JPS5844860A - Processing method for intermediate tone - Google Patents

Abstract

PURPOSE:To reproduce a picture with high gradation and high quality and less recording unevenness, by splitting a screen into blocks of a plurality of picture elements, and assigning recordings of intermediate level and black-and-white binary value in response to an average density level of each block. CONSTITUTION:Picture information read from an original is first stored in a picture memory 9, and when the information for one block's share is recorded, a switch 18 is changed over to the position (b) and the next block is stored in a picture memory 10. When the picture information is stored in the memory 10, the content of the memory 9 is read out and transferred to a block memory 11. From the content of the block memory 11, the density level of each picture is determined with a block processing circuit 13 and written in the 2nd block memory 12. The picture signal thus processed is alternately stored in the 2nd picture memories 15 and 16, and sequentially outputted at each block.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a halftone processing method capable of obtaining high quality recorded images for image processing such as facsimiles.

Conventionally, electrostatic recording or thermal recording, which has been widely adopted as a recording method for facsimiles, etc., can reproduce halftones by changing the amplitude and pulse width of the recording voltage. However, the reproduction of about 10 tones is the limit,
In addition, sufficient recording quality is not always obtained due to uneven recording density and the like.

On the other hand, various methods have been proposed for pseudo-reproducing halftones using black and white binary pixels. Since this method does not include pixels at intermediate density, there is little recording unevenness and high zero tone reproducibility can be achieved, but it generally has the disadvantage of poor resolution.

In order to eliminate these drawbacks, the present invention records only some pixels at an intermediate level and the others at a black and white binary level, and will be specifically explained below.

In the method of the present invention, one screen is divided into blocks each having a plurality of pixels, and the following processing is performed for each block.

I) Find the average value Dav of the density level of each block.

II) Total number of all black pixels in the Prusk'? N as their average concentration level is less than Dαν.

The number N of black pixels in the pull block is determined so as to be the value 1)(lv' that is closest to Daν. That is, it corresponds to the integer part when the total density level in the pull block is converted into black pixels.

III) Allocate N black pixels in descending order of density level of the 4!1 bifurcations in the block.

IV) Intermediate $1 for density level N+1st pixel! I
By assigning a gray level of F and whitening other pixels, the average density of a block consisting of N black pixels and one gray pixel is equal to the average density of the original image. (as close as possible if the concentration level is forceps).

The density level of the gray pixel corresponds to the decimal part when the total density level 't in R,/R is converted into black pixels.

The "density level" here refers to the signal level that has a 1:1 correspondence with the reflection density of the original or recorded image.
It does not necessarily represent the reflection 8 degrees itself or the proportional 'V-' tortoise.

The figure in+ will be explained in detail below.

FIG. 311 is an example of a screen for explaining the present invention.

A small square indicated by a broken line represents one pixel.

Also, large squares l to 8 shown by solid lines each have 1
represents one block. Also, were the numbers shown for each pixel read from the manuscript? It represents the f level P, where 0 is white, 16 is black, and anything in between is gray. In addition,
The density levels in FIG. 1 are expressed as integers for convenience; however, they do not necessarily have to be integers.

Calculating the average density level of each block in Figure 1 is 31
It will look like Figure 2. Each area surrounded by a solid line represents one block, and the number represents its average density level Dav.

The number N of black pixels for each block is determined from the value of the density level P of each pixel using the following equation.

However, m is the number of drawings in the boot stock, and 1) am is the value of the black level. Implementation of Figure A11 and Figure O.2 (-= 4 in ?υ, Al (1) in Dmaz -16 TeDα
ν ・N=integer part of □. In this way, determine f+i f of N and fill
- Figure f213 shows how the levels are assigned in descending order of level.

The blacked out part in the figure is assigned black 'f+,
It's on 5th.

Since the decimal part is discarded in the calculation formula for N described above, the average density of each block of black pixels is smaller than the average density of the original screen. Therefore, the fractions are expressed in halftones so that they are equal. The # degree Dz of the intermediate level is determined by the following formula.

The results of determining the value of D- for FIGS. 1 and 2 in the case of Dz-mDav NJ) gas inert-4, Dmam=16 are also shown in FIG. The diagonally shaded pixels are pixels with the highest density level among the pixels to which black is not assigned, and are assigned a halftone level.

The number indicates the density level to be recorded. In the block indicated by 6, since D11 = 0 (white), there are actually no pixels to be recorded in halftone (1°. Note that in the block indicated by l, pixels at the same density level Since there are multiple pixels, it is not possible to uniquely determine which pixels should be recorded in halftones.

In such a case, a priority order may be determined in advance for the 0 image collection position. The same applies to the black picture cumulative allocation.

