JPH1166300A - Method for emphasizing sharpness - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent grains and quantizing step differences from being emphasized in a sharpness. SOLUTION: A first averaging signal Sma and the second averaging signal Smb in two kinds of mask sizes M1=5×5 and M2=11×11 are calculated and, after that the absolute value signal absSd of their difference signal Sd is calculated. When the difference signal Sd is smaller than a threshold value th1, it is discriminated that a part is the flat part of the image, in other words, the flat part of an image density of the part with few density fluctuation so as to set the value of an emphasizing coefficient (h) to a zero value. In this case, unsharp masking signals (h) and (k)(S-U) is made to be the zero value so that a sharpness emphasizing signal S** is made to be S so as to output the signal without sharpness emphasis. The flat part of the picture is not sharpness-emphasized so that the particles and the quantizing step are not emphasized.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image (image) enhancement, and more particularly to a sharpness enhancement method suitable for application to a printing plate making scanner or the like.

[0002]

2. Description of the Related Art Conventionally, in the field of image processing, a sharpness enhancement method has been adopted in order to improve the sharpness of an image.

[0003] As is well known, this sharpness emphasizing method uses, for example, nxn from the periphery of an image signal (sign is S) before sharpness emphasis obtained from a subject (including an image original). Unsharp signals (also referred to as a local average signal, denoted by U) are obtained by taking out image signals with the number of mask sizes and averaging them.
Create Next, the difference signal between the image signal S and the unsharp signal U, ie, unsharp masking (US
M) Also called a signal. ｝ S−U is obtained, and the difference signal S−U
Is multiplied by a predetermined coefficient k (the value of k is usually selected to be a value between 0 and 1) to obtain unsharp masking (U
Create an SM) signal (code is USM).

As a result, an image signal S ^* (referred to as a sharpness enhanced signal) subjected to a sharpness enhanced process defined by the following equation (1) is obtained.

S ^* = S + USM = S + k · (SU) (1) Then, based on the sharpness enhancement signal S ^* , for example, after binarization processing is performed, halftone dot imaging processing is performed. The halftone image is scanned and formed on a record carrier such as a film. Then, for example, a printing plate is created from the film, and the printing plate is mounted on a printing press to create a printed matter on which a halftone image is formed.

[0006]

By the way, when the image signal is subjected to the sharpness enhancement according to the prior art described above,
On the image, in a flat portion of the image density, in other words, in a portion where the density is constant or gently changing, so-called granular sharpness is applied, for example, on an image formed on the printed matter. There was a problem that the roughness was emphasized.

Further, when the sharpness enhancement according to the above-described prior art is performed, the sharpness enhancement is applied to the quantization step of the image signal in the portion where the image density changes gently, and the sharpness is formed on the printed matter. There was also a problem that streaks like tone jumps appeared on the image.

In order to solve this problem, it is conceivable that sharpness enhancement is not applied to a portion where the density changes about the same as the quantization step, but in such a case, the object itself such as fabric, skin, metal, etc. There is a new problem that sharpness enhancement is not applied to the material texture (feeling as a material), so-called texture (also referred to as texture or texture texture), and the texture texture is lost.

The present invention has been made in view of such a problem, and it is possible to enhance an image having a small contrast such as a texture feeling without sharpness enhancement being applied to granularity and quantization steps. The object of the present invention is to provide a sharpness emphasis method.

[0010]

According to the present invention, an unsharp signal is generated from an input image signal, and a USM (unsharp masking) signal is generated from the generated unsharp signal and the input image signal. In a sharpness enhancement method for generating an image signal after sharpness enhancement processing by combining a signal and the input image signal, an average value in two kinds of ranges is obtained for an image signal around the input image signal, The flatness of the image density is determined by comparing the obtained average values, the magnitude of the USM signal is changed according to the determined flatness, and the USM signal is combined with the input image signal, and the image after the sharpness enhancement processing is performed. It is characterized in that a signal is generated.

According to the present invention, the flatness of the image density can be easily determined, and the sharpness can be enhanced according to the flatness of the determination result.

In this case, when determining the flatness of the image density, the absolute value of the difference between the obtained average values is determined, and if the absolute value of the difference is smaller than a predetermined threshold, U
By setting the magnitude of the SM signal to a zero value, sharpness enhancement is not applied to graininess and quantization steps.

In addition, a first threshold value and a second threshold value larger than the first threshold value are set in advance, and when determining the flatness of the image density, the absolute value of the difference between the obtained average values is determined. , When the absolute value of the difference is smaller than the first threshold,
The magnitude of the USM signal is set to a zero value, and when the absolute value of the difference is larger than the first threshold and smaller than the second threshold, the magnitude of the USM signal is gradually increased according to the absolute value of the difference. , If the absolute value of the difference is greater than a second threshold,
By setting the SM signal to a fixed size, it is possible to enhance an image having a small contrast such as a texture feeling without sharpness enhancement being applied to a granularity or a quantization step.

