JP2001292325A - Edge enhancement device, edge enhancement method and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an edge enhancement device that can adaptively correct various types of edges from steep edges to rather smooth edges. SOLUTION: In the extraction of an edge part and correction of the pixel value of the edge part in its increasing direction to enhance the edge, an image correction processing section 57 generates a variable amplification factor (ΔG) changed corresponding to a pixel value (EBASIC) of the edge part and applies the amplification factor to the pixel value of the edge part to generate an edge enhancement correction value (E). The amplification factor (ΔG) has a characteristic of increasing the edge enhancement correction value (E) when the pixel value of the edge part resides in a large value area (an area where the EBASIC is large) more than when the pixel value of the edge part resides in a small value area (an area where the EBASIC is small).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an edge enhancement device,
The present invention relates to an edge enhancement method and a recording medium. More specifically, the present invention relates to an edge emphasizing apparatus, an edge emphasizing method, and a recording medium that sharpen an edge portion of a digital image (also referred to as sharpness or emphasis) to improve image quality deterioration or process and correct a sharp image.

[0002]

2. Description of the Related Art Edge enhancement is one of the special effects (or image quality improvement) methods for digital images. This is to give visual sharpness to the image quality by making the outline (edge) portion in the image stand out. The edge portion is a portion showing a sudden signal change in a two-dimensional plane of the image. A two-dimensional filter, for example, a first-order differential filter such as Sobel or Prewitt or a second-order differential filter such as Laplacian is used to extract the edge portion. Differentiation is a “difference”, and the principle of any filter is that the value (pixel value) of n × n (generally n = 3) pixels including the pixel of interest is determined according to the position of each pixel. In that the multiplication result is multiplied by a predetermined coefficient, and the respective multiplication results are added to obtain a correction value for edge enhancement with respect to the pixel value of the target pixel. The specific calculation procedure of this correction value will be described later.

[0003] Edge enhancement is one of arbitrary processing methods for digital images (arbitrarily performed to give special effects or artistic effects), but other electronic imaging devices such as digital cameras. It is also used as an inevitable image quality improvement method in a digital image generation device (hereinafter, referred to as an “electronic still camera”) using the image. In an electronic still camera, an optical transfer function (so-called OTF: Optic) of an imaging system such as a photographing lens or an imaging device is used.
al Transfer Function), band limitation in digital signal conversion and various signal processing, etc., the high frequency component of the spatial frequency of the digital image signal is easily lost, and the sharpness of the edge portion of the image is impaired. This is because image correction processing by edge enhancement is indispensable in an important electronic still camera.

FIG. 9A is a conceptual configuration diagram of a main part of a conventional electronic still camera including a configuration for the image correction processing. In this figure, an image of a subject 1 is captured by an imaging device 3 (generally, CCD: Ch) through an optical system 2 such as a photographing lens.
arge Coupled Device)
It is converted to an analog image signal A. The image signal A is converted into a digital image signal D by the analog / digital conversion circuit 4, and then, in the color processing circuit 5, an edge-enhanced luminance signal (hereinafter referred to as “edge-enhanced corrected luminance signal Yed”).
g ”) and a color signal C.

FIG. 9B is a configuration diagram of the color processing circuit 5. The color processing circuit 5 obtains a pixel value lost by the color filter of the imaging device 3 from the pixel values of its surrounding pixels and interpolates the pixel value.
A luminance / color signal conversion processing unit 7 for converting into a color signal C;
An edge detection unit that extracts an edge addition signal E from the luminance signal Y; and an edge addition unit 9 that adds the luminance signal Y and the edge addition signal E and outputs an edge-enhanced corrected luminance signal Yedg. 8 and the edge addition unit 9 constitute an image correction processing unit 10.

FIG. 10A is a configuration diagram of the image correction processing unit 10. The image correction processing unit 10 uses a Laplacian filter or the like to generate (extract) a correction basic signal (edge signal) E _BASIC for edge addition with respect to the pixel of interest, and to generate the correction basic signal E _BASIC in advance. A gain adjustment unit 12 that multiplies a predetermined fixed amplification coefficient G _STATIC to output a gain-adjusted correction basic signal GE _BASIC, and an edge addition signal that _converts a gain-adjusted correction basic signal GE _BASIC equal to or greater than a predetermined threshold value Output as E (Gain adjusted correction basic signal G less than threshold value)
For E _BASIC and a coring unit 13 for outputting) as E = 0.

Here, the operation concept of the edge filter unit 11 is described.
Will be described with reference to FIG.
In FIG. 3A, 3 × 3 cells each represent a pixel.
The square in the center is the pixel of interest (the target image for edge enhancement).
Element), and the eight surrounding pixels are reference pixels. Inside the square
The codes described in are set in advance for each pixel position.
This is the estimated coefficient. The value of this coefficient depends on the filter design.
Above, it can take any value, but basically,
The sum of all coefficients is equal to the coefficient of the pixel of interest
(Or nearly equal), and the correlation is
Coefficients of high reference pixels (for example, top, bottom, left and right pixels)
Satisfy the condition that the coefficient is larger than the coefficient of the outside reference pixel.
Just do it. For example, the coefficient of the pixel of interest is K_Cage,
The coefficient of the reference pixel is sequentially increased clockwise from the upper left by K_NW, K
_N, K_NE, K_E, K_SE, K_S, K_SW, K_WThen, each count
The values can be set as follows: K_C= 1K_NW= -1/12 K_N= -2 / 12K_NE= -1/12  K_E= -2/12  K_SE= -1/12  K_S= -2/12  K_SW= -1/12  K_W= -2/12

Now, consider two pixel arrays. FIG.
FIG. 10B is a first pixel array example, and FIG. 10C is a second pixel array example. In the first pixel array example, the left column has a white level,
The two rows at the center and right are black levels, representing vertical edges. The second example of the pixel array represents a non-edge portion of the entire black level. When the pixel values of the white level and the black level are respectively set to “0” and “1” for convenience, the pixel values of the first pixel array example are as shown in FIG.
The pixel values of the second pixel array example are as shown in FIG. The value of the correction basic signal E _BASIC corresponding to the pixel value Ga of the pixel of interest in FIG.
The value of the correction basic signal E _BASIC corresponding to the pixel value Gb of the target pixel in (e) is obtained by the following equation. E _BASIC = (0 × K _NW ) + (1 × K _N ) + (1 × K _NE ) + (1 × K _E ) + (1 × K _SE ) + (1 × K _S ) + (0 × K _SW) ) + (0 × K _W ) + (1 × K _C ) ≒ 0.333 E _BASIC = (1 × K _NW ) + (1 × K _N ) + (1 × K _NE ) + (1 × K _E) ) + (1 × K _SE ) + (1 × K _S ) + (1 × K _SW ) + (1 × K _W ) + (1 × K _C ) ≒ 0 ...

