JPH05167852A - Image clarifying method - Google Patents

Abstract

PURPOSE:To execute a high speed processing by eliminating roughness of a flat part of density, emphasizing an outline part in accordance with a desire, and also, executing an operation by a look-up table(LUT), at the time of definition processing of an image. CONSTITUTION:In image data inputted from an image input part 310, a value of secondary differentiation of a notice picture element is derived by an arithmetic part 314, and its value is set as an evaluation function F. A defining LUT part 316 stores an LUT of the evaluation function F, and a value of the product kF of a coefficient (k) of strength of definition and the evaluation function F. By referring to the LUT by the evaluation function F of the notice picture element, the value of the product kF is derived. The arithmetic part 314 derives a value h(i, j) of a defined picture element by substituting the value of derived kF in an expression [h(i, j)=f(i, j)-kF].

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of sharpening image data.

[0002]

2. Description of the Related Art Conventionally, a Laplacian processing method has been known as a method for sharpening image data.

This Laplacian process is a sharpening process by quadratic differentiating (Laplacian) the value f (i, j) of each pixel of interest in the image data. That is, if the processed pixel value is g (i, j), the value f (i, j) of the pixel of interest can be sharpened by Laplacian in the following expression (1).

[0004] g (i, j) = f (i, j) -M {d 2 f (i, j) / di 2 + d 2 f (i, j) / dj 2} = f (i, j) - M {f (i + 1, j) + f (i-1, j) + f (i, j + 1) + f (i, j-1) -4.f (i, j)} (1)

Note that M is 1 or a constant.

Such a Laplacian sharpening method corresponds to, for example, an unsharp masking process in a drum type scanner. However, in this method, since the entire image is sharpened uniformly, the roughness in the flat portion becomes conspicuous. This tendency is particularly noticeable when printing image data is input.

Some drum scanners have a function of setting a sharpening effect by a density difference called a starting point in order to prevent such roughness from being conspicuous. Further, in the highlight and shadow portions, there is one having a function of independently controlling the strength of sharpening processing.

[0008]

The image data includes data read by the drum scanner as described above and data read by the plane scanning type scanner. The image data obtained by the flatbed scanner has a lower pixel density than the image data obtained by the drum scanner.

Therefore, if the conventional Laplacian sharpening process is applied to the image data read by the plane scanning type scanner, there is a problem that the roughness in the flat portion becomes noticeable.

The present invention has been made to solve the above-mentioned conventional problems. In the sharpening process of image data, an image which can enhance a contour portion without causing roughness in a flat portion can be obtained. An object is to provide a sharpening method.

[0011]

According to the present invention, in a sharpening method of image data, a value of a secondary characteristic of a target pixel f (i, j) is used as an evaluation function F, and the evaluation function F and the sharpening A lookup table is created with the value of the product k F of the strength coefficient k and the evaluation function F, and the evaluation function F of the target pixel f (i, j) is created.
Is obtained, the value of the product k F is obtained by referring to the lookup table by the obtained evaluation function F, and the obtained product k is obtained.
The above problem was solved by substituting the value of F into the equation h (i, j) = f (i, j) -kF to obtain the sharpened pixel value h (i, j). It is a thing.

[0012]

The inventor has variously studied the sharpening processing of image data having a low pixel density read by a flatbed scanning type scanner so that the roughness in the flat portion is not noticeable.

Here, the drum type scanner has a function of setting a sharpening effect by a density difference called a starting point and a function of independently controlling the strength of highlight and shadow. There is something.

The inventor of the present invention has devised the present invention in order to easily obtain the effect of setting the sharpening intensity and independently controlling the highlight and the shadow.

That is, in the present invention, the value of the second derivative of the pixel of interest is used as the evaluation function.

If the pixel of interest is f (i, j), it is 2
The evaluation function F of the value obtained by the next differentiation is as in the following expression (2).

F = f (i + 1, j) + f (i-1, j) + f (i, j + 1) + f (i, j-1) -4f (i, j) ... (2)

The value of the evaluation function F represents the density difference or smoothness ratio of the target pixel of the image.