By selecting two types of ranges around the input image signal to appropriate values according to the color of the image signal, it is possible to perform optimum sharpness enhancement processing for each color plate.

[0015]

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows a basic configuration of a printing scanner 10 as an example of an image scanning reading and reproducing system to which the present invention is applied. The printing scanner 10 has an input unit 12 in which an image original (not shown) is scanned by light from a light source, and reflected light or transmitted light from the image original is, for example, three primary colors, respectively. The color is separated through a filter, guided to a photoelectric conversion element such as a CCD linear image sensor, and converted into an image signal (pixel signal) as an electric signal through the photoelectric conversion element.

The image signal output from the input unit 12 is A
After being converted into an image signal (pixel signal) as digital data by a control unit 14 having a / D converter and a computer as control means, the control unit 14 converts the image signal as a luminance signal into a density signal. After performing a conversion process to an image signal, a gradation correction process, a sharpness enhancement process, a color correction process, and the like, various image processes such as a halftone image process are performed.

The halftone image signal output from the control unit 14 is electro-optically converted in the output unit 16 and the photosensitive material is scanned with a light beam such as a laser beam to form a latent image on the film. Is done. The film on which the latent image is formed is developed by a developing machine to complete the film on which the image is formed.

From the film on which the image is formed, a printing plate is prepared by a printing plate preparation device, and the prepared printing plate is mounted on a printing press. In the printing press, the transfer of the ink from the printing plate to the printing paper and the multicolor printing are performed, thereby completing a printed matter on which a color image is formed.

FIG. 2 is a block diagram showing an example of the configuration of the sharpness enhancement processing unit 20 in the control unit 14 according to the main part of the present invention. The processing in the sharpness enhancement processing section 20 may be executed by a CPU (not shown) as control means based on software, or may be executed as hardware. In the following description, the term "-unit" can be considered as "-circuit" or "-unit".

In FIG. 2, for example, an image signal (input image signal) S having a resolution of about 10 bits (approximately 1000 gradations) is passed through a port 22 to an unsharp signal generation section 24 and a subtraction section (subtraction signal generation section). 26, an addition unit (addition signal generation unit) 28, a first averaging unit (first averaged signal generation unit) 30, and a second averaging unit (second averaged signal generation unit) 32.

The unsharp signal generator 24 generates an unsharp signal U in accordance with the desired mask size Mu supplied from the port 34 and supplies it to the subtractor 26 (subtraction signal generator).

The first averaging unit 30 and the second averaging unit 32
A predetermined mask size (two types of ranges for the image signal around the image signal S) supplied from the ports 36 and 38 M
1, a signal Sm obtained by averaging the image signal S based on M2
a, Smb (Sma is referred to as a first averaged signal, and Smb is referred to as a second averaged signal) to a subtraction unit (subtraction signal generation unit) 40. In this embodiment, the mask size M1
Is selected for 5 × 5 pixels, and the mask size M2 is 1
It is selected for 1 × 11 pixels. Mask size M1,
M2 may be a different value for each color plate in consideration of the fact that the graininess differs for each color plate.

The image signal S has a mask size Mu and mask sizes M1 and M2 as described later.
Means the pixel signal located at the center of In this embodiment, the size of the pixel signal S, that is, one dot is about 5 μm square. Therefore, the mask size M
The size of 1 is 25 μm square, and the size of the mask size M2 is 55 μm square. Mask size Mu for blur mask
Can be selected to an appropriate value independently of these mask sizes M1 and M2. In this embodiment, 3
× 3 pixels, that is, 15 μm square.

In the subtracting section 40, a difference signal Sd between the first averaged signal Sma and the second averaged signal Smb (Sd = Sma-
Smb) is created. Absolute value signal ab of this difference signal Sd
sSd (absSd = | Sd | = | Sma-Smb |)
Is generated by an absolute value section (also referred to as an absolute value signal generating section or an absolute value calculating section) 42 and supplied to a look-up table (also referred to as a reference table or a search table) 44.

In this case, the first and second averaging units 30,
The subtraction unit 32, the subtraction unit 40, and the absolute value unit 42 constitute a flatness determination unit 29 that determines the flatness of an image.