The gain adjuster 10 performs gain adjustment by multiplying the value of the correction basic signal E _BASIC by a predetermined fixed amplification coefficient G _STATIC . For example, if G _STATIC = 1 for simplicity, the pixel value Ga
Corresponding gain value of adjusted correction base signal GE _BASIC is "0.333", and the value of the gain adjusted compensation base signal GE _BASIC corresponding to the pixel value Gb is "0".

Incidentally, the actual luminance signal Y is multi-graded and takes a discrete value according to the brightness. For example,
In the case of the 2 ^m gradation, one value is obtained by dividing the white level and the black level into 2 ^m steps. For this reason, even in the non-edge portion, it is extremely rare that an ideal state as shown in FIG. 10C is obtained (pixel values are distributed at the same gradation level). Because it is distributed,
The value of the gain-adjusted correction basic signal GE _BASIC corresponding to the target pixel Gb in the non-edge portion hardly becomes “0”. In other words, in many cases, the value becomes a minute value close to “0”. Since the edge emphasis must not be performed on the target pixel Ga of the non-edge portion, the gain adjusted correction basic signal GE _BASIC having a minute value is set to “0”.
A circuit section (coring section 13) is provided for limiting the distance.

FIG. 11 is an input / output characteristic diagram of the coring unit 13. In this figure, a straight line 1 having a slope of “1”
The x-intercepts (intersections with the horizontal axis) of 4 and 15 are
Corresponding to the threshold value. In the illustrated example, the threshold value in the positive direction is “0.25”, and the threshold value in the negative direction is “−0.25”.
, The gain-adjusted correction basic signal GE
_{When BASIC} is in the range of “−0.25” to “+0.25”, E = 0. Therefore, when the gain-adjusted correction basic signal GE _BASIC has a small value in the above range, the gain-adjusted correction basic signal GE _BASIC is limited to “0” so that unnecessary edge enhancement processing is not performed. can do.

[0012]

However, the image correction processing unit 10 has a configuration in which the correction basic signal E _BASIC extracted from the edge filter unit 11 is multiplied by a predetermined fixed amplification coefficient G _{S TATIC.} Had the following problems.

(1) In general, edges contained in an image are
It varies from sharp to gentle ones,
In order to obtain good image quality, it is necessary to perform appropriate enhancement correction according to the degree of these edges,
In the above prior art, since a predetermined fixed amplification coefficient G _STATIC is used, the amplification characteristic is uniform with respect to the correction basic signal E _BASIC and an appropriate amplification characteristic is set for each edge type. There is a disadvantage in that it cannot be given. That is, if the amplification coefficient G _STATIC is set to match a sharp edge, it becomes excessive for a gentle edge, and conversely, if it is set to match a gentle edge, it becomes sharp. As a result, either one of them is actually sacrificed, or the amplification factor G is a compromise between the two.
There is a problem to be solved in that the _{STAT IC} must be set, and various types of edges cannot be adaptively corrected (emphasized) to obtain good image quality.

(2) Also, for example, 3 × 3 where the pixel value of the target pixel is “1” and the pixel values of its peripheral pixels are “0”
Consider the pixel distribution of the configuration (see the conceptual diagram at the left end in FIG. 8). This is a pixel distribution including a so-called isolated point noise in which a central pixel is isolated from the surroundings. In this case, the value of the correction basic signal E _BASIC applied to the pixel value of the target pixel is obtained by the following equation. E _BASIC = (0 × K _NW ) + (0 × K _N ) + (0 × K _NE ) + (0 × K _E ) + (0 × K _SE ) + (0 × K _S ) + (0 × K _SW) ) + (0 × K _W ) + (1 × K _C ) = 1 Therefore, when G _STATIC = 1, GE _BASIC = 1, and GE _{BASIC sets} the threshold value (0.
25), E = GE _BASIC -threshold (ie, E = 0.7) despite non-edge portions
5), and the unnecessary addition processing (0.75 + Y in this example) is performed by the edge addition unit 9, so that the isolated point noise (pixel value of the pixel of interest) is rather emphasized and conspicuous. There is.

The first problem to be solved by the present invention is as follows.
An object of the present invention is to provide an edge emphasizing device capable of adaptively correcting various types of edges from sharp edges to gentle edges. A second problem is that an edge emphasizing apparatus, an edge emphasizing method, and a recording medium capable of adaptively correcting various types of edges from sharp edges to gentle edges without making isolated point noises noticeable. Is to provide.

[0016]

