JPH0991423A - Image processing method and processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the observed image reading performance of a processed image by setting an absolute value of a function depending on a difference signal based on the difference signal between an original image signal and an unsharpened mask signal. SOLUTION: When an original image signal Sorg obtained by a reader or the like is given to a low pass filter 11, the filter 11 sets a prescribed fog mask to the original image signal Sorg and a fog mask signal Sus is outputted to a 1st arithmetic element 13. The arithmetic element 13 calculates a difference signal between the original image signal Sorg and the fog mask signal Sus and provides an output of the result to a conversion table 12. The conversion table 12 converts the received difference signal into an emphasis signal f(Sorg-Sus) and provides an output of it to a 2nd arithmetic means 14. The 2nd arithmetic element 14 adds the original image signal and the emphasis signal to provide an output of a processed image signal Sproc. As a result, the processed image signal is obtained, in which the sharpness of a desired image part is emphasized.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing method and apparatus.

[0002]

2. Description of the Related Art Conventionally, an image signal representing an image obtained by various image acquisition methods has been subjected to image processing such as gradation processing and frequency processing to improve the observation / interpretation performance of the image. It is being appreciated. Particularly in the field of medical images such as radiographic images of the human body, a specialist such as a doctor needs to accurately diagnose the presence or absence of a disease or injury of a patient based on the obtained image, and the image Image processing that improves the image interpretation performance of is becoming indispensable.

As the so-called frequency enhancement processing in this image processing, for example, as shown in Japanese Patent Laid-Open No. 163571/1978, an image signal (original image signal) Sorg such as a density value of an original image is represented by Sproc = Sorg + β × (Sorg−Sus) (4) There is known one for converting into an image signal Sproc. Here, β is a frequency enhancement coefficient that depends on the original image signal Sorg (as an example, a density value), and Sus is a non-sharp mask (so-called blur mask) signal. This blur mask signal Sus sets a mask composed of a pixel matrix of N columns × N rows (N is an odd number) with the original image signal Sorg as the central pixel for pixels arranged two-dimensionally, that is, a blur mask, and Sus = It is an ultra-low spatial frequency component obtained as (ΣSorg) / N ² (5).

The value in the second term parenthesis of the equation (4) (Sorg-S
us) is obtained by subtracting the blur mask signal, which is an ultra-low spatial frequency component, from the original image signal, so the frequency component Ssp higher than the ultra-low spatial frequency of the original image signal from which the ultra-low spatial frequency component has been removed. Means
By multiplying the relatively high frequency component Ssp by the frequency enhancement coefficient β and then adding the original image signal Sorg, the relatively high frequency component Ssp of the original image signal Sorg, for example, the image signal changes sharply. It is possible to obtain the processed image signal Sproc in which the edges and the like are relatively emphasized.

Further, in Japanese Patent Application Laid-Open No. 56-104645, the frequency component Ssp is calculated by performing the operation represented by the following equation (6).
The technique which emphasizes is disclosed.

Sproc = Sorg + F (x) (6) (where x = Sorg−Sus, F (x) is | x ₁ | <|
When x ₂ |, F ′ (x ₁ ) ≧ F ′ (x ₂ ) ≧ 0, and
A value x ₀ (| x ₁ | <| x ₀ | <| x
₂ |) as a boundary, F ′ (x ₁ )> F ′ (x ₂ ) is a monotonically increasing function) This technique uses F ′ (x), that is, the first derivative of the function of the points Sorg and Sus is | x Since | is made smaller as | becomes larger, the rate of increase of frequency emphasis at a large | x | is suppressed. As a result, the degree of frequency emphasis is increased as the difference signal is larger, while that at a large difference signal is increased. This is to prevent the generation of false contours by suppressing the increase in the degree of emphasis.

That is, in a radiographic image, a frequency difference is generally emphasized by a blur masking process in a portion having a small difference signal, and a portion having a large difference signal (for example, a boundary between a bone and a muscle, a boundary between a soft portion and a gas portion, a stomach Ba (barium). ) In the boundary angiography between the filling part and its surroundings, blood vessel shadows, etc.), the increase in the degree of frequency enhancement is suppressed, and the occurrence of false contours can be prevented.

