JP2857906B2 - Halftone binarization processor - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a halftone binarization processing device used in, for example, a facsimile machine.

[Prior Art] Conventionally, an image received by a facsimile apparatus is represented only by white pixels and black pixels, and therefore, when the white and black portions of the image are emphasized too much, they are considerably inferior when viewed qualitatively. Was. Therefore, recently, it has been possible to reproduce halftones by forming data indicating each pixel with a plurality of bits to have a plurality of gradations and changing the appearance frequency of white pixels and black pixels according to the gradations. In some cases, the quality of an image is dramatically improved.

By the way, in the case of an image composed of such a pseudo halftone, for example, when there is another white region having a small level difference in a region close to white, a change between regions when the image is output is obtained. Is also not clear, that is, a so-called underexposure phenomenon occurs. Similarly, when there is another black area having a small level difference in the black area, the change between these areas becomes unclear when printed out. For example, a change in image level is not sufficiently expressed, for example, a blackout phenomenon occurs, and the image quality tends to deteriorate.

[Problems to be Solved by the Invention] As described above, in the conventional image output composed of pseudo halftones, if there is a region having a small level difference between a region close to white and a region close to black, the region between these regions is reduced. There is a tendency that the change of the image level cannot be sufficiently expressed, and the quality of the reproduced image is not sufficient.

The present invention has been made in view of the above circumstances, and provides a halftone binarization processing apparatus capable of emphasizing and expressing image level changes in a region close to white and a region close to black. The purpose is to:

[Means for Solving the Problems] A halftone binarization processing device according to the present invention performs analog-to-digital conversion of a halftone input image signal to quantize it into multi-valued image data. The stored data is stored in the storage means, one pixel data is designated from the storage means, and the average value of the pixel data values around the designated pixel is calculated, and the calculated average value is the white area level or When it is within a predetermined level area close to the black area level, a larger edge enhancement degree is determined for the specified pixel data than when it is outside the predetermined level area. The edge emphasizing process is performed on the designated pixel data on the basis of the specified pixel data, and the edge-emphasized pixel data is binarized.

[Operation] According to the halftone binarization processing device of the present invention, a change in image level can be emphasized and expressed by an edge emphasis degree determined by pixel data around a designated pixel. Alternatively, even in the case where another region having a different level exists in the black region, the edge portion can be emphasized and expressed between these regions, and so-called white loss or black loss on the reproduced image. The phenomenon can be avoided, and a high-quality image can be obtained.

Hereinafter, one embodiment of the present invention will be described with reference to the drawings.

FIG. 1 shows a circuit configuration of the embodiment. In the figure, reference numeral 1 denotes a scanner for reading a document (not shown),
The scanner 1 reads an original line by line in the main scanning direction and outputs each pixel data as an analog signal. The output of the scanner 1 is sent to an A / D converter 2 and converted into a digital signal. In this case, A /
The D converter 2 outputs 8-bit multivalued image data of 0 to 255 for each one pixel.

The output of the A / D converter 2 is sent to the memory 3. The memory 3 stores one byte per pixel, and is capable of storing at least three lines of pixel data.

 The output of the memory 3 is sent to the edge enhancement circuit 4.

Here, the edge emphasizing circuit 4 is configured as shown in FIG. In the edge enhancement processing, the memory 3
FIG. 3 (a) according to the address designation of the address control unit 31
(I-1, j), (i, j-1), centering on the pixel (i, j) edge-emphasized over three lines as shown in FIG.
Data of a total of five pixels is read from (i, j + 1) and (i + 1, j). Then, the central pixel data f (i, j) of the pixel data f is latched.
41, the other four pixel data (i-1, j),
(I, j-1), (i, j + 1), (i + 1, j) are expressed as f
It is provided to the calculation unit 42.

The calculator 42 calculates data f (i−1, j), f (i, j−1), f (f) of four pixels around the pixel data (i, j).
From (i, j + 1) and f (i + 1, j), the average value of these data is determined by the following equation.

(I, j) = 1/4 {f (i-1, j) + f (i, j-1) + f (i, j + 1) + f (i + 1, j)} (1) The output of the latch 41 is added The average value of the calculation unit 42 is provided to the adders 43. Further, the output of the adder 43 is given to an adder 44 via a coefficient circuit 45, and the adder 44 gives g (i, j) = f (i, j) -β ｛(i, j) -f (I, j)... (2) The edge emphasis data g (i, j) is output. In this case, β is a coefficient that determines the degree of edge enhancement,
The coefficient is selected by the coefficient selection circuit 46 in accordance with the average value (i, j) of the four surrounding pixels. Here, assuming that the pixel level at which the blackout phenomenon occurs is L1 and the pixel level at which the whiteout phenomenon occurs is L2, 0 ≦ (i, j) <L1 → β = 10 (3) L1 ≦ (I, j) <L2 → β = 5 (4) L2 ≦ (i, j) ≦ 255 → β = 10 (5)

That is, the edge emphasizing circuit 4 includes the central pixel data f (i, j) shown in FIG. 3 (a) and the four pixel data f (i-1, j), f (i, j-
1), f (i, j + 1), f (i + 1, j)
It functions as an edge enhancement filter shown in FIG.