According to the first aspect of the present invention, there is provided an edge enhancing apparatus for extracting an edge portion of an image and correcting the pixel value of the edge portion in an increasing direction to perform edge enhancement. Generating a variable amplification coefficient that changes in accordance with the magnitude of the pixel value of the edge portion, and applying the amplification coefficient to the pixel value of the edge portion to generate a correction value for edge enhancement, The amplification coefficient has a characteristic that the edge enhancement correction value increases when the pixel value of the edge portion is in a region having a larger value than in a region having a small value.
According to a second aspect of the present invention, there is provided an edge enhancement device for extracting an edge portion of an image, correcting a pixel value of the edge portion in an increasing direction, and performing edge enhancement. Generating a variable amplification coefficient that changes in accordance with the flatness of the pixel value in the pixel plane, applying the amplification coefficient to the pixel value of the edge portion to generate an edge enhancement correction value, The amplification coefficient has a characteristic that the edge enhancement correction value decreases as the flatness increases. Claim 3
The edge enhancement device according to the invention described in the above aspect is characterized in that in the invention according to the first or second aspect, there is provided a two-dimensional filter for extracting the edge portion. An edge enhancing device according to a fourth aspect of the present invention is an edge enhancing device for extracting an edge signal in an image signal, a gain adjusting device for adjusting a gain of the edge signal extracted by the extracting device, Edge emphasizing means for emphasizing an edge part in the image signal using the signal of the edge part whose gain has been adjusted by the gain adjusting means; and the gain adjustment according to the level of the edge part signal extracted by the extracting means. Changing means for changing the degree of gain adjustment by the means. In the edge enhancement device according to a fifth aspect of the present invention, in the invention according to the fourth aspect, the change unit determines that a change amount of a signal level of the edge portion and a change amount of the gain adjustment degree have a predetermined relationship. The degree of the gain adjustment is changed so as to maintain the gain. An edge enhancing device according to a sixth aspect of the present invention is the edge enhancing device according to the fifth aspect, wherein the changing means comprises:
The degree of the gain adjustment is changed so that the predetermined relationship is different between a region where the signal level of the edge portion is low and a region where the signal level is high. According to a seventh aspect of the present invention, in the invention according to any one of the fourth to sixth aspects, when the level of the signal at the edge portion is equal to or less than a predetermined value, the changing unit adjusts the gain. And a means for changing the degree of gain adjustment by the gain adjusting means so that the level of the signal at the edge portion whose gain has been adjusted by the means becomes zero. According to an eighth aspect of the present invention, in the edge enhancement device according to any one of the fourth to seventh aspects, when the level of the signal at the edge portion is equal to or more than a predetermined value, the changing means adjusts the gain. A means for fixing the degree of gain adjustment by the means. An edge enhancing device according to a ninth aspect of the present invention provides an edge enhancing device for extracting an edge signal in an image signal, a gain adjusting device for adjusting a gain of the edge signal extracted by the extracting device, Edge emphasizing means for emphasizing an edge part in the image signal using the signal of the edge part whose gain has been adjusted by the gain adjusting means, and n × n number of weighting coefficients at the center point being smaller than weighting coefficients at the periphery. Evaluation means for evaluating the flatness of the image signal using the flatness evaluation filter, and changing means for changing the degree of gain adjustment by the gain adjustment means according to the flatness of the image signal evaluated by the evaluation means. And characterized in that: The edge enhancement device according to the invention according to claim 10,
According to a ninth aspect of the present invention, the weighting coefficient of the center point is 0. An edge enhancement method according to claim 11 extracts an edge portion of an image,
In an edge emphasizing method for performing edge emphasis by correcting the pixel value of the edge portion in the increasing direction, generating a variable amplification coefficient that changes in accordance with the magnitude of the pixel value of the edge portion, and While applying the correction value for edge enhancement by applying to the pixel value of the edge portion, the amplification coefficient is larger when the pixel value of the edge portion is in a region of a larger value than in a region of a smaller value. It has a characteristic of increasing the edge enhancement correction value.
According to a twelfth aspect of the present invention, in the edge enhancement method of extracting an edge portion of an image and correcting a pixel value of the edge portion in an increasing direction to perform edge enhancement, an n × n pixel including a target pixel is included. Generating a variable amplification coefficient that changes in accordance with the flatness of the pixel value in the pixel plane, applying the amplification coefficient to the pixel value of the edge portion to generate an edge enhancement correction value, The amplification coefficient has a characteristic that the edge enhancement correction value decreases as the flatness increases. Claim 1
An edge emphasizing method according to a third aspect of the present invention is characterized in that, in the eleventh and twelfth aspects, the edge portion is extracted using a two-dimensional filter. 15. The recording medium according to claim 14, wherein a program for extracting an edge portion of an image, correcting a pixel value of the edge portion in an increasing direction, and performing edge enhancement is stored. Generating a variable amplification coefficient that varies in accordance with the magnitude of the value, applying the amplification coefficient to the pixel value of the edge portion to generate a correction value for edge enhancement, and the amplification coefficient Is characterized by having a characteristic of increasing the edge emphasis correction value when the pixel value is in a region having a larger value than in a region having a small value. A recording medium according to a fifteenth aspect of the present invention is a recording medium, comprising: extracting means for extracting a signal of an edge portion in an image signal; gain adjusting means for adjusting a gain of the signal of the edge portion extracted by the extracting means; Edge emphasizing means for emphasizing an edge part in the image signal using the signal of the edge part whose gain has been adjusted by the adjusting means, and the gain adjusting means according to the level of the edge part signal extracted by the extracting means And a changing means for changing the degree of gain adjustment by the program. A recording medium according to the present invention is a recording medium in which a program for extracting an edge portion of an image, correcting a pixel value of the edge portion in an increasing direction, and performing edge enhancement is stored. A variable amplification coefficient that changes in accordance with the flatness of the pixel value in the xn pixel planes is generated, and the amplification coefficient is applied to the pixel value of the edge portion to generate an edge enhancement correction value. The amplification coefficient has a characteristic that the edge enhancement correction value decreases as the flatness increases. Claim 1
7. The recording medium according to claim 7, wherein the extracting means extracts an edge signal in the image signal, the gain adjusting means adjusts the gain of the edge signal extracted by the extracting means, and the gain adjusting means. Edge emphasizing means for emphasizing an edge portion in the image signal using a signal of the edge portion whose gain has been adjusted, and n × n flatnesses in which the weighting coefficient of the center point is smaller than the surrounding weighting coefficient Evaluating means for evaluating the flatness of the image signal using an evaluation filter, and changing means for changing the degree of gain adjustment by the gain adjusting means according to the flatness of the image signal evaluated by the evaluating means are realized. A program for executing the program.

[0017]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below with reference to the drawings, taking an electronic still camera as an example. The application of the present invention is not limited to the following electronic still cameras. Any device or process that handles digital images may be used. For example, the present invention can be applied to various devices such as an image scanner, a facsimile, a printer, a digital video camera, and a digital image editor. Or
The present invention can be applied to application program software including a process for realizing functions equivalent to these devices on hardware such as a personal computer.

FIG. 1 is a block diagram of an electronic still camera. The illustrated electronic still camera has an image generation system 20 for generating an image signal of a subject, a temporary storage system 21 for temporarily storing the image signal to facilitate reproduction processing and other processing, etc. A long-term storage storage system 22 for long-term storage of captured images, an image display system 23 for displaying composition during shooting and displaying a reproduced image, and a long-term storage storage system 22
A compression / expansion processing system 24 for compressing the image signal when storing the image signal or expanding the image signal read from the long-term storage 22; a control system 25 for controlling the entire operation of the electronic still camera; It can be divided into a data transfer system 26 and the like. Hereinafter, the configuration will be described for each system.

