JP2006013751A - Image processing apparatus, image processing method, and its program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing apparatus capable of magnifying a color image at a high speed with high image quality. <P>SOLUTION: The image processing apparatus 2 generates a composite image on the basis of each color component image of a received color image, respectively magnifies the produced composite image (sample image) by a high speed magnification algorithm and a high image quality magnification algorithm, and calculates a difference between the magnified sample images by the respective algorithms. The image processing apparatus 2 distributes the calculated difference in response to a gradation change amount (edge amount) of each color component image of the received color image and uses the distributed difference value (correction amount) to correct each color component image of the received color image magnified by the high speed magnification algorithm thereby obtaining a color magnification image with high image quality. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an image processing apparatus that performs enlargement processing of a color image.

For example, Patent Document 1 discloses an image processing apparatus that uses fine pixel density conversion means for L information and uses nearest neighbor replacement processing for C1 and C2 information. Patent Document 2 discloses an image signal processing apparatus that performs cubic interpolation on a luminance signal and linear interpolation on a chroma signal in the case of 4: 2: 2 format. Patent Document 3 discloses a digital imaging apparatus that performs an interpolation calculation using a Lagrange interpolation formula using nine pixels for Y data and a linear interpolation formula using four pixels for UV data.
Patent Document 4 discloses a color image processing apparatus that performs curved surface interpolation processing for the Y plane, performs linear interpolation processing for the U plane, and performs nearest neighbor processing for the V plane. Patent Document 5 discloses an image processing apparatus that can change a predetermined condition for detecting a similar image block for each color component and luminance component. Further, Patent Document 6 discloses an image data thinning / interpolation apparatus that interpolates luminance components by two or more interpolation methods and interpolates two color difference components by a method below the type of luminance component interpolation method.
Japanese Patent Laid-Open No. 5-120420 JP-A-10-191392 JP-A-11-250239 JP 2000-182022 A JP 2002-163648 A JP 2002-300377 A

  The present invention has been made from the above-described background, and an object thereof is to provide an image processing apparatus that enlarges a color image at high speed and high image quality.

[Image processing device]
In order to achieve the above object, an image processing apparatus according to the present invention is an image processing apparatus that enlarges a color image composed of a plurality of color component images, and applies the first enlargement process to obtain an image. Applying the first enlargement means for enlargement, the second enlargement means for enlarging the image by applying the second enlargement process capable of the enlargement process faster than the first enlargement process, and the input color image A sample creation unit that creates a sample image using at least one color component image, and an image enlarged by the first enlargement unit and the second enlargement for the sample image created by the sample creation unit Difference calculation means for calculating a difference from the image enlarged by the means, each color component image of the color image enlarged by the second enlargement means, and the difference calculated by the difference calculation means When, with on the basis of the gradation change amount of each color component image, and a enlarged image generating means for generating an enlarged image of the input color image.

  Preferably, the image processing apparatus further includes difference distribution means for dividing the difference calculated by the difference calculation means into distribution values according to the gradation change amount of each color component image and distributing the distribution values to the respective color component images. The image generation means is an enlarged image of the input color image based on each of the color component images enlarged by the second enlargement means and the distribution values distributed to the respective color component images by the difference distribution means. Is generated.

  Preferably, the input color image includes RGB color component images, and the sample creation means creates a sample image using the RGB color component images.

  Preferably, when a desired enlarged image is obtained by repeating the enlargement process a plurality of times, for each enlargement process, a color image enlargement process using only the first enlargement unit and a difference calculated for the sample image And a control means for switching and applying the color image enlargement processing based on the above.

  Preferably, the first enlargement process generates an enlarged image with higher image quality than the second enlargement process.

[Image processing method]
An image processing method according to the present invention is an image processing method for enlarging a color image composed of a plurality of color component images, and creates a sample image using at least one color component image, Applying the enlargement process to enlarge the created sample image, applying a second enlargement process capable of an enlargement process faster than the first enlargement process, enlarging the sample image, and The difference between the sample image enlarged by the first enlargement process and the sample image enlarged by the second enlargement process is calculated, and the calculated difference is set for each color according to the gradation change amount of each color component image. An enlarged image of the input color image is generated based on each of the color component images distributed to the component images and enlarged by the second enlargement processing and the difference distributed to each color component image.

  Preferably, when the enlargement process is repeated a plurality of times to obtain a desired enlarged image, at least in the first enlargement process, only the first enlargement process is applied to enlarge the color image, and at least in the last enlargement process The color image is enlarged by applying an enlargement process based on the difference calculated for the sample image.