The evaluation function F and the sharpening strength coefficient k
And a value of the product k F of the evaluation function F and a lookup table are created. The evaluation function F of the pixel of interest is found, and the value of the product k F is found from the obtained evaluation function F with reference to this lookup table. The value of the product k F obtained is calculated by the following equation (3)
To obtain image data h (i, j) obtained by sharpening the image data f (i, j).

H (i, j) = f (i, j) -k {f (i + 1, j) + f (i-1, j) + f (i, j + 1) + f (i, j-1)- 4f (i, j)} = f (i, j) -k * F ... (3)

This equation (3) is the pixel of interest as shown in FIG.
It can be represented by an operator for f (i, j).

When the value of the evaluation function F shown in the equation (2) is small, the density change is small, and when the value of the evaluation function F is large, the density change is large. Further, it can be considered that the shadow gain is obtained when the value of the evaluation function F is negative, and the highlight gain is obtained when the value of the evaluation function F is positive. Therefore, by changing the value of the coefficient k and the product kF of the evaluation function F according to the value of the evaluation function F, the sharpening strength can be set by the density difference, and the shadow gain and the highlight gain can be independently set. It is possible to control and sharpen. Therefore, the flat portion is not roughened, and the contour portion can be emphasized.

Further, if the value of the product k F of the coefficient k of sharpening strength and the evaluation function F is calculated every time the value of the evaluation function F of each target pixel is calculated, it takes a long processing time. Therefore, in the present invention, in order to shorten the processing time, the evaluation function F
And a value of the product k F is to be created as a look-up table. Since the value of the product k F is looked up in the look-up table, it is not necessary to perform each operation individually, and quick sharpening processing can be performed. Therefore, the processing speed does not slow down even if the number of thresholds is increased. This can greatly contribute to increase the processing speed in the image processing computer.

[0024]

Embodiments of the present invention will now be described in detail with reference to the drawings.

A sharpening device for sharpening image data according to an embodiment of the present invention will be described.

This sharpening device has a structure as shown in FIG.

As shown in FIG. 1, the image sharpening apparatus has an image input section 3 for inputting image data to be processed (for example, output from a plane scanning type scanner).
10, an image memory 312 for recording an input image, and an evaluation function F is obtained by performing the operation shown in the formula (2) on the recorded image data, and a look-up table (hereinafter abbreviated as LUT) by the evaluation function. The product k of the sharpening strength coefficient k and the evaluation function F
Substituting the value of F into the equation (3), the pixel value h (i, j)
Calculating unit 314 for sharpening the image by obtaining the above, the evaluation function F, the sharpening strength coefficient k and the evaluation function F
LUT with the value of the product k F of the target pixel and the sharpening LUT unit 316 for obtaining the value of the product k F by referring to the LUT by the evaluation function F of the pixel of interest and stored in the image memory 312. An image output unit 318 that outputs the image data after the sharpening process to an outside (for example, a recording medium such as a magnetic tape).

The equation (3) can be expressed by an operator for the pixel of interest f (i, j) as shown in FIG.

LUT of the sharpening LUT unit 316
Is created in advance by calculating the value of the product k F with respect to the evaluation function F. This LUT outputs various values of the product k F to the evaluation function F in the relationship shown in FIG. 3, for example. The evaluation function F represents the smoothness rate.

That is, in FIG. 3, the slope of the straight line represents the magnitude of the coefficient k of sharpening strength, and x ₀ to x ₂ ,
x ₀ ′ to x ₂ ′ are thresholds, k ₀ to k ₂ , k ₀ ′ to k ₂ ′.
Represents the coefficient k at that threshold. In addition, FIG.
As shown therein, when the evaluation function F is 0 or more and less than x ₀ , k = k ₀ , and when x ₀ or more and less than x ₁ , k =
When k ₁ and x ₁ or more and x ₂ or less, k = k ₂ . Note that there is a relationship of k ₀ <k ₁ <k ₂ .

[0031] Further, the evaluation function F is x ₀ below 0 'when it exceeds a is, k = k _0', 'when it exceeds a, k = k _1' x ₁ in x ₀ below x ₁ ' If less than or equal to x ₂ ′, then k = k ₂ ′.