[0008]

The above-mentioned Japanese Patent Laid-Open No.
According to the technique disclosed in No. 56-104645, since F (x) is a monotonically increasing function, F (x) increases as the difference signal between the original image signal Sorg and the unsharp mask signal Sus increases.
(X) also increases. Here, in the original image, for an image part having a high degree of sharpness and a clear contrast, it is not necessary to originally perform frequency enhancement because the observation and interpretation are sufficient even when the original image is as it is, but the difference signal corresponding to the image part is not necessary. Is a relatively large value, the above equation (6)
According to the emphasizing process according to, the part with clear contrast is also emphasized, and the degree of emphasis becomes stronger than the degree of emphasis of the desired image part to be emphasized because the sharpness of the original image is low. The image portion having a clear contrast causes overshoot or undershoot due to excessive emphasis to generate false contours (artifacts), which deteriorates the observation and interpretation performance of the desired image portion.

On the other hand, the high frequency component, which is the difference signal between the original image signal Sorg and the non-sharp mask signal Sus, contains high frequency noise such as radiation noise in the high frequency region.
It is important not to enhance the high-frequency noise in order to improve the observation / interpretation performance of the image.

The present invention has been made in view of the above circumstances, and while enhancing a desired image portion, high-frequency radiation noise and enhancement of an image portion having a clear contrast to some extent are suppressed to perform observation and interpretation performance. It is an object of the present invention to provide an image processing method and apparatus for obtaining a processed image signal representing a processed image with improved image quality.

[0011]

According to the image processing method of the present invention, an unsharp mask signal Sus of an original image signal Sorg representing an image is obtained, and a difference signal between the original image signal Sorg and the unsharp mask signal Sus is calculated. In the image processing method in which the desired image part is frequency-emphasized by performing the enhancement process according to the following equation (1) using the function f (x) depending on the difference signal, Sproc = Sorg + f (x ) (1) (where f (x) represents a function depending on the difference signal; x
= Sorg-Sus) of the difference signal when the absolute value of the difference signal corresponding to the desired image portion is in a range between a first threshold and a second threshold greater than the first threshold. The absolute value of the value of the function f (x) in the range where the absolute value is smaller than the first threshold is smaller than the value of the function f (x) corresponding to the desired image portion, and the absolute value of the difference signal. Is set to be smaller than the value of the function f (x) corresponding to the desired image portion in the range where is larger than the second threshold value. .

The function f is specifically f ≧ 0 when (Sorg-Sus) ≧ 0, as shown in FIG.
The first difference signal (Sorg-Sus) corresponds to high frequency noise
Within a range up to a predetermined value α ₁ (α ₁ > 0) of the function f, the value of the function f is small, and the difference signal (Sorg-Sus) has a second predetermined value α ₂ (α ₂ >) corresponding to the desired image portion. up to α ₁ ) and monotonically decreases in the range of the second predetermined value α ₂ or more, and (Sorg-Sus) already corresponds to the third predetermined value α corresponding to an image part having a clear contrast. _{The range of 3} (α ₃ > α ₂ ) or more is set so that the value of the function f becomes small.

Further, a function f is used as a correction index function g corresponding to an index value (for example, a variance value σ ² etc.) representing high frequency noise of the image to further suppress the enhancement of the image portion corresponding to this noise. Alternatively, the correction may be performed according to the following equation (2).

F ′ (x) = g (x) × f (x) (2) (where g (x) is a correction index function that monotonically increases in the range 0 ≦ g (x) ≦ 1, f ′ (X) represents a corrected function that is applied in place of the function f (x) in the equation (1)) By performing the correction in this way, the equation (1) may be expressed as the following equation (1 ′). it can.

Sproc = Sorg + g (x) × f (x) (1 ′) Note that the variance value σ ² has a predetermined value (for example, m rows × n).
(Column) pixel mask is set, the average value Smean of the original image signals Sorg of all the pixels in this mask is calculated, and the average value Smean is calculated by the following calculation (see equation (7)). If it is the same as the sharp mask, Smean = Sus and the arithmetic processing can be simplified as shown in the following formula (8).

Σ ² = {Σ (Sorg −Smean) ² } / (m · n) (7) σ ² = (Σ | Sorg −Sus | ² ) / N ² (8) The variance value σ ² is the noise of the image. FIG. 4 shows a specific example of the correction index function g that is an index value representing The correction index function g shown in the figure has a function g within the range in which the variance value σ ² corresponds to the signal region such that the output of the function g is small in the range in which the variance value σ ² corresponds to the high frequency noise region. Is set so that the output of is close to 1.

