JPS63234772A - Binarization method - Google Patents

Abstract

PURPOSE:To reduce the number of an ROM table executing threshold correction and density adjustment by adding restriction to the range of the error or an average error minimum method, simultaneously varying a threshold level, so as to binarize an intermediate tone image and to vary density characteristic. CONSTITUTION:A binarization signal is obtained by finding the error for the correction density value of the binarization density value of a binarization decided picture element near a noticed picture element, multiplying that error by a prescribed weight factor, adding the above result to the actual density value of the noticed picture element, the finding the correction density value of the noticed picture element and by comparing this correction density value with the threshold. As the result of adding the restriction of a prescribed range between the error in the binarization system and the actual density value, the threshold level can be varied and the binarization of the half tone image and the density adjustment can be realized by setting the threshold level at a value out of the range.

Description

[Detailed Description of the Invention] Fields of Application of Rakudo] The present invention utilizes the minimum average error method among the binarization methods used in facsimiles, image scanners, etc. to convert halftone image signals into binarized signals. The present invention relates to a method for binarizing image signals. More specifically, the error in the abrasive binarization decision pixel around the pixel of interest is multiplied by a predetermined weighting coefficient, and this is calculated as
Il! A binarization method is proposed in which a corrected density value of the pixel of interest is obtained by adding it to the actual density value of the pixel ii, and a binarized signal is obtained by comparing the corrected density value with a threshold value.

(Prior art) According to the minimum average error method, halftones can be substantially expressed in two levels, black and white, and the proportion of black dots after binarization to the total number of pixels is equal to the gradation level before binarization. The characteristic is that it is equal to the density average of the data.

[Problem that the invention seeks to solve]

However, according to the minimum average error method, which is a conventional binarization method, in order to correct the γ characteristic of a CRT, printer, etc. that outputs a binarized image, the γ characteristic is stored in ROM in advance, and the 2 A commonly used method is to perform γ correction using a ROM as a table before conversion. In addition to γ correction, there is also a strong desire to adjust the image density according to preference, so in order to achieve fine density adjustment, it is necessary to prepare as many ROM tables as necessary, and the value (δ
The number of tables was decided based on this balance. Therefore, the present invention aims to solve this problem.
The purpose is to provide a binarization method that realizes density adjustment at the same time as binarization processing.

[Means for solving problems]

The binarization method of the present invention is characterized by limiting the error range in the minimum average error method and by making the threshold variable.

〔Example〕

Before explaining the method of the present invention, the minimum average error method will be described. The minimum average false landing method calculates the error of the binarized density value of the binarized determined pixel around the n-th pixel with respect to the corrected density value, multiplies the error by a predetermined weighting coefficient, and calculates the error of the actual binarized density value of the destination pixel. The corrected density value of the pixel of interest is obtained by adding it to the 0 degree value of , and the binarized signal is obtained by comparing the corrected density value with a threshold value. IEIEE-3PecLru in March
Paper published in m r1magcsLlfrom, co
rLocal mentioned in mPutersJ
5paLial average sof
b r i g b L n e s s s J is the basic, then J, F, Jarvis (Bell Laboratories, USA)
In May 1076, Computergrap
bics and image pr.

Paper published in cessing rA 5urve
yo f t e c h n i q u e
s r o r tbe display
of continue.

us tone pictures on b
rmini in ilevel disPlaysJ
mized average error me
This method has been improved in the form of thodJ.

Specifically, as shown in Figure 7112, when the X direction is the main scanning direction and the Y direction is the sub-scanning direction, the actual density value of the pixel (-, y) of the halftone image is expressed as Jxy (0 to R ), the binarized signal of the pixel of interest (x, y) is IxyGO, or the corrected density value of the pixel of interest (x, y) of 1)-1 is J'X
Let the F% error be Exy=J'xx-R@Ixy, and use the weighting coefficient matrix (A.

1), for example: However, if main = pixel of interest, then the corrected density value J'Xy can be obtained from the following equation (2).

J'xy =Jxy +VA+ 7 IIExl-
s +y÷−8...(2) Then, the binary signal 1xy is J'xy(D4ti
It can be determined as follows:

However, before reaching pixel (3, 3), JIXy cannot be calculated, so the corrected density value up to this pixel is, for example, J
' Assuming that XY=JXV, find Ixy from equation (3).