[program]
The program according to the present invention includes a step of creating a sample image using at least one color component image in an image processing apparatus that enlarges a color image composed of a plurality of color component images, and a first enlargement process. Applying the step of enlarging the created sample image, applying a second enlargement process capable of an enlargement process faster than the first enlargement process, and enlarging the sample image; Calculating a difference between the sample image enlarged by the first enlargement process and the sample image enlarged by the second enlargement process, and the difference calculated according to the gradation change amount of each color component image Are distributed to each color component image, each of the color component images enlarged by the second enlargement process, and the difference distributed to each color component image. There are, and a step of generating an enlarged image of the input color image to the image processing apparatus.

  According to the image processing apparatus of the present invention, a color image can be enlarged at high speed and with high image quality.

[First Embodiment]
First, in order to help understanding of the present invention, its background and outline will be described.
When an image enlargement (that is, resolution conversion) is performed and a high-quality enlargement algorithm is applied, a high-quality enlarged image can be obtained, but the load of enlargement processing is large and the processing speed is reduced. For example, the image enlargement algorithm includes a nearest neighbor method (Nearest Neighbor method), a linear interpolation method (Bilinear method), a cubic convolution method (Cubic Convolution method), and a fractal enlargement method based on the concept of fractals. The nearest neighbor method is a high-speed enlargement algorithm capable of high-speed processing, but the image quality of the enlarged image is not so good. Further, the fractal enlargement method can obtain a high-quality enlarged image, but the enlargement process takes time. Under such circumstances, an image enlargement algorithm that can obtain a high-quality enlarged image at high speed is desired. In particular, when the object of enlargement processing is a color image composed of a plurality of color component images, each of the plurality of color component images needs to be enlarged, so that the enlargement processing is accelerated (that is, the processing load is reduced). Becomes more important.

FIG. 1 is a diagram for explaining an image enlargement algorithm that can obtain a high-quality enlarged image at high speed. FIG. 1A shows the entire image obtained by applying the high-quality enlargement algorithm to only some color component images. FIG. 1B shows a method for enlarging a color image based on a difference between sample images enlarged by a high-speed enlargement algorithm and a high-quality enlargement algorithm (this book). Corresponding to the invention).
As shown in FIG. 1A, when a color image is composed of a Y component image, a U component image, and a V component image in the YUV color space, only a high-quality enlargement algorithm (for example, fractal) is used for the Y component image. An enlargement method is applied, and a fast enlargement algorithm (for example, nearest neighbor method) is applied to other U component images and V component images to improve the overall processing speed.
In this case, the Y component image becomes a high-quality enlarged image, and the U component image and the V component image become low-quality enlarged images. Therefore, when viewed as a color image, the image quality is high in a region where the gradation change of the color difference component (U component and V component) is small, but the gradation change of the luminance component (Y component) is small and the color difference component. In the region where the gradation change is large (the edge portion of the U component image or the V component image), the image quality deterioration in the U component image and the V component image becomes obvious. This is due to the property that image quality degradation due to the high-speed enlargement algorithm is likely to occur at the edge portion. In other words, the difference between the enlarged image enlarged by the high-speed enlargement algorithm and the enlarged image enlarged by the high-quality enlargement algorithm is likely to occur at the edge portion and is likely to be manifested.

Therefore, the image processing apparatus 2 according to the present embodiment realizes high-speed and high-quality enlargement processing of a color image by correcting image quality deterioration due to a high-speed enlargement algorithm according to the gradation change amount of each color component image. .
For example, as illustrated in FIG. 1B, the image processing apparatus 2 creates one synthesized image based on each color component image of the input color image, and the created synthesized image (sample image) is processed at high speed. Each of the enlargement algorithm and the high image quality enlargement algorithm is enlarged, and a difference value between the enlarged sample images is calculated. Then, the image processing apparatus 2 distributes the calculated difference value as a correction amount according to the gradation change amount (edge amount) of each color component image of the input color image, and uses the distributed difference value. Then, each color component image of the input color image enlarged by the high-speed enlargement algorithm is corrected to obtain a high-quality color enlarged image.
Thereby, it is possible to speed up the enlargement process as compared with the case where the high-quality enlargement algorithm is applied to all the color component images, and it is possible to realize sufficient image quality.