FIG. 3 shows an example of the LUT.
Since a free coefficient k can be set according to the evaluation function F, for example, when the value of the evaluation function F is negative, it is considered as a shadow gain, and when the value of the F is positive, it is considered as a highlight gain. The value of the coefficient k for F can be set independently. This allows you to control the strength of sharpening on the highlight side and the shadow side.
It is possible to improve the reproducibility of white characters and black characters.

The operation of the image data sharpening device will be described below.

In the image data sharpening device,
First, image data is input from the image input unit 310 and output to the image memory 312. The image memory 312 stores input image data.

The evaluation data F is obtained by substituting the image data output from the image memory 312 into the equation (2).

Then, the sharpening LUT unit 316 refers to the LUT by the evaluation function F to obtain the value of the product k F of the coefficient k of the sharpening strength and the evaluation function F.

Then, the arithmetic unit 314 calculates the product k
Substituting the value of F into equation (3) above, the pixel of interest f (i,
The value of the pixel h (i, j) obtained by sharpening j) is obtained and the image data is sharpened.

Since this image sharpening apparatus has the LUT shown in FIG. 3, the sharpening strength can be set by the evaluation function F, that is, the density difference, and therefore the flat portion (evaluation function F
If the value of k is reduced near 0), the flat image after processing will not be rough, and if the value of k is increased at the contour (the positive or negative large part of the evaluation function F), the processed image will be rough. It is possible to enhance the image. Also in the highlight portion and the shadow portion, as shown in FIG. 3, they can be controlled independently by changing the slope of the straight line of the coefficient k.

Here, an example of image data processed by the image sharpening device will be described. FIG. 4 shows an example of image data before processing, and FIG. 5 shows an example of image data processed by the sharpening device. For comparison, FIG. 6 shows an example of image data sharpened by the conventional Laplacian of the formula (1).

In the conventional method shown in FIG. 6, the flat portion fluctuates upward and downward, resulting in large roughness. On the other hand, as shown in FIG. 5, in the sharpening device, it can be seen that such a variation is small and the roughness is small.

In the sharpening processing apparatus of the above embodiment, x ₀ to x ₂ which determine the coefficient k by the evaluation function F, and
Although x ₀ ′ to x ₂ ′ are used as threshold values for determining the coefficient k, the number of threshold values can be set arbitrarily.
Even if the number of the thresholds is increased, the LUT method of referring to the LUT to obtain the value of the product k F of the coefficient k of the sharpening strength and the evaluation function F is adopted. It won't be late.

[0042]

As described above, according to the present invention, when the image is sharpened, it is possible to prevent the roughness from occurring in the flat portion where the density changes little and to emphasize the contour portion arbitrarily. Moreover, since the gains of highlight and shadow can be controlled independently, a desired image expression can be obtained. Further, since the look-up table is used, it is not necessary to obtain the sharpening strength by calculation, and therefore, the excellent effect that the image processing can be speeded up can be obtained.

[Brief description of drawings]

FIG. 1 is a block diagram showing an overall configuration of an image sharpening device according to the above embodiment.

FIG. 2 is a plan view showing an operator example of an image sharpening process according to the present invention.

FIG. 3 is a diagram showing an example of a look-up table of the image sharpening device.

FIG. 4 is a diagram showing an example of an original image for explaining the effect of the sharpening processing of the present invention.

FIG. 5 is a diagram showing an example in which the original image is sharpened by the image sharpening device.

FIG. 6 is a diagram similarly showing an example in which the original image is sharpened by a conventional method.

[Explanation of symbols]

 310 ... Image input unit, 312 ... Image memory, 314 ... Calculation unit, 316 ... Sharpening look-up table unit, 318 ... Image processing unit.

Claims (1)

[Claims] 1. A sharpening method of image data, wherein a value of a secondary characteristic of a pixel of interest f (i, j) is used as an evaluation function F, an evaluation function F, a coefficient k of sharpening strength, and an evaluation function. A lookup table with the value of the product k F of F is created, the evaluation function F of the pixel of interest f (i, j) is obtained, and the lookup table is referred to by the obtained evaluation function F to refer to the product k The value of F is obtained, and the value of the obtained product k F is substituted into the equation h (i, j) = f (i, j) -k F to obtain the sharpened pixel value h (i, j). An image sharpening method characterized by obtaining.