JP3480547B2 - Image correction method and apparatus, and storage medium storing program - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image correction method for correcting a digital original image, an image correction apparatus for executing the method, and a storage medium recording a program for causing a computer to execute the method.

[0002]

2. Description of the Related Art As a conventional image correction process of this type,
Noise removal filter processing and sharpness processing are known.

As is well known, the noise removal filter process uses each pixel of the original image as a target pixel in order, and the density data of the target pixel and the density data of each pixel forming a peripheral pixel group around the target pixel. It is a process of using the weighted average value as the density data of the pixel of the corrected image with respect to the pixel of interest,
By this processing, the noise component of the original image can be reduced.

As is well known, in the sharpness processing, a blurred image (unsharp mask) of the original image is obtained, and a difference between the original image and the blurred image is obtained for each corresponding pixel to obtain an edge component of each pixel, This is a process of adding the edge component of the corresponding pixel to each pixel of the original image. By this process, the sharpness of the edge component of the original image can be emphasized.

[0005]

However, the conventional example having such a structure has the following problems. For example, the original GG shown in FIG. 14, when capturing an image in a state in which each pixel of the imaging device is positioned, such as a CCD, as shown by the dotted line in the figure, captured image (original image) OG Figure 15 As shown in. It should be noted that, as shown in FIG. 1
5 (a) shows the original image OG, and FIG. 5 (b) shows the original image OG.
The density data of each pixel are numerical values, and FIG. 7C is a graph showing the density data, and the density data is 0.00 to 1.0.
It is assumed to be normalized to the range of 0.

As shown in FIG. 15 , in the original image OG, noise components ((t _X2 ,
t _Y ) pixels) and the edge E of the original GG.
Since G passes through the pixel of (t _X6 , t _Y ), the sharpness of the edge EG is lower than that of the original GG.

[0007] with respect to the original image OG shown in FIG. 15,
When the left and right pixels of the target pixel subjected to the noise removal filter as peripheral pixel group becomes as shown in FIG. 16 (a),
When subjected to sharpness processing as shown in FIG. 16 (b).

As shown in FIG. 16 , in the case of the noise removal filter processing, the noise component of the pixel of (t _X2 , t _Y ) of the original image is reduced, but at the same time, the density data of the pixel in the vicinity of t _X6 is reduced. As a result, the sharpness of the edge EG portion of the original GG is further lowered than that of the original image OG. Further, in the case of the sharpness processing, the sharpness of the edge EG portion of the document GG is improved, but at the same time, (t _X2 , t of the original image).
The noise component of the pixel of _{Y 2} ) is also emphasized and the image quality is degraded.

As described above, in the conventional image correction, there is a problem that the sharpness of the edge is lowered at the same time as the noise component is reduced, while the noise component is emphasized while the sharpness of the edge is improved. there were.

The present invention has been made in view of the above circumstances, and an image correction method and apparatus capable of improving the sharpness of an edge and reducing a noise component by a single image correction, and an apparatus thereof. It is an object to provide a storage medium recording a program.

[0011]

The present invention has the following constitution in order to achieve such an object. That is, the invention according to claim 1 is, in an image correction method for correcting a digital original image, (1-1) a step of calculating a contrast amount near a target pixel of the original image, and (1-2)
Among the density data of each pixel forming the adjacent pixel group adjacent to the target pixel, the density data closest to the density data of the target pixel (hereinafter, this density data is referred to as “correction value”), and the target pixel And the weighted average value (hereinafter, this weighted average value is simply referred to as “weighted average value”) of the density data of each of the pixels forming the peripheral pixel group around the pixel of interest and It is determined according to the value of the contrast amount so that the mixing ratio of the correction value increases as the value of the amount increases, and the mixing ratio of the weighted average value increases as the value of the contrast amount decreases. And (1-3) the result of mixing the correction value and the weighted average value with the determined mixing ratio as the density data of the pixel of the correction image with respect to the target pixel. , It is characterized in that to obtain a corrected image obtained by correcting the original image as a sequential pixel of interest of each pixel of the original image comprises a.

[0012]

According to a second aspect of the present invention, in an image correction device that corrects a digital original image, an original image storage unit that stores the original image, and a corrected image storage unit that stores the corrected image obtained by correcting the original image. A contrast amount calculating means for calculating a contrast amount in the vicinity of the target pixel of the original image, and a density closest to the density data of the target pixel among the density data of each pixel forming an adjacent pixel group adjacent to the target pixel. Data (Hereafter, this density data is called "correction value")
And a weighted average value of the density data of the target pixel and the density data of each pixel forming a peripheral pixel group around the target pixel (hereinafter, this weighted average value is simply referred to as “weighted average”). Value)), a mixing ratio of the correction value is increased as the value of the contrast amount is increased, and a mixing ratio of the weight average value is decreased as the value of the contrast amount is decreased. So as to increase the amount, a mixing ratio determining means for determining a mixing ratio between the correction value and the weighted average value according to the value of the contrast amount, and the correction value and the weighted average value at the determined mixing ratio. And a mixing unit that stores the result of mixing the corrected image storage unit in the corrected image storage unit as density data of the pixel of the corrected image with respect to the target pixel, And control means for controlling so as to obtain the serial corrected image, is characterized in that it comprises a.