Specifically, the correction function g can be expressed by the following equation (3).

G (x) = G (Σ | Sorg −Sus |) (3) The image processing apparatus of the present invention is a specific apparatus for carrying out the above-described image processing method of the present invention. The non-sharp mask signal calculating means for obtaining the non-sharp mask signal Sus of the original image signal Sorg represented, the difference signal calculating means for obtaining the difference signal between the original image signal Sorg and the non-sharp mask signal Sus, and the difference signal An image processing that includes a conversion table that converts a function f (x) depending on a difference signal, and an emphasis processing unit that performs an emphasis process according to the following expression (1), and frequency-enhances a desired image portion of the image. In the apparatus, Sproc = Sorg + f (x) (1) (where f (x) represents a function dependent on the difference signal; x
= Sorg-Sus) of the difference signal when the absolute value of the difference signal corresponding to the desired image portion is in a range between a first threshold and a second threshold greater than the first threshold. The absolute value of the value of the function f (x) in the range where the absolute value is smaller than the first threshold value is smaller than the value of the function f (x) corresponding to the desired image portion, and the absolute value of the difference signal. Is set to be smaller than a value of the function f (x) corresponding to the desired image portion in a range in which is larger than a second threshold value. Is.

In addition, the difference signal is used as a correction index function g that represents high-frequency noise in the image that depends on the difference signal.
A conversion table for correction to be converted into (x) and the following formula (2)
Accordingly, a configuration further including the function f (x) can be adopted.

F ′ (x) = g (x) × f (x) (2) (where g (x) is a correction index function that monotonically increases in the range of 0 ≦ g (x) ≦ 1, f ′ (X) represents a corrected function that is applied in place of the function f (x) of the equation (1). Note that the enhancement function f (x) is the enhancement function in the image processing method of the present invention. Similarly, specifically, the one shown in FIG. 2 or the like can be applied.

Further, the enhancement function f (x) is converted into an image portion corresponding to the noise by a correction index function g (x) corresponding to an index value such as a variance value σ ² representing noise in the original image signal. It is also possible to adopt a configuration further including a correction unit that corrects according to the following formula (2) so that the emphasis of 1 is suppressed.

F ′ (x) = g (x) × f (x) (2) (where g (Sorg-Sus) is an index function; 0 ≦ g (S
org-Sus) ≤ 1, f '(Sorg-Sus) represents a corrected function applied in place of the function f (Sorg-Sus) of the equation (1)) Specifically, depending on the index value The function g can be expressed by the following equation (3).

G (x) = G (Σ | Sorg−Sus | ² ) (3)

[0024]

According to the image processing method and apparatus of the present invention,
In the range where the difference signal between the original image signal Sorg and its unsharp mask signal is small, the high-frequency noise occupies a large part of the difference signal. Improve
On the other hand, since the original image in a range where the difference signal is larger than the difference signal corresponding to the desired image portion has a clear contrast to some extent, by suppressing the enhancement processing of the original image, it is possible to prevent the false contour from being generated in the portion. It is possible to improve the observation / interpretation performance.

[0025]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of an image processing apparatus for specifically implementing the image processing method of the present invention will be described below with reference to the drawings.

FIG. 1 is a schematic block diagram showing a first embodiment of the image processing apparatus of the present invention. The illustrated image processing apparatus uses an original image signal (density value) Sorg representing an image.
, A low-pass filter 11 for obtaining an unsharp mask signal Sus corresponding to an ultra-low spatial frequency, and this low-pass filter 11
First, a difference signal (Sorg-Sus) between the unsharp mask signal Sus and the original image signal Sorg obtained by
An operator 13 and a conversion table 12 for inputting the obtained difference signal (Sorg-Sus) and outputting a predetermined emphasis signal f (Sorg-Sus) associated with the difference signal (Sorg-Sus). , The enhancement signal f (Sorg-Sus) and the original image signal Sorg and the processed image signal Spr.
and a second operator 14 for calculating oc.

The low-pass filter 11 sets a blur mask consisting of, for example, a pixel matrix of 3 columns × 3 rows for the input original image signal Sorg, and the following equation (5)
The blur mask signal Sus obtained according to (set to N = 3) is output to the image enhancing means 16.