[Hardware configuration]
Next, the hardware configuration of the image processing apparatus 2 will be described.
FIG. 2 is a diagram illustrating a hardware configuration of the image processing apparatus 2 to which the image processing method according to the present invention is applied, centering on the control apparatus 20.
As illustrated in FIG. 2, the image processing apparatus 2 includes a control device 20 including a CPU 202 and a memory 204, a communication device 22, a recording device 24 such as an HDD / CD device, an LCD display device or a CRT display device, and a keyboard. A user interface device (UI device) 26 including a touch panel and the like is included.
The image processing device 2 is, for example, a general-purpose computer in which an image enlargement program 5 (described later) is installed, acquires color image image data via the communication device 22 or the recording device 24, and the acquired image data is acquired. Enlarge according to the output resolution. For example, the image processing apparatus 2 converts the resolution to 600 dpi or 2400 dpi when outputting image data to the printer apparatus 10, and 75 dpi or the like when outputting image data to the UI apparatus 26. Convert to the resolution.

[Image enlargement program]
FIG. 3 is a diagram illustrating a functional configuration of the first image enlargement program 5 which is executed by the control device 20 (FIG. 2) and implements the image processing method according to the present invention.
As illustrated in FIG. 3, the first image enlargement program 5 includes a storage unit 500, a sample creation unit 510, a high-speed enlargement unit 520, a high-quality enlargement unit 530, a difference calculation unit 540, a gradation change determination unit 550, a difference A distribution unit 560, an enlarged image generation unit 570, and a control unit 580 are included.
In the image enlargement program 5, the storage unit 500 controls the memory 204 (FIG. 2) and the recording device 24 (FIG. 2), and stores the input color image image data. The storage unit 500 includes a sample creation unit 510, a high-speed enlargement unit 520, a high-quality enlargement unit 530, a difference calculation unit 540, a gradation change determination unit 550, a difference distribution unit 560, an enlarged image generation unit 570, and a control unit 580. Secure a work area.

The sample creation unit 510 creates a sample image based on the color component image included in the input color image. For example, the sample creation unit 510 creates a sample image using at least one of a plurality of color component images. The sample creation unit 510 in the present embodiment synthesizes all the color component images constituting the color image at a predetermined synthesis ratio, and creates one synthesized image as a sample image. When there are three color component images such as RGB or CMY, the composition ratio of the color component images may be set to one third, or “0.5”, “0.25”, and “0.25”. The ratio may be such that a color component image for synthesis can be calculated only by bit shift. Hereinafter, a case where the color component of the input color image is RGB will be described as a specific example. The synthesis ratio of R, G and B is expressed as pR, pG and pB, respectively.
Note that the sample creation unit 510 may change the composition ratio for each image region. For example, the sample creation unit 510 preferentially suppresses image quality degradation of important color components in each object by changing the composition ratio in accordance with the type of object displayed in each image area.

  The high-speed enlargement unit 520 (second enlargement unit) and the high-quality enlargement unit 530 (first enlargement unit) both enlarge the image. The high-speed enlargement unit 520 can enlarge the image faster than the high-quality enlargement unit 530, and applies, for example, the nearest neighbor method. The high image quality enlargement unit 530 can enlarge the image with higher image quality than the high speed enlargement unit 520, and applies, for example, a fractal enlargement method. Note that it is desirable that the high-quality enlargement unit 530 applies an enlargement algorithm having higher edge reproducibility than the high-speed enlargement unit 520.

  The difference calculation unit 540 calculates a difference between the sample image enlarged by the high-quality enlargement algorithm and the sample image enlarged by the high-speed enlargement algorithm. Specifically, the difference calculation unit 540 compares the sample image enlarged by the high-speed enlargement unit 520 and the sample image enlarged by the high image quality enlargement unit 530 for each pixel, and calculates a difference value dif for each pixel. calculate.

The gradation change determination unit 550 determines a gradation change amount for each color component image included in the input color image. Specifically, the gradation change determination unit 550 calculates the amount of change in the pixel value by comparing the target pixel and the pixels around the target pixel in each color component image.
For example, the tone change determination unit 550 calculates the tone change amount dif_R of the R component for each pixel by the following equation.
dif_R = | fr (x, y) -fr (x + 1, y) | + | fr (x, y) -fr (x, y + 1) | + | fr (x, y) -fr (x- 1, y) | + | fr (x, y) -fr (x, y-1) |
Here, fr (x, y) is the pixel value of the target pixel.
Similarly, the gradation change determination unit 550 calculates the gradation change amount dif_G of the G component image and the gradation change amount dif_B of the B component image for each pixel.
In this example, the gradation change amount is calculated by comparing the pixel of interest with four pixels that are vertically and horizontally adjacent. However, the present invention is not limited to this. For example, the pixel of interest and the surrounding 8 pixels are calculated. The gradation change amount may be calculated by comparing two pixels.

