JPS613568A - Intermediate tone area identification system - Google Patents

Abstract

PURPOSE:To identify an intermediate gradation area at each small area with high accuracy on a real time processing ny discriminating it as the intermediate gradation area when a prescribed set of picture elements or below exist where the density difference exceeds a prescribed value. CONSTITUTION:Density level information for 16 picture elements in the block area extracted by a matrix register section 35 is outputted to selectorsd 36, 37 at the same time. The selectors 36, 37 select the density level information of the picture element to obtain the density difference and the accumulated sum according to selector signals A, B. Then a comparator circuit 40 compares the density difference with a prescribed value and the block area having a prescribed number of sets of below of the picture element coupled where the density difference exceeds a prescribed value in excess of the prescribed value is discriminated as the intermediate area. Thus, the intermediate gradation area identification is executed in the minute block area unit with high accuracy.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field] The present invention relates to a method for identifying halftone regions of a multivalued image.

[Prior art]

In image processing, it is advantageous to be able to identify whether an image is a halftone image or a binary image, since an appropriate processing method can be selected. Particularly, in the case of document images containing photographs, etc., it is desirable to distinguish between halftone areas and binary areas. For example, in the case of binarization processing of single-current image data.

If the halftone region and the binarization region can be distinguished, the dither method can be used for the halftone region, and the binary region can be binarized using a fixed threshold. Also, in the case of data compression,
Binary region data can be compressed using the MH method or MR method suitable for binary images, and halftone regions can be compressed using other suitable methods. Furthermore, when editing a single image by applying certain processing to a certain area of the image bit by bit, it is necessary to decide in advance whether contours or surface atmosphere are important, or the visual quality of the single image will change. The atmosphere is outstanding,
It deteriorates rapidly.

Now, as a typical technique for identifying halftone images, there is a special technique called "Parallel Two-Yield Segmentation f Method for Free Format Documents" in the Proceedings of the 20th National Conference of the Information Processing Society of Japan, 1974, pages 453 and 454. 1977-
No. 1005'49 "Image area signal generation method" is known.

However, in the former case, real-time processing cannot be expected because it requires the intervention of one person's judgment, such as checking the processing results and repeating the processing if dissatisfied. Another problem is that a buffer memory with a capacity for at least one screen is required.

Although the latter method is only disclosed for processing binary digital images, it can basically be extended and applied to multi-value digital images as well.

But when I try to apply it to multivalued digital images.

Creating and maintaining a mask pattern is a huge burden.

Furthermore, since one image is not processed by mesh-dividing it, it is difficult to simultaneously reduce processing time per pixel and improve identification accuracy. In other words, in order to shorten the processing time, it is necessary to reduce the number of scan lines stored in the buffer5, but if we do this, we have no choice but to make the mask smaller.
Identification accuracy decreases.

On the other hand, if the accuracy is to be improved, it is necessary to increase the number of scanning lines to be stored in the buffer and to make the mask larger, which reduces the processing speed and increases the burden of creating and holding the mask pattern.

〔the purpose〕

Therefore, an object of the present invention is to provide an intermediate area identification method that can accurately identify intermediate tone areas for each small area in real time processing with a simple device configuration.

〔composition〕

In a halftone image, the density level does not vary dynamically in a microscopic manner, but in a macroscopic manner. Therefore, in a halftone image, it is extremely rare for the density difference between adjacent pixels to exceed a predetermined value, compared to the case of a binary image. A large density difference occurs only in one contour portion. The present invention focuses on the difference in density variation between a halftone image and a binary image to identify a halftone region in a multivalued image.

According to the present invention, a multivalued image is divided into block regions each having a predetermined number of pixels. Within each block.

A plurality of pairs of pixels or image elements larger than pixels are selected, and the density difference between each pair of pixel elements is determined. Then, a block area in which the number of image element pairs whose density difference exceeds a predetermined value is less than or equal to a predetermined value is determined to be a halftone area.

The present invention will be specifically explained below based on one embodiment.

