JPS63205783A - Picture tone discriminating device - Google Patents

Abstract

PURPOSE:To accurately discriminate a continuous image tone from a linear image at every block unit by dividing a digital image into picture element blocks calculating their mean value and a standard deviation value, setting up threshold corresponding to the standard deviation, and comparing the threshold with the means value. CONSTITUTION:When the mean value (m) of the picture element blocks is in the vicinity of the mean (Max+Min)/2 of the maximum value Max and the minimum value Min of image data, the threshold of the reference deviation is set up to large value Th1. As the mean value (m) reaches the Max or Min, the threshold is set up to a small TH2-TH5. Since the threshold for discriminat ing the picture tone is changed such as TH1-TH5 in accordance with the mean value (m) of the picture element density of the picture element blocks, picture element blocks 302, 304 are decided as picture element block including a linear image, i.e. an edge and a picture element block corresponding to a position 406 is accurately decided as a plane block.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Field of Application] The present invention relates to an image tone identification device that identifies the image tone of digital image data having gradation.

[Prior Art] Various methods have been proposed for distinguishing digital image data having gradation into line drawings such as characters and continuous tone images such as photographs. This performs image processing based on the image tone identification results.
The aim is to obtain good processing results by performing the processing on each image area. For example, when image data is binarized and reproduced with a dot printer such as a laser beam printer, image reproduction with emphasis on resolution is achieved by simply binarizing the line drawing portion of the image data, while continuous tone image Image parts where gradation is important, such as the above, are reproduced by performing dither processing, for example, to obtain a high-quality reproduced image. This method is especially
This is effective when binarizing an image in which text and photographs coexist because the binarization method can be automatically switched according to the image tone identification result.

As a method for identifying this image tone, there is a method that identifies each block of multiple pixels, and this method identifies the image tone from quantities that represent the contrast within each block, such as the variance value and standard deviation value. be.

In general, pixel blocks in line drawing portions include edges, so the contrast is high when viewed on a block-by-block basis, and the standard deviation of pixel data of such pixel blocks becomes a large value. On the other hand, since the contrast within the pixel splitter of an image such as a photograph is low, the standard deviation has a small value.

Taking the letter "A" in FIG. 3 as an example, the deviation value takes a large value in blocks including edges such as pixel blocks 302 to 304, but
In a block such as , since the entire block is included within the line, its pixels have almost the same density level, and the standard deviation value is small.

Therefore, some methods compare the standard deviation value of each pixel block with a predetermined threshold value and use the comparison result as the image tone identification result.

[Problems to be Solved by the Invention] However, setting a constant value as the threshold in this way may lead to erroneous determination. That is, for example, pixel blocks such as pixel block 302 and pixel block 304 in FIG. 3 have a smaller standard deviation value than a pixel block such as pixel block 303 in which black and white pixels are equally present. . That is, when comparing the standard deviation value with a certain threshold value, pixel blocks 302 and 304 are determined to be continuous tone pixel blocks depending on the threshold value. Conversely, if the threshold value is set low to prevent this, a pixel block that should be determined as continuous tone will be determined to be a line drawing pixel block. In particular, in the case of image data read by a reading system with a low MTF, the standard deviation value is low overall, so the possibility of misjudgment increases.

The present invention has been made in view of the above-mentioned conventional example, and an object of the present invention is to provide an image tone identification device that can accurately identify the image tone of digital image data having gradation in units of pixel blocks.

[Means for Solving the Problems] In order to achieve the above object, the image tone identification device of the present invention has the following configuration. That is, an extraction means for manually extracting a pixel block consisting of a plurality of pixels from image data; a calculation means for calculating a characteristic value based on a representative density value of the pixel block and a density difference between each pixel of the pixel block; The image forming apparatus includes a setting means for setting a threshold value corresponding to the feature amount based on the representative density value, and a comparison means for comparing the threshold value and the feature amount.

