JPS59148465A - Quantizing circuit of picture signal - Google Patents

Abstract

PURPOSE:To attain the quantization of an analog picture signal emphasizing the intermediate tone reproducibility by connecting a resistance between the intermediate tap of a reference voltage divider constituting a parallel comparison type A/D converter and a tap giving a lower limit reference voltage. CONSTITUTION:The intermediate tap M is led from the reference voltage divider 1 constituting the parallel comparison type A/D converter together with a high speed comparator group 2 and an encoder 3 and a resistance Rc is connected between the tap and the tap L to which the lower limit reference voltage VREF2 is given. A resistance value by half of the full resistance between the tap L and the tap H to which the upper limit reference voltage is given is set to be a resistance value between the taps M and L. The picture signal quantizing circuit is obtained, from which a reproduced picture with high quality in which the intermediate tone at the black color surplus part is emphasized is obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION Technical Field The present invention relates to an image signal quantization circuit that performs multi-element conversion by converting an analog image signal into a digital image signal with an index bit using an AD converter.

[Prior Art General] As shown in Fig. 1, a quantization circuit for the image signal of Φ consists of a reference voltage divider l consisting of 21 ladder resistors, and the outputs of each of the dividers and the analog image signal. A group of 21-1 high-speed comparators 2 to be compared, and a group of n bits 2 according to each comparator output.
Using a parallel ratio flexible AD converter consisting of an encoder 3 that creates a digital signal based on a base code, the upper limit reference 1 VRKFI and the lower limit reference voltage VRKF2 are used.
The analog picture signal f: is converted into an n-bit digital one-page signal g.

However, such conventional lI'! In the quantization circuit of No. 1i (2V), both of the multi-1 direct digitized signals are binarized by dither processing to express halftones (44
In order to emphasize the contrast, the difference between the upper and lower limit reference voltages in the AD converter is reduced to reduce the A of the analog image signal.
If the D conversion range is narrowed, the range for expressing midtones will be narrowed, resulting in poor reproducibility of photographs, etc. On the other hand, in order to improve the reproducibility of midtones, [if the range of ADf conversion is widened, There is a problem in that it is no longer suitable for quantizing image signals such as characters.

FIG. 2+t, the analog image signal read by the CCD image sensor 4 and level amplified by the amplifier 5 is quantized by n-value in the quantizer 6, and the quantized digital image signal is converted to 2 by the dither processor 10. A specific example of the configuration when converting into values is shown.

The quantization unit 6 includes a background level detector 7 that detects and holds the peak value of the analog image signal, and a background level detector 7 that divides the detected background level of the original to determine the highest level and lowest level for quantization. and a quantization circuit 9 that performs n-value amplification of the analog image signal in the determined digitization range. The dither processing unit 10 also includes a dither matrix circuit 11 in which a threshold value for binarizing a digital image signal is set, and a dither matrix circuit 11 that calculates a predetermined value from a dither matrix while counting main scanning and sub-scanning clocks. A dither position it polarization generator 12 consisting of a dither counter that generates a signal for sequentially reading a threshold value, and a digital comparator 13 that binarizes a digital image signal using a threshold value that is sequentially applied to the dither matrix circuit 11. It is constructed behind the scenes.

The analog image signal obtained when the original image is read is converted into an n-bit digital image signal by the conventional quantization circuit 9, and the converted digital single signal is converted into a dithered signal representing halftones by the digital comparator 13. When a dot printer records and reproduces an image based on the image signal that has been binarized through processing, the characteristics of the reflectance of the reproduced image with respect to the reflectance of the original image become third-order.
The result will be as shown in the figure. In other words, the reflectance of the reproduced image is calculated from the reflectance T MIN of the point portion to the reflectance T M of the white portion.
It changes linearly up to AX, and the middle tone part is T MI
It will be reproduced to have a reflectance between N and TMAX. Note that sTMXkl is the reflectance at a single point where all dots in the unit matrix are printed, and TMAX is the reflectance at the white area (background part of the recording paper) where the number of printed dots in the unit matrix is zero. rate. The reflectance of the original image is a relative value with respect to the reflectance i1 of the background. In addition, in Figure 3, the range of AD conversion is between 5% and 80% of the background level of the detected document, and the characteristic is shown in one word.
% to close % is the intermediate tone, 0 to 5% of the background level
is completely black, and 80% to 100% of the background level is completely white. Therefore, the quality of the reproduced image, such as the character part of a document that is not halftone, is good. In order to achieve this, it is necessary to emphasize the reproducibility of intermediate tones when the relative reflectance is between 0.1 and 0.3 when the background level of the original is 1.0. There is. However, with the characteristics shown in Figure 3, the relative reflectance is 0,0
Halftone reproduction is uniformly performed over the entire range from 5 to 0.8712 r, and the relative reflectance is 0.1 to 0.3.
The reproducibility of halftones between the two positions is not emphasized at all, and the reproducibility of halftone images is poor.