FIG. 2 is a conceptual diagram showing a pixel arrangement within a block area, and FIG. 3 is a flowchart of a process for determining whether a block area is a halftone area or a binary area. The details of the halftone area identification process will be explained with reference to these figures.

A multivalued image is divided into 4×4 pixel block areas as shown in FIG. There are only n C'/pairs of combinations of two pixels in the block area, and in this embodiment there is 1
The next four pairs of pixels are selected, and the absolute values D1 to D4 of the density differences between the pixels of each pair are determined. That is, in this embodiment, one pixel is used as the above-mentioned image element.

DI=I P (0,0)-P (2,2) ID2
=IP (0,3)-P (2,1)ID3=IP
(3,0)-P (1,2) ID4=IP
(3,3)-P (1,1) 11::,P
(i, j) is the density level of one pixel (i, j), which corresponds to FIG. 1, respectively.

Furthermore, the cumulative sum S of the density levels of the eight pixels is determined as a parameter representing the average density of the block area.

S=P (0, O) 10P (2,2) 10P (0,
3)+p (2,1)+P (3,0)+P (1,2
)+P (3, 3)+P (1+1) The following description will be made with reference to FIG. First, counter M is cleared (step 10). The density difference D1 is compared with the density difference judgment value (step 11), and if Di>k, 9 is added and 1 is added to the counter M (step 12).
. Next, the density difference D2 and the determination value are compared (step 13), and if D2>k, 1 is added to the counter M (step 14). Similarly, the density difference D3.04 and the judgment value are compared (step 15.17), and D3>k, D
If i>k, then 1 is added to the counter M (step 16.18).

The value of the counter M at the time when the judgment of Di is completed is.

This is the number of pixel pairs whose density difference exceeds k.

Next, the value of the counter M and the determination value m are compared (step 19). If M > m, the block area is 2
The value area is determined (step 20).

If M≦m, the block area is likely to be a halftone area. Although it is possible to immediately determine such a block area as an intermediate m area, in this embodiment, the cumulative sum S is further compared with the determination values u and v (step 21). and.

If U≦S≦V, the block area is determined to be a halftone area (step 22); otherwise, it is determined to be a binary area (step 20).

This kind of judgment regarding the cumulative sum S is performed when all pixels in the block area are at a nearly uniform density level when the block area is part of a spatially large thick black line or a background area. This is to avoid determining that such a block area is a halftone area. In other words, it is determined whether or not it is a white background by comparing it with U,
By comparing it with V, it is determined whether it is a solid black part or not.

FIG. 1 shows an example of the configuration of an image processing apparatus according to this embodiment. In this figure, multivalued input image signals are alternately input to buffer memories 3 and 32 in units of four sub-scanning lines and stored therein. for example.

During the period when the multi-value input image signal is input to the buffer memory 31, the image signal of four sub-scanning lines four sub-scanning lines before is read out from the buffer memory 32.

It is input to the matrix register section 35 via the changeover switch 34, and the 2
It is input to a value image processing circuit 46 and a halftone image processing circuit 47.

The matrix register section 35 is a 4×4 matrix register shown in FIG.
This is for extracting pixel density information of a pixel block area, and is composed of a 4-digit, 4-stage shift register. The image signal output from the buffer memory 31 or 32 is directly input to the first stage shift register of the matrix register section 35, and is input to the first, second, and third stage shift registers.
The accumulated information of the digits is stored in line buffers 49, 50, .
2nd, 3rd, 4th stage after being delayed by 51
) entered into the eyes. The density level information of 16 pixels in the block area extracted by this matrix register section 35 is simultaneously output to selectors 36 and 37.

Select j3s, 37 is a select signal A.

According to B, the density level information of the pixels for calculating the density differences D1 to D4 and the cumulative sum S is selected, and the difference calculation circuit 38
and is input to the cumulative sum calculation circuit 39. The difference calculation circuit 38 calculates the density difference DI-D4. Comparison circuit 4
0 compares the density difference Di-D4 with the judgment value and outputs a signal when the density difference exceeds the judgment value (corresponds to steps 11, 13, 15, and 17 in FIG. 3). counter 43
counts the output signal of the comparison circuit 40 (corresponding to step 12.li 1B, 1B in FIG. 3).