The image generation system 20 captures an image of a subject with an imaging device, converts the image into an electric signal, and generates and outputs an image signal (also referred to as a frame image signal) having a predetermined period from the electric signal. Although a color image signal is generated in the present embodiment, a monochrome image signal may be used. The image generation system 20 includes an optical system 2 including a photographic lens and an aperture mechanism.
9. CCD that converts light from a subject that has passed through the optical system 29 into an electric signal and outputs a frame image signal of a predetermined period
Imaging device (hereinafter, represented by "CCD") 30, such as
Driver 31 for driving CCD 30, CCD 30
Timing generator 3 for generating various timing signals such as a signal for controlling the imaging time (electronic shutter time) of the camera
2. Includes a sample-and-hold circuit 33 for sampling (eg, correlated double sampling) the frame image signal output from the CCD 30 to remove noise, and an analog-to-digital converter 34 for converting the frame image signal after noise removal into a digital signal. And a luminance signal Y and a chrominance signal C using the output from the analog-to-digital converter 34.
And a color process circuit 35 that generates a luminance / color difference combined signal (hereinafter, also referred to as a “YUV signal”) composed of the optical system 29 and the CC.
An image correction processing unit (an image correction processing unit 57 in FIG. 2) for compensating for a high-frequency component in an image signal lost due to a pre-stage circuit such as D30 and internal processing of the color processing circuit 35,
The details will be described later).

The temporary storage system 21 includes a buffer memory 37 having a predetermined storage capacity formed of a rewritable storage medium (for example, a semiconductor memory such as a DRAM or an SRAM). It has a buffer area of a sufficient size (storage capacity) to be able to develop the image signal fetched from the image generation system 20 via the system 26 or the image signal fetched from the long-term storage system 22 via the data transfer system 26. . The long-term storage system 22 is a rewritable nonvolatile storage medium, for example,
The flash memory 39 has a capacity capable of storing tens to hundreds of image files of a predetermined format that have been compressed by the compression / decompression processing system 24. The flash memory 39 may have a removable shape (for example, a card type).

The image display system 23 reproduces and displays an image signal output from the image generation system 20 at a predetermined cycle for confirming the composition (so-called through image display), and reproduces and displays an image recorded in the flash memory 39. To do,
A digital video encoder 42 for converting the size of a reproduced image to a display size, a color liquid crystal display 43 for displaying an output from the digital video encoder 42 on a screen, and detecting touch coordinates on the display screen of the liquid crystal display 43. Touch panel 44, touch panel 4
4 includes a touch panel I / F (interface) 45 that converts the output signal of the control unit 4 into a predetermined format and outputs the converted signal to the control system 25. The compression / expansion processing system 24 converts the image signal stored in the buffer memory 37 into a predetermined format (for example, JPE).
G: joint photographic expe
(rts group), and decompresses the compressed image file stored in the flash memory 39 in the same format. The control system 25 executes a predetermined control program to control the entire operation of the electronic still camera. The control system 25 generates a required key operation signal in response to the operation of various keys including a shutter key. A key input unit 41 that outputs a signal to the CPU 40 is included. The data transfer system 26 includes a video transfer circuit 36 that arbitrates the flow of data between the systems, and a bus (general term for a data bus, an address bus, and a control line) 45 that connects the systems.

FIG. 2 is a configuration diagram of the color process circuit 35. The color process circuit 35 obtains the pixel value lost by the color filter of the CCD 30 from the pixel values of the surrounding pixels and interpolates the pixel value. The color processing circuit 35 converts the image signal obtained by the pixel interpolation into a luminance signal Y (having brightness information Y). Signal) and color signal C
(A signal having tint information; generally, a color difference signal), and an edge filter unit 53 (which generates a correction basic signal E _BASIC corresponding to the pixel of interest using a Laplacian filter or the like). two-dimensional filter according to the gist of the invention, the equivalent) to the extraction means, a level conversion unit 54 for generating a variable amplification factor ΔG based on the correction basic signal E _bASIC, the correction basic signal E _bASIC and ΔG times The gain adjustment unit 55 that adjusts the gain (E _BASIC after the gain adjustment is E), adds the edge addition signal E (corresponding to the edge enhancement correction value described in the gist of the invention) and the luminance signal Y, and sets an edge. An edge adder 56 for extracting the enhanced corrected luminance signal Yedg. Edge filter unit 53, level conversion unit 54, gain adjustment unit 55, and edge addition unit 56
Integrally constitutes an image correction processing unit 57 (corresponding to the edge enhancement device, the gain adjustment unit, and the change unit described in the gist of the invention).

FIG. 3A is a conceptual diagram of the filter of the edge filter unit 53. In FIG.
Each of the three cells represents a pixel, the central cell is the pixel of interest (the pixel to be edge-emphasized), and the eight surrounding pixels are reference pixels. As described at the beginning, the symbols (K _NW , K _N , K _NE , K _E , K _SE , K _S ,
K _SW , K _W , K _C ) are predetermined coefficients (filter coefficients) corresponding to each pixel position, and the value of this coefficient is the sum of all the coefficients of the reference pixel and the coefficient of the target pixel. Peripheral pixels (e.g., upper, lower, left, and right pixels) having a value equal to (or substantially equal to) and having high correlation with the target pixel
Is arbitrary as long as the condition that the coefficient of is larger than the coefficients of the other peripheral pixels is satisfied.

FIG. 3B shows an example of the filter coefficient. That is, this is an example in which the coefficients of the four pixels in the upper, lower, left, and right directions having a high correlation with the target pixel are “−2/12”, and the coefficients of the four corner pixels having a low correlation are “−1/12”. The coefficient of the target pixel is “1”, and the total coefficient value of the eight reference pixels is (−2/12) × 4 + (− 1/12) × 4 = 1.0. Are almost the same and satisfy the above conditions. The edge filter unit 53 takes in the luminance signal Y output from the luminance / color signal conversion unit 52, and applies the filter coefficient of FIG. 3B to each 3 × 3 pixel of the luminance signal Y to apply the luminance signal Y. Correction base signal E for the pixel of interest
Generate _BASIC . For example, when the distribution of the luminance signal Y is the above-described first pixel array example (see FIGS. 10B and 10D), the correction basic signal E _BASIC is given by
It becomes about “0.333”. Alternatively, the distribution of the luminance signal Y is determined by using the above-described second pixel array example (FIG. 10C and FIG.
(See (e)), the correction basic signal E _BASIC becomes approximately “0” according to the previous equation.