The difference distribution unit 560 distributes the difference value dif calculated by the difference calculation unit 540 to each color component image according to the gradation change amount of each color component image. Specifically, the difference distribution unit 560 compares the gradation change amount of each color component image for each pixel, and uses the difference value calculated for each pixel by the difference calculation unit 540 as a ratio of the gradation change amount (RGB Distribution for each pixel in accordance with the ratio between them. More specifically, the difference distribution unit 560 distributes the difference value dif to each color component image according to the ratio of the gradation change amount and the synthesis ratios pR, pG, and pB in the sample image.
For example, the difference distribution unit 560 calculates a distribution value vR distributed to the R component image for each pixel by the following equation.
vR = {(dif) * dif_R} / {pR * dif_R + pG * dif_G + pB * dif_B}
Similarly, the difference distribution unit 560 calculates the distribution value vG of the G component image and the distribution value vB of the B component image for each pixel.

The enlarged image generation unit 570 generates enlarged color image data based on each color component image enlarged by the high speed enlargement unit 520 and the difference value distributed to each color component by the difference distribution unit 560. Specifically, the enlarged image generation unit 570 adds or subtracts the distribution value calculated by the difference distribution unit 560 with respect to the R component image, the G component image, and the B component image enlarged by the high speed enlargement unit 520. Then, the pixel value of each color component image is corrected to obtain a color component image constituting a high-quality color image.
The control unit 580 controls each component included in the image enlargement program 5 so that the input color image is enlarged according to the output resolution.

[Overall operation]
Next, the first image enlargement process (S10) by the image processing apparatus 2 will be described.
FIG. 4 is a flowchart showing the overall operation of the first image enlargement process (S10). In this example, a case where a color image composed of an R component image, a G component image, and a B component image is input is taken as a specific example. Further, in this example, a specific example is a case where the high-speed enlarging unit 520 applies the cubic convolution method and the high-quality enlarging unit 530 applies the fractal enlargement method.
As shown in FIG. 4, in step 100 (S100), the image processing apparatus 2 receives image data of a color image (RGB) to be output, and when an output destination is specified, the image processing apparatus 2 responds to the output resolution of the output destination. Then, the enlargement magnification is determined, and the image data is input to the image enlargement program 5 (FIG. 3) to enlarge to the determined enlargement magnification.
In the image enlargement program 5, each color component image (R component image, G component image, and B component image) of the input color image is stored in the storage unit 500.
The high-speed enlargement unit 530 applies a cubic convolution method to enlarge each color component image of the input color image.

  In step 105 (S105), the sample creation unit 510 creates an RGB composite image by synthesizing the input color component image (before enlargement) with a predetermined composition ratio (1/3). Specifically, for each pixel, the sample creation unit 510 multiplies the pixel value of each color component by the composition ratio.

In step 110 (S110), the high image quality enlargement unit 530 applies the fractal enlargement method to enlarge the RGB composite image created by the sample creation unit 510 to the output resolution.
In step 115 (S115), the high-speed enlargement unit 520 applies the cubic convolution method to enlarge the RGB composite image created by the sample creation unit 510 to the output resolution.

  In step 120 (S120), the difference calculation unit 540 calculates the difference value between the RGB composite image enlarged by the high-quality enlargement unit 530 and the RGB composite image enlarged by the high-speed enlargement unit 520 for each corresponding pixel. calculate.

In step 125 (S125), the gradation change determination unit 550 calculates a gradation change amount for each pixel for each color component image of the input color image.
In step 130 (S130), the difference distribution unit 560 is calculated by the difference calculation unit 540 according to the gradation change amount ratio calculated by the gradation change determination unit 550 and the composition ratio in the RGB composite image. The difference values are distributed, and the distribution values of the R component image, the G component image, and the B component image are determined for each pixel.

In step 135 (S135), the enlarged image generation unit 570 determines the distribution values determined by the difference distribution value 560 for the R component image, the G component image, and the B component image respectively enlarged by the high speed enlargement unit 520. Add or subtract.
The image processing apparatus 2 transfers the R component image, the G component image, and the B component image obtained by adding / subtracting the distribution value by the enlarged image generation unit 570 to the output destination, and outputs the enlarged color image.