[0014]

According to a third aspect of the present invention, there is provided a storage medium in which a program for causing a computer to execute a process of correcting a digital original image is recorded. A step of calculating the amount, and (1-2) the density data closest to the density data of the pixel of interest (hereinafter referred to as density data) among the density data of each pixel forming the adjacent pixel group adjacent to the pixel of interest. Is referred to as a “correction value”) and a density average value of the density data of the pixel of interest and density data of each pixel forming a peripheral pixel group around the pixel of interest (hereinafter, this weighted average value is simply referred to as a “weighted average”). Value)) as the mixing ratio of the correction value increases as the value of the contrast amount increases, and the weighted average value of the weighted average value decreases as the value of the contrast amount decreases. A step of determining according to the value of the contrast amount so as to increase the mixing ratio,
(1-3) The step of setting the result of mixing the correction value and the weighted average value at the determined mixing ratio as the density data of the pixel of the correction image with respect to the pixel of interest, A program for causing a computer to execute a process for obtaining a corrected image in which an original image is corrected by sequentially setting pixels as a target pixel is recorded in a storage medium.

[0016]

[0017]

The image correction method according to the invention described in claim 1 is
Using each pixel of the original image as a target pixel in sequence, the following steps (1-1) to (1-3) are performed for each target pixel to obtain a corrected image in which the original image is corrected.

In step (1-1), the contrast amount in the vicinity of the current pixel of interest of the original image is calculated.

In the step (1-2), among the density data of each pixel forming the adjacent pixel group adjacent to the current target pixel, the density data (correction value) closest to the density data of the target pixel is obtained. , The weighted average value (weighted average value) of the density data of the target pixel and the density data of each pixel forming the peripheral pixel group around the target pixel is (1-
As the value of the contrast amount calculated in the step 1) increases, the mixing ratio of the correction value increases, and as the value of the contrast decreases, the mixing ratio of the weighted average value increases. Determine according to the value of.

In step (1-3), the result of mixing the correction value and the weighted average value with the mixing ratio determined in step (1-2) is used as the density of the pixel of the corrected image with respect to the current pixel of interest. Data.

If the edge passes through the target pixel of the original image, the contrast amount near the target pixel becomes large, while the density data of the pixels near the target pixel of the original image are substantially the same (flat portion). Alternatively, the contrast amount in the vicinity of the target pixel becomes small in the portion where the density data changes smoothly. Further, if there is a noise component in the flat portion or the portion where the density data changes smoothly, the contrast amount in the vicinity thereof will be slightly larger, but it is smaller than the contrast amount due to the edge.

Therefore, as described above, by determining the mixture ratio of the correction value and the weighted average value according to the contrast amount in the vicinity of the pixel of interest, in the correction of the edge portion of the original image, the mixture ratio of the correction value is increased. Then, the correction value and the weighted average value are mixed (if necessary, the mixing ratio of the weighted average value is set to “0”) to obtain the density data of the pixel of the corrected image. The correction value is the density data closest to the density data of the target pixel among the density data of each pixel forming the adjacent pixel group adjacent to the target pixel. By increasing the mixture ratio of the correction values, the density data of the pixel of the corrected image with respect to the pixel of interest becomes substantially the same as the density data of the adjacent one pixel. If the edge passes through the target pixel of the original image, the density of the adjacent pixel on one side of the target pixel is dark, the density of the adjacent pixel on the other side is light, and the density of the target pixel is in the middle, dark, The sharpness of the edge of the original image was reduced by arranging the middle and light pixels, but by the above image correction, the corrected image is dark, dark, light (or dark,
Light (light) pixels are arranged, and the sharpness of the edge can be improved. Further, since the density data of the adjacent pixel closest to the density data of the pixel of interest is selected as the correction value, it is possible to emphasize the sharpness of the edge by reflecting the position of the edge in the pixel of interest.

On the other hand, in the correction of the flat part of the original image or the part where the density data is changing smoothly, the mixing ratio of the weighted average value is increased (if necessary, the mixing ratio of the correction value is changed to " (“0”), the correction value and the weighted average value are mixed to obtain the density data of the pixel of the corrected image, so that the noise component of the flat part of the original image or the part where the density data is smoothly changed is the pixel of interest. The weighted average of the density data of a plurality of pixels around the pixel is averaged and reduced.

When the contrast amount in the vicinity of the pixel of interest is an intermediate value, the correction value and the weighted average value are mixed at a mixing ratio according to the value of the contrast amount, so that a natural gradation change is impaired. The image can be corrected without causing any tone jump or the like.