Sus = (ΣS org) / N ² (5) As the blur mask, a simple average of the pixel values in the mask is used as shown in Expression (5), or the mask in the mask is calculated according to the distance from the center pixel. It is also possible to use one in which the weighting of the pixel value of is changed.

In the conversion table 12, the difference signal (Sorg-Sus) and the enhancement signal f (Sorg-Sus) are associated with each other in accordance with the function shape as shown in FIG.

This function shape is such that f ≧ 0 when (Sorg-Sus) ≧ 0, and (Sorg−Sus) corresponds to the high frequency noise of the original image, the first predetermined value α ₁ (α ₁ > 0). )
The value of the function f is small in the range up to (Sorg
-Sus) is the second predetermined value α ₂ corresponding to the desired image portion
It increases monotonically until (α ₂ > α ₁ ) and reaches the second predetermined value α ₂
In the range above, the value of the function f is monotonically decreasing, and in the range above the third predetermined value α ₃ (α ₃ > α ₂ ) where (Sorg-Sus) already corresponds to an image part having a clear contrast, the value of the function f is It is set to be small.
The function shape when (Sorg-Sus) ≦ 0 is (So
The function shape in the range of rg-Sus) ≧ 0 is symmetrical about the origin.

Next, the operation of the image processing apparatus of this embodiment will be described.

First, the original image signal Sorg obtained by a predetermined image information reading device is the low-pass filter 1
1, the first operator 13 and the second operator 14, respectively.

The low-pass filter 11 sets, for the input original image signal Sorg, a blur mask consisting of, for example, a pixel matrix of 3 columns × 3 rows, and outputs the blur mask signal Sus obtained according to the above equation (5) as the first blur mask signal Sus. Operator 13
Output to

As the blur mask, a simple average of pixel values in the mask is used as shown in equation (5),
As described above, the weighting of pixel values in the mask may be appropriately changed according to the distance from the central pixel.

The first operator 13 calculates the difference signal (Sorg-Sus) based on the input original image signal Sorg and blur mask signal Sus, and outputs the result to the conversion table 12.

The conversion table 12 receives the difference signal (Sor
g-Sus) is an emphasized signal f (Sorg-
Sus) and output to the second operator 14.

This emphasized signal f (Sorg-Sus) is a difference signal (Sorg-Su) according to the function shape as described above.
s) and the enhancement signal f (Sorg-Sus) are associated with each other, the difference signal has a very small range such as radiation noise, that is, a range of the first predetermined value (-α ₁ <
The enhancement signal in Sorg-Sus <α ₁ ) is suppressed to a small value close to zero, and this difference signal is larger than that due to radiation noise, and depends on the signal component corresponding to the spatial frequency of the desired structure. The range, that is, the range of the difference signal between the first predetermined value and the third predetermined value (α ₁ ≦ Sorg −
The enhancement signal in Sus ≦ α ₃ or −α ₃ ≦ Sorg −Sus ≦ −α ₁ ) outputs a relatively large value, and the range (α ₃ <which corresponds to an image portion having a sharpness higher than that of the desired structure is used.
Sorg-Sus or Sorg -Sus <-α ₃₎ the enhancement signal is suppressed to a small value close to approximately zero.

The second operator 14 performs addition processing on the input original image signal Sorg and the emphasized signal f (Sorg-Sus), and outputs a processed image signal Sproc shown in the following equation (1).

Sproc = Sorg + f (Sorg−Sus) (1) As a result, in the processed image signal Sproc, high-frequency radiation noise and enhancement of an image portion having a clear contrast to some extent are suppressed, and a desired image portion is suppressed. The sharpness of is emphasized.

Therefore, according to the image processing apparatus of the present embodiment, the enhancement is suppressed for the image portion of the original image which already has a clear contrast to some extent.
It is possible to prevent overshoot and undershoot due to excessive enhancement of the image portion, reduce the formation of artifacts, and obtain a processed image signal suitable for reproducing a visible image with excellent observation and interpretation performance.

FIG. 3 is a schematic block diagram showing a second embodiment of the image processing apparatus of the invention. The image processing apparatus shown in the figure is different from the image processing apparatus of the first embodiment shown in FIG. 1 in that the variance value calculating means 15 and the correction value table 16 are correction means for correcting the enhancement signal f (Sorg-Sus). And a third operator 17 are further provided.

