JP2011193073A - Image processing apparatus, printing system, image processing method, and program - Google Patents

Abstract

It is possible to appropriately perform red-eye correction.
A red-eye region specifying information generation unit 25 combines a correction intensity coefficient determined in a determination process and a correction intensity coefficient of image data of a predetermined number of lines determined in a previous determination process in a vertical blur process. And a horizontal blurring process for performing Gaussian blurring on the correction intensity coefficient obtained as a result using a plurality of correction intensity coefficients adjacent in the horizontal direction. The present invention can be applied to a printer.
[Selection] Figure 2

Description

  The present invention relates to an image processing apparatus, a printing system, an image processing method, and a program.

  When a person is photographed using a flash, a red-eye phenomenon occurs in which the eyes are photographed in red or gold. Some apparatuses having an image processing function such as a digital camera, a computer, or a printer can correct an image in which the red-eye phenomenon occurs by image processing (see Patent Document 1).

JP-A-10-233929

  However, in the conventional method, the region where the red-eye phenomenon occurs cannot be accurately identified, and as a result, there are cases where the red-eye correction cannot be performed appropriately.

  For example, if a large-capacity memory cannot be installed due to printer design limitations, a memory area necessary for image processing cannot be secured, and red-eye correction may not be performed properly.

  The present invention has been made in view of such a situation, and makes it possible to appropriately perform red-eye correction.

  An image processing apparatus according to an aspect of the present invention is an image processing apparatus that performs processing for correcting red eyes for each color image data having a predetermined number of lines. The distribution of the pixels constituting the red-eye region and the distribution of the other pixels are separated in the acquisition means for acquiring the image and the distribution of the feature amount of the correction target region in a space centered on the feature amount of at least two colors. The red-eye area specifying information generating means for generating the separation curve, the value of each pixel in the correction processing target area, and the magnitude of the correction of the pixel based on the value corresponding to the separation curve A correction intensity determining unit that performs a determination process for determining a correction intensity coefficient for each image data of a predetermined number of lines in the correction process target region in raster order, and the correction intensity determination unit is determined by the determination process. A vertical blur process is performed to synthesize the correction intensity coefficient and the correction intensity coefficient of the predetermined number of lines of image data determined in the previous determination process, and the resulting correction intensity coefficient is adjacent in the horizontal direction. A horizontal blurring process for performing Gaussian blurring using a plurality of correction intensity coefficients is performed.

  With such a configuration, for example, even when the work area is small and only one line of image data can be processed, red-eye correction can be performed appropriately.

  The correction intensity determining means can make the degree of blurring in the horizontal blurring process stronger than the degree of blurring in the vertical blurring process.

  Since it has such a structure, red eyes can be corrected more appropriately.

  The correction strength determination means can combine the predetermined ratio in the vertical blurring process, and can change the combination ratio depending on the difference between the correction intensity coefficients adjacent in the vertical direction.

  Since it has such a structure, red eyes can be corrected more appropriately.

  The correction intensity determination means can change the reference range in the Gaussian blurring process as the horizontal blurring process according to the radius of the reddish area of the correction process target area.

  Since it has such a structure, red eyes can be corrected more appropriately.

  An image processing method according to one aspect of the present invention is an image processing method of an image processing apparatus that performs a process of correcting red eyes for each color image data having a predetermined number of lines. In the acquisition step of acquiring as a correction process target area, and the distribution of the feature quantity of the correction process target area in a space centered on the feature quantity of at least two colors, the distribution of pixels constituting the red-eye area and other pixels A red-eye region specifying information generation step for generating a separation curve for separating the distribution of the image, a value of each pixel in the correction processing target region, and a correction of the pixel based on the value corresponding to the separation curve A correction strength determination step including a correction strength determination step for determining a correction strength coefficient indicating the strength of each of the image data for a predetermined number of lines in the correction processing target region in raster order. The step is obtained by performing a vertical blurring process that combines the correction intensity coefficient determined in the determination process and the correction intensity coefficient of the image data of the predetermined number of lines determined in the previous determination process, and the result is obtained. With respect to the correction intensity coefficient, horizontal blurring processing for performing Gaussian blurring is performed using a plurality of correction intensity coefficients adjacent in the horizontal direction.

  A program according to an aspect of the present invention is a program that causes a computer to execute image processing, and in the image processing method of an image processing apparatus that performs a process of correcting red eyes for each color image data having a predetermined number of lines. In the acquisition step of acquiring the region including the eye image of the image as the correction processing target region, and the distribution of the feature amount of the correction processing target region in the space centered on the feature amount of at least two colors, the red eye region is A red-eye region specifying information generation step for generating a separation curve for separating the distribution of the constituting pixels and the distribution of the other pixels, the value of each pixel in the correction processing target region, and the value corresponding to the separation curve A determination process for determining a correction strength coefficient indicating the correction strength for the pixel based on the size is performed in the raster order in a predetermined number of lines of image data in the correction target area. A correction intensity determination step that is performed for each image, and the correction intensity determination step combines the correction intensity coefficient determined in the determination process and the correction intensity coefficient of the predetermined number of lines of image data determined in the previous determination process. The computer performs image processing for performing horizontal blur processing for performing Gaussian blur using a plurality of correction strength coefficients adjacent in the horizontal direction for the correction strength coefficient obtained as a result of performing vertical blur processing. It is made to perform.