This counter 43 is cleared at the time of block area switching. The comparison circuit 44 compares the value of the counter 43 and the judgment value m, and when the counter value is less than or equal to the judgment value m, the I
Output the I 1 #l signal (corresponding to step 19 in FIG. 3). On the other hand, the comparison circuit 42 includes the cumulative sum calculation circuit 3
The cumulative sum S obtained in step 9 is compared with the 1 judgment values u and v, and if U≦S≦V, the II l #l signal is output (corresponding to step 21 in FIG. 3). The output signal of the comparison circuit 42 is ANDed with the output signal of the comparison circuit 44, and the AND signal is stored in the image characteristic storage circuit 45 in a one-to-one correspondence with the block area.

By the way, during the period when the multi-value input image signal is stored in the buffer memory 31, the stored image signal in the buffer memory 32 is read out and inputted to the matrix register section 35, but after that, the buffer memory 32 is in the input mode. When switching to , the accumulated image signal is read out again at the same time as the multi-value input image signal is input.

It is input to a binary image processing circuit 46 and a halftone image processing circuit 47. This 1: means buffer memory 31.
The same applies to any of 32.

The binary image processing circuit 46 receives the input image signal.

Perform processing suitable for binary images. The halftone image processing circuit is
Perform processing suitable for halftone images on the input image signal. For example, in the case of binarizing an image signal, the binary image processing circuit 46 binarizes the input image signal using a fixed threshold value, and the halftone image processing circuit 47 processes both input signals using the de-Isa method. In synchronization with the second process, data for the corresponding block area is read from the image characteristic storage circuit 45, and is input to the selector 48 as a selection signal. When the selection signal is LL O11, that is, in the case of a binary region, the selector 48 selects and outputs the processing result of the binary image processing circuit 46.

If the selection signal is 1'', this is because it is in the intermediate tone area.

The selector 48 selects and outputs the output signal of the halftone image processing circuit 47.

Although one embodiment has been described above, the present invention is not limited only to the above embodiment.

For example, the shape and number of pixels of the block area.

What are the combinations of pixels for determining the density difference, the number of such combinations, and the number of pixels for determining the cumulative sum?

You may change it as necessary. Moreover, the average value may be used instead of the cumulative sum.

Furthermore, instead of finding the density difference between two pixels, the density difference between an image element larger than the pixel may be found.

For example, two adjacent pixels may be considered as one image element, and the density difference may be determined between the image elements.

By doing this, when a certain pixel has an abnormal density level compared to surrounding pixels, it is possible to prevent erroneous determination due to the abnormal density level.

Furthermore, the processing may be executed by software as appropriate.

〔effect〕

As is clear from the above description, according to the present invention, it is possible to perform high-accuracy halftone region identification of minute block regions of the order of m, and it is possible to perform high-precision halftone region identification of minute block regions. When processing, etc., it is now possible to perform appropriate processing according to each image, and real-time processing is possible with a simple device equipped with a relatively small buffer memory. You can get

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an example of a device according to an embodiment of the present invention, FIG. 2 is a conceptual diagram for explaining the pixel arrangement of a block area in the same embodiment, and FIG. 3 is an intermediate diagram in the same embodiment. 5 is a flowchart showing key area determination processing. 31.32...Buffer memory, 35...Matrix register section, 49,50.51...Line buffer, 36.37.48...Selector. 38... Difference calculation circuit, 39... Cumulative sum calculation circuit. 40.42.44...Comparison circuit, 43...Counter, 45...Image characteristic storage circuit, 46...2
Value image processing circuit, 47... Halftone image processing circuit. Figure 1

Claims (1)

[Claims] (1) Divide the multivalued image into block areas of a predetermined number of pixels,
In each block area, the density difference between pairs of image elements that are equal to or larger than a plurality of pixels is determined, and the block area where the number of image element pairs whose density difference exceeds a predetermined value is less than or equal to a predetermined number is determined as a halftone area. A halftone area identification method characterized by determining that.