FIG. 3 shows a configuration example of the lookup table 44. The horizontal axis is the absolute value signal abs as input.
Sd, and the vertical axis is a coefficient (emphasis coefficient) h that takes a value from 0 to 1 as an output. Absolute value signal abs
When the value of Sd is smaller than the first threshold th1, the coefficient h is set to h = 0. Also, the absolute value signal ab
When the value of sSd is larger than the first threshold th1 and smaller than the second threshold th2, the coefficient h becomes h
From 0 to h = 1, the value gradually increases, and the value of the coefficient h proportional to the value of the absolute value signal absSd is output. On the other hand, the value of the absolute value signal absSd is equal to the second threshold th.
When the value is larger than 2, the coefficient h is set to h = 1. The threshold values th1 and th2 are supplied through the port 46, and the values can be changed. The look-up table 44 can be rephrased as the generation unit 43 of the sharpness enhancement coefficient depending on the flatness of the image.

The coefficient h output from the lookup table 44 as a value corresponding to the value of the absolute value signal absSd is:
As shown in FIG. 2, the signal is supplied to a multiplier (multiplier signal generator) 48.

On the other hand, in the subtraction unit 26, an unsharp masking (USM) signal SU, which is a difference signal, is created from the image signal S and the unsharp signal U, and supplied to the multiplication unit (multiplication signal generation unit) 50. You. Note that the unsharp masking signal SU can also be calculated by calculating the Laplacian of the image signal S. When calculating the Laplacian, the unsharp signal generator 24 and the subtractor 2
6 may be replaced by a Laplacian generator as a secondary differential signal generator.

The multiplication unit 50 multiplies the coefficient (also referred to as a sharpness enhancement coefficient) k supplied from the port 52 by the unsharp masking signal SU to generate an unsharp masking signal k (SU). .

The multiplication section 48 multiplies the unsharp masking signal k (SU) by the coefficient h to create a new unsharp masking signal h · k (SU). To supply.

The adder 28 adds the unsharp masking signal h · k (SU) to the image signal S as an input signal, and adds a sharpness enhancement signal S ^** shown in the following equation (2).
And outputs it to the port 54.

S ^** = S + h · k (S−U) (2) In FIG. 2, the mask sizes Mu, M1, M2 supplied from the ports 34, 36, 38, 46, and 52, and the threshold value th
Normally, default values are set for the values of 1, th2, and coefficient k through a CPU (not shown). However, the user uses input means such as a keyboard (not shown) to set desired values through the CPU. It is also possible to set.

To determine the default value,
By outputting an image from the output unit 16 in accordance with the type of the original image, for example, whether it is a reversal original or a digital camera signal, the resolution, or the type of the subject, etc. Mask size Mu, M1,
The values of M2, thresholds th1 and th2, and coefficient k can be determined, and can be stored in a ROM (write-only memory) (not shown).

Next, the operation of the main part of the above embodiment will be described in more detail.

Assuming that the image signal S is composed of pixel data as shown in FIG. 4 (data obtained by decomposing an image original into pixels), consideration will be given to the optimal sharpness enhancement of the desired image signal S = a66. I do. It should be noted that the image signal S shown in FIG. 4 schematically corresponds to an image original after decomposition of pixels (the size of one pixel is, for example, 5 μm square as described above).

As described above, when the mask size Mu for forming the unsharp signal U is set to Mu = 3 × 3, the unsharp signal U surrounds the image signal S = a66 (including a66). S = a55, a56, a5
7, a65, a66, a67, a75, a76, a77
Can be calculated, for example, as an average value of the nine pixel data.

In the first averaging section 30, since the mask size M1 is set to M1 = 5 × 5, the first averaging signal Sma includes the pixel data a44, a48, a8 at the four corners.
4, can be calculated as an average value of 25 pixel data surrounded by a88.

Further, in the second averaging section 32, the mask size M2 is set to M2 = 11 × 11.
The averaged signal Smb includes pixel data a11, a11 at the four corners.
It can be calculated as an average value of 121 pieces of pixel data surrounded by 1, a111, and a1111.

The absolute value absSd of the difference signal between the first averaged signal Sma and the second averaged signal Smb is calculated by the subtractor 40 and the absolute value calculator 42.

If the absolute value absSd of the difference signal is smaller than the threshold th1, the value of the emphasis coefficient h is set to h = 0 from the look-up table 44 shown in FIG. In this case, the unsharp masking signal h.
Since k (SU) is set to a zero value, the sharpness-enhanced signal S ^** is set to S ^** = S, and a signal that is not sharpened is output. That is, the absolute value ab of the difference signal between the first averaged signal Sma and the second averaged signal Smb
If sSd is smaller than the threshold th1, the image is determined to be a flat portion of the image, in other words, a flat portion of the image density or a portion where the density variation is small, and the sharpness is not enhanced, so that the granularity or quantization is reduced. Sharpness is not applied to steps. Thereby, the roughness of the output image can be reduced.