[0025]

The image correction apparatus according to the second aspect of the present invention is for suitably implementing the image correction method according to the first aspect of the invention, and its operation is as follows.

The contrast amount calculation means calculates the contrast amount in the vicinity of the target pixel of the original image using the density data of necessary pixels of the original image stored in the original image storage means, and the mixture ratio determination means, Corrected according to the value of the contrast amount so that the mixed ratio of the correction value increases as the value of the calculated contrast amount increases, and the mixing ratio of the weighted average value increases as the value of the contrast amount decreases. Determine the mixing ratio of the value and the weighted average.
On the other hand, the correction value extraction means extracts the correction value using the density data of the target pixel of the original image stored in the original image storage means and the density data of each pixel forming the adjacent pixel group,
The weighted average value calculation means uses the density data of the target pixel of the original image stored in the original image storage means and the density data of each pixel forming the peripheral pixel group to calculate the weighted average value of the density data of those pixels. calculate. Then, the mixing unit stores the result of mixing the correction value and the weighted average value at the determined mixing ratio in the correction image storage unit as the density data of the pixel of the correction image with respect to the target pixel. The control unit controls each unit so that each pixel of the original image is sequentially set as a target pixel to obtain a corrected image.

[0028]

According to the third aspect of the present invention, by causing the computer to read the program recorded in the storage medium, the computer executes the image correction processing by the method of the first aspect of the present invention.

[0030]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an example of the apparatus configuration when the image correction apparatus according to the present invention is configured by a computer system.

In the configuration of FIG. 1, the CPU 10 which executes the image correction process for correcting the digital original image to obtain the corrected image.
Is connected to the internal memory 30 configured by a RAM or the like via the bus line 20. In this internal memory 30, a program memory 31 into which a program is ported when executing image correction processing, an original image memory 32 for storing an original image, a corrected image memory 33 for storing a corrected image, a contrast value / mixture ratio table. Each area such as a table memory 34 for storing is set. Also,
The CPU 10 is connected via an input / output interface 40 to a driver 41 for an external storage device, an original image input device 42, an output device 43, a setting device 44, and the like.

An external storage device 45 is loaded in the driver 41. The external storage device 45 stores an image correction processing program created in advance. External storage device 4
Reference numeral 5 comprises a magneto-optical disk, a CD-ROM, a floppy disk, a hard disk, a magnetic tape and other storage media. When the power of the apparatus is turned on and the image correction processing according to the present embodiment is activated, the image correction processing program is read from the external storage device 45 and the program memory 3 is read.
1 and the CPU 10 executes the image correction processing described later according to the image correction processing program. The external storage device 45 in which this image correction processing program is stored is
It corresponds to a storage medium in which the program according to the invention described in claims 5 and 6 is recorded.

The original image input device 42 is a device for inputting an original image, and is, for example, a digital camera, an input scanner, an image capturing device using a line sensor such as a CCD disclosed in Japanese Patent Laid-Open No. Hei 3-16367. Composed of. The original image input by the original image input device 42 is stored in the original image memory 32, the original image is subjected to image correction processing described later, and the obtained corrected image is stored in the corrected image memory 33. . An original image input in advance is stored in an external storage device, the original image is read from the external storage device, stored in the original image memory 32, and an image correction process is performed on the original image. You may comprise.

The original image memory 32, as shown in FIG.
The two-dimensional digital original image is configured to be stored, and the modified image memory 33 stores the two-dimensional digital original image as shown in FIG.
It is configured to store a dimensional digital modified image. Note that t _X and t _Y in the figure represent coordinate values of each pixel of the original image and the corrected image, P (t _X , t _Y ) is density data of each pixel constituting the original image, and Q (t _X , T _Y ) indicates the density data of each pixel forming the corrected image.

The output device 43 is for outputting the corrected image generated in the corrected image memory 33, and is composed of, for example, a display device, a recording scanner, a printer and the like. The modified image may be stored in the external storage device.

The setting device 44 is for performing various settings, designations, instructions, etc., and is composed of a keyboard, a mouse, a switch, a setting monitor, and the like.

FIG. 4 is a functional block diagram of a first image correction device for executing a first image correction process described later. A contrast amount calculation unit 51, a mixture ratio determination unit 52,
Correction value extraction unit 53, weighted average value calculation unit 54, multiplication unit 5
5, 56, the addition unit 57, and the control unit 58 are C in the configuration of FIG.
Corresponds to PU10. In addition, a plurality of CPUs 10 are provided, and the processes of the respective units 51 to 58 are performed in parallel, for example, the process of the contrast amount calculation unit 51, the process of the mixture ratio determination unit 52, the process of the correction value extraction unit 53, and the weighted average. Value calculator 5
4 may be performed in parallel, and thereafter, the processing of the multiplication unit 56 and the processing of the multiplication unit 57 may be performed in parallel. The first image correction device corresponds to the image correction device according to the third aspect of the invention.

