A part where lines are closely spaced...A part where dark and thin lines or thin, thick and thin lines are widely spaced.Now, first, connect the power supply connection terminals 14 and 16, the volumes 15 and 17, and the diode. 12 and 13 do not exist, when the analog image signal a as described above is input to the analog image signal input terminal 1, the DC component of the analog image signal a is removed by the capacitor 2, and the second A signal b as shown in column b of the figure is generated at both terminals of the resistor 4. but,
Diode 5 is connected to the connection point between capacitor 2 and resistor 4.
is connected, the signal b is regenerated by direct current, and the discrimination signal d as shown by the solid line in column d of Fig. 2 is generated.
is the common-mode input terminal (+) of analog comparator 3
is input. Note that in column d, the signal b is also shown with a broken line for ease of comparison.

The DC level of the discrimination signal d is the voltage Vb shown in column c of FIG.
The voltage is obtained by subtracting the forward voltage of the diode 5 from [V], but in this embodiment, the voltage Vb is made equal to the forward voltage of the diode 5, so the DC level of the discrimination signal d is 0. [V].

Further, the discrimination signal d is converted by the functions of the diodes 6 and 7 and the capacitor 10 into a signal e as shown by a solid line in column e of FIG. Note that in column e, the signal d is also shown with a broken line for ease of comparison. The signal e is generated at a common connection point between the diodes 6 and 7, the capacitor 10, and the resistor 11, and is inputted to the negative phase input terminal (-) of the analog comparator 3 via the resistor 11.

Although a detailed explanation of the operation of converting the discrimination signal d into the signal e as described above will be omitted, two types of intersections W and X occur between the signals d and e. That is, one intersection point W occurs when the discrimination signal d rises to a predetermined level from the minimum value of the trough of the waveform, and the other intersection point X occurs when the discrimination signal d rises to a predetermined level from the maximum value of the peak of the waveform. Occurs when falling. While the discrimination signal d goes from the intersection point W at the rising edge to the intersection point X at the falling edge, the discrimination signal d is larger than the signal e; While heading toward the vehicle, the signal e is larger than the discrimination signal d.

Since the signals d and e described above are input to the in-phase and anti-phase input terminals of the analog comparator 3, respectively, the intersection points W and X act as trip points of the comparator 3, and the output terminals of the comparator 3 18, a digital image signal is obtained by binarizing the discrimination signal d using each intersection point W, X as a threshold value.

Here, the discrimination signal d is obtained by removing the DC component from the analog image signal a (this DC component includes the density level Va [V] of the background of the original) and performing DC reproduction. , to which an appropriate DC level has been applied, so if a binarized digital image signal is obtained by comparing the discrimination signal d with an appropriate threshold value as described above, the density relative to the background of the original can be adjusted. There is no need to perform correction. Furthermore, for the same reason, there is no need to perform density correction for the fluorescent lamp drift problem.

Furthermore, as in this embodiment, instead of using only one fixed threshold value as in the conventional binary discrimination method, two types of threshold values are used whose level automatically changes as described above according to the discrimination signal d. By comparing the threshold value (intersection W, X) and the same signal d, the same signal d is
If it is converted into a value, the resolution can be improved as will be explained in detail later, and all parts of the analog image signal a can be reproduced as the received image temperature.

However, as it is, even if the reading device reads a part of the document that has a slightly higher density than the background, such as very thin dirt on the background of the document, the intersection point W at the rise point will be occurs,
There is a possibility that this portion will appear black in the received image, resulting in a very ugly received image.

In addition, the reading device includes a photodiode array,
If a CCD array is used,
Due to the bit dispersion of these elements, the analog image signal a always has a small level fluctuation at any level. Furthermore, even if parts of a document, such as lines, appear to have uniform density to the naked eye, there are some density irregularities when viewed at the pixel level. The analog image signal a has a small level fluctuation. There is a possibility that such small level fluctuations in the analog image signal a may cause the received image to appear black or white, resulting in a very unsightly received image.

