JP4840295B2 - Image processing apparatus and image processing program - Google Patents

Description

  The present invention relates to an image processing apparatus and an image processing program for performing image correction processing.

2. Description of the Related Art Conventionally, an image processing apparatus that performs image correction according to user preferences is known.
For example, image adjustment is performed based on the adjustment data specified by selecting one adjustment data from a plurality of types of adjustment data prepared in advance or by setting arbitrary adjustment data. There exists a structure to perform (refer patent document 1).
JP 2007-89179 A

  However, in order to set adjustment data so that desired image correction is performed with the above-described configuration, specialized knowledge about images is required, and it is difficult to set intuitively.

  The present invention has been made in view of these problems, and an object thereof is to provide an image processing apparatus and an image processing program capable of intuitively performing image correction settings.

In order to achieve the above object, an image processing apparatus of the present invention includes a first image input unit capable of inputting a plurality of types of first images, and a feature of each first image input by the first image input unit. A first feature quantity specifying unit that specifies a first feature quantity that represents a plurality of types of features of the plurality of types of first images based on the first feature quantities specified by the first feature quantity specifying unit. An overall first feature amount specifying means for specifying an overall first feature amount including a representative value of the image, a second image input means for inputting a second image that is an image to be corrected, and the second image A second feature amount specifying means for specifying a second feature amount including a plurality of representative values representing features of the second image inputted by the input means; a plurality of representative values of the total first feature amount; 2 The second image is matched with the second image based on the correspondence relationship with the plurality of representative values. Comprising an image correcting means for performing image correction processing that, the. The image processing apparatus having the first configuration of the present specification represents a first image input unit capable of inputting a plurality of types of first images, and features of the plurality of types of first images input by the first image input unit. Overall first feature quantity specifying means for specifying the overall first feature quantity, second image input means for inputting a second image as an image to be corrected, and second input by the second image input means. A second feature quantity specifying unit that specifies a second feature quantity that represents a feature of the image; and a second first feature quantity that is specified by the second feature quantity specifying unit, and an overall first feature that is specified by the overall first feature quantity specifying unit. Image correction means for performing image correction processing on the second image so as to approach the amount.

  According to such an image processing apparatus, the user can intuitively set the image correction by using the first image. In particular, in this image processing apparatus, by using a plurality of types of first images, it is possible to perform image correction processing based on those characteristics. Therefore, compared with the case where only one first image is used, image correction settings are set. Can be done in more detail.

Specifically, for example, as in the second configuration , the comprehensive first feature amount specifying unit represents the characteristics of one whole image composed of a plurality of types of first images input by the first image input unit. The feature amount may be specified as the comprehensive first feature amount. In this way, for example, when one image is divided and read, the same image correction as when one image is read as it is can be performed.

Further, for example, as in the third configuration , the first feature amount specifying means for specifying the first feature amount representing the feature of each first image input by the first image input means is provided, and the comprehensive first feature amount specifying The means may specify an overall first feature quantity that represents the features of the plurality of types of first images from the first feature quantities specified by the first feature quantity specifying means. In this way, since it is not necessary to store all the input first images at the same time, it is possible to save the storage capacity necessary for storing the first images.

  And when specifying the comprehensive first feature quantity based on the first feature quantity of each first image in this way, for example, the relative sizes of the plurality of types of first images input by the first image input means are used. It is conceivable to specify the comprehensive first feature amount without taking into account the difference. In this way, even when the first images having different sizes are used, image correction can be performed without being affected by the difference in size.

On the other hand, for example, as in the fourth configuration , the first feature amount specified from each first image and the first feature amount for the plurality of types of first images input by the first image input means, It is also conceivable to specify the total first feature value from the size of the range of the first image used to specify one feature value. In this way, it is possible to perform appropriate image correction when the size of the first image should be taken into account, such as when one image is divided into a plurality of different sizes and read. .

An image processing apparatus having a fifth configuration includes an output unit that outputs a corrected image representing an image correction process result by the image correction unit, and a print that prints the second image on which the image correction process has been performed by the image correction unit. Control means. According to such an image processing device, the user can easily determine whether or not a desired correction result is obtained based on the corrected image.