  Note that the high-speed enlarging unit 520 and the high-quality enlarging unit 530 of the present example apply the third-order convolution method and the fractal enlarging method, but are not limited to this, for example, the nearest neighbor method and the fractal enlarging method A combination or a combination of a linear interpolation method and a fractal expansion method may be used. In particular, when a high-quality enlarged image is required, it is desirable that the high-quality enlargement unit 530 apply a local collage enlargement method disclosed in Japanese Patent Application Laid-Open No. 2003-283811. This enlargement method by local collage (Japanese Patent Application Laid-Open No. 2003-283811) is a general fractal enlargement method for comprehensively solving image quality problems such as blur, jaggy and block distortion caused by application of the fractal enlargement method. An enlarged image with higher image quality can be obtained.

  Further, the image processing apparatus 2 of the present example enlarged the color image composed of the R component image, the G component image, and the B component image, but the color space of the image to be enlarged is not limited to this, for example, , YUV color space, Lab color space, or CMYK color space.

  As described above, the image processing apparatus 2 in the present embodiment determines the gradation change amount for each color component, and displays the color component image enlarged by the high-speed enlargement algorithm according to the determined gradation change amount. Since the correction is performed, the color image can be enlarged at a high speed while suppressing the image quality degradation that easily occurs at the edge portion.

[Modification]
Next, an image enlargement method that does not require the high-speed enlargement unit 520 will be described as a modification of the above-described embodiment.
FIG. 5 is a diagram for explaining an image enlargement method by the nearest neighbor method.
As shown in FIG. 5, the pixel value Pout of the output image based on the nearest neighbor method is the pixel value on the input image closest to the point Pr obtained by inversely mapping the center point of Pout on the input image. From the pixel value Pin and the horizontal / vertical magnification (rx, ry).
Pout (X, Y) = Pin (floor ((X + 0.5) / rx), floor ((Y + 0.5) / ry))
The floor (a) is a function for obtaining an integer value obtained by rounding down the decimals of a.
When processing is performed as in this equation, an output image is generated by performing nearest neighbor enlargement processing on the input image at a magnification (rx, ry).
In this way, the nearest neighbor method applies the pixel value of the pixel closest to the back-projected point as the pixel value after enlargement when the pixel of interest after enlargement is back-projected on the original image. is there. Therefore, the pixel value of the image enlarged by the nearest neighbor method can be calculated based on the position of the enlarged pixel and the original image without actually enlarging the image data. This makes it possible to calculate the pixel value of the composite image enlarged by the nearest neighbor method and the pixel value of each color component image enlarged by the nearest neighbor method without developing the entire image enlarged by the nearest neighbor method in the memory. it can.
Therefore, as a modified example of the above-described embodiment, an image processing apparatus 2 in which the high-speed enlargement unit 520 is replaced with a reverse projection calculation unit 590 will be described.

FIG. 6 is a diagram for explaining a functional configuration of the second image enlargement program 52. It should be noted that among the components shown in the figure, the same reference numerals are given to the components that are substantially the same as those shown in FIG.
As shown in FIG. 6, the second image enlargement program 52 has a configuration in which the high-speed enlargement unit 520 is replaced with a reverse projection calculation unit 590 among the components of the first image enlargement program 5.
The reverse projection calculation unit 590 calculates a pixel value when the original image is enlarged by the nearest neighbor method based on the position of the enlarged pixel and the original image. Specifically, the reverse projection calculation unit 590 calculates the coordinates of a reverse projection point obtained by reversely projecting the position of the enlarged pixel according to the magnification, and the pixel on the original image closest to the calculated reverse projection point. (Hereinafter referred to as reverse projection pixel) is specified, and the pixel value of the specified reverse projection pixel is read out. In this case, the difference calculation unit 540 compares the pixel value of the reverse projection pixel read by the reverse projection calculation unit 590 with the pixel value of the image enlarged by the high image quality enlargement unit 530, and calculates the difference value. calculate. Further, the enlarged image generation unit 570 calculates the pixel value of each color component image based on the pixel value of the reverse projection pixel read by the reverse projection calculation unit 590 and the distribution value distributed by the difference distribution unit 560. calculate.