In this embodiment, these problems can be solved by supplying a bias voltage from the power supply connection terminal 14 to the negative phase input terminal (-) of the analog comparator 3 via the volume 15 and the diode 12. ), and by connecting the power supply connection terminal 16 to the negative phase input terminal (-) of the analog comparator 3 via the volume 17 and diode 13, the same negative phase input terminal (- ) This problem is solved by also setting an upper limit on the signal voltage input to the terminal.

That is, if the voltage at the cathode of the diode 13 is set to Vc [V] shown in column f of FIG. 2, and the voltage at the anode of the diode 12 is set to Ve [V] shown in column g of FIG. , the signal h generated at the negative phase input terminal (-) of comparator 3
The voltage is the voltage Vf [V] obtained by subtracting the forward voltage Vb' [V] of the diode 12 from the Ve [V].
If the diode 12 becomes conductive, and Vf [V] is supplied to the negative phase input terminal (-) through the diode 12, the signal h will never become below Vf [V]. do not have. Also,
When the signal h tries to become higher than the voltage Vd [V] which is the sum of the forward voltage Vb' [V] and the Vc [V],
Since the diode 13 becomes conductive, the negative phase input terminal (-) of the comparator 3 is connected to the same diode 13.
is maintained at the anode voltage Vd [V], and the signal h
never exceeds Vd [V].

Furthermore, when the signal h is between the voltages Vf and Vd, the diodes 12 and 13 are both non-conductive, so the signal h becomes the same as the discrimination signal e.

Therefore, the signal h has a waveform as shown by the solid line in the column h of FIG. 2 (the broken line in the column h indicates the discrimination signal d). Note that by selecting an appropriately high resistance value for the resistor 11, the DC component supplied or maintained via the diodes 12 and 13 as described above will affect the common-mode input terminal (+) of the analog comparator 3 and the capacitor 2. It is designed to prevent

Since the signal h has the waveform as described above, there is a gap between the discrimination signal d and the signal h, as shown in column h of FIG. 2, when the discrimination signal d rises to the bias voltage Vf [V]. intersection point Y that occurs when the discrimination signal d reaches the bias voltage Vd [V] at the time of falling; intersection point Z that occurs when the discrimination signal d reaches the bias voltage Vd [V] at the time of falling; Four types of intersections, W and X (however, only those at the level between intersections Y and Z) occur.
Therefore, the analog comparator 3 converts the discrimination signal d into a digital image signal i using these intersections W, X, Y, and Z as trip points.

The minimum density of the document at which the received image is black is determined by the intersection point Y whose level is fixed, and all portions of the discrimination signal d whose level is lower than this intersection point Y become white. This intersection point Y can be set to any level using the volume 15 depending on the document. Also, the maximum density of the original where the received image changes from black to white is:
The level is determined at the fixed intersection Z, and the portion of the discrimination signal d whose level is higher than this intersection Z is
All will be black. This intersection point Z can be set to an arbitrary level using the volume 17 according to criteria described later.

Therefore, as mentioned above, there are the disadvantages that even the extremely low density of the original is reproduced as black in the received image, and the disadvantage that small level fluctuations in the analog image signal a are reproduced as black or white in the received image. can be prevented.

Also, if the level of the intersection Z is determined appropriately,
By providing the same intersection point Z, the resolution of the facsimile apparatus as a whole does not deteriorate.

Next, the criteria for determining the level of the same intersection point Z will be explained.

In a facsimile machine, an analog image obtained when a reading device that reads with only one photoelectric conversion element such as a photodiode reads a white eyelid line (crossing perpendicular to the reading direction) in a black area. When looking at the image signal, if the width of the eyelash white line is gradually made thinner from a thick line, the analog image signal will go from the eyelash white level to the halftone level and reach a point that is almost the same as the eyelash black level. changes up to.