The first image correction process with the above configuration will be described below with reference to the flowchart shown in FIG. The first image correction process corresponds to the image correction method according to the first aspect of the invention and the image correction process according to the fifth aspect of the invention.

It is assumed that the original image is stored in the original image memory 32 at the time when the processing is started, and a contrast value / mixing ratio table, which will be described later, is set in advance by the setting device 44 and stored in the table memory 34. It is assumed that

Step S1 of FIG. 5: The control unit 58 sets initial values (coordinate values of the pixel to start processing) to the coordinate values t _X and t _{Y of the} target pixel of the original image.

Step S2 of FIG. 5 ((1-1) of the present invention)
(Corresponding to the process of)): The contrast amount calculation unit 51 reads out the density data of necessary pixels from the original image memory 32,
The contrast amount Ct in the vicinity of the current pixel of interest (pixel of (t _X , t _Y ) of the original image) is calculated, for example, by the following arithmetic expression (1)
Calculate using any formula from ~ (10).

Ct = Ct _UD (1) Ct = Ct _LR (2) Ct = Ct _LU-RD (3) Ct = Ct _RU-LD (4) Ct = (Ct _UD + Ct _LR ) / 2 (5) Ct = (Ct _LU-RD + Ct _RU-LD ) / 2 (6) Ct = (Ct _UD + Ct _LR + Ct _LU-RD + Ct _RU-LD ) / 4 (7) Ct = MAX [Ct _UD , Ct _LR ] (8) Ct = MAX [Ct _LU-RD , Ct _RU-LD ] (9) Ct = MAX [Ct _UD , Ct _LR , Ct _LU-RD , Ct _RU-LD ] (10) , MAX [A, B, ...] Is a function for extracting the maximum value from A, B ,.

Ct _UD is the amount of contrast in the up-down direction near the pixel of interest, Ct _LR , Ct _LU-RD , Ct.
_RU-LD is in the left-right direction near the pixel of interest, upper left-lower right direction,
It is the contrast amount in the upper right-lower left direction and is calculated as follows.

Ct _UD = ABS {([sum of density data of all pixels in upper neighborhood area AU] / [number of pixels in upper neighborhood area AU])-([density data of all pixels in lower neighborhood area AD] Of total] / [the number of pixels in the lower neighborhood area AD])} Ct _LR = ABS {([total of density data of all pixels in the left neighborhood area AL] / [number of pixels in the left neighborhood area AL])-
([Sum of density data of all pixels in right neighborhood area AR] /
[Number of pixels in right neighboring area AR])} Ct _LU-RD = ABS {([sum of density data of all pixels in upper left neighboring area ALU] / [number of pixels in upper left neighboring area ALU])-([ Sum of density data of all pixels in lower right neighborhood area ARD] / [number of pixels in lower right neighborhood area ARD]) Ct _RU-LD = ABS {([of density data of all pixels in upper right neighborhood area ARU] Total] / [number of pixels in upper right neighborhood area ARU])-([sum of density data of all pixels in lower left neighborhood area ALD] / [number of pixels in lower left neighborhood area ALD])} where ABS {} is , Absolute value.

Each neighboring area is set, for example, as shown in FIG. 6A to 6D show setting examples of the upper neighborhood area AU, the right, bottom, and left neighborhood areas AR and A are shown.
D and AL are 90 degrees clockwise in the upper neighborhood area AU of FIGS. 6A to 6D with the center CP of the target pixel AP as a rotation axis.
It is a region rotated by 180 °, 180 °, 270 °. Also,
6E to 6H show setting examples of the upper left neighborhood area ALU, the upper right neighborhood, the lower right neighborhood, and the lower left neighborhood neighborhood ARU, A
RD and ALD are the upper left neighborhood area A in FIGS. 6 (e) to 6 (h).
LU is a region rotated 90 °, 180 °, and 270 ° clockwise around the center CP of the pixel of interest AP as a rotation axis.

As shown in FIG. 6, each neighboring area may be set in the peripheral area (peripheral pixel) of the target pixel AP, or may be set including the target pixel AP and its peripheral pixels.
In addition, it is preferable to set the neighboring regions of each pair of top-bottom, left-right, upper left-lower right, and upper right-lower left to be symmetrical, and each neighboring region of each pair includes the same number of pixels. It is preferable to set as follows. Further, the distance LK that defines the outward spread of each neighborhood area from the center CP of the pixel of interest AP depends on the extent of blurring of the original image, and as the extent of blurring of the original image increases, the distance LK becomes longer, and each neighborhood is larger. The area of the area (the number of pixels included in the neighborhood area) depends on the degree of noise contained in the original image, so that the area of each neighborhood area becomes wider as the extent of noise contained in the original image increases. Set the area.