[Operation] In the above configuration, image data is input, a pixel block consisting of a plurality of pixels is extracted, and a representative density value of the pixel block and a feature quantity based on the density difference of each pixel of the pixel block are calculated. A threshold value is set based on this representative density value, and the comparing means compares the threshold value with the feature amount to perform image tone identification.

[Embodiments] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[Explanation of Principle (FIGS. 3 to 5)] FIG. 4 is a diagram showing the relationship between the density average value m and the standard deviation value σ of image data.

The maximum value of the density of the target image data is Ma: (, the minimum value is Min, the average value within a predetermined pixel block is m,
Assuming that the standard deviation value is σ, these values exist within the range shown by the following formula.

If this is shown on an orthogonal two-dimensional plane, it becomes an area surrounded by a semicircle 401 in FIG. Here, it is considered that pixel blocks including edges exist in a region close to the arc of the semicircle 401, and pixel blocks not including edges exist in a region close to the m-axis.

Each pixel block 301 to 304 in FIG. 3 is shown in FIG.
Corresponding to each position 402 to 405 indicated by ', position 4
02 is pixel block 301, position 403 is pixel block 302, position 404 is pixel block 303, position 405
correspond to the pixel blocks 304, respectively.

5(A) to 5(D) are diagrams showing the relationship between each pixel and the density corresponding to each pixel. corresponds to each concentration of As is clear from these figures, even blocks having edges of the same contrast have different average values M depending on the ratio of the number of black pixels to white pixels. For example, the pixel block 303 (position 404) consisting of uniform pixels in FIG. It can be determined that the pixel block contains the pixel block.

On the other hand, pixel blocks 302 (position 403) and 304 (position 405) are located near the arc of semicircle 401;
It is located close to m1li[b because the standard deviation is small.

FIG. 6 is a one-dimensional representation of pixel data of a block included in a continuous tone picture such as a photograph, where the density level is around the average value m of the maximum and minimum levels of pixels in the pixel block. It is. This pixel block is
It has an average value and a standard deviation value shown at position 406 within semicircle 401 in FIG. This pixel block is at position 40
3 (pixel block 302) and position 405 (pixel block 304).

Therefore, as shown in FIG. 4, if the threshold value 410 is set as the discrimination threshold, the pixel block having a standard deviation value larger than the threshold value 410 is treated as a line drawing area, and the threshold value 410
7. Pixel block 302 and pixel block 304 are determined to be line drawings, but the pixel block that should be determined to be continuous tone image (corresponding to position 406) ) is also incorrectly determined to be a line drawing. Therefore, if the threshold value is raised to 411, the pixel block corresponding to position 406 is correctly determined to be a continuous tone image, but this time pixel block 302 and pixel block 304 are incorrectly determined to be continuous tone images.

[Explanation of Image Identification in this Embodiment (FIG. 2)] FIG. 2 is a diagram for explaining image tone identification in this embodiment, and parts common to FIG. 4 are indicated by the same symbols.

As shown in the figure, the average value m of the pixel block is the average value (Max+Mi) of the maximum value Max and minimum value Min of the image data.
n)/2, the threshold value of the standard deviation value is increased to THI. Then, as the average value m approaches the maximum value Max or the minimum value Min, the threshold value is decreased from TH2 to TH5. In this way, the average value m of the density of pixels in a pixel block
By changing the threshold for image tone identification from THI to TH5, pixel block 302 and pixel block 304 are determined to be line drawings, that is, pixel blocks containing edges, and the pixel block corresponding to position 406 is correctly determined to be a flat block. can do.

[Description of Image Processing Section (FIG. 1)] FIG. 1 shows the configuration of the image processing section of this embodiment, which performs image tone identification according to the above-mentioned criteria.