In the case of the embodiment shown in Fig. 3, it is possible to record 16 levels of intermediate tones, but when reproducing 16 gradations for each block, it is not necessarily necessary for each pixel to be able to record 16 levels (5 It's good to be able to record the level.

FIG. 4 shows that in the example described above, D- is
Level 4 for 3-6. Level 8゜lO-13 for 7-9.
This shows an example of recording in black at levels 12.14 to 16. In actual practice, the method of the present invention allows for six or more halftone recording levels for each pixel in order to correct the recording characteristics even when reproducing 16 gradations (I) by selecting an appropriate density rW. It is desirable to
゜Also, in the above-mentioned embodiment, the number N of black pixels is determined in proportion to the average density level of each block, but it is not necessarily necessary to do so, and it is determined to be non-linear, and the configuration of each block is 2×2. As described above, it can be arbitrarily selected depending on the number of recordable halftone levels and the number of gradations to be reproduced for each pixel.

FIG. 5 shows two views of the boot for implementing the method of the present invention.

In Fang 5, 9. IO is N’1 image memory.

114i'A11 block memory, 12 M-2 block memory, 13 block processing circuit, 14 signal generation circuit, 15.16 N'2 image memory, 17, 18゜1
9. The image information read from the original is stored in the image memory 9.

For example, if 1 x 4 x 4 16 compartments are configured, after 4 scanning lines are stored, the switch 17 is
The switch 18 changes from the side b to the side b, and the contents of the image memory 9 are sequentially processed while storing the primary image information in the image memory 10. The contents of mii R memory 9 are as follows:
will be forwarded to.

Next, based on the contents of the N'l book memory 11, the density level of each pixel is determined by the block processing circuit 13 using the method described above, and is written into the second block memory 12.

The signal generation circuit 14 is a circuit that generates a write signal. The processed image signals are sequentially stored in the two-image memory 15 for each block. After the data corresponding to 4 scanning lines have been stored, the switch 19 is switched from a to a, and the switch 201'ib is switched to a, and while the processed image signal is stored in the second image memory 16, the contents of the 0 image memory 15 are outputted one-by-one. be done.

As explained above, the method of the present invention is one in one.

Since fewer pixels are recorded at the intermediate level, there is an advantage that there is less variation in recording unevenness, and since gradation is reproduced block by block, high gradation can be reproduced even if the number of gradations is small for each pixel. . Furthermore, since black pixels are assigned in descending order of density level of original image information, high-quality reproduction is possible even when high resolution is required, such as for one character.

Roughly speaking, if recording in halftones - ρ11 prime number is small, if applied to high-speed thermal recording, etc., it is ?IiA degree control furnace control of the thermal head.In other words, if high-speed recording is performed by thermal recording, For black recording, the cooling time after heating the thermal head is shortened, heat accumulates and the substrate temperature rises.To avoid this, when black pixels continue.

In the case of binary recording, the pulse width is reduced to control the temperature after heating to be constant. In the case of halftone recording, it is difficult to perform such control because the pulse width differs depending on the recording density. The method of the present invention, which is simple, allows control close to black and white binary. In other words, it is possible to take a method such as controlling only black pixels by considering only the history of black pixels.

[Brief explanation of the drawing]

Figure 221 is a configuration diagram of a double-sided example for detailed explanation of the present invention, Figure 172 is a configuration diagram showing the average density level of each block in Figures 3 and 1, Figure 213 is a configuration diagram showing an example of a pattern after processing, Figure 3t4 is a configuration diagram showing another pattern example after processing,
FIG. 5 shows a book diagram for implementing the method of the invention. In the figure, 9. IO is an open image memory, 11 is an optical block memory, 12 is a second block memory, 13 is a pull processing circuit, 14 is a signal generation circuit, 15° and 16 are second image memories, and 17 to 20 are switching Showing a switch. Patent Applicant Nippon Telegraph and Telephone Public Corporation Patent Attorney Mori 1) Hiroshi Figure 1 Figure 2 Prisoner $3 Figure 4

Claims (1)

[Claims] The screen is divided into blocks each consisting of a plurality of pixels. The pixels to be recorded at the intermediate level in each block are less than or equal to f1 pixels, and the number of black pixels for each block is determined by the integer part when the total density level of the block is converted by black pixels, and at the intermediate level of 1llf. The density level of the pixels to be recorded is determined by the decimal part when converting the total density level of the R1r write block into black pixels, and the black pixels are assigned in order to the pixels with the highest density level on the original screen, and the black pixels are assigned in order of the pixels with the highest density level on the original screen. Intermediate level for the image collection with the highest density level among the pixels. A halftone processing method characterized by assigning pixels and making the remaining pixels white.