  In one aspect of the present invention, a region including an eye image in a color image is acquired as a correction processing target region, and the feature amount of the correction processing target region in a space having at least two color feature amounts as axes. In this distribution, a separation curve for separating the distribution of the pixels constituting the red-eye region from the distribution of the other pixels is generated, and the value of each pixel in the correction processing target region and the value corresponding to the separation curve are generated. A determination process for determining a correction intensity coefficient indicating the correction intensity for the pixel based on the size is performed for each image data of a predetermined number of lines in the correction process target region in raster order, and determined by the determination process. A vertical blurring process that combines the correction intensity coefficient and the correction intensity coefficient of the image data of the predetermined number of lines determined in the previous determination process is executed, and the correction intensity coefficient obtained as a result is Blurring in the horizontal direction to apply a Gaussian blur by using a plurality of correction intensity coefficient adjacent in the horizontal direction is executed.

1 is a perspective view illustrating a configuration of a printer 1 according to an embodiment of the present invention. FIG. 2 is a block diagram illustrating a functional configuration example of a printer 1 in FIG. 1. It is the figure which showed typically the printing process target image TI. FIG. 4 is a diagram schematically showing an enlarged image region including a left eye of a person at the center of the print processing target image TI in FIG. 3. 6 is a flowchart illustrating a flow of printing processing. It is a flowchart which shows the detail of a feature-value extraction process. It is the figure which plotted each pixel of correction | amendment process object area | region Pe1 on the plane which uses R value and R value / G value as an axis according to the R value and R value / G value. It is a flowchart which shows the flow of a correction | amendment intensity map production | generation process. It is a figure which shows the example of a correction | amendment intensity | strength coefficient table. It is a flowchart which shows the flow of an output data generation process.

  Embodiments of the present invention will be described below. Correspondences between the configuration requirements of the present invention and the embodiments described in the detailed description of the present invention are exemplified as follows. This description is to confirm that the embodiments supporting the present invention are described in the detailed description of the invention. Accordingly, although there are embodiments that are described in the detailed description of the invention but are not described here as embodiments corresponding to the constituent elements of the present invention, It does not mean that the embodiment does not correspond to the configuration requirements. Conversely, even if an embodiment is described here as corresponding to a configuration requirement, that means that the embodiment does not correspond to a configuration requirement other than the configuration requirement. Not something to do.

  An image processing apparatus according to an aspect of the present invention is an image processing apparatus that performs processing for correcting red eyes for each color image data having a predetermined number of lines. In the distribution of the feature quantity of the correction process target area in the space centered on the feature quantity of at least two colors and the acquisition means (for example, the correction process target area acquisition unit 24 in FIG. 2) A red-eye region specifying information generating unit (for example, the feature amount extraction unit 31 and the separation curve generation unit 32 in FIG. 2) that generates a separation curve for separating the distribution of the constituent pixels and the distribution of the other pixels, and a correction processing target region Based on the value of each pixel and the magnitude of the value corresponding to the separation curve, a determination process for determining a correction intensity coefficient indicating the correction intensity for the pixel is performed in the order of the correction process. Correction intensity determination means (for example, the correction intensity map generation unit 33 in FIG. 2) for each predetermined number of lines of image data, and the correction intensity determination means includes a correction intensity coefficient determined in the determination process, A vertical blurring process for combining the correction intensity coefficients of the image data of the predetermined number of lines determined in the previous determination process is executed (for example, step S37 in FIG. 8), and the correction intensity coefficient obtained as a result is horizontal. A horizontal blurring process (for example, step S38 in FIG. 8) for performing Gaussian blurring using a plurality of correction intensity coefficients adjacent to each other in the direction is performed.

[Description of Printer 1 Configuration]
FIG. 1 is a block diagram illustrating a hardware configuration example of a printer 1 according to an embodiment of the present invention. The printer 1 is a color inkjet printer that supports so-called direct printing, in which an image is printed based on image data acquired from a memory card 18 or the like. The printer 1 automatically detects a red-eye portion of a color image to be printed, and corrects the portion to a natural color (for example, black) (hereinafter referred to as an automatic red-eye correction function). have.

  The printer 1 includes a CPU 11, an internal memory 12, an operation unit 13, a display unit 14, a printer engine 15, and a card interface (card I / F) 16.

  A CPU (central processing unit) 11 controls each unit and executes various processes according to a program stored in the internal memory 12.

  For example, when the automatic red-eye correction mode is set, the CPU 11 selects an eye such as a person, a dog or other pet from a color image to be printed (hereinafter referred to as a print processing target image TI) prior to printing. Processing for detecting an area including an image or the like (hereinafter referred to as detection processing Z0) is executed. The CPU 11 also generates information for specifying a red-eye region in which a red-eye phenomenon has occurred (hereinafter referred to as red-eye region specifying information generation processing Z1) from the detected region (hereinafter referred to as a correction processing target region Pe1). And a process of replacing the color of the red-eye area specified by the information with a natural color (for example, black) (hereinafter referred to as a correction process Z2).

  The internal memory 12 includes a ROM (Read Only Memory) that stores various programs executed by the CPU 11 and various data, and a RAM (Random Access Memory) that temporarily stores programs and data to be executed by the CPU 11. Has been.

  The operation unit 13 includes one or a plurality of buttons and a touch panel, and notifies the CPU 11 of the operation content by the user.

  The display unit 14 includes a liquid crystal display and displays an image corresponding to the display data supplied from the CPU 11.

  The printer engine 15 is a printing mechanism that performs printing based on print data supplied from the CPU 11.