Step S3 of FIG. 5 ((1-2) of the present invention)
The mixing ratio determination unit 52 refers to the contrast value / mixing ratio table stored in the table memory 34, and sets the correction value (CN) and the weighted average value (HN). Mixing ratio [km: (1-km)]
Is increased with an increase in the contrast amount Ct calculated in step S3, and is increased with a decrease in the contrast amount Ct as a mixture ratio (1-km) of the weighted average value HN. Is determined according to the value of the contrast amount Ct. In addition,
Details of the correction value CN and the weighted average value HN will be described later.

The contrast value / mixing ratio table is, for example, a table showing the relationship between the value of the contrast amount Ct and km shown in FIG. In FIGS. 7A to 7C, when the value of the contrast amount Ct is considerably small (Ct1 or less in the figure), it is regarded as the contrast due to noise, and km is as small as possible (“0.
0 ”), and the value of the contrast amount Ct is a predetermined value (C in the figure
When t2) is exceeded, it is regarded as the contrast due to the edge, and km is set to the maximum value (1.0). Ct1, above
Ct2 may be experimentally obtained in advance. Note that FIG.
If the mixing ratio is determined as shown in (b), the image can be corrected more smoothly than the mixing ratio is determined as shown in FIG. 7 (a). Moreover, as shown in FIG. 7D, the maximum km is km.
By suppressing the value to 1 (km1 <1.0, for example, 0.5), it is possible to prevent the corrected image from becoming unnatural when the natural image or the like is corrected. Note that Ct _MAX in the figure is the maximum value of the contrast amount Ct. The mixing ratio of the correction value CN and the weighted average value HN is increased by the mixing ratio of the correction value CN as the value of the contrast amount Ct increases.
In order to increase the mixing ratio of the weighted average value HN as the value of the contrast amount Ct becomes smaller in accordance with the value of the contrast amount Ct, the values shown in FIG.
The contrast amount C according to the graphs other than the relationship between the value of the contrast amount Ct and km as shown in each graph of (e)
The mixing ratio may be determined according to the value of t.

Step S4 of FIG. 5: Correction value extraction unit 53
Extracts the density data (correction value CN) closest to the current density data of the target pixel from the density data of each pixel forming the adjacent pixel group adjacent to the current target pixel.

As shown in FIG. 8A, the pixel of interest AP is
On the other hand, upper left, upper, upper right, left, right, lower left, lower, lower right 8
Adjacent pixel NP_LU, NP_U, NP_RU, NP_L, NP_R,
NP _LD, NP_D, NP_RDFrom, for example, as shown in FIG.
Set multiple pixels at appropriate positions so that these pixels
Adjacent pixel group NPG. Then, the set adjacent pixel group
Density data of each pixel forming the NPG (see FIG. 8 (c))
Of the density data (P (t_X, T
_Y)) The density data closest to
It For example, if the adjacent pixel group is NPG = {NP_L, NP_R}
If set to ABS {P (t_X, T_Y) -P (t_X−
1, t_Y)} And ABS {P (t_X, T_Y) -P (t_X+
1, t_Y)} Is compared with the size, and if the former is small, P
(T_X-1, t_Y) Is a modified value CN, and the latter is small
If P (t_X+1, t_Y) Is the correction value CN.

There may be a plurality of density data having the same proximity to the density data of the target pixel in the density data of each pixel forming the adjacent pixel group in the following three forms. The first form is larger than the density data of the pixel of interest 2
In the second mode, two or more density data smaller than the density data of the pixel of interest are the density of the pixel of interest. This is the case when they are the same close to the data.
The third mode is a case in which the density data larger than the density data of the pixel of interest and the density data smaller than the density data of the pixel of interest have the same proximity to the density data of the pixel of interest. In the first and second modes, since the values of two or more density data that have the same proximity to the density data of the pixel of interest are all the same,
The density data may be used as the correction value CN. On the other hand, the third
In the above form, since the density data of two types, large and small, are close to the density data of the pixel of interest across the density data of the pixel of interest, the density data of the larger one is selected or the density data of the smaller one is selected. Whether to select is determined in advance, or a random function or the like is used to randomly select each process.

Step S5 of FIG. 5: Weighted average value calculator 5
4 is the density data (P (t _X ,
t _Y )) and the weighted average value HN of the density data of each pixel forming the peripheral pixel group around the current target pixel are calculated.

For example, as shown in FIG. 9A, in the case of targeting a range of 3 × 3 pixels centering on the pixel of interest AP, 8 adjacent to the pixel of interest AP as in the above adjacent pixel group. A plurality of pixels at appropriate positions are set from the pixels (pixels indicated by diagonal lines), and these pixels are set as a peripheral pixel group. As the pixels of the peripheral pixel group, pixels at the same position as the pixels of the adjacent pixel group may be selected, or pixels at different positions may be selected. Further, although the target pixel of the adjacent pixel group is limited to eight pixels adjacent to the target pixel, in the case of this peripheral pixel group, pixels further outside thereof, for example, as shown in FIG. 9B. In addition, a plurality of pixels at appropriate positions are set from the 24 pixels (pixels indicated by diagonal lines) around the pixel of interest AP in the range of 5 × 5 pixels around the pixel of interest AP, and these pixels are set as a peripheral pixel group. Good.