The above method is the minimum average error method, and by setting the sum of the m image numbers in equation (I) to 1 and setting the threshold to R/2 as in equation (3), the error during binarization can be reduced. The ratio of the number of pixels whose density becomes R as a result of binarization to the total number of pixels is equal to the average density of the gradation data before binarization. Characteristics. but,
When actually outputting a binarized image to a printer or the like, it is necessary to correct the seven characteristics of the output device. In other words, if a small part of the input gradation data is, for example, 11 degrees of 0.3R, it is accurate in terms of 2' value conversion theory to make exactly 30% of the binarized image of that part black dots, but the printer The size of one dot is usually a little larger than the size of an 11if cable, so in order to make it look like a density of 0.3R, the proportion of black dots should be set to 0.25, for example.
must be divided.

This density correction is performed prior to binarization using a ROM table as shown in FIG. This ROM table is also used to perform density conversion, such as making a black part of an image even darker, or making a near-white part completely white.

Therefore, in the method of the present invention, in order to perform density correction or density conversion at the same time as binarization, the error range is changed from -R/2 to R.
/2, and the threshold value can be set to a value other than R/2. According to this method, the density conversion characteristic when changing the threshold value is 9 as shown in FIG. It seems natural to lower the threshold value when you want to increase the density, for example, but by limiting the range of error, you can only obtain the density conversion characteristics shown in Figure 1. This is because the minimum average error method is a method that refers to errors, so simply changing the threshold value will not make the expected value of the error 0.For example, when the threshold value is set to 0.3R, the average value will be 0.3R. Assuming that a 5R random analog image signal is input, the range of possible error values is E knee 0.7R to +0.3R, and since similar occurrences can occur within this range, the expected value is fo',' tXax)=-0.2R. Therefore, in equation (2), the expected value of the second term on the right side is 10.2R, and the average value of the first term JMν is 0.2R.
Even in 5R, the average value of the corrected density values J','x'V compared with the threshold value is 0.3R, and even if this is compared with the threshold value of 0.3R, the pixels whose binarization at is R are all 50%, and the density characteristics remain unchanged. In this way, the density characteristics do not change just by changing the threshold value, but the 0 degree characteristics can be changed by adding a limit to the error range. The error range needs to be wider than -R/2 to R/2 in order to obtain an image that is no different from the conventional minimum average error method when the threshold is R/2, and when the threshold is changed from R/2. (In order to obtain the corresponding density conversion characteristics, -R/2 to R/
It needs to be narrower than 2, and just -R/2 to R/2 is optimal.

〔Effect of the invention〕

As explained above, according to the above configuration of the present invention, the density characteristics can be changed at the same time as the halftone image is binarized, and the number of ROM tables for performing γ correction and density adjustment can be reduced. Furthermore, if the same number of ROM tables as in the past are shifted by Δ, the number of types of density adjustment will be greatly increased. Furthermore, if this binarization method does not refer to errors in pixels that have already been binarized around the pixel of interest, it will be the same as the binarization method using the simple binarization method, but by making the threshold variable, the threshold This means that a simple binary method that can freely set the values can be realized, and when actually building a binarization processing device, a halftone processing binarization processing device and a simple binary processing device can be used simultaneously. There is a special effect.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing density conversion characteristics that can be realized by the binarization method of the present invention. FIG. 2 is an explanatory diagram of a pixel arrangement in a conventional binarization method. FIG. 3 is a configuration diagram showing the relationship between a ROM table and binarization processing in a conventional binarization method. Applicant: Seiko Epson Corporation Figure 1 Figure ζ Figure 3

Claims (1)

[Claims] The error of the binarized density value of the binarized determined pixel around the pixel of interest with respect to the corrected density value is calculated, the error is multiplied by a predetermined weighting coefficient, this is added to the actual density value of the pixel of interest, and the pixel of interest is calculated. In a binarization processing method that obtains a corrected density value of and obtains a binarized signal by comparing the corrected density value and a threshold value,
A binarization method characterized by limiting the error range and making the threshold variable.