  The card interface (I / F) 16 is an interface for exchanging data with the memory card 18 inserted in the card slot 17. In this example, image data as RGB data is stored in the memory card 18, and the printer 1 acquires the image data stored in the memory card 18 via the card interface 16.

  The components of the printer 1 are connected to each other via a bus 19.

  The printer 1 may further include an interface for performing data communication with other devices (for example, a digital still camera or a computer).

[Description of Functional Configuration Example of Printer 1]
FIG. 2 is a block diagram illustrating a functional configuration example of the printer 1. This function can be realized by the CPU 11 executing a predetermined program stored in the internal memory 12.

  The image reading unit 21 reads the print processing target image TI from, for example, the memory card 18 via the card I / F 16.

  The read control unit 22 reads a predetermined number of lines of image data (in this example, one line of image data) from the print processing target image TI read by the image reading unit 21 and stores it in the line buffer 23. In the printer 1, print data to be printed is generated for each line of image data stored in the line buffer 23.

  The correction process target area acquisition unit 24 performs a detection process Z0 on a reduced image obtained by thinning a predetermined image from the print process target image TI read by the image reading unit 21. Specifically, for example, using a known detection method such as a pattern matching method using a template representing a basic shape of a person's eyes (for example, Japanese Patent Application Laid-Open No. 2006-279460), eyeballs, white eyes, eyelids, A process of acquiring a region including a part such as the eyes and the corners of the eyes is executed. The correction process target area acquisition unit 24 further executes a process of acquiring a region including a reddish pupil and iris from the area acquired as a result of this process (that is, the correction process target area Pe1). Note that the region acquired as a result of this processing is referred to as a red-eye candidate region Pe2.

  FIG. 3 is a diagram schematically showing the print processing target image TI. Here, the red-eye phenomenon occurring in the left eye portion of the person at the center of the print processing target image TI is assumed to be the processing target. FIG. 4 is a diagram schematically showing an enlarged image region (region surrounded by a square frame in FIG. 3) including the left eye of the person at the center of the print processing target image TI in FIG. FIG. 4 shows a correction processing target area Pe1 and a red-eye candidate area Pe2. In the case of this example, a rectangular area that includes parts such as eyeballs, white eyes, eyelids, eyes, and corners of the eye is detected as the correction processing target area Pe1, and a circular area that includes the eyeball in which the red-eye phenomenon occurs is a red-eye candidate. It is detected as a region Pe2.

  Returning to FIG. 2, the red-eye area specifying information generating unit 25 executes the red-eye area specifying information generating process Z1 using the correction process target area Pe1 and the red-eye candidate area Pe2 acquired by the correction process target area acquiring unit 24. In this example, as information for specifying the red-eye region, in the distribution of the feature amount of the correction processing target region Pe1 in the space having the color feature amount as an axis, the distribution of the pixels constituting the red-eye region and the other pixels A separation curve is generated that separates the distributions of. For example, a correction intensity map Ma that stores a correction intensity coefficient corresponding to the distance from the separation curve for each pixel is generated. The correction intensity map Ma has a storage unit corresponding to each pixel of the correction processing target region Pe1, and in each storage unit, the corresponding pixel of the correction processing target region Pe1 may be a pixel constituting a red-eye region. A correction strength coefficient corresponding to the sex is stored.

  The red-eye region specifying information generation unit 25 includes a feature amount extraction unit 31, a separation curve generation unit 32, and a correction intensity map generation unit 33. The operation of these components will be described later together with the description of the red-eye area specifying information generation process Z1.

  The correction unit 26 replaces the color of the print processing target image TI with black or the like with the correction intensity according to the correction intensity coefficient stored in the correction intensity map Ma generated by the red-eye area specifying information generation unit 25 (red-eye). Execute the process.

  The output control unit 27 generates print data from the image data corrected by the correction unit 26, supplies the print data to the printer engine 15, and executes printing of an image based on the print data.

  FIG. 5 is a flowchart showing the flow of the printing process. In step S <b> 11, the correction process target area acquisition unit 24 generates a reduced image in which predetermined pixels are thinned out from the print process target image TI read by the image reading unit 21. In step S12, the correction process target area acquisition unit 24 performs a detection process Z0 on the generated reduced image (hereinafter also referred to as a reduced print process target image TI as appropriate). That is, the correction processing target area Pe1 and the red-eye candidate area Pe2 are detected (FIG. 4). The detected positions and the like of the correction processing target area Pe1 and the red-eye candidate area Pe2 are stored in the internal memory 12.

  Next, in step S <b> 13, the feature amount extraction unit 31 of the red-eye region specifying information generation unit 25 extracts the feature amount of the correction processing target region Pe <b> 1 acquired by the correction processing target region acquisition unit 24. When there are a plurality of correction processing target areas Pe1, the feature amounts of the respective correction processing target areas Pe1 are extracted.

  FIG. 6 is a flowchart showing details of the feature amount extraction processing. When there are a plurality of correction process target areas Pe1, this process is executed for each correction process target area Pe1.

  In step S21, the feature amount extraction unit 31 determines whether or not the size of the correction processing target region Pe1 is an appropriate size. Specifically, for example, it is determined whether or not the width in the longitudinal direction of the correction processing target region Pe1 is twice or more the diameter of the red-eye candidate region Pe2. The reason for making such a determination on the size of the correction processing target area Pe1 will be described later.