The weighted average of these pixels may be performed by setting the weights of the respective pixels used for the weighted average calculation to be the same (1 / [the number of pixels of the pixels used for the weighted average calculation]) or the weighted pixel of interest. It may be performed by making it larger than the weight of other pixels. For example, when the peripheral pixel group is set to {NP _L , NP _R } (see FIG. 8), the weighted average value HN is calculated as follows.

HN = (1/3) × P (t _X− 1, t _Y ) + (1/3) × P (t _X , t _Y ) + (1/3) × P (t _X +1, t _Y) ... (11) or, HN = (1/4) × P (t X -1, t Y) + (2/4) × P (t X, t Y) + (1/4) × P ( t _X +1, t _Y ) (12)

Step S6 of FIG. 5 ((1-3) of the present invention)
(Corresponding to the process of step S): The multiplication units 55 and 56 and the addition unit 57 and the correction value CN extracted in step S4 and step S4.
The weighted average value HN calculated in step 5 is mixed at the mixing ratio determined in step S3, and the result is stored as Q (t _X , t _Y ) in the corrected image memory 33. The formula is as follows. Q (t _X , t _Y ) = km × CN + (1-km) × HN

In step S7 of FIG. 5, the control unit 58 determines the end of the image correction process, and if not, sets t _X and t _Y to the coordinate values of the next target pixel in step S8, and then proceeds to step S2. And the steps S2 to S8 are repeated until the image correction process is completed, and each Q (t _X ,
t _Y ) is sequentially obtained, and a corrected image obtained by correcting the original image is generated.

The end of the image correction processing in step S8 is determined by whether or not all the pixels of the original image OG have been set as the target pixel. For example, each neighboring area for calculating the contrast amount near the target pixel is determined. Depending on the setting method of the pixel group, the adjacent pixel group, and the peripheral pixel group, the original image O
In some cases, it may not be possible to correct the image by using each pixel in the outer peripheral portion of G as the target pixel. In such a case, it is determined whether each pixel inside the diagonal line is the processing range and all the pixels within these processing ranges are the target pixels. Then, it is determined whether the image correction processing in step S8 is completed.

When the original image of FIG. 15 is image-corrected by the first image correction process under the following processing conditions, the result shown in FIG. 11 is obtained.

<Processing Condition> (A) The contrast amount near the target pixel is obtained by the above equation (2), and the left neighboring area is set to one pixel on the left side of the target pixel and the right neighboring area is set to one pixel on the right side of the target pixel. To do. (B) The mixing ratio is determined based on Fig. 7 (c). However, C
Let t1 = 0.10 and Ct2 = 0.50. (C) The adjacent pixel group and the peripheral pixel group are set to a total of two pixels, one pixel on the left side of the target pixel and one pixel on the right side. (D) The weighted average value HN is calculated by the above equation (11).

FIG. 11A shows the calculation result of the contrast amount Ct when each pixel is the target pixel, and FIGS. 11B to 11D show the processing results. In addition, in FIG. 11, the density data of the pixel on the left side of (t _X1 , t _Y ) is calculated as “0.30”, and the density data of the pixel on the right side of (t _X8 , t _Y ) is calculated as “0.80”. ing.

As is clear from FIG. 11, in the processing in which the pixel of (t _X2 , t _Y ) to which the noise component is added at the time of capturing the image is the target pixel, the mixture ratio is km = 0.0.
Therefore, the noise component of the pixel of (t _X2 , t _Y ) is reduced by the weighted average with the pixels of (t _X1 , t _Y ) and (t _X3 , t _Y ) forming the peripheral pixel group. Further, in the processing when the pixel of (t _X6 , t _Y ) through which the edge EG of the document GG has passed is the _target pixel, the mixture ratio becomes km = 1.0, and the pixel of (t _X5 , t _Y ) Concentration data of the corrected value CN
As the correction value CN is used as the density data of the pixel of (t _X6 , t _Y ) of the correction image, the density data of the pixel of (t _X6 , t _Y ) and the pixel of (t _X7 , t _Y ) The density change with the density data becomes sharp, and the sharpness of the edge EG portion of the original GG is improved as compared with the original image OG. Furthermore, (t _X3 , t _Y ), (t _X5 , t _Y ), (t _X7 , t _Y ).
In the processing when the target pixel is, the contrast value Ct
Since the mixing ratio of the correction value C and the weighted average value HN is determined according to the value of, and the correction value C and the weighted average value HN are mixed at the mixing ratio, the gradation change is smoothly corrected. .