In the image processing apparatus having the sixth configuration , on the condition that the corrected image is output by the output unit, the first image input unit inputs the first image, and the image correction unit receives the input first image. The image correction process is performed on the second image from the overall first feature quantity in the first image used for the image correction process of one image and the corrected image output by the output unit. According to such an image processing apparatus, it is possible to finely adjust the correction by adding the first image. For example, when the user who has seen the output result determines that the desired correction result has not been obtained based on the corrected image image, the user inputs an instruction to newly input the first image by the image input means and finely adjust the correction. Can do.

Further, in the image processing apparatus having the seventh configuration , the image correction unit includes a plurality of types used in the image correction process of the corrected image image output by the output unit on condition that the corrected image image is output by the output unit. Among the first images, image correction processing is performed on the second image from the overall first feature amount based on the first image other than the predetermined first image. According to such an image processing apparatus, it is possible to finely adjust the correction by excluding the unnecessary first image. For example, when the user who has seen the output result from the output means determines that a desired correction result has not been obtained based on the corrected image image, the correction may be finely adjusted by selecting an unnecessary first image or the like. it can.

Here, the corrected image image may be obtained by performing image correction processing on the reduced image of the second image, for example, as in the eighth configuration . In this way, it is possible to simplify the process of generating a corrected image compared to performing the image correction process on the second image itself.

On the other hand, in the image processing apparatus having the ninth configuration , the first image input means inputs an image optically read from a print medium set at a predetermined reading position, and prints set at the reading position. When a plurality of images are read from the medium, each image can be input as an independent first image. According to such an image processing apparatus, the user sets a plurality of print media at a reading position at a time or sets a printing medium on which a plurality of images are printed at a reading position. One image can be easily read, and the working time can be shortened.

Specifically, as in the tenth configuration , for example, the first image input means can input a plurality of images read in each reading area obtained by dividing the reading position into a plurality of images as independent first images. Can be recognized relatively accurately.

In particular, as in the eleventh configuration , for example, when the average luminance value of the image read in the reading area is greater than or equal to a predetermined value, the first image input unit determines that there is no image in the reading area. By doing so, it is possible to prevent a portion where no image exists from being input as the first image.

Next, an image processing program having a twelfth configuration includes a first image input unit capable of inputting a plurality of types of first images, and a comprehensive feature representing the characteristics of the plurality of types of first images input by the first image input unit. Overall first feature quantity specifying means for specifying the first feature quantity, second image input means for inputting a second image that is an image to be corrected, and a second image input by the second image input means A second feature quantity specifying unit that specifies a second feature quantity that represents a feature of the first feature quantity, and a second feature quantity that is specified by the second feature quantity specifying unit, and a total first feature quantity that is specified by the total first feature quantity specifying unit. The computer is caused to function as an image correction unit that performs an image correction process on the second image so as to approach the image.

According to such an image processing program, it is possible to cause a computer to function as the image processing apparatus having the first configuration , thereby obtaining the above-described effects.

Embodiments to which the present invention is applied will be described below with reference to the drawings.
[1. First Embodiment]
[1-1. overall structure]
FIG. 1 is a perspective view showing an appearance of a multifunction machine 10 as an image processing apparatus according to the first embodiment.

  The multifunction machine 10 has a printer function, a scanner function, a color copy function, and the like, and includes an image reading unit 20 used for reading a document at an upper position in the main body casing 11.

  The image reading unit 20 is a so-called flatbed scanner that optically reads an image from a document set (placed) on a document placement surface (glass table). Here, the upper surface of the document placement surface is covered with a thin plate-like document cover 21, and the document cover 21 is opened upward to set the document on the document placement surface and remove the set document. (In other words, it is possible to put a document in and out). Further, the surface of the document cover 21 opposite to the document placement surface is white, and the document is read when the image is read with the document cover 21 closed (the state shown in FIG. 1). The portion of the placement surface where no document is placed is read in white.

  On the other hand, the multifunction device 10 includes an operation unit 31 in which various operation buttons are arranged at a front position (front side position) of the image reading unit 20 and a display unit (for example, a liquid crystal display) 32 that displays an image such as a message. A panel 30 is provided.

  In addition, the multifunction machine 10 includes an image printing unit 40 that can print a color image on a printing medium such as paper at a position below the image reading unit 20. The sheet on which the image is printed by the image printing unit 40 is discharged from the opening 12 formed on the front surface of the main body casing 11.