  In step S21, if the size of the correction processing target region Pe1 is not an appropriate size (that is, in this example, the longitudinal width of the correction processing target region Pe1 is shorter than twice the diameter of the red-eye candidate region Pe2). If it is determined, in step S22, the feature amount extraction unit 31 enlarges the correction processing target region Pe1 to an appropriate size. In the case of this example, the pixels of the correction processing target region Pe1 are complemented with an enlargement ratio at which the width in the longitudinal direction of the correction processing target region Pe1 is twice the diameter of the red-eye candidate region Pe2.

  When it is determined in step S21 that the size of the correction processing target area Pe1 is an appropriate size, or when the correction processing target area Pe1 is enlarged to an appropriate size in step S22, in step S23, The feature amount extraction unit 31 detects the R component value (R value), the G component value (G value), and the B component value (B value) of each pixel in the correction processing target region Pe1, and corrects the correction processing target region Pe1. The average values of R value, G value, and B value are respectively calculated.

  Next, in step S24, the feature amount extraction unit 31 calculates the R value / G value standard deviation and the R value / B value standard deviation of the correction processing target region Pe1. Then, in step S25, the feature amount extraction unit 31 calculates the R value (axis) range (0 to 255) every 16 sections (hereinafter referred to as red component sections Lr) obtained by equally dividing the range. The average value of the R value / G value and the average value of the R value / B value of each pixel having the R value in the red component section Lr are obtained.

  In this way, “average values of R value, G value, and B value of correction processing target region Pe1”, “standard deviation of R value / G value and standard deviation of R value / B value of correction processing target region Pe1” And “average value of R value / G value and average value of R value / B value for each red component section Lr” are extracted as feature amounts. The extracted feature amount is stored in the internal memory 12 as appropriate. Thereafter, the processing proceeds to step S14 in FIG.

In step S14, the separation curve generation unit 32 of the red-eye area specifying information generation unit 25 obtains a separation curve used for generating the correction intensity map Ma. In this example, F _{R / G} (r) shown in Formula (1) and F _{R / B} (r) shown in Formula (2) are obtained. When there are a plurality of correction process target areas Pe1, a separation curve is generated for each correction process target area Pe1.

In the formula (1), H _{R / G} (r) is a curve that passes through the average value of the R value / G value of the red component section Lr. r is the R value. σ _{R / G} is a standard deviation of the R value / G value of the correction processing target region Pe1. α _{R / G} and β _{R / G} are predetermined coefficients. In Equation (2), H _{R / B} (r) is a curve that passes through the average value of the R value / B value of the red component section Lr. σ _{R / B} is a standard deviation of the R value / B value of the correction processing target region Pe1. α _{R / B} and β _{R / B} are predetermined coefficients.

  In other words, the shape of any separation curve is determined by the first term of the equations (1) and (2), and the R value / G value or the average value of the R value / B value in the red component section Lr is used as a reference. , And changes in irregularities having a size corresponding to the standard deviation of the R value / G value or the R value / B value. The change is finely adjusted by α. The height of the curve is determined by the second term of the equations (1) and (2), and the ratio of the average value of the R values to the average value of the G values or the average value of the R values in the correction processing target region Pe1. The ratio of the average value of B values is the standard. Note that the height is finely adjusted by β.

  Here, the meanings of the expressions (1) and (2) will be described.

  FIG. 7 is a diagram in which each pixel in the correction processing target region Pe1 is plotted on a plane having the R value and the R value / G value as axes, at positions corresponding to the R value and the R value / G value. A white square mark (□ mark) indicates a pixel in which a red red-eye phenomenon occurs (hereinafter, referred to as a red-red eye pixel as appropriate), and a black square mark (■ mark) indicates a pixel other than a red-red-eye pixel. A pixel (that is, a pixel in which a red red-eye phenomenon has not occurred (hereinafter, appropriately referred to as a non-red red-eye pixel)) is shown. In addition to the case where the pupil portion is red, the red eye phenomenon includes a gold phenomenon (so-called gold eye). A pixel in which the gold red eye phenomenon occurs will be described later.

  In the example of FIG. 7, most of the non-red red-eye pixels are gathered on the lower side. On the other hand, the red red-eye pixels are distributed and distributed in comparison with the non-red red-eye pixels, and many of them exist above the region where the non-red red-eye pixels are gathered.

By the way, when the size of the correction processing target region Pe1 (FIG. 4) is sufficiently large, the correction processing target region Pe1 includes a large number of pixels of the image other than the pupil. That is, the number of non-red red eye pixels in the correction processing target region Pe1 is overwhelmingly larger than the number of red red eye pixels. Therefore, when the size of the correction processing target region Pe1 is sufficiently large, the average value of the R value / G value for each R value (for example, for each red component section Lr) represents the distribution of non-red red-eye pixels. Become. Therefore, the curve H _{R / G} (r) passing through the average value of the R value / G value of the red component section Lr shown in the equation (1) is a curve along the distribution of the non-red red-eye pixels.

Further, the curve H _{R / G} (r) along the distribution of the non-red red-eye pixels is moved upward by a value corresponding to [ratio of the average value of the R value and the average value of the G value in the correction processing target region Pe1]. And (ie, when added), a curve along the upper side of the non-red red-eye pixel distribution can be obtained. Since many of the red-red-eye pixels exist above the area where the non-red-red-eye pixels are gathered, the curve along the upper side of this non-red-red-eye pixel distribution is the distribution of the non-red-red-eye pixel distribution and the red-red-eye pixel distribution. It becomes a curve indicating the boundary, that is, a curve separating the two. Therefore, F _{R / G} (r) in Expression (1) is a curve along the upper side of the distribution of the non-red red-eye pixels, and is a curve that separates the distribution of the non-red red-eye pixels and the distribution of the red-red eye pixels.