FIG. 7 is a flowchart for explaining the second image enlargement process (S12). Note that among the processing steps shown in this figure, the same reference numerals are assigned to the processing steps that are substantially the same as those shown in FIG. In this example, a specific example is a case where a color image composed of an R component image, a G component image, and a B component image is input.
As shown in FIG. 7, first, in S105 and S110, when image data of a color image (RGB) to be output is input to the image processing apparatus 2, the sample creation unit 510 displays the input color image (before enlargement). ) Color component images are mixed at a predetermined composition ratio (1/3) to create an RGB composite image, and the high-quality enlargement unit 530 applies a high-quality enlargement algorithm (for example, a fractal enlargement method), The RGB composite image created by the sample creation unit 510 is enlarged to the output resolution.

  In step 140 (S140), the reverse projection calculation unit 590 performs reverse projection on the position of the pixel after enlargement according to the enlargement magnification (output resolution) to calculate the coordinates of the reverse projection point, and the calculated reverse projection point. The reverse projection pixel on the RGB composite image closest to is specified, and the pixel value of the specified reverse projection pixel is read out from the RGB composite image.

  In step 145 (S145), the difference calculation unit 540 calculates the difference between the RGB composite image enlarged by the high image quality enlargement unit 530 and the pixel value (pixel value of the reverse projection pixel) read by the reverse projection calculation unit 590. A value is calculated for each pixel.

  Next, in S125 and S130, the gradation change determination unit 550 calculates a gradation change amount for each pixel for each color component image of the input color image, and the difference distribution unit 560 determines the gradation change determination unit 550. The difference value calculated by the difference calculation unit 540 is distributed according to the ratio of the gradation change amount calculated by the above and the composition ratio in the RGB composite image, and the R component image, the G component image, and the B component image are distributed. A value is determined for each pixel.

In step 150 (S150), the inverse projection calculation unit 590 calculates the coordinates of the inverse projection point by inversely projecting the position of the enlarged pixel according to the enlargement magnification (output resolution), and the calculated inverse projection point. Are identified in the pre-enlargement image, and the pixel values of the identified reverse projection pixels are read out from the R component image, the G component image, and the B component image, respectively.
The enlarged image generation unit 570 adds or subtracts the distribution values determined by the difference distribution value 560 to the pixel values read from the R component image, the G component image, and the B component image by the back projection calculation unit 590. To do.
The image processing apparatus 2 transfers the R component image, the G component image, and the B component image obtained by adding / subtracting the distribution value by the enlarged image generation unit 570 to the output destination, and outputs the enlarged color image.

  As described above, the image processing apparatus 2 according to the present modification has the pixel value of the synthesized image enlarged by the nearest neighbor method and the nearest neighbor without expanding the entire image enlarged by the nearest neighbor method in the memory. The pixel value of each color component image enlarged by the method can be calculated.

[Second Embodiment]
Next, a second embodiment will be described.
When enlarging an original image to a desired output resolution, the enlarging process may be divided into a plurality of stages to obtain a desired enlarged image. For example, when it is desired to enlarge the original image by 8 times in the main scanning direction and the sub-scanning direction, 8 times enlargement is realized by applying the process of enlarging to 2 times in three stages.
In such a case, the image processing apparatus 2 in the second embodiment applies an image enlargement algorithm that prioritizes image quality in the enlargement process in the previous stage, and in the enlargement process in the subsequent stage, in the enlargement process in the previous stage. Apply an image enlargement algorithm that prioritizes processing speed. Specifically, when the enlargement process is configured in n stages (n: a natural number equal to or greater than 2), the image processing apparatus 2 performs the high image quality enlargement unit in the enlargement process from the first to the (n−1) stage. All color component images are enlarged by 530 (FIG. 3), and in the final n-th stage enlargement process, the enlargement process (FIG. 4) based on the difference value is applied and enlarged.
As a result, the image processing apparatus 2 applies the high-quality enlargement algorithm in the previous stage where the size of the target image is small and the processing load is relatively small, thereby suppressing image quality degradation as much as possible, and the size of the target image increases and the processing load increases. By applying an enlargement process based on the difference value at a large subsequent stage, it is possible to improve the overall processing speed while suppressing image quality deterioration.

FIG. 8 is a flowchart for explaining the multi-stage enlargement process (S20).
As shown in FIG. 8, in step 200 (S200), the control unit 580 (FIG. 3) in the image enlargement program 5 receives the image data of the color image (RGB) to be output to the storage unit 500, and the output destination is When specified, the number of stages of the enlargement process is determined based on the resolution of the input image data and the output resolution of the output destination, and an instruction to start the enlargement process is given to each component of the image enlargement program 5. .