  Furthermore, the multifunction machine 10 includes a card slot 50 into which various memory cards (portable storage media) such as an SD card and a CF card can be inserted at a position above the opening 12 on the front surface of the main casing 11. The multifunction device 10 has a function (so-called direct print function) for reading an image (such as an image taken with a digital still camera) directly from a memory card and printing the image without using an information processing device such as a personal computer. is doing.

Next, a control system of the multifunction machine 10 will be described.
FIG. 2 is a block diagram illustrating a schematic configuration of a control system of the multifunction machine 10.
As shown in the figure, the multifunction machine 10 includes the image reading unit 20, the operation panel 30, the image printing unit 40, the card slot 50, the communication unit 60, and the control unit 70 described above. They are connected via a line 80.

  The communication unit 60 performs data transmission / reception processing via the communication cable in a state where the communication cable (LAN cable) is connected. That is, it is for performing data communication with an external device. For example, data communication can be performed with a personal computer in a LAN or a web server on the Internet.

  The control unit 70 is configured mainly with a microcomputer including a CPU 71, a ROM 72, a RAM 73, and the like, and performs overall control of each unit constituting the multifunction machine 10. The ROM 72 stores a program for causing the CPU 71 to execute a color conversion process described later.

[1-2. Overview of color conversion process]
Next, an outline of color conversion processing performed by the multifunction machine 10 will be described.
The MFP 10 according to the present embodiment performs color conversion processing based on an image (hereinafter also referred to as “example image”) serving as a sample for color conversion on an image to be subjected to color conversion. Here, first, the basic flow of such color conversion processing will be described with reference to FIG.

  When a user sets a document (for example, a photograph) on which a model image is printed on the document placement surface of the image reading unit 20 and performs a document reading operation with the operation unit 31 (1), the multifunction machine 10 places the document on the document. A model image is read from a set range on the surface (a range set by the user, such as L size and A4 size) (2). As a result, the model image is read from the document set on the document placement surface.

  Next, when the user inserts a memory card storing an image to be color-converted into the card slot 50 (3), the multi-function device 10 recognizes the inserted memory card and notifies the user of the memory card. The image to be subjected to color conversion is selected from the images stored in (4). In addition, as a process for making a user select an image, a well-known process (For example, the process which displays each image memorize | stored in the memory card on the display part 32, and makes it select by operation in the operation part 31) suitably. It can be adopted.

  When the user selects a color conversion target image (5), the multifunction machine 10 reads the selected image (6). In the following description, the image may be referred to as an “original image”.

Thereafter, the multifunction machine 10 performs a process of correcting the original image read from the memory card using the model image read from the image reading unit 20 as a sample (7).
Here, the procedure for reading the original image from the memory card after reading the model image from the image reading unit 20 is illustrated, but the present invention is not limited to this, and the original image is read from the memory card first, and then A model image may be read from the image reading unit 20.

According to such color conversion processing, the user can perform color conversion of the original image with a simple operation and sensuously by using the model image.
Specifically, for example, if you want to convert the sky blue to the vivid ocean blue for the original image showing the building and the sky, you can use the model image that captures the vivid ocean, Blue can be converted into vivid sea blue.

  Further, when it is desired to brightly convert the skin color with respect to the original image in which the human face is reflected, the skin color of the original image can be converted into a bright skin color by using a model image showing a bright skin color.

  As described above, the user can perform desired color conversion only by causing the image reading unit 20 to read the model image without requiring any specialized knowledge. Furthermore, since the area for color conversion is automatically selected, it is possible to stop or reduce the conversion of areas that are difficult to perceive, and to convert only the areas that are easily perceived.

However, it is conceivable that a desired color conversion process cannot be performed with only one model image.
Therefore, in the MFP 10 according to the present embodiment, color conversion processing using a plurality of types of model images is possible.

[1-3. Specific contents of color conversion processing]
Hereinafter, specific contents of the color conversion process performed by the multifunction machine 10 of the present embodiment will be described.
[1-3-1. Color conversion process]
FIG. 4 is a flowchart of the color conversion process executed by the CPU 71.

  When the CPU 71 starts the color conversion process, first, in S101, the CPU 71 reads a set range on the document placement surface of the image reading unit 20 as a model image. Although the format of the read image is not particularly limited, the present embodiment will be described on the assumption of the RGB format.