That is, the variance value calculating means 15 obtains the variance value σ ² as an index value including the high frequency noise component of the image based on the above difference signal (Sorg-Sus), and the correction value table 16 uses this variance value σ. correction coefficient corresponding to ² g (0 ≦ g ≦
1) and the third operator 17 multiplies the enhancement signal f (Sorg-Sus) by the obtained correction coefficient g to obtain the product. The variance value σ ² is obtained by the following equation (8), which can be expressed as a function g (Sorg-Sus) corresponding to the difference signal (Sorg-Sus).

Σ ² = (Σ | Sorg −Sus | ² ) / N ² (8) Here, the correction coefficient g output from the correction value table 16 is as shown in FIG.
As shown in, in the range where the variance value σ ² is extremely small, 1
Is significantly lower than, and the variance value σ ² is associated with the variance value σ ² as a function shape that takes a value substantially equal to 1 in a range of a certain magnitude or more. That is, in the range in which the variance value σ ² is small, the ratio of high-frequency noise is larger than that in the image signal, so that the emphasis signal is suppressed small in this range.

Next, the operation of the image processing apparatus of this embodiment will be described.

First, the original image signal Sorg obtained by a predetermined image information reading device is the low-pass filter 1
1, the first operator 13 and the second operator 14, respectively.

The low-pass filter 11 sets, for the input original image signal Sorg, a blur mask consisting of, for example, a pixel matrix of 3 columns × 3 rows, and outputs the blur mask signal Sus obtained according to the above equation (5) as the first blur mask signal Sus. Operator 13
Output to

The first operator 13 calculates a difference signal (Sorg-Sus) based on the input original image signal Sorg and blur mask signal Sus, and the result thereof is converted into a conversion table 12 and a variance value calculating means 12. Output to.

The conversion table 12 displays the input difference signal (Sor
g-Sus) is an emphasized signal f (Sorg-
Sus) and output to the second operator 14.

This emphasized signal f (Sorg-Sus) has a range (-α) in which the difference signal has a very small magnitude such as radiation noise.
The emphasized signal in ₁ <Sorg-Sus <α ₁ ) is suppressed to a small value close to zero, and this difference signal is larger than that due to radiation noise and corresponds to the signal component due to the structure of the desired spatial frequency. Range (α ₁ ≤ Sorg
The enhancement signal in −Sus ≦ α ₃ or −α ₃ ≦ Sorg −Sus ≦ −α ₁ ) outputs a relatively large value, and a portion corresponding to an image portion where the contrast is clearer than that of the desired structure (α ₃ <Sorg -Sus or Sorg-Sus <-α ₃ )
Then, the emphasis signal is suppressed to a small value close to zero and is input to the third operator 17.

On the other hand, the variance value calculating means 15 obtains the variance value σ ² according to the input difference signal (Sorg-Sus). Here, the mask that is locally set for obtaining the variance value σ ² is the same as the non-sharp mask for the sake of simplifying the arithmetic processing. This gives the mean value Smean in the variance mask.
Is equal to the blur mask signal Sus, and the obtained variance value σ ² is input to the correction value table 16.

The correction value table 16 stores the input variance value σ ²
The correction coefficient g (Sorg-Sus) corresponding to is output. That is, in the range where the variance value σ ² is extremely small, the correction coefficient g (Sorg-Sus) that is significantly less than 1 is set to the variance value σ
^In the range where ² is larger than a certain level, the correction coefficient g (Sorg-Sus) which is substantially equal to 1 is output.

The third operator 17 inputs the correction coefficient g (S
org-Sus) and the emphasized signal f (Sorg-Sus) are obtained, and the obtained product is output to the second operator 14.

The second operator 14 is the enhancement signal f (Sorg-Sus) .g (Sorg-Sus) corrected by the input original image signal Sorg and the correction coefficient g (Sorg-Sus).
And are subjected to addition processing, and the processed image signal Sproc shown in the following equation (1 ′) is output.

Sproc = Sorg + g (Sorg−Sus) × f (Sorg−Sus) (1 ′) As a result, the processed image signal Sproc has high-frequency radiation noise and enhancement of an image portion which already has a clear contrast to some extent. It is suppressed and the sharpness of the desired image portion is emphasized.