In step 205 (S205), the control unit 580 determines whether or not the next enlargement process to be performed is the last stage. If it is the last stage enlargement process, the control unit 580 proceeds to the process of S10. In this case, the process proceeds to S210.
In S10, the enlargement process based on the difference value is performed by the enlargement process described with reference to FIG.

In step 210 (S210), the control unit 580 instructs the high quality enlargement unit 530 to perform enlargement processing of each color component image.
The high image quality enlargement unit 530 enlarges each color component image by applying a high image quality enlargement algorithm (for example, a fractal enlargement method) under the control of the control unit 580.

  In step 215 (S215), the control unit 580 determines whether or not the determined number of stages of enlargement processing has been completed. If all the stages have been completed, the multistage enlargement processing is terminated and image data is output. If all the stages have not been completed, the process returns to S205 and the enlargement process is repeated.

  As described above, the image processing apparatus 2 according to the second embodiment includes an enlargement process using a high-quality enlargement algorithm and an enlargement process based on a difference value according to each stage in the enlargement process including multiple stages. By switching between and, it is possible to improve the overall processing speed while suppressing image quality deterioration.

  Note that the image processing apparatus 2 may combine an enlargement process using a high-speed enlargement algorithm in addition to an enlargement process using a high-quality enlargement algorithm and an enlargement process based on a difference value. For example, when the image processing apparatus 2 performs an enlargement process including n stages (n is a natural number of 3 or more), a high-quality enlargement algorithm (for example, a fractal enlargement method) is performed from the forefront stage to the (n-2) stage. Is applied, and the (n-1) -th stage performs an enlargement process based on the difference value (FIG. 4 and the like), and the last stage, the n-th stage, applies a high-speed enlargement algorithm (for example, nearest neighbor method). Then enlarge. In this case, the image processing apparatus 2 switches between an enlargement process using a high-quality enlargement algorithm, an enlargement process based on a difference value, and an enlargement process using a high-speed enlargement algorithm based on a predetermined switching criterion. The switching reference is, for example, the actual resolution of the output destination (hereinafter, actual resolution). When outputting with a printer or the like, the actual resolution (actual resolution) may be lower than the output resolution due to the influence of the screen or the like. Therefore, the image processing apparatus 2 applies the enlargement process based on the difference value immediately before the resolution of the target image becomes the actual resolution, and uses the high-speed enlargement algorithm after the resolution of the target image becomes equal to or higher than the actual resolution. Apply magnification processing.

FIG. 9 is a flowchart of a multi-stage enlargement process (S22) combining an enlargement process using a high-quality enlargement algorithm, an enlargement process based on a difference value, and a high-speed enlargement algorithm. Note that among the processing steps shown in the figure, the same reference numerals are given to the processing steps substantially the same as those shown in FIG.
As shown in FIG. 9, first, in S200, the control unit 580 (FIG. 3) inputs color image (RGB) image data to be output when it is input to the storage unit 500 and an output destination is specified. Based on the resolution of the image data and the output resolution of the output destination, the number of stages of the enlargement process is determined, and each component of the image enlargement program 5 is instructed to start the enlargement process.

In step 220 (S220), the control unit 580 determines whether or not the resolution of the target image is just before the actual resolution of the output destination. If the resolution is just before the actual resolution or more, the process of S10 is performed. In the case of other places, the process proceeds to S210.
In S10, the enlargement process based on the difference value is performed by the enlargement process described with reference to FIG. In S210, each color component image is enlarged by the high image quality enlargement unit 530 (FIG. 3) in response to an instruction from the control unit 580.

In step 225 (S225), the control unit 580 determines whether or not the resolution of the target image is equal to or higher than the actual resolution. If the resolution is equal to or higher than the actual resolution, the control unit 580 proceeds to the process of S230 and is less than the actual resolution. If there is, the process returns to S220 and the enlargement process is repeated.
In step 230 (S230), the control unit 580 instructs the high-speed enlargement unit 520 (FIG. 3) to perform enlargement processing of each color component image.
The high-speed enlargement unit 520 enlarges each color component image by applying a high-speed enlargement algorithm (for example, nearest neighbor method) under the control of the control unit 580.
As a result, the image processing apparatus 2 can improve the overall processing speed while suppressing the deterioration of the image quality in consideration of the actual resolution of the output destination.