Returning to FIG. 1, the output of the edge enhancement circuit 4 is stored in the memory 5. The memory 5 also stores one byte per pixel, as in the memory 2 described above.

The data in the memory 5 is sent to the error diffusion processing circuit 6. The error diffusion processing circuit 6 executes a process by an error diffusion method. The data subjected to error diffusion processing by the error diffusion processing circuit 6 is sent to a binarization circuit 7.

Next, the operation of the embodiment configured as described above will be described.

Now, when the original is read line by line by the scanner 1, the scanner 1 outputs an analog signal corresponding to the pixel.

The analog signal is sent to the A / D converter 2, and for each pixel, 0 to 0 in which the gray level of black to white is 8 bits.
The image data is converted into 255-valued image data and stored in the memory 3.

The pixel data stored in the memory 3 is stored in an address control unit.
By the address designation of 31, as shown in FIG.
Around the pixel of (i, j) over the line, (i-1,
j), (i, j-1), (i, j + 1), (i + 1,
The data of the four pixels j) are read out and sent to the edge enhancement circuit 4.

From this state, the edge enhancement circuit 4 performs the edge enhancement processing shown in FIG.

First, in step A1, the central pixel data f (i, j) of the pixel data f read from the memory 3 is latched by the latch 41. At the same time, the four pixel data f (i−i) around the central pixel data f (i, j)
1, j), f (i, j-1), f (i, j + 1), f
(I + 1, j) is sent to the calculation unit 42, and the average value (i, j) of these pixel data is obtained from the above equation (1) (step A2).

Next, the process proceeds to step A3. In step A3, the magnitude of the average value (i, j) calculated by the calculation unit 42 is determined. In this case, the average value (i, j) is equal to the above (3) to
It is determined which of the equations (5) is applicable. Now, the average value (i, j) is between levels L1 and L2, and (4)
When it is determined that the expression is applicable, “5” is selected as the coefficient β for determining the degree of edge enhancement by the coefficient selection circuit 46,
In step A4, β = 5 is given to the coefficient circuit 45.

Next, the process proceeds to step A5. In step A5, edge emphasis data g (i, j) is calculated. In this case, first, the outputs of the latch 41 and the calculation unit 42 are given to the adder 43, and (−f) is calculated. And this adder 43
Is sent to the adder 44 together with the output of the latch 41 via the coefficient circuit 45 in which the coefficient β is set. Thus, the adder 44 calculates g (i, j) = f (i, j) -β {(i, j) -f (i, j)}, and the edge emphasis data g (i, j) is obtained. Is output. Then, the process proceeds to step A6, where the edge emphasis data g
(I, j) is written to the memory 5.

The data g (i, j) in this case is an average value (i,
Since j) to (4) are determined to be applicable and “5” is selected as the coefficient β for determining the degree of edge enhancement,
It becomes something that does not emphasize the edge extremely.

The above is the edge enhancement processing for one pixel, but the same processing is executed for all the pixel data f included in the j-th line of the memory 3 every time a new line is read by the scanner 1.

In step A3, when the average value (i, j) of the pixel data calculated by the calculation unit 42 is between the level L1 and “0”, and the equation (3) is applicable, the edge selection is performed by the coefficient selection circuit 46. “10” as coefficient β to determine the degree of emphasis
Is given to the coefficient circuit 45 in step A7, and the edge emphasis data g (i, j) is calculated in step A5 based on the coefficient β = 10. In this case, the data f (i, j) is determined to be a portion closer to black than surrounding pixels, and “10” is selected as the coefficient β for determining the degree of edge enhancement, so that strong edge enhancement is performed. .

Similarly, when the average value (i, j) of the pixel data calculated by the calculation unit 42 is between the level L2 and “255” and the equation (5) is applicable, the edge selection is performed by the coefficient selection circuit 46. “10” is selected as the coefficient β for determining the degree, and β = 10 is given to the coefficient circuit 45 in step A8. Based on the coefficient β = 10, the edge emphasis data g (i, j) is obtained in step A5. Is calculated. In this case, the data f (i, j) is determined to be a part close to white, and “10” is selected as the coefficient β for determining the degree of edge enhancement, and strong edge enhancement is performed.

Thereafter, the edge emphasis processing is executed in a similar manner, and the edge emphasis data g (i, j) obtained by this processing is sequentially written to the memory 5.

Next, the edge enhancement data written in the memory 5 is
The data is sent to the error diffusion processing circuit 6, where the error diffusion processing is performed.

Here, the error diffusion processing will be briefly described. Now, the third
As shown in FIG. 5C, when considering the data x of (i, k) in the memory 5, (i + 1, k), (i-1, k +) over two lines including the data x (i, k)
Attention is paid to the data of each coordinate of 1), (i, k + 1), and (i + 1, k + 1), and a diffusion coefficient as shown in FIG.

Then, first, the value of the data x is 128 ≦ x ≦ 255, 0 ≦ x
It is determined whether or not ≦ 127. If the value is “128” or more, the value of the data x is regarded as “255” (white). If the value is “127” or less, the value of the data x is set to “0”. (Black).