The level conversion section 54 outputs the correction basic signal E
An amplification coefficient ΔG that changes according to the size of the _BASIC is generated. FIG. 4A is a correlation diagram showing the correspondence between the correction basic signal E _BASIC and the amplification coefficient ΔG. The vertical axis represents the magnitude of the amplification coefficient ΔG, and the horizontal axis represents the magnitude of the correction basic signal _EBASIC . The size of the amplification factor ΔG any time, magnitude and characteristic line 61a of the correction base signal E _BASIC at that time is derived from the intersection of the 61b. For example, in the illustrated example, when E _BASIC = 0.5, ΔG = 0.33
3, when E _BASIC = 0.75, ΔG = 0.66
7, and when E _BASIC = 1.0, ΔG = 1.0. However, the x-intercept (intersection with the horizontal axis) of the characteristic lines 61a and 61b is SL in the positive region and -SL in the negative region. For example, SL = 0.25 and -SL = -0.25. . Therefore, the magnitude of the correction basic signal E _BASIC is SL to -SL
Is within the range, ΔG is limited to “0”.

FIG. 4B shows the corrected basic signal E_BASICAnd d
FIG. 7 is a correlation diagram showing a correspondence relationship with a margin addition signal E. Vertical axis
Represents the magnitude of the edge addition signal E, and the horizontal axis represents the correction basic signal.
No. E _BASICRepresents the size of. Edge addition at any time
The magnitude of the arithmetic signal E is the correction basic signal E at that time._BASIC
Derived from the intersection of the size of the characteristic with the characteristic lines 62a and 62b
It is. For example, in the illustrated example, E_BASIC= 0.5
0.5 × ΔG_(0.5)= 0.167, and E_BASIC=
When 0.75, 0.75 × ΔG_(0.75)= 0.5
E_BASIC= 1.0 G when = 1.0_(1.0)= 1.
It becomes 0. Where ΔG_(0.5)Is the smell in FIG.
And E_BASIC= 0.5 when ΔG = 0.5 (0.33
3) and ΔG_(0.75)Is E in FIG._BASIC=
ΔG at 0.75 (0.667),
ΔG_(1.0)Is E in FIG._BASIC= 1.0
The magnitude of ΔG (1.0). Shown above the characteristic line 62a
Three positions (P_0.5, P_0.75And P_1.0) It
E = 0.167, E = 0.5 and E = 1.0 respectively
Correction basic signal E_{BASI C}Represents the intersection with
From these intersections, the characteristic line 62a indicates that the “correction basic signal E
_BASICChange tendency that increases exponentially as
Have ". This trend is
The characteristic line 62b in the negative region only differs in polarity.
It is the same even if you Position P_0.5Is described in the summary of the invention.
It corresponds to the “small value area” and P_0.75And P_1.0Is the same
This corresponds to the “large value area” described in the summary.

The tendency of the characteristic lines 62a and 62b that the edge addition signal E increases exponentially as the correction basic signal E _BASIC increases as the correction basic signal E _BASIC increases is the first object of the present invention. To be able to adaptively correct various types of edges, from edges to gentle edges). In the sharp edge portion, the correction basic signal E _{BA SIC} is large, and therefore, the edge addition signal E is also large. Thus, the strong edge correction is performed by the large edge addition signal E.
In a gentle edge portion, the correction basic signal E _BASIC is small, and therefore, the edge addition signal E is also small.
This is because weak edge correction is performed by the small edge addition signal E.

The electronic still camera according to the present embodiment has a CC
The image signal captured at D30 is converted to an A / D signal via S / H33.
34, the digitally converted image signal is taken into a color processing circuit 35, and the color processing circuit 35 generates a luminance signal Y and a color signal C from the image signal. (Correction basic signal E _BASIC ) is taken out, and an edge addition signal E corresponding to the magnitude of the edge component is added to the luminance signal Y to perform a necessary edge enhancement correction (optical system 29 or CCD 3).
To correct the deterioration of the image by compensating for the high-frequency components lost due to OFT such as 0 and various band restrictions), and transmits the corrected luminance signal (corrected luminance signal Yedg) and color signal C to the video transfer circuit 36. Buffer memory 3 via
7 to be displayed on the liquid crystal display 43 as a through image or recorded in the flash memory 39.

Here, a first object of the present invention is to make it possible to adaptively correct various types of edges from sharp edges to gentle edges.
The above embodiment can solve this problem. The amplification coefficient G _STATIC of the conventional example was uniform regardless of the magnitude of the correction basic signal E _BASIC . Now, E _B
Consider three examples where _ASIC = 0.5, E _BASIC = 0.75, and E _BASIC = 1.0. If the amplification coefficient G _STATIC of the conventional example is set to “1” for convenience, the edge addition signal E of the conventional example is given by “E = E _BASIC × G _{STATIC −the} threshold value of the coring unit 13”. Each of the two examples
= 0.25, E = 0.5, and E = 0.75. That is, the corrected luminance signal Yedg when E _BASIC = 0.5
Becomes “Yedg = Y + 0.25”, and E _BASIC = 0.
The corrected luminance signal Yedg at 75 is “Yedg = Y
+0.5 ", and the corrected luminance signal Yedg when E _BASIC = 1.0 is" Yedg = Y + 0.75 ".

On the other hand, in this embodiment, the amplification coefficient ΔG
Changes according to the magnitude of the correction basic signal E _BASIC . To be precise, the amplification coefficient ΔG has a changing tendency to increase exponentially as the correction basic signal E _BASIC increases. An example of a specific value of the amplification coefficient ΔG is shown in FIG.
As shown in, when E _BASIC = 0.5, ΔG = 0.33
3. When E _BASIC = 0.75, ΔG = 0.667, E
_{When BASIC} = 1.0, ΔG = 1.0. The edge addition signal E in the present embodiment is “E = E _BASIC × ΔG”
, Each of the above three examples has E =
0.5 × 0.333, E = 0.75 × 0.667, E =
1.0 × 1.0. That is, when E _BASIC = 0.5, the corrected luminance signal Yedg is “Yedg = Y + 0.
5 × 0.333 ≒ Y + 0.167 ”, and E _BASIC =
The corrected luminance signal Yedg at the time of 0.75 is “Yedg
= Y + 0.75 × 0.667 ≒ Y + 0.5 ”, and E
_{When BASIC} = 1.0, the corrected luminance signal Yedg is “Y
edg = Y + 1.0 × 1.0 = Y + 1.0 ”.