In the figure, reference numeral 1 denotes an image processing section that manually inputs digital image data, and is comprised of, for example, a memory that stores digital image data or a manual section that receives external image data and manually inputs it. 2 is input image data extracted in units of a predetermined number of pixels; for example, in this embodiment, it is extracted in 4×4 pixel blocks. 3 is the average value 4 (m) and standard deviation value 5 (σ) of each pixel density value of pixel block 2
This is a calculation unit that calculates . The average value 4 from the calculator 3 is R
It is input as the address of OM6. The average value 4 is manually entered into the ROM6 as an address, and the threshold value THI shown in Fig. 2 is set.
~TH5 is stored as data. Therefore, the threshold value 7 from the ROM 6 and the standard deviation value 5 from the arithmetic unit 3 are compared by the comparator 8, and when the standard deviation value 5 is larger, the selection signal 9 becomes 0.1''.

The pixel data of the pixel block 2 is processed by the image processing unit 10 for line drawing.
and the image processing unit 11 for continuous tone images, and image processing corresponding to each tone is performed in parallel.

The selector 14 selects the A manual input when the selection signal 9 is 1°, and selects and outputs the B input when the selection signal 9 is 0. As a result, when the standard deviation value 5 is larger than the threshold value 7, it is assumed that the pixel block 2 is a line drawing block, and the image data 12 processed by the line drawing image processing unit 10 is sent to the image output unit 16 as image data 15. Output. Conversely, when the standard deviation value 5 is smaller than the threshold value 7, the selection signal 9 becomes "o", and the output 13 of the image processing section 11 for processing the gradation image is selected and outputted to the image output section 16. The image output section 16 is composed of a printer such as a laser beam printer, for example, and reproduces the image by manually inputting the image data 15.

In this embodiment, the standard deviation value is used as the amount representing the contrast of the pixel block, but the invention is not limited to this. For example, instead of the standard deviation value, the variance value, which is the square value of the standard deviation value, or the pixel block The difference value between the maximum level and the minimum level of pixels within the range may be used.

In addition, although the average value of the pixel block was used as the value to be compared with the contrast amount, the sum of the maximum level and the minimum level or one of them
An average value of /2 may be used.

Generally, it is easier to detect the maximum and minimum values within a pixel block than to calculate the standard deviation value or average value, and when trying to configure an inexpensive device, it is necessary to detect the maximum and minimum values within a pixel block. It is preferable to use the difference value and the sum value of the pixel block as the feature value and the representative value, respectively.

As explained above, according to this embodiment, a digital image is divided into pixel blocks, the average value and standard deviation value are calculated, a threshold value is set corresponding to the standard deviation value, and the threshold value is compared with the average value. This has the effect of being able to accurately distinguish between continuous tone and line drawings on a block-by-block basis.

[Effects of the Invention] As described above, according to the present invention, there is an effect that the tone of image data can be identified with high accuracy.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of the image processing unit of this embodiment, FIG. 2 is a diagram showing the criteria for determining image tone in this embodiment, and FIG. 3 is a diagram showing an example of a pixel block of a character image. Figure 4 is a two-dimensional plane representing the existence range of the average value and standard deviation value of a pixel block, and Figures 5 (A) to (D) represent the density levels of pixels in each pixel block in Figure 3. FIG. 6 is a diagram showing density levels of pixels in a pixel block that can be considered as a part of a continuous tone image. In the figure, l: image input section, 2: pixel block, 3
...Arithmetic unit, 4...Average value, 5...Standard deviation value,
6...ROM, 7...Threshold value, 8...Comparator, 9...
... Selection signal, 10.11... Image processing section, 14...
・Selector, 16... Image output section, 301 to 304・
...It is a pixel block. Figure 2 I-ju → Figure 5 Figure 6

Claims (3)

[Claims] (1) Extraction means for inputting image data and extracting a pixel block consisting of a plurality of pixels; and calculation means for calculating a characteristic amount based on the representative density value of the pixel block and the density difference of each pixel of the pixel block. An image tone identification device comprising: a setting means for setting a threshold value corresponding to the feature amount based on the representative density value; and a comparison means for comparing the threshold value and the feature amount. (2) The image tone identification device according to claim 1, wherein the representative density value is an average density value of pixels in a pixel block. (3) The image tone identification device according to claim 1, wherein the feature amount is a standard deviation value of pixel density of a pixel block.