On the other hand, the absolute value absSd of the difference signal between the first averaged signal Sma and the second averaged signal Smb is equal to the threshold th1.
If the value is greater than 0, the value of the enhancement coefficient
The sharpness enhancement signal S ^**
Means that the second term h · k (SU) on the right side of the equation (2) has a finite value, and sharpness enhancement is applied. In this case, that is, when absSd> th1, the image signal S is determined to be a contrast portion, for example, a texture portion, and sharpness enhancement is applied to the image signal S. Sharpness is effectively applied to a portion having contrast.

As described above, according to the above-described embodiment,
Flatness (flatness) of image density based on the difference between the average values of each image calculated by two types of mask sizes M1 and M2
Thus, it is possible to achieve an effect that sharpness enhancement can be effectively applied to an image having a small contrast while sharpness enhancement can be prevented from being applied to roughness and quantization steps.

In the example shown in FIG. 2, the flatness of the image density is determined based on the average value difference between the images calculated by the two types of mask sizes M1 and M2. Then, two types of mask sizes M1 and M2
The flatness (flatness) of the image density is determined based on the ratio of the average values of the images calculated by the following formula. If the value of the ratio is close to 1, the value of the ratio is 1 which is a flat portion. If the value is more than a predetermined value away from, it is determined that it is not a flat portion, and the value of the coefficient h can be determined.

After all, the flatness of the image density can be determined by comparing the average values calculated by the two types of mask sizes M1 and M2.

The present invention is not limited to the above-described embodiment, but can adopt various configurations without departing from the gist of the present invention.

[0047]

As described above, according to the present invention, an image signal around an input image signal is averaged in two types of ranges, and the calculated averages are compared with each other. The flatness of the density is determined, and the magnitude of the USM signal is changed according to the determined flatness, and the USM signal is combined with the input image signal to generate an image signal after sharpness enhancement processing.

Therefore, the flatness of the image density can be easily determined. Therefore, desired sharpness enhancement can be applied according to the flatness of the determination result.

For example, sharpness enhancement can be performed on an image having a small contrast such as texture, while preventing sharpness enhancement from being applied to graininess (roughness) and quantization steps. As a result, it is possible to obtain a high-quality image output free from a feeling of roughness or a line related to tone jump.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a basic configuration of a printing scanner to which the present invention is applied.

FIG. 2 is a block diagram illustrating a configuration example of a sharpness enhancement processing unit.

FIG. 3 is a diagram illustrating the contents of a look-up table in the example of FIG. 2;

FIG. 4 is a schematic diagram of image data.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 10 ... Printing scanner 12 ... Input part 14 ... Control part 16 ... Output part 20 ... Sharpness enhancement processing part 29 ... Image flatness determination part 30 ... First averaging part 32 ... Second averaging part 42 ... Absolute value part 43: a generation unit of a sharpness enhancement coefficient depending on the flatness of an image 44: a lookup table absSd: an absolute value signal of a difference signal h, k: a sharpness enhancement coefficient M1, M2, Mn
... Mask size S ... Image signal Sd ... Difference signal Sma ... First averaged signal Smb ... Second averaged signal SU, k (SU), hk (SU) ... Unsharp masking (USM) ) Signal S ^* , S ^** ... Sharpness emphasis signal U ... Unsharp signal

Claims (4)

[Claims] An unsharp signal is generated from an input image signal, a USM signal is generated from the generated unsharp signal and the input image signal, and the generated USM signal and the input image signal are combined to enhance sharpness. In a sharpness enhancement method for generating an image signal after processing, an image signal around the input image signal is averaged in two types of ranges, and the calculated averages are compared with each other. Sharpness, wherein the flatness is determined, the magnitude of the USM signal is changed in accordance with the determined flatness, and the USM signal is combined with the input image signal to generate an image signal after sharpness enhancement processing. Emphasis method. 2. The sharpness enhancement method according to claim 1, wherein when determining the flatness of the image density, an absolute value of a difference between the obtained average values is obtained, and the absolute value of the difference is determined by a predetermined threshold value. If it is smaller than the threshold value, the magnitude of the USM signal is set to a zero value. 3. The sharpness enhancing method according to claim 1, wherein a first threshold value and a second threshold value larger than the first threshold value are set in advance, and when determining the flatness of the image density, The absolute value of the difference between the determined average values is determined. If the absolute value of the difference is smaller than the first threshold, the magnitude of the USM signal is set to a zero value, and the absolute value of the difference is determined by the absolute value of the difference. When the difference is larger than the first threshold and smaller than the second threshold, the magnitude of the USM signal is gradually increased according to the absolute value of the difference, and the absolute value of the difference is larger than the second threshold. The USM signal has a constant magnitude when the magnitude is also large. 4. The sharpness enhancing method according to claim 1, wherein the two ranges of the image signal around the input image signal are set to appropriate values according to the color of the image signal. A sharpness enhancement method characterized by selecting.