Next, when the value of the data x (i, k) is “128” or more, (x−255) × (an / s) + y = Y (6) When the value of the data x (i, k) is “127” In the following case, error diffusion is performed on the data of each coordinate by (x−0) × (an / s) + y = Y (7) Here, an is a diffusion coefficient given to each coordinate in advance, s is a total value of each diffusion coefficient, y is a data value of each coordinate,
Y is a data value to be stored in each coordinate after the calculation according to the above equation.

Now, assuming that the value of data x is "200", 1
28 ≦ x ≦ 255 holds and the value of the data x (i, k) is “2
55 ”. Next, error diffusion is performed on the data at each coordinate according to equation (6). In this case, (i +
Assuming that the coordinate data of (1, k) is, for example, “150”, “7” is added in advance as a diffusion coefficient, so that (200−255) × (7/16) + 150 = 126, and (i + 1) , K) with new data “126”
Is stored. Hereinafter, similarly, the remaining coordinates (i + 1,
k), (i-1, k + 1), (i, k + 1), (i +
For (1, k + 1), a new data value is calculated based on each data value and the diffusion coefficient, and is stored again at each coordinate.

In this way, for the coordinates (i, k), “255” or “0” is determined based on the data value x at this time, and one of the arithmetic expressions (6) and (7) is selected, and based on these expressions, Coordinates (i + 1, k), (i-1, k + 1),
For (i, k + 1) and (i + 1, k + 1), an error diffusion process in which a new data value is calculated based on each data value and diffusion coefficient is sequentially performed on each data in the memory 5.

The data subjected to the error diffusion processing by the error diffusion processing circuit 6 is sent to a binarization circuit 7, where the data is binarized, transferred to a printer (not shown), and printed out.

Therefore, in this case, the input image signal obtained from the scanner 1 is converted into multi-valued image data via the A / D converter 2, and the image data is stored in the memory 3. Pixel data and its surroundings 4
Is read out, the pixel data is supplied to the edge emphasizing circuit 4, and a coefficient β for determining the degree of edge emphasis is determined from the average value of the values of the surrounding four pixel data, and the determined coefficient β is used. Since edge-enhanced output data is obtained, it is difficult to express small changes in the image level in areas close to white and areas close to black in the past. The portion can be emphasized and expressed, so that the so-called underexposure or underexposure phenomenon on the reproduced image can be reliably avoided, and a high-quality image can be reproduced.

Note that the present invention is not limited to the above-described embodiment, and can be appropriately modified and implemented without changing the gist. For example, in the above-described embodiments, a facsimile apparatus has been described consistently, but the present invention can be applied to a copying apparatus.

[Effects of the Invention] The halftone binarization processing device of the present invention quantizes the halftone input image signal into multi-valued image data by analog-to-digital conversion, and stores the quantized data. Means for designating one pixel data from the storage means, and calculating an average value of pixel data around the specified pixel data, and calculating the calculated average value as a white area level or a black area level. When the pixel data is within a predetermined level area close to the predetermined level area, a larger edge enhancement degree is determined for the specified pixel data than when the pixel data is outside the predetermined level area, and the edge enhancement degree is determined based on the determined edge enhancement degree. Since the edge enhancement processing is performed on the designated pixel data and the edge-enhanced pixel data is binarized, the edge set by the pixel data around the designated pixel. Since the change in image level can be emphasized and expressed according to the edge enhancement level, even if there are other areas with a small level difference in an area close to white or an area close to black, these areas can be used. This makes it possible to emphasize the edge portions between the images, avoiding the so-called underexposure or underexposure phenomenon on the reproduced image, and obtaining a high-quality image that is easy to see visually. The image quality of the device can be improved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a circuit configuration of an embodiment of the present invention, FIG. 2 is a block diagram showing an edge emphasizing circuit of the embodiment, and FIG. 3 is a diagram for explaining the embodiment. FIG. 4 is a flow chart for explaining the operation of the embodiment. 1, scanner 2, A / D converter 3, 5, memory 4, edge emphasis circuit 41, latch 42, f calculator 43, 44, adder 45 ... coefficient circuit, 46 ... coefficient selection circuit, 6 ... error diffusion processing circuit, 7 ... binarization circuit.

──────────────────────────────────────────────────続 き Continuation of front page (56) References JP-A-58-88969 (JP, A) JP-A-61-144973 (JP, A) JP-A-61-123274 (JP, A) JP-A-1- 222575 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) H04N 1/40-1/409

Claims (1)

(57) [Claims] A quantizing means for converting a halftone input image signal from analog to digital to quantize it into multi-valued image data; a memory means for storing the quantized data; and a memory means for storing the quantized data. One of the calculated pixel data, and calculates an average value of pixel data around the specified pixel, and calculates the calculated average value within a predetermined level area close to the white area level or the black area level. When
Edge emphasis degree determining means for deciding a larger edge emphasis degree with respect to the specified pixel data than in the case where the edge emphasis degree is outside the predetermined level area; A halftone binarization processing device comprising: edge enhancement processing means for performing edge enhancement processing on designated pixel data; and binarization means for binarizing the edge-enhanced pixel data.