As described above, according to the first image correction processing, each pixel of the corrected image is changed while changing the mixture ratio of the correction value CN and the weighted average value HN according to the value of the contrast amount Ct near the target pixel. Since the density data Q (t _X , t _Y ) of is obtained, the noise component of the original image OG can be reduced and the sharpness of the edge can be improved by one correction process.

The edge EG of the original GG is, for example,
As shown in FIG. 12A, when passing through the pixels of the original image OG (pixels of (t _X , t _{Y in the} figure) so that the area of high density increases, the position of the edge EG of the document GG Is
Since it is on the right side of the center line CL of the pixel (t _X , t _Y ) of the original image OG, the edge on the corrected image corresponding to the edge EG of the original GG is (t _X -1, t of the original image OG. pixel and (t _X of _Y), t _Y) and when located at the boundary between the pixels of the original image OG of (t _X, the pixel of t _Y) (t _X +1,
Compared with the case of being located at the boundary with the pixel of (t _Y ), the former has a larger deviation from the position of the edge EG of the document GG. Therefore, in the case of FIG. 12A, the edge EG ′ on the corrected image corresponding to the edge EG of the original GG is
As shown in (b), the one located at the boundary between the (t _X , t _Y ) pixel and the (t _X +1, t _Y ) pixel of the original image OG is closer to the position of the edge EG of the original GG. It is more preferable to reflect the position of the edge EG of the document GG in the corrected image. Similarly, in the case of FIG. 12 (c), it is preferable to modify as shown in FIG. 12 (d). Here, in the first image correction process, as the correction value CN, the density data closest to the density data of the target pixel among the density data of each pixel forming the adjacent pixel group, that is, the density of the target pixel Since the density data closer to the intensity is selected,
The correction result for the original image OG in FIG.
12B, and the correction result for the original image OG in FIG. 12C is shown in FIG. 12D, and the sharpness of the edge can be emphasized by reflecting the position of the edge EG of the document GG. Wear.

In the first image correction process,
The processing of steps S2 and S3, the processing of step S4,
The processing of step S5 but it may also be performed in parallel.

Further, when km can be 1.0 or 0.0 as the mixing ratio, the following processing may be performed. That is, as a result of determining the mixture ratio in step S3, when 0.0 <km <1.0, the correction value CN is extracted and the weighted average value HN is calculated as in the flowchart of FIG. , Correction value C with the determined mixture ratio
By mixing N and the weighted average value HN, Q (t _X , t _Y ) is obtained. On the other hand, as a result of determining the mixing ratio in step S3, k
When m = 1.0, the mixed calculation of the correction value CN and the weighted average value HN is substantially Q (t _X , t _Y ) = CN, and the weighted average value HN is not used in this calculation. Therefore, in this case, the weighted average value HN is not calculated, and the correction value C
When only N is extracted and the mixing ratio is determined in step S3 and km becomes 0.0, the correction value C
The mixed operation of N and the weighted average value HN is substantially Q (t _X ,
Since t _Y ) = HN, and the correction value CN is not used in this calculation, in this case, the correction value CN is not extracted and only the weighted average value HN is calculated. When the image correction processing is implemented by software, unnecessary processing can be omitted by performing the above processing, the load on the CPU 10 can be reduced, and the processing time can be shortened when processing is performed by one CPU 10. it is Ru can.

[0067]

[0068]

[0069]

[0070]

[0071]

[0072]

[0073]

[0074]

[0075]

[0076]

[0077]

[0078]

[0079] Incidentally, for example, in the case of image correction to the pixel column as soon as the pixel columns of each main scan direction is captured by the image capture apparatus using a line sensor, an original image memory 32 as shown in FIG. 13 The density data for one pixel column in the main scanning direction can be stored in the original image memory 32 shown in FIG. 13 every time the density data for one pixel column in the main scanning direction is captured by the image capturing device. The image may be stored and sequentially corrected in the first image correction process and stored in the corrected image memory 33.

[0080]

As is apparent from the above description, the claims
According to the invention described in 1 , the correction value or the maximum / minimum selection value and the weighted average value are mixed at a mixing ratio determined according to the contrast amount in the vicinity of the pixel of interest of the original image to correct the image. Therefore, the sharpness of the edge part of the original image can be emphasized, and the noise component is reduced in the flat part of the original image and the part where the density data changes smoothly, so that the image can be displayed without impairing the natural gradation change. It can be corrected, and a corrected image with good quality can be obtained.

According to the invention of claim 2 ,
It is possible to realize an apparatus capable of suitably implementing the image correction method according to the invention described in item 1 .

[0082] According to the invention described in claim 3, claim
The computer can be made to execute the image correction processing by the method of the invention described in 1 .

[Brief description of drawings]

FIG. 1 is a block diagram showing an example of a device configuration when an image correction device according to the present invention is configured by a computer system.

FIG. 2 is a diagram showing a configuration of an original image memory.

FIG. 3 is a diagram showing a configuration of a modified image memory.

FIG. 4 is a functional block diagram of a first image correction device.