Subsequently, in S102, a process of converting each pixel constituting the model image read in S101 into an HSV parameter (H value: 0 to 360, S value and V value: 0 to 1) is performed.
Note that the conversion from RGB to HSV parameters and the conversion from HSV to RGB can be performed according to the known conversion formulas shown below.
(1) RGB → HSV conversion formula
max (a, b, c) represents the largest value among a, b, and c.
min (a, b, c) represents the smallest value among a, b, and c.
V = max (R / 255, G / 255, B / 255)
When V is not 0
S = [V-min (R, G, B)] ÷ V
When V is 0
S = 0
When [V-min (R, G, B)] is not 0,
r = (V-R / 255) ÷ (V-min (R, G, B)
g = (V-G / 255) / (V-min (R, G, B)
b = (V-B / 255) / (V-min (R, G, B)
When [V-min (R, G, B)] is 0,
r = 0
g = 0
b = 0
When V = R / 255
H = 60 × (bg)
When V = G / 255
H = 60 × (2 + rg)
When V = B / 255
H = 60 × (4 + gr)
However, when H <0
H = H + 360
Are converted from RGB to HSV. Also, from HSV to RGB,
(2) HSV-> RGB conversion formula
(In, fl, m, and n shown below are parameters used in the process of calculating RGB from HSV)
in the integer part of (H / 60)
Let fl be the fractional part of (H / 60).
If in is an even number
fl = 1-fl
m = V × (1-S)
n = V × (1-S × fl)
When in is 0
R = V × 255
G = n × 255
B = m × 255
When in is 1
R = n × 255
G = V × 255
B = m × 255
When in is 2
R = m × 255
G = V × 255
B = n × 255
When in is 3
R = m × 255
G = n × 255
B = V × 255
When in is 4
R = n × 255
G = m × 255
B = V × 255
When in is 5
R = V × 255
G = m × 255
B = n × 255
Is converted as

  Subsequently, in S103, based on the HSV parameter obtained by the conversion process in S102, a first feature quantity specifying process for specifying a first feature quantity that is a feature quantity representing the feature of the model image is performed. The specific processing content of the first feature quantity specifying process will be described later (FIG. 6).

  Subsequently, in S104, the first feature amount specified in S103 is stored in the RAM 73. Here, if the first feature value for another model image is already stored in the RAM 73, the first feature value is stored in a newly added form while maintaining the storage state of the first feature value. Specifically, the first feature amount stored in the RAM 73 is erased at the start or end of the main color conversion process, and is stored between the start and end of the main color conversion process. State is maintained. Therefore, by repeating the processing of S101 to S104, a plurality of first feature values for a plurality of model images are stored in the RAM 73. Note that the storage destination of the first feature value is not limited to the RAM 73. For example, when the multifunction device 10 includes a hard disk, the first feature value may be stored in the hard disk.

In step S105, it is determined whether there is a document that has not been read. Specifically, it can be determined by any one of the following methods (1) to (3), for example.
(1) The user sets the number of documents in advance and determines whether or not the number has been reached.

(2) When all the originals have been read, the user is allowed to press a specified button (for example, a print button), and it is determined whether or not the specified button has been pressed.
(3) The user sets in advance a paper size that is the sum of the sizes of all the originals, and determines whether or not the size has been reached. This method is effective when an original having a paper size larger than the original placing surface is divided into a plurality of parts and read.

  If it is determined in S105 that there is a model image that has not yet been read, the process returns to S101. That is, by performing the document reading operation again, a plurality of types of model images can be read.

  On the other hand, if it is determined in S105 that there is no model image that has not yet been read (all model images have been read), the process proceeds to S106, and all the first images stored in the RAM 73 in S104. Feature values (hereinafter also referred to as “first feature value group”) are read out.

  Subsequently, in S107, based on the first feature value group read in S106, a first feature value (hereinafter, also referred to as “total first feature value”) that represents a comprehensive feature of a plurality of types of model images is specified. . A specific identification method will be described later.

  Subsequently, in S108, the color conversion target image (original image) stored in the memory card is read into the RAM 73. Although the format of the image to be read is not particularly limited, the present embodiment will be described on the assumption that the RGB format is used.

  Subsequently, in S109, a thumbnail image of the original image read in S108 is generated. That is, a size change process (reduction process) is performed on the original image so as to have a preset thumbnail image size. Since the original image is usually larger than the thumbnail image, it has been described here that the reduction process is performed as the size change process. However, when the original image is smaller than the thumbnail image size, the size change process is performed. An enlargement process is performed.