  On the other hand, a pixel in which the golden red-eye phenomenon occurs (hereinafter referred to as a golden red-eye pixel) is a yellowish color, and therefore, the golden red-eye pixel has a plane with R value and R value / G value as axes. It will be included in the distribution of non-red red-eye pixels. However, on the plane with the R value and the R value / B value as the axes, the distribution of the golden red eye pixels is distributed upward in the same manner as the red red eye pixels on the plane having the R value and the R value / G value as the axes. Distributed. Pixels other than the golden red-eye pixels (hereinafter referred to as non-golden red-eye pixels) are distributed so that many of them are gathered on the lower side in the same manner as the non-red red-eye pixels in the plane having the R value and R value / G value as axes. To do. Accordingly, the golden red-eye pixel is present above the region where the non-golden red-eye pixels are gathered.

Therefore, _{FR / B} (r) in equation (2) is also a curve along the upper side of the non-golden red-eye pixel distribution for the same reason as FR _{/ G} (r) in equation (1). This is a curve that separates the distribution of non-golden red-eye pixels and the distribution of golden-red-eye pixels.

Note that a curve F _{R / G} (r) that separates the distribution of non-red red-eye pixels and the distribution of red-red eye pixels and a curve F _{R / B} (r) that separates the distribution of non-gold red-eye pixels and the distribution of golden red-eye pixels are used. To obtain, the average value of the R value / G value for each R value (in this example, for each red component section Lr) represents the distribution of non-red red eye pixels, and the average value of the R value / B value is It is necessary to represent the distribution of golden red-eye pixels. That is, it is necessary that the correction processing target region Pe1 has an image size that includes a large number of pixels of an image other than the pupil.

  For this reason, in the feature amount extraction process (step S13 in FIG. 6) of the correction process target area Pe1, it is determined whether or not the correction process target area Pe1 is large enough to include many images other than the pupil. If it is determined (step S21 in FIG. 6) and it is determined that the size is not such, the correction processing target area Pe1 is enlarged (step S22). In this example, the correction processing target area Pe1 is enlarged so that the longitudinal width of the correction processing target area Pe1 is twice the diameter of the red-eye candidate area Pe2.

The separation curve generation unit 32 stores the obtained separation curves F _{R / G} (r) and F _{R / B} (r) in the internal memory 12.

In the case of this example, R = 0 / R = 225, and the average value of R / G values in association with the median value (16 points) of each red component section Lr is determined as a separation curve F _{R / G} (r). It is memorized as a control point, and the space between the control points is complemented based on the adjacent control points. Further, R = 0 / R = 225 and the average value of R / B values are stored as control points of the separation curve F _{R / B} (r) in association with the median value (16 points) of each red component section Lr. And between the control points, it complements based on the control points on both sides. By storing only the control points in this way, the storage capacity can be saved.

  When the separation curve is generated in this way, the process proceeds to step S15 in FIG. In step S <b> 15, the read control unit 22 reads image data for one line from the print processing target image TI and stores it in the line buffer 23.

  When the data for one line is stored in the line buffer 23, in step S16, the correction intensity map generating unit 33 of the red-eye area specifying information generating unit 25 stores one line stored in the line buffer 23 based on the separation curve. It is specified whether or not the pixel of the image data is an image constituting the red-eye area. In the case of this example, a correction intensity map Ma is generated as information indicating the specific result.

  In this example, each storage unit (that is, each pixel) of the correction intensity map Ma has a value of 128 or more when it can be estimated that the corresponding pixel in the correction processing target region Pe1 is a pixel constituting the red-eye region. Otherwise, a value smaller than 128 is stored as the correction intensity coefficient.

  FIG. 8 is a flowchart showing the flow of the correction intensity map generation process.

  In step S31, the correction intensity map generation unit 33 determines whether or not the image data stored in the line buffer 23 includes a pixel in the correction processing target area Pe1 (FIG. 4). If it is determined that the image data stored in the line buffer 23 includes a pixel in the correction processing target area Pe1, the correction intensity map generation unit 33 stores the correction data in the line buffer 23 in step S32. One pixel of interest is selected from the pixels in the correction processing target area Pe1.

  In step S33, the correction intensity map generation unit 33 acquires the R value, R value / G value, and R value / B value of the target pixel.

  In step S34, the correction intensity map generation unit 33 acquires a value of a separation curve corresponding to the R value of the target pixel (hereinafter referred to as a separation threshold). Specifically, an R value / G value (hereinafter referred to as a separation threshold R / G) as a separation threshold for the R value of the pixel of interest is obtained from Equation (1), and the R value of the pixel of interest is obtained from Equation (2). R value / B value (hereinafter referred to as separation threshold value R / B) is obtained as a separation threshold value.

  In step S35, the correction intensity map generation unit 33 determines a correction intensity coefficient corresponding to the separation threshold obtained in step S34. For example, referring to the correction intensity coefficient table shown in FIG. 9, the correction intensity coefficient corresponding to the ratio between the R value / G value of the target pixel and the separation threshold R / G, the R value / B value of the target pixel, and the separation threshold A correction intensity coefficient corresponding to the R / B ratio is obtained, and the larger correction intensity coefficient is set as the correction intensity coefficient of the target pixel. Then, the correction intensity map generation unit 33 stores the obtained correction intensity coefficient in the storage unit of the correction intensity map Ma corresponding to the target pixel.