It is a figure explaining the image enlargement algorithm which can obtain a high quality enlarged image at high speed, (A) is a method of applying the high quality enlargement algorithm to only a part of the color component images and speeding up as a whole. (B) shows a method of enlarging the color image based on the difference between the sample images enlarged by the high-speed enlargement algorithm and the high-quality enlargement algorithm. It is a figure which illustrates the hardware constitutions of the image processing apparatus 2 with which the image processing method concerning this invention is applied centering on the control apparatus 20. FIG. It is a figure which illustrates the functional structure of the 1st image expansion program 5 which is performed by the control apparatus 20 (FIG. 2), and implement | achieves the image processing method concerning this invention. It is a flowchart which shows the whole operation | movement of a 1st image expansion process (S10). It is a figure explaining the expansion method of the image by the nearest neighbor method. It is a figure explaining the function structure of the 2nd image expansion program. It is a flowchart explaining a 2nd image expansion process (S12). It is a flowchart explaining a multistage expansion process (S20). It is a flowchart of the multistage expansion process (S22) which combined the expansion process using a high image quality expansion algorithm, the expansion process based on a difference value, and the high-speed expansion algorithm.

Explanation of symbols

DESCRIPTION OF SYMBOLS 2 ... Image processing apparatus 5 ... Image expansion program 500 ... Memory | storage part 510 ... Sample preparation part 520 ... High speed expansion part 530 ... High quality image enlargement part 540 ... Difference calculation part 550 ... gradation change determination unit 560 ... difference distribution unit 570 ... enlarged image generation unit 580 ... control unit 590 ... reverse projection calculation unit

Claims (8)

An image processing apparatus for enlarging a color image composed of a plurality of color component images,
Applying a first enlargement process to enlarge an image;
Applying a second enlargement process capable of an enlargement process faster than the first enlargement process, and a second enlargement means for enlarging the image;
Sample creation means for creating a sample image using at least one color component image of the input color images;
For the sample image created by the sample creation means, a difference calculation means for calculating a difference between the image enlarged by the first enlargement means and the image enlarged by the second enlargement means;
Based on each of the color component images of the color image enlarged by the second enlargement unit, the difference calculated by the difference calculation unit, and the gradation change amount of each color component image, the input color image An image processing apparatus comprising: an enlarged image generating unit that generates an enlarged image. A difference distribution unit that divides the difference calculated by the difference calculation unit into distribution values according to a gradation change amount of each color component image, and distributes the distribution value to each color component image;
The enlarged image generation unit is configured to output an input color image based on each color component image enlarged by the second enlargement unit and a distribution value distributed to each color component image by the difference distribution unit. The image processing apparatus according to claim 1, wherein an enlarged image is generated. The input color image includes RGB color component images,
The image processing apparatus according to claim 1, wherein the sample creation unit creates a sample image using each color component image of RGB. When the enlargement process is repeated a plurality of times to obtain a desired enlargement image, for each enlargement process, a color image enlargement process using only the first enlargement means and a color image based on the difference calculated for the sample image The image processing apparatus according to claim 1, further comprising a control unit that switches and applies the enlargement process. The image processing apparatus according to claim 1, wherein the first enlargement process generates an enlarged image with higher image quality than the second enlargement process. An image processing method for enlarging a color image composed of a plurality of color component images,
Create a sample image using at least one color component image,
Apply the first enlargement process to enlarge the created sample image,
Applying a second enlargement process capable of an enlargement process faster than the first enlargement process, the sample image is enlarged,
Calculating a difference between the sample image enlarged by the first enlargement process and the sample image enlarged by the second enlargement process;
In accordance with the gradation change amount of each color component image, the calculated difference is distributed to each color component image,
An image processing method for generating an enlarged image of an input color image based on each color component image enlarged by the second enlargement processing and a difference distributed to each color component image. When obtaining a desired enlarged image by repeating the enlargement process a plurality of times, at least the first enlargement process applies only the first enlargement process to enlarge the color image, and at least the last enlargement process, the sample image The image processing method according to claim 6, wherein a color image is enlarged by applying an enlargement process based on the difference calculated for. In an image processing apparatus for enlarging a color image composed of a plurality of color component images,
Creating a sample image using at least one color component image;
Applying the first enlargement process to enlarge the created sample image;
Applying a second enlargement process capable of an enlargement process faster than the first enlargement process, and enlarging the sample image;
Calculating a difference between the sample image enlarged by the first enlargement process and the sample image enlarged by the second enlargement process;
Distributing the calculated difference to each color component image according to the gradation change amount of each color component image;
Generating an enlarged image of the input color image based on each of the color component images enlarged by the second enlargement processing and the difference distributed to each of the color component images; The program to be executed.

Images