Subsequently, in S110, a process of converting each pixel constituting the original image read in S108 (not the thumbnail image generated in S109) into an HSV parameter is performed.
Subsequently, in S111, based on the HSV parameter obtained by the conversion process in S110, a second feature amount specifying process for specifying a second feature amount that is a feature amount representing the feature of the original image is performed. The specific processing content of the second feature amount specifying processing will be described later (FIG. 7).

  Subsequently, in S112, a representative value resetting process is performed in which the values of the first feature value and the second feature value are reset according to conditions. The specific processing contents of the representative value resetting process will be described later (FIG. 9).

Subsequently, in S113, the thumbnail image is corrected based on the total first feature amount and the second feature amount. A specific correction method will be described later.
Subsequently, in S114, as shown in FIG. 5, the corrected thumbnail image corrected in S113 (hereinafter also referred to as “corrected image image”) and confirmation of whether there is no problem in printing with this correction result are confirmed. A message for prompting is displayed on the display unit 32.

Subsequently, in S115, it is determined whether to perform printing according to the operation performed on the operation unit 31 with respect to the content displayed on the display unit 32 in S114.
If it is determined in S115 that printing is not performed, the process proceeds to S116, where it is determined whether to increase (decrease) the model image. Specifically, the user's intention about whether to increase or decrease the model image is confirmed by displaying a message on the display unit 32 and prompting an operation on the operation unit 31.

If it is determined in S116 that the model images are to be increased, the process returns to S101, and a new model image is read.
On the other hand, if it is determined in S116 that the model image is not increased (decreased), the process proceeds to S117, and the color patches corresponding to each of all the first feature values stored in the RAM 73 in S104 are displayed on the display unit 32. Display. Specifically, as will be described later, the first feature amount includes a representative value of each region obtained by dividing the constituent pixels of the sample image into six regions according to the hue. Here, each sample image 6 representative values are displayed as color patches.

  Subsequently, in S118, the first feature amount of the color patch selected by the operation on the operation unit 31 is set to be invalid. That is, the model image for which the color patch is selected is excluded from the plurality of first feature amounts (a plurality of types of model images read by the image reading unit 20) stored in the RAM 73.

  Subsequently, in S119, based on the first feature quantity other than the first feature quantity set to invalid among the first feature quantities stored in the RAM 73, the overall first feature quantity is obtained by the same method as in S107. Identify. Thereafter, the process returns to S112.

  On the other hand, if it is determined in S115 that printing is to be performed, the process proceeds to S120, and the original image (not the thumbnail image) is corrected based on the total first feature value and the second feature value specified in S111. A specific correction method will be described later.

Subsequently, in S122, after the original image corrected in S121 is printed on the image printing unit 40, the main color conversion process is terminated.
[1-3-2. First feature amount specifying process]
Next, the first feature amount specifying process performed in S103 in the color conversion process (FIG. 4) will be described with reference to the flowchart of FIG. In the first feature amount specifying process, the H value takes a value of −30 to 330. If the H value is not within this range, the H value is appropriately converted (for example, “ H value + 360 × n ”or“ H value−360 × n ”, where n is an integer), and adjust within this range.

When starting the first feature amount specifying process, the CPU 71 first divides the model image to be processed into a plurality of regions in S201. In this embodiment, the image is divided based on six commonly used hues. Specifically, based on the H value of each pixel,
-R region: -30 or more and less than 30-Y region: 30 or more and less than 90-G region: 90 or more and less than 150-C region: 150 or more and less than 210-B region: 210 or more and less than 270-M region: Divide into 270 or more and less than 330. That is, a process of classifying the constituent pixels of the model image into six classification items according to the classification standard corresponding to the hue value is performed. The correspondence relationship between these areas and the H value is merely an example, and can be changed as appropriate.

  Subsequently, in S202, for each region divided in S201, the ratio of each region in the model image and the representative value (HSV value) representing the characteristics of the constituent pixels belonging to each region are calculated as the first feature amount. To do.

Here, the representative value (HSV value) of each region is defined as follows.
-Typical values of R region: sHr, sSr, sVr
-Typical values in the G region: sHg, sSg, sVg
-Typical values of the B region: sHb, sSb, sVb
-Typical value of C region: sHc, sSc, sVc
-Typical values in the M region: sHm, sSm, sVm
-Typical values of the Y region: sHy, sSy, sSy
In this embodiment, the average value of each HSV value of the constituent pixels belonging to each region is specified as a representative value. The representative value is not limited to the average value, and for example, an intermediate value can be used.