A pixel in which the ratio of the R value / G value of the pixel of interest to the separation threshold R / G is 1 or more is on the separation curve FR _{/ G} on a plane having the R value and the R value / G value as an axis (FIG. 7). Or it exists above it. A pixel having a ratio of the R value / B value of the target pixel to the separation threshold R / B of 1 or more is on or above the separation curve FR _{/ B} on a plane having the R value and the R value / B value as an axis. Present above. Therefore, when the ratio between the R value / G value of the target pixel and the separation threshold R / G, or the ratio between the R value / B value of the target pixel and the separation threshold R / B is 1 or more, the target pixel is a red-eye pixel. Are estimated as pixels constituting the red-eye region, and a value of 128 or more is assigned to the pixel.

On the other hand, a pixel in which the ratio of the R value / G value of the pixel of interest and the separation threshold R / G is smaller than 1 is a separation curve FR _{/ G on} a plane having the R value and the R value / G value as an axis (FIG. 7). It exists below. A pixel having a ratio of the R value / B value of the pixel of interest and the separation threshold R / B smaller than 1 exists below the separation curve FR _{/ B} on a plane having the R value and the R value / B value as an axis. . Therefore, when the ratio between the R value / G value of the target pixel and the separation threshold R / G, and the ratio of the R value / B value of the target pixel and the separation threshold R / B are both smaller than 1, the target pixel is a non-red-eye Estimated as a pixel in the pixel distribution, that is, a pixel that does not constitute a red eye, a value less than 128 is assigned.

  Even in a pixel estimated to be a pixel that does not constitute red eye, the ratio of the R value / G value to the separation threshold R / G (or the ratio of the R value / B value to the separation threshold R / B) is 0.8 or more. In some cases (ie, for pixels distributed 20% below the separation curve), a certain magnitude of correction intensity factor is assigned. As a result, the color of the pixel is corrected by the correction process Z2 with the correction intensity corresponding to the correction intensity coefficient. This is to apply a certain amount of correction to a portion that is not affected by the red-eye phenomenon even if it is not a pupil.

  Next, in step S36, the correction intensity map generation unit 33 determines whether or not all the pixels in the correction processing target region Pe1 stored in the line buffer 23 have been selected as the target pixel. If there is a pixel that has not yet been selected as the target pixel, the correction intensity map generation unit 33 returns to step S32 and selects another pixel of the correction processing target area Pe1 stored in the line buffer 23 as the target pixel. The processes after step S33 are similarly executed.

  If it is determined in step S36 that all the pixels in the correction processing target area Pe1 stored in the line buffer 23 have been selected as the target pixel, the correction intensity map Ma for one line is generated. In the correction intensity map Ma, a value of 128 or more is stored for a pixel estimated to be a pixel constituting a red-eye region, and a value smaller than 128 is corrected for a pixel estimated to be a pixel not constituting a red-eye. Stored as an intensity factor.

  In this way, the correction intensity map generation unit 33 determines the correction intensity coefficient for each line of image data in raster order. The correction intensity map generation unit 33 has a memory area for storing correction intensity coefficients for two lines obtained by the correction intensity coefficient determination process as a work area, and there are adjacent lines (for example, , (N-1th line and nth line) correction intensity coefficients are stored for each line. When the correction intensity coefficient of the next line (for example, the (n + 1) th line) is obtained, the correction intensity coefficient of the previous line (for example, the (n-1) th line) is erased and obtained in the memory area. The corrected intensity coefficient of the (n + 1) th line is stored.

In step S37, vertical blurring is performed. This vertical blurring process combines the correction intensity coefficient of the currently obtained line (for example, the nth line) and the correction intensity coefficient of the previous line (for example, the (n-1) th line). Specifically, equation (3) is calculated. In Expression (3), Map _n (x) represents the xth correction intensity coefficient in the horizontal direction from the origin in the correction processing target area Pe1 of the line (nth line) obtained now. Map _n-1 (x) shows the xth correction intensity coefficient in the horizontal direction from the origin in the correction processing target region Pe1 of the previous line (the (n-1) th line). α is a blend ratio (synthesis ratio), for example, a coefficient of 1 or less. In this example, it is 0.75. Map _new is a correction strength coefficient obtained as a result of the vertical blurring.

When Equation (3) is calculated for all the correction intensity coefficients for one line that have been obtained, horizontal blurring processing is performed in step S38. This horizontal blurring process performs Gaussian blurring in the horizontal direction, and specifically, equation (4) is calculated. In Expression (4), P _i represents the correction intensity coefficient of interest in the correction intensity coefficient for one line that has been subjected to the vertical blurring process, and P _new represents the horizontal blurring. Is a corrected intensity coefficient obtained as a result. In this example, eight correction intensity coefficient is referenced Gaussian blur around the correction intensity coefficient P _i of interest aligned in the horizontal direction is performed. The coefficient r is a coefficient that determines the correction strength.

  As described above, when the horizontal blur process is performed in step S38, the correction intensity coefficient obtained as a result is stored in the correction intensity map Ma. Thereafter, the processing proceeds to step 17 in FIG. When it is determined in step S31 that the image data stored in the line buffer 23 does not include the pixel in the correction processing target area Pe1, the processing in steps S31 to S38 is skipped, and similarly, Proceed to step S17 of FIG. In step S17, output data generation processing is executed.