If the signal e whose signal level has no upper limit is input to the negative phase input terminal (-) of the analog comparator 3, even if the white line becomes quite thin, the output terminal 18 of the comparator 3
The digital image signal obtained from the image reproduces the white eyelash line. However, since the resolution of the facsimile receiver is finite, as the eyelash white line becomes thinner and thinner, even though the digital image signal itself reproduces the eyelash white line, the resolution of the facsimile receiver becomes smaller. For this reason, there is a limit to which the white eyelash line will not be reproduced in the received image.

Therefore, it is meaningless to reproduce the eyelash line in a digital image signal to a thinness that is higher than the resolution of the facsimile receiver. It is sufficient if it can be reproduced in the image signal.

According to this embodiment, the limit of the thinness of the eyelash white line that can be reproduced in the received image due to the resolution of the receiver is that the level of the analog image signal a when the eyelash white line is read is The level was approximately 50% of the eyelash white level (minimum value) to the eyelash black level (maximum value) of the image signal a.

In addition, if the reading device has multiple photoelectric conversion elements such as a photodiode array, even if you read the eyelid white line (intersecting perpendicularly to the reading direction) in the black area as described above, the photoelectric conversion elements will not be detected. Conversion element is 1
Unlike the case where the analog image signal changes to almost the same level as the eyelash black level, there are very few cases where the analog image signal changes to almost the same level as the eyelash black level. Most of them are.

Furthermore, when looking at an analog image signal obtained when a reading device reads multiple black lines that have the same width and interval between each line and intersect perpendicularly to the reading direction, the width and interval of each line are the same. As long as is thick, the analog picture signal becomes a rectangular wave with the same cycle as each line, but as the width and interval of each line gradually becomes thinner, the analog picture signal changes to a sine wave with the same cycle as each line, Moreover, the amplitude gradually decreases, and eventually there is no change in the amplitude and it reaches a constant level. The rectangular wave and the sine wave change around the constant level.

Therefore, if the trip point of the analog comparator 3 is generated at the constant level, the plurality of lines can be best reproduced in the digital image signal.

According to this embodiment, the certain level is about 50% of the eyelash white level to the eyelash black level of the analog image signal a, similar to the case where the eyelash white line can be reproduced. .

For the above reasons, by setting the intersection point Z to the level of approximately 50%, it is possible to prevent small level fluctuations in the analog image signal a from being interpreted as black or white in the received image as described above. At the same time, even if the same intersection point Z is provided, there will be no problem such as a decrease in resolution, and lines in the received image will become significantly thicker, thinner, or not reproduced compared to the lines in the original, or multiple lines in the original will not be reproduced. It won't collapse into one piece.

Therefore, as shown in the received image column of Fig. 2,
All of the above portions of the analog image signal a can be reproduced.

Additionally, conventionally, when using multiple photoelectric conversion elements such as photodiode arrays in a reading device,
Since it was necessary to select and use a device with little bit variation, the yield was poor and the cost was high. However, according to this embodiment, even a device with a lot of bit variation can be used, so it is possible to perform photoelectric conversion. The cost of the element can be reduced.

Next, the improvement in resolution achieved by this embodiment will be explained in more detail.

In Figure 3, the reading device detects a dark and thick line A, a dark and thin line B, a dark and extremely thin line C (thinner than the size of one dot), a thin and thick line D, a thin and thin line E, and a thin and extremely fine line. 3 shows a signal conversion state and a received image according to the conventional binary discrimination method when reading a document in which a line F exists. As is clear from the figure, in the conventional binary discrimination method, an analog image signal is converted into a binary digital image signal only by determining the level indicated by the broken line T relative to a fixed threshold value. However, even if the density of each line of the original is the same, if the thickness of each line is different, it may or may not be reproduced in the received image (if it is thin, it will not be reproduced).

FIG. 4 shows waveforms and received images in this embodiment when the reading device reads the same document as in FIG. 3. As is clear from this figure, according to this embodiment, the discrimination signal d is set using the intersection points W, X, whose level automatically changes according to the discrimination signal d, and the intersection points Y, Z, whose level is fixed, as thresholds. Since it is binarized, all the lines A to F can be reproduced in the received image, and the above-mentioned drawbacks of the conventional method can be solved.