Further, the ratio of each area in the model image is defined as follows.
The ratio of the R region in the model image: sRateR
-Ratio of the G area in the model image: sRateG
-Ratio of area B in the model image: sRateB
-Ratio of the C area in the model image: sRateC
-Ratio of M area in the model image: sRateM
The ratio of the Y area in the model image: sRateY
For example, for the R region:
sRateR = (number of pixels in the R region in the model image) / (total number of pixels in the model image)
It can be. In addition, you may define by another type | formula.

[1-3-3. Processing for specifying the overall first feature amount]
Next, a process for specifying the comprehensive first feature value (first feature value taking into account the features of all model images) performed in S107 in the color conversion process (FIG. 4) described above will be described.

Here, a case where the image reading unit 20 reads three example images A to C will be described as an example. Here, the first feature amount of each of the model images A to C is defined as follows.
<First feature amount of model image A>
-Representative values of R region: sHr (A), sSr (A), sVr (A)
-Representative values of G region: sHg (A), sSg (A), sVg (A)
-Representative value of B region: sHb (A), sSb (A), sVb (A)
-Representative value of C region: sHc (A), sSc (A), sVc (A)
-Typical values of M region: sHm (A), sSm (A), sVm (A)
-Representative values of the Y region: sHy (A), sSy (A), sSy (A)
The ratio of the R region in the model image: sRateR (A)
-Ratio of the G area in the model image: sRateG (A)
-Ratio of area B in the model image: sRateB (A)
-Ratio of the C area in the model image: sRateC (A)
-Ratio of M area in model image: sRateM (A)
-Proportion of the Y area in the model image: sRateY (A)
<First feature amount of model image B>
・ Representative value of R region: sHr (B), sSr (B), sVr (B)
-Typical values in the G region: sHg (B), sSg (B), sVg (B)
-Representative value of B region: sHb (B), sSb (B), sVb (B)
-Representative value of C region: sHc (B), sSc (B), sVc (B)
-Typical values in the M region: sHm (B), sSm (B), sVm (B)
-Representative values of the Y region: sHy (B), sSy (B), sSy (B)
The ratio of the R area in the model image: sRateR (B)
-Ratio of the G area in the model image: sRateG (B)
-Ratio of area B in the model image: sRateB (B)
-Ratio of the C area in the model image: sRateC (B)
The ratio of the M area in the model image: sRateM (B)
-Proportion of the Y area in the model image: sRateY (B)
<First feature amount of model image C>
-Typical values of R region: sHr (C), sSr (C), sVr (C)
-Representative value of G region: sHg (C), sSg (C), sVg (C)
-Representative value of B region: sHb (C), sSb (C), sVb (C)
-Representative value of C region: sHc (C), sSc (C), sVc (C)
-Representative values of M region: sHm (C), sSm (C), sVm (C)
-Representative values of the Y region: sHy (C), sSy (C), sSy (C)
The ratio of the R region in the model image: sRateR (C)
-Ratio of the G area in the model image: sRateG (C)
-Ratio of area B in the model image: sRateB (C)
-Ratio of the C area in the model image: sRateC (C)
-Ratio of the M area in the model image: sRateM (C)
The ratio of the Y area in the model image: sRateY (C)
Further, the representative value (HSV value) of each area of the comprehensive first feature amount specified in S107 and the ratio of each area in the model image are defined as follows.
-Typical values of R region: sHr, sSr, sVr
-Typical values in the G region: sHg, sSg, sVg
-Typical values of the B region: sHb, sSb, sVb
-Typical value of C region: sHc, sSc, sVc
-Typical values in the M region: sHm, sSm, sVm
-Typical values of the Y region: sHy, sSy, sSy
The ratio of the R region in the model image: sRateR
-Ratio of the G area in the model image: sRateG
-Ratio of area B in the model image: sRateB
-Ratio of the C area in the model image: sRateC
-Ratio of M area in the model image: sRateM
The ratio of the Y area in the model image: sRateY
In S107, the average value of the first feature values of the model images A to C is specified as the comprehensive first feature value.