Purpose The present invention has been made with the above points in mind, and it is possible to improve the reproducibility of characters when converting an analog image into a multi-valued image using a parallel comparison type AD converter. Both signals effectively expand the range in which midtones can be reproduced without any deterioration, and emphasize areas that are difficult to reproduce in midtones to improve reproducibility. It provides a quantization circuit.

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

As shown in FIG. 4, the quantization circuit according to the present invention has a reference voltage divider l, a high-speed comparator group 2, an encoder 3 or 7, and a parallel comparison type AD converter. Pull out the intermediate tap Mi from the divider l and set the lower limit to 1. (・Resistance R6 is connected between the tap L to which the voltage V, . . . 2 is applied.

In this structure, the resistor between the tag H and the intermediate tap M, to which the upper limit reference voltage is applied, is R/2 (R is the total resistance of the reference voltage divider 1). ), so that the resistance value between the intermediate tap M and the tap L is R/2, and the voltage applied between the taps H and M is the tap M.

Add between L. Rc=R/ so that it is twice the in+ pressure
Set to 2. Therefore, when such a circuit is used as the quantization circuit 9 in FIG. 2 to perform dither processing on the digital image signal converted into multi-level vertical characters, and then perform all image reproduction, the relative The characteristics of the reflectance of the reproduced image 1 with respect to the reflection 4 are shown in FIG.

In the characteristics shown in Fig. 5, the reflectance of one reproduced image is T M
It becomes a polygonal line at the point Tv of A'X/2, and /? ! μ/21ii respectively with different reference direct pressures in the r range
, and the background level of the manuscript is in the ranges of 0 to 5% and 80% to 100% (all black and all days are set to 5), which is the same as in the case of the factory. ◎Therefore, As is clear from this characteristic, it is possible to obtain a high-quality reproduced image in which the midtones in the low relative reflectance of the original between 0.1 and 0.3, that is, the dark parts, are emphasized. In the case of character parts of a document, the reproduced image is not impaired, and a reproduced image of high quality can be obtained as before.

Note that the reference voltage V to the image signal quantization circuit according to the present invention
To provide RIeνI VRICF2, a voltage divider 8 may be used as shown in FIG. 2, but a configuration as shown in FIG. 6 may be used instead. That is, by connecting the abrasive iR1 to the tag H of the reference voltage divider 1 and applying the peak value of the analog image (g) detected and held by the ground level detector 7 through the resistor R1 to the reference voltage divider 1, A resistor R2 is connected to the tap L1 (111) of the voltage divider 1, and the end of the resistor R2 is grounded. Also, this shows the configuration when converting the analog No. 4 symbol into a 4-sew stitcher. .

Further, in this embodiment, switches SWI and SWZ are provided in parallel with the resistors R1 and R2, and a switch SWB is provided in series with the resistor Rc.
By turning off switch SW3 and turning on switch SW3, halftone emphasis is performed.

It is set to R5, which can de-emphasize midtones by turning on SW2 and turning off SW3.

As described above, in the image signal quantization circuit according to the present invention, which performs multi-value quantization of an analog image signal using a parallel comparison type AD converter, from the reference voltage divider in the AD converter. The tap is pulled out and a resistor is connected between the tag and the tap that applies the base pressure to the reference pressure divider, which causes the linearity of the AD conversion to be lost and An analog image signal that emphasizes the reproducibility of halftones in a part of the conversion range: Chilling has the excellent advantage of C giro.

[Brief explanation of the drawing]

Figure 1 is an electrical wiring diagram showing a quantization circuit for both signals using a conventional parallel comparison type AD converter, and Figure 2 is a dithering circuit that performs multi-level quantization of both signals of the original captured by an image sensor. A Glock diagram showing a general configuration example for processing, and Figure 3 shows the reflectance of the reproduced image when the digital image signal quantized by a conventional quantization circuit is dithered and reproduced. Figure 4 is an electrical wiring diagram showing an embodiment of the quantization circuit according to the present invention, and Figure 5 is a diagram showing the characteristics of the quantization circuit according to the present invention. FIG. 6 is an electrical wiring diagram showing another embodiment of the present invention. 1... Reference direct pressure divider 2... High floor comparator group 3... Machine/coder 4... CCD image sensor 5... Amplifier 7... Background level detector 8.
... Voltage divider 9... Quantization circuit 11... Dematrix circuit 12... Dither position number generator 13
・・・Digital image signal is used as a digital image signal. Anti-light hand's lower eyes

Claims (1)

[Claims] When a parallel comparison type AD KM device is used to convert an analog image signal into a multivalued skeleton, the tag sound is extracted from the reference voltage divider in the AD converter, and the reference voltage is applied to the tap and base voltage divider. A dual-signal ψ chill circuit characterized by connecting a resistor between the input tag and the E5 tag.