As described above, according to the image processing apparatus of this embodiment, the enhancement is suppressed for the image portion of the original image which already has a clear contrast to some extent.
It is possible to prevent overshoot and undershoot due to excessive emphasis, reduce the formation of artifacts, and obtain an image signal suitable for reproducing a visible image with excellent observation and interpretation performance.

[Brief description of drawings]

FIG. 1 is a schematic block diagram showing a first embodiment of an image processing apparatus of the present invention.

FIG. 2 shows a difference signal (Sorg-Sus) and an emphasis signal f (Sorg).
-Sus) is a graph showing a conversion table in which

FIG. 3 is a schematic block diagram showing a second embodiment of the image processing apparatus of the invention.

FIG. 4 is a graph showing a conversion table that associates a variance value σ ² with a correction coefficient g (Sorg-Sus).

[Explanation of symbols]

 11 Low-pass filter 12 Conversion table 13, 14, 17 Operator 15 Variance value calculation means 16 Correction value table

Claims (6)

[Claims] 1. An unsharp mask signal Sus of an original image signal Sorg representing an image is obtained, and a difference signal between the original image signal Sorg and the unsharp mask signal Sus is obtained,
In the image processing method for frequency-enhancing a desired image portion by performing an enhancement process according to the following formula (1) using the function f (x) depending on the difference signal, Sproc = Sorg + f (x) ( 1) (where f (x) represents a function dependent on the difference signal; x
= Sorg-Sus) When the absolute value of the difference signal corresponding to the desired image portion is in a range between a first threshold value and a second threshold value larger than the first threshold value, the difference signal The absolute value of the value of the function f (x) in the range where the absolute value is smaller than the first threshold is smaller than the value of the function f (x) corresponding to the desired image portion, and the absolute value of the difference signal. 2. The absolute value of the value of the function f (x) in the range where is larger than the second threshold value is set smaller than the value of the function f (x) corresponding to the desired image portion. The described image processing method. 2. The function f (x) is corrected according to the following equation (2) using a correction index function g (x) representing high frequency noise in the image based on the difference signal. The image processing method according to claim 1. f ′ (x) = g (x) × f (x) (2) (where g (x) is a correction index function that monotonically increases in the range of 0 ≦ g (x) ≦ 1, f ′ (x ) Represents a corrected function that is applied in place of the function f (x) in equation (1)) 3. The correction index function g (x) is a function G (x) corresponding to the variance value of the difference signal shown in the following equation (3).
The image processing method according to claim 2, wherein g (x) = G (Σ | Sorg-Sus | ² ) (3) 4. A non-sharp mask signal calculating means for obtaining a non-sharp mask signal Sus of an original image signal Sorg representing an image, and a difference signal calculating means for obtaining a difference signal between the original image signal Sorg and the non-sharp mask signal Sus. And a conversion table that converts the difference signal into a function f (x) that depends on the difference signal, and an emphasis processing unit that performs an emphasis process according to the following equation (1). In an image processing device for frequency-enhancing a part, Sproc = Sorg + f (x) (1) (where f (x) represents a function depending on the difference signal; x
= Sorg-Sus) When the absolute value of the difference signal corresponding to the desired image portion is in a range between a first threshold value and a second threshold value larger than the first threshold value, the difference signal The absolute value of the value of the function f (x) in the range where the absolute value is smaller than the first threshold value is smaller than the value of the function f (x) corresponding to the desired image portion, and the absolute value of the difference signal. Where the absolute value of the value of the function f (x) in the range where is larger than the second threshold value is set smaller than the value of the function f (x) corresponding to the desired image portion. Processing equipment. 5. The correction index function g (x) representing the difference signal, which represents high-frequency noise in the image that depends on the difference signal.
5. The image processing apparatus according to claim 4, further comprising: a correction conversion table for converting into the following expression, and a correction unit that corrects the function f (x) according to the following expression (2). f ′ (x) = g (x) × f (x) (2) (where g (x) is a correction index function that monotonically increases in the range of 0 ≦ g (x) ≦ 1, f ′ (x ) Represents a corrected function that is applied in place of the function f (x) in equation (1)) 6. The correction index function g (x) is a function G (x) corresponding to the variance value of the difference signal shown in the following equation (3).
The image processing apparatus according to claim 5, wherein g (x) = G (Σ | Sorg-Sus | ² ) (3)