FIG. 5 is a flowchart showing a procedure of first image correction processing.

FIG. 6 is a diagram for explaining a neighboring area used for calculating a contrast amount.

FIG. 7 is a diagram showing an example of a contrast value / mixing ratio table.

FIG. 8 is a diagram for explaining an adjacent pixel group.

FIG. 9 is a diagram for explaining a peripheral pixel group.

FIG. 10 is a diagram illustrating an example of a processing range of image correction processing.

FIG. 11 is a diagram showing a result of correcting the original image of FIG. 15 by a first image correction process.

FIG. 12 is a diagram for explaining the effect of the first image correction processing.

FIG. 13 is a diagram showing a configuration of a modified example of an original image memory.

FIG. 14 is a diagram showing an example of a positional relationship between a document and each pixel of an image pickup device when an image is captured.

15 is a diagram showing an original image captured in a positional relationship of FIG. 14.

FIG. 16 is a diagram for explaining a problem of the conventional method.

[Explanation of symbols]

10: CPU 31: Program memory 32: Original image memory 33: Modified image memory 34: Table memory 45: External storage device 51: Contrast amount calculation unit 52: Mixing ratio determination unit 53: Correction value extraction unit 54: Weighted average value calculation part 55: multiplier 57: adder unit 58: control unit _{P (t X, t Y)} : original image pixel density data _{Q (t X, t Y)} : the pixels in the corrected image density data Ct: contrast Quantity CN: corrected value HN: weighted average value km : mixed value of corrected value (1-km): mixed ratio of weighted average value

─────────────────────────────────────────────────── ─── Continuation of front page (56) Reference JP-A-7-203193 (JP, A) JP-A-61-157162 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) H04N 1/40-1/409

Claims (3)

(57) [Claims] 1. An image correction method for correcting a digital original image, comprising: (1-1) calculating a contrast amount in the vicinity of a pixel of interest of the original image; and (1-2) adjoining the pixel of interest. Of the density data of each pixel forming the pixel group, the density data closest to the density data of the target pixel (hereinafter, this density data is referred to as “correction value”), the density data of the target pixel and the target pixel The mixing ratio of the density data of each pixel forming the surrounding peripheral pixel group and the weighted average value (hereinafter, this weighted average value is simply referred to as “weighted average value”) is set as the contrast amount increases. Increasing the mixing ratio of the correction value and increasing the mixing ratio of the weighted average value as the value of the contrast decreases, a step of determining according to the value of the contrast, (1-3) a step of setting a result of mixing the correction value and the weighted average value at the determined mixing ratio as density data of a pixel of the correction image with respect to the pixel of interest, An image correction method characterized by obtaining a corrected image in which an original image is corrected by sequentially setting pixels as a target pixel. 2. An image correction apparatus for correcting a digital original image, wherein an original image storage means for storing the original image, a corrected image storage means for storing a corrected image obtained by correcting the original image, and a target pixel vicinity of the original image Contrast amount calculating means for calculating the contrast amount of the target pixel, and density data closest to the density data of the target pixel among the density data of each pixel forming the adjacent pixel group adjacent to the target pixel (hereinafter, this density data Is referred to as a "correction value"), and a weighted average value of density data of the pixel of interest and density data of each pixel forming a peripheral pixel group around the pixel of interest (hereinafter, this weight) (The average value is simply called "weighted average value")
And a weighted average value calculating means for calculating a mixture ratio of the correction value as the value of the contrast amount increases, and a mixture ratio of the weighted average value as the value of the contrast amount increases. As described above, a mixing ratio determining unit that determines a mixing ratio of the correction value and the weighted average value according to the value of the contrast amount, and the correction value and the weighted average value are mixed at the determined mixing ratio. Mixing means for storing the result in the corrected image storage means as density data of the pixel of the corrected image with respect to the target pixel; control means for controlling so that each pixel of the original image is sequentially used as the target pixel to obtain the corrected image; An image correction device comprising: 3. A storage medium having a program for causing a computer to execute a process of correcting a digital original image, comprising: (1-1) calculating a contrast amount near a target pixel of the original image; (1-2) Of the density data of each pixel forming the adjacent pixel group adjacent to the target pixel, the density data closest to the density data of the target pixel (hereinafter, this density data is referred to as “correction value”) And a weighted average value (hereinafter, this weighted average value is simply referred to as “weighted average value”) of the density data of the pixel of interest and the density data of each pixel forming a peripheral pixel group around the pixel of interest. The proportion is increased as the value of the contrast amount is increased by the mixing ratio, and is increased as the value of the contrast amount is increased by the mixing ratio of the weighted average value. As described above, (1-3) the result of mixing the correction value and the weighted average value at the determined mixing ratio is used as a result of the pixel of the corrected image for the target pixel. A storage medium in which a program for causing a computer to execute a process of obtaining density data, in which each pixel of the original image is sequentially corrected as a target pixel to obtain a corrected image, is recorded.