This is converted to the corrected luminance signal Yed of the prior art.
g Calculation Example (Yedg = Y + 0.25, Yedg = Y +
0.5, Yedg = Y + 0.75)
First, when the value of the correction basic signal E _BASIC is small (E _BASIC
= 0.5) is "Yedg ≒ Y + 0.167" in the present embodiment, as opposed to "Yedg = Y + 0.25" in the conventional example.
Therefore, the difference between the two (about 0.25-0.167 =
0.083 ”), the degree of edge enhancement is weakened. Next, in the case where the value of the correction basic signal E _BASIC is moderately small (E _BASIC = 0.75), “Yedg =
Y + 0.5 ”, the present embodiment also provides“ Yedg ≒ Y
+0.5 ", there is no difference between the two, and the required strength is emphasized. Finally, the correction basic signal E
_{In the} case where the value of _BASIC is large (E _BASIC = 1.0), the present embodiment is "Yedg = Y + 1.0" in contrast to "Yedg = Y + 0.75" of the conventional example. The degree of edge enhancement is increased by about “1−0.75 = 0.25”).

Therefore, in the present embodiment, the amplification coefficient ΔG is changed in accordance with the magnitude of the correction basic signal E _BASIC , so that “various types of edges from sharp edges to gentle edges are adapted. The electronic still camera can solve the first problem of the present invention.

In the above embodiment, the amplification coefficient Δ
The change in G is linear (characteristic line 61a in FIG.
61b), but is not limited to this. For example, an increase may be made with a predetermined inclination from the intersection of the vertical axis and the horizontal axis in FIG. 4A, and then coring may be performed. Alternatively, as shown in FIG. It may be a straight line having several bending points. FIG.
In (a), the correction basic signal E _BASIC and the amplification coefficient ΔG
Is a characteristic line representing the relationship with the first straight line portion 6 in this case.
3a (corresponding to a "small value area" described in the gist of the invention), a second straight portion 63b (corresponding to "large value region" described in the gist of the invention), and a third straight portion 63c. , The first linear portion 63a is in charge of the small value area of the correction basic signal E _BASIC , and the second linear portion 63b is
Responsible for the intermediate value region of _BASIC, the third straight portion 63c is responsible for maximum (limit) region to be corrected basic signal E _BASIC.

As can be understood from the drawing, the inclinations of these three straight lines are steepest in the second straight line portion 63b, second in the first straight line portion 63a, and lowest in the third straight line portion 63c (zero). Has become. It should be noted that there is a difference in inclination between the first straight portion 63a and the second straight portion 63b. Due to this difference, a small amplification coefficient ΔG can be obtained in the small value region of the correction basic signal E _BASIC , while the correction basic signal E _BASIC can be obtained.
_BASIC relatively in the middle value area (compared to the minute value area)
A large amplification coefficient ΔG can be obtained, and as a result, it is possible to “adaptively correct various types of edges from sharp edges to gentle edges”, thereby solving the first problem of the present invention. This is because an electronic still camera can be provided.

The number of straight lines constituting the characteristic line is not limited to three (the first straight portion 63a to the third straight portion 63c) as illustrated. There may be two or three or more. In short, as the correction basic signal E _BASIC increases,
It is only necessary that the characteristic line structure be such that a large amplification coefficient ΔG can be obtained. In the figure, the inclination of the third straight portion 63c is set to zero, but this is a practical measure to prevent excessive edge enhancement. When the magnitude of the correction basic signal E _BASIC exceeds the second linear portion 63b, the magnitude of the amplification coefficient ΔG can be limited to a predetermined value, and excessive edge enhancement (overcorrection)
Can be prevented.

In the above embodiment, the correction basic signal E _BASIC extracted from the edge filter unit 53 is directly input to the gain adjustment unit 55. However, the present invention is not limited to this configuration. A part may be provided. FIG. 5B shows an example of the configuration, in which the coring unit 66 provided between the edge filter unit 53 and the gain adjustment unit 55 is the coring unit 13 in the above-described conventional example.
Works the same as. That is, the illustrated coring portion 6
Reference numeral 6 designates a correction basic signal E _{BASIC equal} to or greater than a predetermined threshold value as a correlated correction basic signal CE _BASIC to the gain adjustment unit 55 and limits the correction basic signal E _BASIC below the threshold value to “0”. And CE _BASIC = 0
To the gain adjustment unit 55. Also in the above embodiment, the same effect is obtained by setting the value of SL to “0.25”.
And the same effect as in the above embodiment can be obtained by a configuration in which the value of SL is set to “0”.

Next, a second object of the present invention (to make it possible to adaptively correct various types of edges from sharp edges to gentle edges without making isolated point noise conspicuous). Another embodiment that can achieve the above will be described. FIG. 6A is a configuration diagram of a main part thereof, and is a modified configuration diagram of the image correction processing unit 57 of the embodiment. In the figure, the same reference numerals are given to the same components as those in the above-described embodiment. Gain adjuster 5
5 is a correction basic signal E _BASIC from the edge filter unit 53
And the flatness evaluation filter unit 67 (“n” described in the gist of the invention).
× n pixel planes ”and an amplification coefficient ΔG ′ from the evaluation means).
Is output. The flatness evaluation filter unit 67 is, for example, a filter having a 3 × 3 configuration as shown in FIG.
The coefficient of the center pixel (pixel of interest) of the filter is “0”, the coefficient of the four upper, lower, left, and right pixels that have a large effect on the pixel of interest among peripheral pixels is “1/4”; The coefficient is “− ／”. As explained at the beginning, the relationship between these coefficients is such that the addition result of all the coefficients becomes “0” (or becomes almost zero).

It will be verified that isolated point noise is suppressed by such a configuration. 7 and 8 are explanatory diagrams thereof. First, in FIG. 7, the example of the pixel distribution shown at the left end is a normal edge portion (in this example, a vertical edge portion). For such a pixel array, intended edge enhancement must be performed. Edge filter unit 53
As shown in the figure, the edge filter unit 53
The product between the elements in is the level 1 pixel (upper, upper right, right, lower right, lower, and center pixels) of the pixel distribution example.
"1 x (-2/12)" and "1 x (-1/1
2) "," 1 x (-2/12) "," 1 x (-1/1)
2) "," 1 × (−2/12) ”and“ 1 × 1 ”, and the correction basic signal E _BASIC that is the result of adding them is“ 0.333 ”. On the other hand, the flatness evaluation filter unit 6
7 are “1 × (１ ／)” and “1” for level 1 pixels (upper, upper right, right, lower right, lower, and center pixels) of the pixel distribution example, respectively. × (-1 /
4), “1 × (１ ／)”, “1 × (− ／)”,
“1 × (1/4)” and “1 × 0”, and the amplification coefficient ΔG ′ resulting from these additions is “0.25”.
Therefore, the output (edge addition signal E) of the gain adjustment unit 55 in this case is “0.083”, and the value “0.083” is added to the luminance signal Y in the edge addition unit 56 (see FIG. 2).
As a result of adding 83 ", edge enhancement correction corresponding to the addition can be performed without any trouble.