FIG. 5 shows each waveform and a received image in this embodiment when a document in which a plurality of thin lines are present at narrow intervals and the lines have four different densities is read.

In the conventional binary discrimination method, an analog image signal is converted into a binary digital image signal based only on the level relative to a single fixed threshold as described above. There have been disadvantages in that lines cannot be reproduced in the received image, and thin lines lined up at close intervals are turned black, but in this embodiment, for the reasons mentioned above, lines of all densities are reproduced as shown in Fig. 5. can be reproduced, and the above-mentioned conventional drawbacks can be solved.

Further, according to this embodiment, the levels of two of the four types of trip points of the analog comparator 3 automatically change according to the analog image signal a, and the levels of the remaining two types of trip points are fixed. However, since subtle changes in the setting level do not have a large effect on the received image, it is easier to adjust the circuit, and there is no need to make detailed adjustments for each individual document. is obtained.

In addition, once the standard is determined when using the circuit of the above embodiment, volumes 9, 15, 17
Instead, it is also possible to use a voltage divider circuit with fixed resistors to completely fix the DC reproduction level of the analog image signal a and the levels of the intersections Y and Z. In this way, there is no need to make any adjustments at all. become.

Further, in the embodiment shown in FIG. 1, after removing the DC component from the analog image signal a, DC regeneration is performed, and by comparing the discrimination signal d resulting from the DC regeneration with a threshold value, the analog image signal a is However, in the present invention, it is not necessary to perform DC reproduction, and by directly comparing the signal obtained by removing the DC component from the analog image signal with a threshold value, the analog image signal is determined to be binary. It may be determined that However, if DC reproduction is not performed in this way, the reproduction of thin lines or thin lines on the original may deteriorate, although the probability is small; however, if DC reproduction is performed as in the above embodiment, Such inconveniences can be prevented.

Furthermore, in the above embodiment, as thresholds for binary discrimination, two types of thresholds (intersection points W, However, in the present invention, any threshold value may be used for binary discrimination, and even if only one threshold value with a fixed level is used as in the past, good.

However, if the above embodiment is adopted, the resolution can be improved as described in detail above.

As described above, the binary discrimination method and device according to the present invention creates a discrimination signal by removing the DC component originally contained in an analog signal, etc., and compares this discrimination signal with an appropriate threshold. Since the analog signal is discriminated into binary values, there is no need to separately provide a density correction circuit for the background of the original or a density correction circuit for the drift problem of the fluorescent light that is the light source of the reading device, and it is also possible to improve the resolution. It is possible to obtain this excellent effect.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of an analog signal binary discrimination device according to the present invention, FIG. 2 is a waveform diagram of main parts showing the discrimination operation of the embodiment, and FIG. 3 is a conventional binary discrimination method. FIGS. 4 and 5 are waveform diagrams showing other discrimination operations of the embodiment, respectively. 2... Capacitor, 3... Analog comparator, 4... Resistor, 5, 6, 7... Diode, 9
...Volume, 10...Capacitor, 12,1
3...Diode, 15, 17...Volume.

Claims (1)

[Scope of Claims] 1. Discrimination signal generation means for creating a signal from which a DC component contained in the analog signal is removed from an analog signal to be discriminated into binary values; Threshold generation means outputs a signal obtained from the voltage of the capacitor charged and discharged by the output of the discrimination signal generation means as a threshold, and the output of the discrimination signal generation means is compared with the threshold, and a binary value is generated according to the comparison result. 1. A binary discriminator, comprising a comparison means for outputting either one of the outputs, and a voltage limiting circuit for limiting fluctuations in the voltage of the capacitor within a certain range within the threshold value creation means. 2. The discrimination signal generation means includes a DC component removal circuit that removes a DC component from an analog signal to be discriminated into binary values, and a DC regeneration circuit that regenerates the output of this DC component removal circuit into DC and outputs it to the comparison means. A binary discriminator according to claim 1, comprising: