JP2005267410A - Image processing method and device, and printer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To change a degree of color conversion for a color of a larger area and a color of a smaller area and to change only a color of hair into brown hair without changing the color of the eye in particular. <P>SOLUTION: A contrast signal Y is generated from R, G, B signals showing digital color images and a contrast signal YL of a blurred image is generated by putting the contrast signal Y through an IIR type low-pass filter. By the contrast signal YL of the blurred image, color correction data are read from a brown hair LUT 123 which stores color correction data differing among R, G, B. The color correction data for each R, G, B read from the brown hair LUT 123 are added to the R, G, B signals. Thus, R', G', B' signals which have changed only a color balance of a larger dark area (an area of hair) are obtained. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to an image processing method and apparatus, and a printer, and more particularly to a technique for changing the color of hair.

  Conventionally, it is often performed to convert a designated color area into another color in a digital image. For example, after a color area is designated using digital image editing software (such as Adobe Photoshop), the color of the designated area is often converted to a desired color.

In Patent Document 1, a histogram is created for each of R, G, and B from data of a specific portion of a digital image, a color conversion table is created based on the created histogram, and this color conversion table is used. A digital camera for adjusting the color of shooting data is described. The digital camera has a function of changing the color conversion table by an input operation while referring to a preview image.
JP-A-11-88906

  In recent years, the age of users using amusement sticker printing machines that synthesize and print photographed face photos and template images has been declining, and in the past, high school girls were the main users, but the number of junior high school girls has increased. Yes.

  Since junior high school students are prohibited from dyeing their hair according to school regulations, if they print with brown hair using an amusement sticker printing machine, they can fulfill their desire for female junior high school students.

  When converting the hair in the digital image to brown hair, use the conventional digital image editing software to specify the hair area, and when converting the color of the specified area to brown, specify the hair area accurately. There is a problem that it is difficult.

  In addition, when adjusting the color tone of an image using a color conversion table as described in Patent Document 1, if the color of black or black-brown hair is adjusted to brown, the color balance of other areas such as the face is adjusted. There is a problem of significant deviation. Furthermore, there is a technique for converting a specific color (black or black brown) with darkness into a brown color. In this case, however, there is a problem that the pupil color also becomes brown.

  The present invention has been made in view of such circumstances, and the degree of color conversion can be changed between the color of a large area region and the color of a small area region. It is an object of the present invention to provide an image processing method and apparatus, and a printer that can automatically change only the color of the hair area of a large area to the color of brown hair without changing.

  In order to achieve the above object, an image processing method according to claim 1 includes a step of generating a blurred image from a digital color image, and the color image based on brightness of each pixel of the generated blurred image. A step of generating color correction information for changing the degree of color conversion between a color of a large area and a color of a small area, and a step of correcting the color image based on the generated color correction information; It is characterized by including.

  That is, a blurred image is created from a digital color image. Using this blurred image, an image (pixel) of a large-area color region and an image (pixel) of a small-area color region can be separated. Since the brightness of a plurality of pixels is averaged in a blurred image, the brightness of the blurred image in a large-area color region has little change with respect to the brightness of the original image, whereas the brightness of a small-area color The brightness of the blurred image in the region varies greatly with respect to the brightness of the original image. Therefore, for each pixel of the blurred image, it is possible to determine whether the pixel is a pixel in a large area color region or a small area color pixel based on the brightness of the pixel. Then, color correction information for changing the degree of color conversion is generated between a pixel in a large area color region and a pixel in a small area color region in the color image, and the color image is generated based on the color correction information. The color of the corresponding pixel is corrected.

  The image processing method according to claim 1, wherein the color image is an image including a person having black or black-brown hair, and the large area includes a hair region. The small area region includes a pupil region. Thus, in the original color image, even if the color of the hair region is close to the color of the pupil region, it is possible to change only the hair color without changing the pupil color.

  In the image processing method according to claim 2, in the image processing method according to claim 2, in the step of generating the blurred image, the brightness of the blurred image with respect to the original image is higher in the pupil region than in the hair region. A feature is that a blurred image that is brighter is generated. Accordingly, the hair region and the pupil region can be distinguished from the brightness of the blurred image, and the pupil color can be prevented from being changed in the same manner as the hair color. That is, the color balance can be changed only in a dark area having a large area.

  The image processing method according to any one of claims 1 to 3, wherein the step of generating the blurred image includes a step of generating an image indicating lightness from the color image, and the lightness. And a step of filtering the image with a predetermined digital filter to generate a blurred image.

  First, an image showing lightness is created from a color image, and the image showing lightness is blurred by a predetermined digital filter such as an IIR (Infinite Impulse Response) filter or FIR (Finite Impulse Response) filter.

  5. The image processing method according to claim 1, wherein a step of extracting an image near an achromatic color including an achromatic color from the color image, and the image near the extracted achromatic color according to claim 5. And a step of weighting the color correction information so as to change the color of the image near the achromatic color.

  Thereby, it is possible to change only the color of an image of a large-area color area such as hair and an image near an achromatic color.

  An image processing apparatus according to claim 6, wherein a blurred image generation unit that generates a blurred image from a digital color image, and a large-area area of the color image based on the brightness of each pixel of the generated blurred image Color correction information generating means for generating color correction information for changing the degree of color conversion between the color of the image and the color of the small area, and color correction means for correcting the color image based on the generated color correction information It is characterized by having.

  As shown in claim 7, in the image processing apparatus according to claim 6, the color correction information generating means has a look-up table for storing color correction information corresponding to brightness information of input pixels in advance, Each time the brightness information of each pixel of the generated blurred image is input, the corresponding color correction information is read from the look-up table and output. Note that a plurality of lookup tables may be provided according to the contents of color conversion, and an appropriately selected lookup table may be used.

  The image processing apparatus according to claim 6 or 7, wherein the color correction information generating means generates color correction information for making black or black-brown hair brown. .

  A printer according to a ninth aspect includes the image processing apparatus according to any one of the sixth to eighth aspects, and a printing unit that prints an image processed by the image processing apparatus on photographic paper. Yes.

  According to a tenth aspect of the present invention, the printer according to the ninth aspect further includes a camera for photographing a person and acquiring a digital color image.

  According to the present invention, the blur image is generated from the color image to be processed, and the image of the large color area and the image of the small color area can be identified using the blur image. The degree of color conversion can be changed between the image of the large area color area and the image of the small area color area of the image, and the color of the small area pupil area can be changed without changing the color of the small area area. Only the color of the hair area can be automatically changed to the color of brown hair.

  Hereinafter, preferred embodiments of an image processing method and apparatus and a printer according to the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 1 is a block diagram showing an embodiment of a printer according to the present invention.

  As shown in the figure, the printer 100 of this embodiment is an amusement sticker printing machine installed at a game center or a store, and mainly includes a camera unit 110, an image processing unit 120, a display control unit 130, The display 140 includes a central processing unit (CPU) 150, a main memory 160, a storage 170, a print engine 180, and an operation unit 190.

  The camera unit 110 captures a user's face and outputs digital color image data (R, G, B signals). The image processing unit 120 is prepared in advance with a function for performing image processing according to the present invention (hereinafter referred to as “brown hair processing”) based on the R, G, and B signals output from the camera unit 110 and a user's face image. For example, a function for synthesizing a template image or graffiti, and a function for creating data for printing from the synthesized image data. Details of the brown hair treatment will be described later.

  The image processing unit 120 converts R, G, and B signals that have undergone various types of image processing into Y, M, and C (yellow, magenta, and cyan) signals, and converts the converted Y, M, and C signals to the print engine 180. Output to.

  The display control unit 130 creates image data for display for displaying the user's face image on the display 140 from the image data (image data indicating a moving image) input from the camera unit 110, and this is displayed on the display 140. An image is output and displayed, or a composite image obtained by combining the user's face image and a template image prepared in advance is displayed on the display 140.

  The CPU 150 is a processing unit that performs overall control of the entire system. By the CPU 150, a shooting instruction to the camera unit 110, reading of image data from the camera unit 110, processing instructions for the read image data, and each control Various instructions such as delivery of image data to the block and execution of printing are given. Various processing instructions to the CPU 150 are given from the operation unit 190 based on user operations. For example, the operation unit 190 is configured to select a template image to be used for image composition while viewing a template image displayed on the display device 140, a button for selecting a print type to be printed, and whether or not brown hair processing is performed. It has buttons for selecting and selecting the type of brown hair treatment. Note that the operation unit 190 may be realized by a soft button of a touch panel on the display 140.

  The main memory 160 stores a control program for the printer and is used as a work area when executing the program. The storage 170 is a non-volatile storage unit, and includes various template images and a look-up table (LUT) for brown hair processing. ) Etc. are stored.

  The print engine 180 employs, for example, a TA (Thermo Autochrome) system as a printing system, and heats color photographic paper (hereinafter referred to as “TA paper”) having C, M, and Y color layers. It is colored and fixed by irradiation with light of a predetermined wavelength, and has means for conveying TA paper, a thermal head, a fixing lamp, and the like. When printing a color image on TA paper, the TA paper is first transported and the thermal head is controlled by the Y signal to develop the yellow layer of the TA paper, and then the yellow color is fixed by the fixing lamp. Color development of the magenta layer and cyan layer of the TA paper is similarly performed based on the M signal and the C signal, thereby printing a color image on the TA paper. The printer of this embodiment is a TA printer. However, the present invention is not limited to this, and the present invention can also be applied to other types of printers such as other thermal printers and inkjet printers.

  Next, brown hair processing in the image processing unit 120 will be described.

  FIG. 2 is a block diagram showing an embodiment of the brown hair processing unit of the image processing unit 120.

  As shown in FIG. 2, the brown hair processing unit includes a light / dark signal generation unit 121, a blur image generation unit 122, a brown hair LUT 123, a multiplier 124, a saturation calculation unit 125, a weight calculation unit 126, and an adder 127. It consists of and.

The light / dark signal generator 121 generates a light / dark signal Y according to the following equation based on the input R, G, B signals.
[Equation 1]
Y = (5R + 6G + 5B) / 16 (1)
The light / dark signal Y is not limited to the equation (1), but may be obtained by an equation such as Y = 0.3 R + 0.6 G + 0.1 B, Y = (R + G + B) / 3.

  The blurred image generation unit 122 generates a light / dark signal YL indicating a blurred image based on the light / dark signal Y input from the light / dark signal generation unit 121.

  FIG. 3 is a flowchart of a processing procedure in the blurred image generation unit 122, and FIG. 4 is a diagram illustrating an image of each of the processing steps of the blurred image generation unit 122.

  The blurred image generation unit 122 receives the light / dark signal Y from the light / dark signal generation unit 121. An example of the image (1) of the light / dark signal Y is shown in FIG.

  The blurred image generation unit 122 generates a reduced image (2) by performing 1/2 thinning processing on the image (1) (step S10 in FIG. 3, FIG. 4B). As shown in FIG. 4B, the image size (number of pixels) of the reduced image (2) is ¼ of the number of pixels of the original image (1), so the subsequent processing reduces the computation time. can do.

  Subsequently, the blurred image generation unit 122 generates a blurred image (3) by applying an IIR low-pass filter to the reduced image (2) (step S12, FIG. 4C).

  FIG. 5 shows a model of an IIR low-pass filter. As shown in the figure, the IIR low-pass filter includes an adder 132, a one-pixel delay element 134, and multipliers 136 and 138. The multipliers 136 and 138 multiply the predetermined filter coefficients α and (1−α).

  The blurred image (3) is obtained by applying the IIR low-pass filter having the above configuration to the input image (2) four times in the vertical and horizontal directions. Further, the filter coefficient α is appropriately 31/32, 15/16, 7/8, or the like.

  Note that the filter for generating the blurred image (3) is not limited to the IIR low-pass filter, but may be, for example, an FIR low-pass filter. In short, any filter may be used as long as it digitally blurs the image.

  Next, a blur image (4) having the same number of pixels as the original image (1) obtained by magnifying the blur image (3) by a factor of two is generated (step S14, FIG. 4D). The enlargement process of the blurred image (3) is performed by interpolating the brightness data of adjacent pixels to generate a new pixel. Note that the reduction and enlargement processes in steps S10 and S14 are 1/2 thinning and x2 enlargement in FIG. 3, but not limited to this, 1/3 thinning, x3 enlargement, or 1/4 thinning, × 4 enlargement may be used.

  6A shows an example of the histogram of the light / dark signal Y of the image input to the blurred image generation unit 122, and FIG. 6B shows the histogram of the light / dark signal YL of the blurred image generated by the blurred image generation unit 122. An example is shown.

  As shown in the histogram of FIG. 6 (A), the brightness of the hair and the pupil in the light / dark signal Y of the unblurred image is close, but as shown in the histogram of FIG. 6 (B), in the light / dark signal YL of the blurred image. The difference between the brightness of the hair and the pupil widens. This is because the area of the hair has a large area, so the brightness of the area of the hair in the blurred image has little change, while the area of the pupil is small and the surrounding area is a white eye or bright skin color area. This is because the brightness of the pupil region in the image varies greatly (the brightness increases).

  Returning to FIG. 2, the brown hair LUT 123 includes three LUTs of R, G, and B colors, and each of the three LUTs stores color correction data corresponding to the magnitude of the light / dark signal YL. The three LUTs store different color correction data for correcting the R, G, B signals indicating black or black-brown hair into R ′, G ′, B ′ signals indicating brown hair. . The three LUTs constituting the brown hair LUT 123 read out the color correction data corresponding to the light / dark signal YL based on the light / dark signal YL of the blurred image generated by the blurred image generating unit 122, and each R, G, B signal Are output to the multiplier 126. Details of the brown hair LUT 123 will be described later.

Chroma calculation unit 125, R to enter, G, by equation (2) based on the B signal, the chrominance signal C _b, to generate a C _r, the color difference signal C _b, wherein based on the C _r (3) To calculate the saturation C.
[Equation 2]
C _b = B−Y, C _r = R−Y (2)
(However, Y is a light / dark signal)
[Equation 3]
C = √ (C _b ² + C _r ² ) (3)
The weight calculation unit 126 calculates a weight W _gray corresponding to the magnitude of the saturation C input from the saturation calculation unit 125.

That is, when the saturation C is calculated by the saturation calculation unit 125 as shown in FIG. 7A, the weight calculation unit 126 calculates the saturation C as shown in FIG. The corresponding weight W _gray is output.

Weight calculating unit 126 as shown in FIG. 7 (B), when the range chroma C is 0 to C ₁ outputs the weight W _gray = 1, when the chroma C is a C ₂ or higher The weight W _gray = 0 is output, and when the saturation C is in the range of C _{1 to} C ₂ , the weight W _gray that continuously changes from 1 to 0 according to the magnitude of the saturation C is output. Note that the saturation C in the range of 0 to C ₁ is a range that is recognized as an almost achromatic color, C _{2 and} higher are relatively large chromatic colors, and the saturation in the range of C _{1 to} C ₂ is an achromatic color. The saturation is between chromatic and chromatic.

The weight W _gray calculated by the weight calculation unit 126 in this way is output to the multiplier 124.

The multiplier 124 multiplies the color correction data input from the brown hair LUT 123 by the weight W _gray input from the weight calculation unit 126. The color correction data obtained by this multiplication is data for correcting only the color of pixels near the achromatic color. This is because only the color of the pixel corresponding to the hair of the achromatic color or near the achromatic color is corrected.

  The multiplication result of the multiplier 124 is output to the adder 127. R, G, B signals are added to the other inputs of the adder 127, and the adder 127 adds the two and outputs R ', G', B 'signals after color correction (brown hair processing). To do.

  FIG. 8 is a main block diagram of the brown hair processing unit shown in FIG.

  As shown in FIG. 8, the brown hair LUT 123 includes three LUT_R, LUT_G, and LUT_B, and a light / dark signal YL of a blurred image is added to each LUT_R, LUT_G, and LUT_B.

  As shown in FIGS. 9A, 9B, and 9C, the three LUT_R, LUT_G, and LUT_B store color correction data corresponding to the light / dark signal YL, and correspond to the input light / dark signal YL. Read the color correction data. As shown in FIG. 9A, LUT_R reads larger color correction data as the light / dark signal YL becomes smaller when the light / dark signal YL becomes smaller than the predetermined light / dark signal YL ′. Here, the light / dark signal YL smaller than the predetermined light / dark signal YL ′ corresponds to the light / dark signal of the blurred image of the hair region, and color correction data is output when the light / dark signal YL is input. In this embodiment, LUT_G and LUT_B store 0 color correction data over all the light / dark signals YL as shown in FIGS.

The color correction data read from the LUT_R, LUT_G, and LUT_B are output to the multipliers 124R, 124G, and 124B, respectively. The other inputs of the multipliers 124R, 124G, and 124B are added with a weight W _gray that gives a large weight to the vicinity of the achromatic color, and the multipliers 124R, 124G, and 124B respectively multiply the two inputs that are input. The multiplication results are output to the adders 127R, 127G, and 127B.

  The other inputs of the adders 127R, 127G, and 127B are added with R, G, and B signals, respectively. The adders 127R, 127G, and 127B are weighted with the R, G, and B signals, respectively. The color correction data is added and R ′, G ′, B ′ signals indicating the addition result are output.