Next, in FIG. 8, the example of the pixel distribution shown on the left end includes isolated point noise (only the pixel of interest has a level significantly different from the surroundings). A pixel array including isolated point noise is a non-edge portion, and edge enhancement should not be performed on such a pixel array. It is needless to say that the isolated point noise is emphasized and the image quality is rather deteriorated. In the above-mentioned conventional example, this measure is insufficient, and there is a problem in terms of image quality deterioration. Assuming that the coefficients of the edge filter unit 53 are as shown in FIG. 7 (the same as in FIG. 7), the product between the elements in the edge filter unit 53 is as follows:
“1 × 1” and the correction basic signal E _BASIC is “1.0”
Becomes On the other hand, the product between the elements in the flatness evaluation filter unit 67 is “1 × 0” for the pixel at level 1 (center pixel) in the pixel distribution example, and the amplification coefficient ΔG ′ is also “0”. Therefore, in this case, the gain adjustment unit 5
5 (edge addition signal E) becomes “0”, and even if this value “0” is added to the luminance signal Y in the edge addition section 56 (see FIG. 2), the luminance signal Y = the corrected luminance signal Yed
Because of g, the problem of enhancing isolated point noise can be solved without performing unnecessary edge enhancement correction.

As described above, the main functions of the present embodiment are realized by hardware by the color process circuit 35, but the idea of the present invention is not limited to this realized form. That is, a microcomputer (for example, FIG. 1)
It can also be functionally realized by an organic combination of a hardware asset including the CPU 40) and software assets such as an OS and various programs, and such a software implementation is also included in the concept of the present invention. Is done. In this case, since the hardware resources and the OS can be general-purpose ones, essential items indispensable for the present invention are substantially the application programs (or driver programs, etc.) which have already described the main functions. It can be said that they are aggregated. Accordingly, the present invention provides a recording medium, such as a floppy disk, an optical disk, a compact disk, a magnetic tape, a hard disk, or a semiconductor memory, or a component (unit) including these recording media, in which all of the program or a main part of the program is stored. Product, finished product or semi-finished product). In addition, the recording medium or the component includes not only a recording medium or a component on a distribution channel itself but also a recording medium or a component on a network that provides only recorded contents.

[0041]

According to the first, fourth, fifteenth, eleventh, or fourteenth aspects of the present invention, when the pixel value of an edge portion is in a region having a small value, the edge portion has a small pixel value. Edge enhancement is performed by correcting the pixel value to be weaker. On the other hand, when the pixel value of the edge portion is in a region where the pixel value is large, edge enhancement can be performed by correcting the pixel value of the edge portion more strongly. Therefore, it is possible to provide an edge emphasizing apparatus that can adaptively correct various types of edges from sharp edges (pixel values of edge portions are large) to gentle edges (pixel values of edge portions are small). . According to the invention described in claim 2, claim 9, claim 12, claim 16, or claim 17, when the flatness of the pixel value in the n × n pixel planes including the target pixel is large, for example, In the case of pixel distribution such as isolated point noise, a small amplification coefficient can be used, and the degree of edge enhancement can be reduced to eliminate the isolated point noise enhancement problem. According to the third or thirteenth aspect, an edge portion can be extracted by performing an operation between elements of a two-dimensional filter, and the two-dimensional filter can be easily configured by an image memory or the like. Can be simplified. According to the fifth or sixth aspect of the present invention, various types of edges are adaptively selected from sharp edges (pixel values of edge portions are large) to gentle edges (pixel values of edge portions are small). An edge enhancement device that can be corrected can be provided. According to the invention as set forth in claim 7 or claim 10, the problem of enhancing isolated point noise can be solved. According to the invention described in claim 8, excessive edge enhancement can be suppressed.

[Brief description of the drawings]

FIG. 1 is a block diagram of an electronic still camera.

FIG. 2 is a configuration diagram of a color process circuit 35.

FIG. 3 is a diagram illustrating a conceptual filter of an edge filter unit 53 and an example of a filter coefficient.

FIG. 4 is a correlation diagram showing a correspondence between a correction basic signal E _BASIC and an amplification coefficient ΔG, and a correlation diagram showing a correspondence between a correction basic signal E _BASIC and an edge addition signal E.

FIG. 5 is a diagram showing another characteristic line (having a bending point) representing the relationship between the correction basic signal _EBASIC and the amplification coefficient ΔG.

FIG. 6 is a main part configuration diagram of another embodiment.

FIG. 7 is an explanatory view (1/1/2) for verification of suppression of isolated point noise;
2).

FIG. 8 is an explanatory diagram for verifying suppression of isolated point noise (2 /
2).

FIG. 9 is a conceptual configuration diagram of a main part of a conventional electronic still camera and a configuration diagram of a color processing circuit 5.

FIG. 10 is a diagram illustrating a configuration diagram of an image correction processing unit 10, and a first pixel array example and a second pixel array example.

11 is an input / output characteristic diagram of the coring unit 13. FIG.

[Explanation of symbols]

E Edge addition signal (edge enhancement correction value) P _0.5 area of small value P _0.75 , area of P _1.0 large value ΔG Amplification coefficient ΔG 'Amplification coefficient 53 Edge filter unit (two-dimensional filter, extraction means) 57 Image correction processing unit (Edge enhancement device, gain adjusting means, changing means) 63a First linear portion (region of small value) 63b Second linear portion (region of large value) 67 Flatness evaluation filter unit (n × n pixel planes, evaluation) means)

Continued on front page F-term (reference) 5B057 CA08 CA12 CA16 CB08 CB12 CB16 CE03 CE06 CH01 CH11 DA08 DB02 DB09 DC16 5C021 PA17 PA53 PA58 PA66 PA67 PA79 PA99 RA02 RA08 RB08 RC06 SA22 SA25 XB03 5C077 LL06 MP07 MP08 PP03 PP10 PP03 PP10 PQ12 PQ18 TT09

Claims (17)