[Description of output data generation processing]
FIG. 10 is a flowchart showing the flow of output data generation processing. In step S41, the correction unit 26 determines whether or not the pixel of the red-eye candidate region Pe2 (FIG. 4) is stored in the line buffer 23, and if the pixel of the red-eye candidate region Pe2 is stored in the line buffer 23. If determined, in step S42, one pixel of interest is selected from the pixels of the red-eye candidate region Pe2 stored in the line buffer 23.

  In step S43, the correction unit 26 reads the correction intensity coefficient of the target pixel from the correction intensity map generation unit 33 (its correction intensity map Ma).

  In step S44, the correction unit 26 adjusts the color of the target pixel according to the read correction intensity coefficient. For example, a dark black color corresponding to the magnitude of the correction intensity coefficient is selected, and the color of the target pixel is replaced with the black color.

For example, according to Expression (5), the target pixel i of the red-eye candidate region Pe2 and the RGB of the pixels i-1 and i + 1 on both sides thereof are blended. Then, the gray color is obtained as the target color (Target Color in Expression (6)) using the blended RGB (R ′, G ′, B ′ in Expression (5)) according to Expression (6).

Further, according to the equation (7), the target color and the RGB of the pixel of interest are blended using the correction intensity coefficient of the pixel of interest (Intensity in equation (7)), and the color to be replaced (Rnew, Gnew, Bnew) is determined. Then, the color of the target pixel is replaced with the color obtained here.

  Next, in step S45, the correction unit 26 determines whether or not all the pixels in the red-eye candidate region Pe2 have been selected as the target pixel. If there is a pixel that has not yet been selected as the target pixel, the process returns to step S42, and the correction unit 26 performs the subsequent processing in the same manner.

  If it is determined in step S45 that all the pixels in the red-eye candidate region Pe2 have been selected as the target pixel, or in step S41, the red-eye candidate region Pe2 (FIG. 4) is included in the image data stored in the line buffer 23. ), The process proceeds to step S46. In step S <b> 46, the output control unit 27 generates print data from the image data for one line of the print processing target image TI stored in the line buffer 23. The generated print data is supplied to the printer engine 15, where an image is printed based on the print data.

  When the output data generation process is executed as described above, the process proceeds to step S18 in FIG. 5, and the read control unit 22 determines whether or not data for all lines has been read from the print process target image TI. However, if it is determined that there are still unread lines, the process returns to step S15, and the subsequent processes are performed in the same manner.

  If it is determined in step S18 that data for all lines (that is, all image data of the print processing target image TI) has been read, the printing process ends.

[Effect in the embodiment]

  1. As described above, since the correction intensity coefficient is set for each line of image data in the correction processing target area Pe1, the work area of the internal memory 12 is small. Even when only line image data is supplied, red-eye correction can be performed appropriately.

  2. In the above, since the vertical and horizontal blur correction is applied to the correction intensity coefficient of the generated one-line image data, for example, the red eye has a good appearance such as smoothing the edge portion of the red-eye region. It can be corrected.

[Modification]
1. In the above, we did not mention the relative difference between the degree of blurring in the vertical direction blurring process and the degree of blurring degree in the horizontal direction blurring process. The degree of blur in the vertical blur process can be made stronger. Note that the degree of blurring is confirmed by human vision, for example, the coefficient r in the horizontal blurring process so that the so-called blurring degree of the horizontal blurring process is stronger than the so-called blurring degree of the vertical blurring process. The value of the coefficient α in the vertical blurring process is determined.

  2. In the above description, the blend rate α used in the vertical blurring process is 0.75, but other values may be used. Also, blending is performed in accordance with the magnitude of the difference between the correction intensity coefficient of interest in the nth line and the correction intensity coefficient corresponding to the correction intensity coefficient of interest of the previous (n−1) th line. It is also possible to change the value of the rate α. For example, when the difference between the two is 10 or more, the blend rate can be changed to another value.

  3. In the above description, the reference range of the Gaussian blur processing as the horizontal blur processing is set to a predetermined value, but can be obtained from the size of the diameter of the red-eye candidate region Pe2.

  4). In the above description, the case where processing is performed for each line of image data has been described as an example. However, for example, processing may be performed for image data of a predetermined number of lines of two or more lines.

  5. In the above description, the case where the printer 1 is used has been described as an example. That is, it can be implemented in red-eye correction software or the like installed in the image processing apparatus. In this case, a function for generating the correction intensity map Ma is implemented in the image processing apparatus, and the correction intensity map Ma is provided from the image processing apparatus to the printer or the display side so that color correction is executed on the printer or the display side. You can also.

6). When the functions described above are realized by a computer or the like, a program describing the processing contents of the functions described above is provided. The processing function is realized on the computer by executing the program on the computer. The program describing the processing contents can be recorded on a computer-readable recording medium. Examples of the computer-readable recording medium include a magnetic recording device, an optical disk, a magneto-optical recording medium, and a semiconductor memory. Examples of the magnetic recording device include a hard disk device (HDD), a flexible disk (FD), and a magnetic tape. Optical discs include DVD (Digital Versatile Disc), DVD-RAM, CD-ROM (Compact
Disc ROM) and CD-R (Recordable) / RW (ReWritable). Magneto-optical recording media include MO (Magneto-Optical Disc).

