JPH0334272B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention uses a method of expressing gradation by associating multiple dots with one pixel of a gradation image, and is capable of printing text on a low-contrast character document recorded on a paper surface with a high background density. This invention relates to a dot pattern generation method that can record characters with high intelligibility.

As shown in Figure 1, this type of conventional method scans a character area with an image sensor, detects changes from light to dark or from dark to light, and generates a black right-aligned or left-aligned pattern. was. If this method is used, as shown in Figure 2, the dots generated on originals with high background density and low contrast, such as originals recorded on blueprints or originals recorded on colored paper, can be reduced. The pattern consisted of gray and black, and the difference in density at the outline of the characters became small, making it impossible to record the characters clearly.

In order to eliminate these drawbacks, the present invention scans a character area with an image sensor and detects a change point from bright to dark or dark to light, and then displays white and black corresponding to the change point regardless of the paper density. The drawings will be described in detail below.

FIG. 3 is a diagram showing the principle of the present invention, where 1 is a character pattern, 2 is a scanning direction of the image sensor, 3 is paper density, 4 is print density, and 5 is a change point from bright to dark in the two scanning directions. , 6 is a pattern corresponding to the change point from dark to light, 7 is a pattern corresponding to change point 5, 8 is a pattern corresponding to change point 6, 9 is a pattern corresponding to light (paper density), 10 is dark (print density) The pattern corresponding to , 11 is the density of white. The example in FIG. 3 is representative of the case where a two-dot (1×2) dot pattern is associated with one pixel of the original image.
4 dots (2×2), 8 dots (2×
4), 9 dots (3×3), etc. are placed over a corresponding period.

In the method of expressing gradation by associating multiple dot patterns of three values (white, gray, black) to one pixel of a grayscale image, the paper is blue-printed or colored, and the difference between paper density and print density is may be small. In such a case, in order to record the character pattern in the character area, first record the character pattern 1 in the character area.
Scan 2 is performed on. A density distribution is obtained by scanning 2, and paper density 3, print density 4, a change point 5 from bright to dark, and a change point 6 from dark to bright in the two scanning directions are determined. Based on this concentration distribution, change point 5
Corresponding to this, a pattern 7 is generated in which the right half is black and the left half is white, and a pattern 8 is generated in which the left half is black and the right half is white in correspondence to the change point 6. If the light is continuous, a pattern 9 in which all the dots are made of gray corresponding to the paper density is generated, and if the dark is continuous, a pattern 10 in which all the dots are black is generated. By generating the pattern in this manner, the white portion 11 is obtained along the outline of the character pattern. As a result, even if the difference between paper density and print density is small,
The outline of character lines becomes clearer.

FIG. 4 shows an embodiment of the present invention, in which 12 is an image memory for a character area, 13 is an address register for reading data from the memory 12, 14 is a shift register consisting of two registers, and 15 is an ExclusiveOR
16 is a comparator, 17 is a storage memory for the value "1" indicating black, 18 is a pattern 8 generation circuit, 19 is a pattern 7 generation circuit, 20 is a comparator, 21 is a pattern 9 generation circuit, 22 is a pattern 10 generation circuit,
23 is output buffer memory, 24 is register 1
A signal 25 increases the content of 3 by 1, and 25 is a circuit that binarizes into black (=1) and non-black (=0). The operation of this circuit is as follows: First, the density information of the character area is stored in the memory 12.
Store in. The density information of each pixel in one line is read out one by one from left to right from the memory 12 by the address register 13 as shown in FIG.
The digitization circuit 25 performs binarization and the shift register 1
Store in R2 of 4. At this time, the information of the previous pixel that was already stored in R2 is shifted to R1. The contents of R1 and R2 are ExclusiveOR circuit 15
, and if the contents of R1 and R2 are different, comparator 16 is routed. The comparator 16 uses the information of R1 and the black value (=1) stored in advance in the memory 17.
When R1 is "1" (black), pattern 8 is generated, and when R1 is not "1" (paper density 3), pattern 7 is generated in circuits 18 and 19, respectively. R in ExclusiveOR circuit 15
If it is determined that the contents of 1 and R2 are equal, the comparator 20 is activated. The comparator 20 compares the information of R2 with the black value "1" from the memory 17, and if R2 is not "1", pattern 9 is executed.
When “1”, pattern 10 is connected to circuit 21.
and occurs at 22. Here, it is assumed that the density corresponding to the gray color of pattern 9 has been detected in advance.
The patterns generated by the circuits 18, 19, 21 or 22 are stored in the output buffer memory 23, and are recorded when storage for one line is completed. When one pattern has been generated, the signal 24
is sent to the register 13, its contents are incremented by 1, and the value of the next pixel in the memory 12 is sent to the binarization circuit 25.
It is binarized and read out to R2. With the above operation, the third
It is possible to generate an output pattern as shown in the figure.

As explained above, since white dots are always generated in contact with the outside of the outline of characters, there is an advantage that character patterns recorded on paper with high density can be clearly recorded.

[Brief explanation of drawings]

Figure 1 shows dots and dots for conventional character patterns.
FIG. 2 is an explanatory diagram of a pattern generation method, FIG. 2 is an explanatory diagram of a dot pattern when paper density is high according to a conventional method, FIG. 3 is a diagram of the principle of the present invention, and FIG. 4 is a diagram of an embodiment of the present invention. . In the figure, 1 is the character pattern, 2 is the scanning direction of the image sensor, 3 is the paper density, 4 is the print density, 5 is the point of change from bright to dark in the two scanning directions, and 6 is the point of change from dark to light. , 7 is a pattern corresponding to change point 5, 8 is a pattern corresponding to change point 6, 9 is a pattern corresponding to light (paper density), 10 is a pattern corresponding to dark (print density), 11 is white density It is.

Claims (1)

[Claims] 1.Multiple dots correspond to one pixel of image information,
In a method of expressing gradation by changing the areas of the three values of white, gray, and black, if the detected value obtained by scanning the character area with an image sensor is that the right side is darker than the left, the black dots are shifted to the right, and the white dots are shifted to the right. So that the dots are aligned to the left,
Also, if the right side is brighter than the left, a dot pattern is generated so that the black dots are aligned to the left and the white dots are aligned to the right. If the brightness is continuous, all dots corresponding to the paper density are generated in a gray pattern. 1. A dot pattern generation method for image processing, characterized in that when a dot is generated and the darkness is continuous, a pattern in which all dots are black is generated.