JPS6335074A - Picture processing method - Google Patents

Abstract

PURPOSE:To reduce the scale of circuit and to obtain density expression equal to an original image by determining values to be allotted to each picture element to make the total value equal to density difference information for each value of density difference information. CONSTITUTION:An RAM 19 is looked by the density of a decisive picture element P and values of error a', c', D' to be allotted to picture elements A, C, D and binarization output are obtained. The value of binarization outpnt is outputted as it is. Output c' is added to a register 12 and goes to a buffer memory 16. Output D' is added to a register 11 and goes to the register 12, and output a' is added to a register 13 and goes to a register 14. New density data 23 enters the register 11, and density data before one line enters the register 13 from the memory 16. By repeating this, the errors due to binarization coding is eliminated and pseudo half tone expression can be obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to image processing, and particularly to an image processing method suitable for expressing a halftone image in black and white binary.

[Conventional technology]

Conventionally, in addition to the dither method, Niss I Day 1975 Digest No.
Items 6 to 37 (SID 75 Digest, pp.
.

There is an error diffusion method discussed in 36-37.1975).

In Fig. 2, this method calculates the density of the determined pixel P by P+P.
The density of pixels A, B, C, and D around the point is a
ng b((HCn, dn, If the threshold value for binarization at point P is T, then the binarization error E at the determined pixel P is empirically determined as in equations (1) to (5). With weights I, J, and L, any pixels around point P
n+ Qn* dn, and the average density seen from a macro perspective is made equal to the density of the original image.Compared to the DISA method, the fine structure of the image can be expressed, and the average density seen from a macro perspective is It has the advantage that it can be made to be the same as the original image.

(However, N+++ax is the maximum concentration, Nll1in is the minimum concentration) an"x=an+EXI ・
...(2)1), +1=bn+E X J
−(3) CIl”l= Cn + E X K
...(4) Dn+z=Dn'
+ E
From equations 1) to (5), one multiplier and one divider are required, which increases the size of one circuit, and when integer operations are performed, due to rounding errors, for example, if the binarization error is E = 5, EXI=5X7/1 is 2 in integer operation, EX
J=5X1/1 is 0, EXK=5X5/16S:
This is 1, and EXL=5X3/1, and this is O1. When performing integer arithmetic, the value of EXI+EXJ+EXK+EXL becomes 3, which is not equal to the binarization error value E=5. Therefore, the average density is not equal to the original image density, and there is a drawback that the density is expressed as being changed compared to the original image.

[Problem that the invention seeks to solve]

The conventional technique is due to an increase in circuit scale for one multiplier and a divider, and rounding errors due to division.

There were problems such as changes in the average concentration seen from a macroscopic perspective.

An object of the present invention is to reduce the circuit scale and eliminate rounding errors by eliminating the need for single multipliers and dividers.

[Means for solving problems]

The above purpose is to apply (1
)~(4) instead of calculating and sorting,
This is achieved by determining the error distribution method for each error value as shown in FIG.

[Effect]

In the conventional method, multiplication and division were performed to determine how to allocate errors, but as shown in Figure 1, there is a
If you decide on a numerical value to allocate the error according to the error value, you only need to add that value and the pixel density, so you can configure the circuit with only an adder. This also eliminates rounding errors. Furthermore, since the distribution method can be freely changed depending on the error value, it is possible to use a more beautiful expression method and simplify the circuit.

〔Example〕

An embodiment of the present invention will be described below with reference to FIG. 23
is the 4-bit density data 11, 12, 13°14 is the register containing the density of the pixel, and the codes inside are A and C.
, D, and P match the symbols in Figure M. 16 is a line memory that stores one line of image data, 15 is an adder (maximum value is 15°, minimum value is O), and 19 is a RAM that stores error distribution method. , it was also assigned to B in Figure 1, but was omitted for the sake of simplicity.1
7 is the address of the RAM, and the address is specified by the density value of 14. In order to make this table even simpler, instead of specifying the error value as shown in Figure 1, it is possible to sort by density value. It is set at 8. 22,2
1.20 is the error value distributed to pixels A, C, and D, 1
8 is.

The value 24 of the binarized output is the binarized output.

First, subtract 19 RA M from the density of the determined pixel P.

Error value distributed to pixels A, C, and D: 1, C/.

D' and a binarized output are obtained. The binarized output value is output as is. D' is added to 12 and the buffer memory 16
D' is added to 11 and goes to 12, a' is added to 13 and goes to 14, and new density data 23 is added to 11.
The density data for the previous line is input from 16 to 13, respectively. By repeating this process, pseudo-halftone expression can be achieved that eliminates the error caused by binarization.

According to this embodiment, the multiplier and divider of the conventional example are not required, and the scale of one circuit can be reduced. Furthermore, the weight for distributing errors can be freely changed depending on the error value, making it possible to express even more beautiful halftones. There is a certain effect.

〔Effect of the invention〕

According to the present invention, a circuit can be configured with only adders. Since the influence of rounding errors can also be eliminated, there is an effect of reducing the circuit scale and expressing density equal to that of the original image.

[Brief explanation of the drawing]

FIG. 1 is a circuit configuration diagram of an embodiment of the present invention, FIG. 2 is an explanatory diagram of pixels, and FIG. 3 is a diagram showing an example of an error distribution method. 11.12, 13.14...Register for pixel density,
16... Line memory for one line, 15... Adder, 19... RAM for error distribution.

Claims (1)

[Claims] 1. In a pseudo halftone expression method that distributes the original image density and the density difference information after binarization determination to surrounding pixels so that the average density is equal to the original image density, for each value of the density difference information,
An image processing method characterized by determining numerical values for distributing the values to each pixel so that the total value is equal to density difference information.