  When distributing the program, for example, portable recording media such as a DVD and a CD-ROM in which the program is recorded are sold. It is also possible to store the program in a storage device of a server computer and transfer the program from the server computer to another computer via a network.

  The computer that executes the program stores, for example, the program recorded on the portable recording medium or the program transferred from the server computer in its own storage device. Then, the computer reads the program from its own storage device and executes processing according to the program. The computer can also read a program directly from a portable recording medium and execute processing according to the program. Further, each time the program is transferred from the server computer, the computer can sequentially execute processing according to the received program.

  In the above embodiment, a part of the configuration realized by hardware may be replaced with software, and conversely, a part of the configuration realized by software may be replaced by hardware. .

  7). The process shown in the flowchart is executed in parallel or individually even if the steps are not necessarily processed in time series, as well as processes performed in time series in the order described. It includes processing that. Further, the processing can be omitted as appropriate. For example, in FIG. 5, the process of generating the separation curve, the process of generating the correction intensity map Ma, and the process of generating the output data (including the correction process) are continuously performed. It can also be done automatically.

  DESCRIPTION OF SYMBOLS 1 Printer, 11 CPU, 12 Internal memory, 13 Operation part, 14 Display part, 15 Printer engine, 16 Card I / F, 17 Card slot, 18 Memory card, 21 Image reading part, 22 Reading control part, 23 Line buffer, 24 correction processing target region acquisition unit, 25 red-eye region specific information generation unit, 26 correction unit, 27 output control unit, 31 feature amount extraction unit, 32 separation curve generation unit, 33 correction intensity map generation unit

Claims (6)

In an image processing apparatus that performs processing for correcting red eyes for each color image data of a predetermined number of lines,
Acquisition means for acquiring an area including an eye image in the color image as a correction processing target area;
Generates a separation curve that separates the distribution of the pixels that make up the red-eye region from the distribution of other pixels in the distribution of the feature amounts in the correction target region in a space that has at least two color feature amounts as axes. Red-eye area specifying information generating means for
Based on the value of each pixel in the correction processing target area and the magnitude of the value corresponding to the separation curve, a determination process for determining a correction strength coefficient indicating the strength of correction for the pixel is performed in raster order. Correction intensity determination means for each image data of a predetermined number of lines in the correction processing target area,
The correction intensity determining means executes a vertical blur process for combining the correction intensity coefficient determined in the determination process and the correction intensity coefficient of the predetermined number of lines of image data determined in the previous determination process. An image processing apparatus that performs horizontal blurring processing that performs Gaussian blurring using a plurality of correction strength coefficients adjacent in the horizontal direction with respect to the correction intensity coefficient obtained as a result. The image processing apparatus according to claim 1.
The image processing apparatus according to claim 1, wherein the correction intensity determination unit makes the degree of blurring in the blurring process in the horizontal direction stronger than the degree of blurring in the blurring process in the vertical direction. The image processing apparatus according to claim 1.
The correction intensity determination means combines the predetermined ratio in the vertical blurring process, and changes the combination ratio according to the difference between the correction intensity coefficients adjacent in the vertical direction. An image processing apparatus. The image processing apparatus according to claim 1.
The image processing apparatus according to claim 1, wherein the correction intensity determination unit changes a reference range in the Gaussian blur process as the horizontal blur process according to a radius of a reddish area of the correction process target area. In an image processing method of an image processing apparatus that performs processing for correcting red eyes for each color image data of a predetermined number of lines,
An acquisition step of acquiring an area including an eye image in the color image as a correction processing target area;
Generates a separation curve that separates the distribution of the pixels that make up the red-eye region from the distribution of other pixels in the distribution of the feature amounts in the correction target region in a space that has at least two color feature amounts as axes. A red-eye region identification information generation step,
Based on the value of each pixel in the correction processing target area and the magnitude of the value corresponding to the separation curve, a determination process for determining a correction strength coefficient indicating the strength of correction for the pixel is performed in raster order. A correction intensity determination step performed for each image data of a predetermined number of lines in the correction processing target area,
The correction intensity determination step executes a vertical blurring process for combining the correction intensity coefficient determined in the determination process and the correction intensity coefficient of the predetermined number of lines of image data determined in the previous determination process. An image processing method characterized by executing a horizontal blurring process for performing Gaussian blurring using a plurality of correction intensity coefficients adjacent in the horizontal direction with respect to the correction intensity coefficient obtained as a result. A program for causing a computer to execute image processing,
In an image processing method of an image processing apparatus that performs processing for correcting red eyes for each color image data of a predetermined number of lines,
An acquisition step of acquiring an area including an eye image in the color image as a correction processing target area;
Generates a separation curve that separates the distribution of the pixels that make up the red-eye region from the distribution of other pixels in the distribution of the feature amounts in the correction target region in a space that has at least two color feature amounts as axes. A red-eye region identification information generation step,
Based on the value of each pixel in the correction processing target area and the magnitude of the value corresponding to the separation curve, a determination process for determining a correction strength coefficient indicating the strength of correction for the pixel is performed in raster order. A correction intensity determination step performed for each image data of a predetermined number of lines in the correction processing target area,
The correction intensity determination step executes a vertical blurring process for combining the correction intensity coefficient determined in the determination process and the correction intensity coefficient of the predetermined number of lines of image data determined in the previous determination process. A program for causing a computer to execute image processing for executing horizontal blurring processing for performing Gaussian blurring using a plurality of correction strength coefficients adjacent in the horizontal direction with respect to the correction intensity coefficient obtained as a result.

Images