  Here, since the color correction data is added to the R signal when the light / dark signal YL of the blurred image of the hair region is obtained as shown in FIG. 9A, the R ′ signal after the addition is added. Will grow. On the other hand, since the color correction data is not added to the G and B signals as shown in FIGS. 9B and 9C, the G and B signals do not change (G ′ = G, B ′ = B). . As a result, the color balance of the R, G, and B signals changes in the R direction, and black or black-brown hair becomes brown hair.

  In the example of the LUT shown in FIG. 9, only LUT_R has color correction data. However, different color correction data is stored in each of the three LUT_R, LUT_G, and LUT_B so that desired color correction can be performed. May be. Alternatively, a plurality of sets of three LUT_R, LUT_G, and LUT_B may be prepared so that various brown hair processing such as brown hair processing 1 and brown hair processing 2 can be realized.

  In this embodiment, the light / dark signal YL of the blurred image is generated from the light / dark signal Y, and the light / dark signal YL is used as a common input signal for the three LUT_R, LUT_G, and LUT_B. , G, and B may be generated for each of the light and dark signals RL, GL, and BL of the blurred image, and these light and dark signals RL, GL, and BL may be used as input signals for three LUT_R, LUT_G, and LUT_B, respectively.

FIG. 1 is a block diagram showing an embodiment of a printer according to the present invention. FIG. 2 is a block diagram showing an embodiment of the brown hair processing unit of the image processing unit shown in FIG. FIG. 3 is a flowchart showing a processing procedure in the blurred image generation unit shown in FIG. FIGS. 4A to 4D are diagrams showing images of the images at each processing stage of the blurred image generation unit. FIG. 5 is a diagram showing a model of the IIR low-pass filter. 6A and 6B are histograms of the brightness signal of the image to be processed and the brightness signal of the blurred image, respectively. FIGS. 7A and 7B are diagrams used to explain saturation calculation and weights corresponding to the saturation. FIG. 8 is a main block diagram of the brown hair processing unit shown in FIG. FIGS. 9A to 9C are diagrams used for explaining the color correction data of the brown hair LUT.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Printer, 110 ... Camera part, 120 ... Image processing part, 121 ... Light / dark signal generation part, 122 ... Blurred image generation part, 123 ... Brown hair LUT, 124, 124R, 124G, 124B ... Multiplier, 125 ... Saturation calculation 126, weight calculation unit, 127, 127R, 127G, 127B ... adder, 130 ... display control unit, 140 ... display unit, 150 ... central processing unit (CPU), 160 ... main memory, 170 ... storage, 180 ... Print engine, 190 ... operation unit

Claims (10)

Generating a blurred image from a digital color image;
Generating color correction information for changing the degree of color conversion between a color of a large area and a color of a small area of the color image based on the brightness of each pixel of the generated blurred image When,
Correcting the color image based on the generated color correction information;
An image processing method comprising:   The color image is an image including a person having black or black-brown hair, the large area region includes a hair region, and the small area region includes a pupil region. The image processing method according to claim 1.   The step of generating the blurred image generates a blurred image in which the brightness of the blurred image with respect to the original image is brighter in the pupil region than in the hair region. An image processing method described in 1.   The step of generating the blur image includes a step of generating an image indicating brightness from the color image, and a step of generating a blur image by filtering the image indicating brightness with a predetermined digital filter. The image processing method according to claim 1.   A step of extracting an image near the achromatic color including an achromatic color from the color image, and weighting the color correction information to change a color of the image near the achromatic color based on the extracted image near the achromatic color. The image processing method according to claim 1, further comprising a step. A blurred image generating means for generating a blurred image from a digital color image;
A color for generating color correction information for changing the degree of color conversion between a color of a large area and a color of a small area of the color image based on the brightness of each pixel of the generated blurred image Correction information generating means;
Color correction means for correcting the color image based on the generated color correction information;
An image processing apparatus comprising:   The color correction information generation means has a lookup table that stores color correction information in advance corresponding to the brightness information of the input pixel, and each time the brightness information of each pixel of the generated blurred image is input, The image processing apparatus according to claim 6, wherein the corresponding color correction information is read out from the lookup table and output.   The image processing apparatus according to claim 6, wherein the color correction information generation unit generates color correction information for making black or black-brown hair brown.   9. A printer comprising: the image processing apparatus according to claim 6; and a printing unit that prints an image processed by the image processing apparatus on a photographic paper. The printer according to claim 9, further comprising a camera that captures a digital color image by photographing a person.

Images