[Claims] 1. An edge emphasizing device for extracting an edge portion of an image and correcting the pixel value of the edge portion in an increasing direction to perform edge emphasis, wherein the edge value changes according to the magnitude of the pixel value of the edge portion. A variable amplification coefficient is generated, and the amplification coefficient is applied to the pixel value of the edge portion to generate a correction value for edge enhancement. The amplification coefficient is in an area where the pixel value of the edge portion is small. An edge emphasizing apparatus characterized in that the edge emphasizing device has a characteristic of increasing the edge emphasis correction value when it is located in a region having a larger value than when. 2. An edge enhancement device for extracting an edge portion of an image, correcting the pixel value of the edge portion in an increasing direction, and performing edge enhancement, comprising: a pixel value in an n × n pixel plane including a target pixel; Generating a variable amplification coefficient that changes in accordance with the flatness of the edge portion, and applying the amplification coefficient to the pixel value of the edge portion to generate an edge enhancement correction value. An edge emphasizing device having a characteristic of decreasing the edge emphasis correction value as the value increases. 3. The edge enhancement device according to claim 1, further comprising a two-dimensional filter for extracting the edge portion. 4. An extracting means for extracting a signal of an edge part in an image signal, a gain adjusting means for adjusting a gain of a signal of an edge part extracted by the extracting means, and a gain adjusted by the gain adjusting means. Edge enhancing means for enhancing an edge part in the image signal using the signal of the edge part, and a degree of gain adjustment by the gain adjusting means according to a level of the edge part signal extracted by the extracting means. An edge emphasizing device comprising: 5. The method according to claim 1, wherein the change unit changes the gain adjustment degree such that a change amount of a signal level of the edge portion and a change amount of the gain adjustment degree maintain a predetermined relationship. The edge enhancement device according to claim 4, wherein 6. The apparatus according to claim 5, wherein the changing unit changes the degree of the gain adjustment so that the predetermined relationship is different between an area where the signal level of the edge portion is low and an area where the signal level is high. An edge enhancement device as described. 7. The gain adjusting means, wherein the level of the signal at the edge portion, the gain of which is adjusted by the gain adjusting means, becomes zero when the level of the signal at the edge portion is equal to or less than a predetermined value. 7. The edge emphasizing device according to claim 4, further comprising a unit for changing a degree of gain adjustment based on the threshold value. 8. The apparatus according to claim 4, wherein said changing means includes means for fixing a degree of gain adjustment by said gain adjusting means when a signal level of said edge portion is equal to or more than a predetermined value. 7. The edge enhancement device according to any one of 7. 9. An extracting means for extracting a signal of an edge part in an image signal, a gain adjusting means for adjusting a gain of a signal of an edge part extracted by the extracting means, and a gain adjusted by the gain adjusting means. Edge enhancing means for enhancing an edge portion in the image signal using the signal of the edge portion, and n × n flatness evaluation filters in which the weighting coefficient of the center point is smaller than the surrounding weighting coefficient. Evaluation means for evaluating the flatness of the image signal, and changing means for changing the degree of gain adjustment by the gain adjustment means according to the flatness of the image signal evaluated by the evaluation means. Edge enhancement device. 10. The edge emphasizing apparatus according to claim 9, wherein a weighting coefficient of the center point is 0. 11. An edge emphasizing method for extracting an edge portion of an image and correcting the pixel value of the edge portion in an increasing direction to perform edge emphasis, wherein the edge value changes in accordance with the pixel value of the edge portion. A variable amplification coefficient is generated, and the amplification coefficient is applied to the pixel value of the edge portion to generate a correction value for edge enhancement. The amplification coefficient is in an area where the pixel value of the edge portion is small. An edge emphasizing method characterized by having a characteristic of increasing the edge emphasis correction value when in an area having a larger value than when. 12. An edge enhancement method for extracting an edge portion of an image, correcting the pixel value of the edge portion in an increasing direction, and performing edge enhancement, comprising: a pixel value in an n × n pixel plane including a target pixel; Generating a variable amplification coefficient that changes in accordance with the flatness of the edge portion, and applying the amplification coefficient to the pixel value of the edge portion to generate an edge enhancement correction value. An edge emphasizing method characterized by having a characteristic of decreasing the edge emphasis correction value as the value increases. 13. The edge enhancement method according to claim 11, wherein the extraction of the edge portion is performed using a two-dimensional filter. 14. A recording medium storing a program for extracting an edge portion of an image, correcting a pixel value of the edge portion in an increasing direction, and performing edge enhancement, wherein a program corresponding to the magnitude of the pixel value of the edge portion is provided. Generating a variable amplification coefficient that changes by applying the amplification coefficient to the pixel value of the edge portion to generate a correction value for edge enhancement, and the amplification coefficient is a pixel value of the edge portion having a small value. A recording medium storing a program, characterized in that it has a characteristic of increasing the edge emphasis correction value in an area having a larger value than in an area. 15. An extracting means for extracting a signal of an edge part in an image signal, a gain adjusting means for adjusting a gain of a signal of an edge part extracted by the extracting means, and a gain adjusted by the gain adjusting means. Edge enhancing means for enhancing an edge part in the image signal using the signal of the edge part, and a degree of gain adjustment by the gain adjusting means according to a level of the edge part signal extracted by the extracting means. A storage medium storing a program for realizing a changing means for performing the above. 16. In a recording medium storing a program for extracting an edge portion of an image, correcting a pixel value of the edge portion in an increasing direction, and performing edge enhancement, the recording medium includes n × n pixel planes including a pixel of interest. Generate a variable amplification coefficient that changes in accordance with the flatness of the pixel value in, generate an edge enhancement correction value by applying the amplification coefficient to the pixel value of the edge portion, the amplification coefficient is A recording medium storing a program, wherein the recording medium has a characteristic of decreasing the edge enhancement correction value as the flatness increases. 17. An extracting means for extracting a signal of an edge part in an image signal, a gain adjusting means for adjusting a gain of a signal of an edge part extracted by the extracting means, and a gain adjusted by the gain adjusting means. Edge enhancing means for enhancing an edge portion in the image signal using the signal of the edge portion, and n × n flatness evaluation filters in which the weighting coefficient of the center point is smaller than the surrounding weighting coefficient. A program for implementing evaluation means for evaluating the flatness of the image signal; and changing means for changing the degree of gain adjustment by the gain adjustment means in accordance with the flatness of the image signal evaluated by the evaluation means. A recording medium characterized by being stored.