Specifically, for example, the representative value of the R region is obtained as follows.
sHr = ((sHr (A) × sRateR (A))
+ (SHr (B) × sRateR (B))
+ (SHr (C) × sRateR (C)))
/ (SRateR (A) + sRateR (B) + sRateR (C))
sSr = ((sSr (A) × sRateR (A))
+ (SSr (B) × sRateR (B))
+ (SSr (C) × sRateR (C)))
/ (SRateR (A) + sRateR (B) + sRateR (C))
sVr = ((sVr (A) × sRateR (A))
+ (SVr (B) × sRateR (B))
+ (SVr (C) × sRateR (C)))
/ (SRateR (A) + sRateR (B) + sRateR (C))
In this way, the average value of the representative values is calculated by taking into account the proportion of each model image without considering the relative size difference of the model images. It should be noted that the representative values of other areas are calculated in the same manner.

The ratio of each area in the model image is obtained as follows.
sRateR
= (SRateR (A) + sRateR (B) + sRateR (C)) / 3
sRateG
= (SRateG (A) + sRateG (B) + sRateG (C)) / 3
sRateB
= (SRateB (A) + sRateB (B) + sRateB (C)) / 3
sRateC
= (SRateC (A) + sRateC (B) + sRateC (C)) / 3
sRateM
= (SRateM (A) + sRateM (B) + sRateM (C)) / 3
sRateY
= (SRateY (A) + sRateY (B) + sRateY (C)) / 3
[1-3-4. Second feature value specifying process]
Next, the second feature amount specifying process performed in S111 in the above-described color conversion process (FIG. 4) will be described with reference to the flowchart of FIG. In the second feature quantity specifying process, the same process as that performed on the model image in the first feature quantity specifying process (FIG. 6) described above is performed on the original image.

  That is, when the CPU 71 starts the second feature amount specifying process, first, in S301, the original image is divided into six regions. Since the processing content is the same as the processing of S201 in the first feature amount specifying processing, a detailed description thereof will be omitted.

Subsequently, in S302, the second feature value is calculated by performing the same process as the process of S202 in the first feature value specifying process on the original image. Here, the representative value (HSV value) of each region is defined as follows.
・ Representative value of R region: iHr, iSr, iVr
・ Representative value of G region: iHg, iSg, iVg
・ Representative value of B area: iHb, iSb, iVb
-Typical value of C region: iHc, iSc, iVc
-Typical values in the M region: iHm, iSm, iVm
・ Representative value of Y area: iHy, iSy, iSy
Further, the ratio of each area in the original image is defined as follows.
-Ratio of R area in original image: iRateR
-Ratio of the G area in the original image: iRateG
-Ratio of B area in original image: iRateB
-Ratio of the C area in the original image: iRateC
-Ratio of M area in original image: iRateM
The ratio of the Y area in the original image: iRateY
[1-3-5. Original image correction processing]
Next, a specific method of the correction process of the original image (the thumbnail image of the original image in S113, which will be read similarly hereinafter) performed in S113 and S120 in the color conversion process (FIG. 4) described above will be described. This process is performed by converting the H value, S value, and V value of each pixel of the original image.

First, the conversion process for the H value will be described.
The representative value of the H value of the second feature amount is plotted on the X axis, and the representative value of the H value of the total first feature amount is plotted on the Y axis, and the representative value of the H value for each region is plotted. Then, a hue correction table shown in FIG. 8 is created by, for example, linear interpolation between the plotted points. Here, the H value after correction by the hue correction table (H value of the Y-axis) is H ′. When H ′ <0, H ′ = H ′ + 360, and when H ′> 360, H ′> 360. '= H'-360.

Then, the H value is corrected by applying the hue correction table to each pixel of the original image. Specifically, H ′ after correction can be defined by the following equation.
H ′ = (y2−y1) ÷ (x2−x1) × H
-(y2-y1) ÷ (x2-x1) x x2 + y2
... (Formula 1)
Here, x1, x2, y1, and y2 are defined as follows.

When H <iHr,
(X1, y1) = (iHm-360, sHm-360)
(X2, y2) = (iHr, sHr)
When iHr ≦ H <iHy,
(X1, y1) = (iHr, sHr)
(X2, y2) = (iHy, sHy)
When iHy ≦ H <iHg,
(X1, y1) = (iHy, sHy)
(X2, y2) = (iHg, sHg)
When iHg ≦ H <iHc,
(X1, y1) = (iHg, sHg)
(X2, y2) = (iHc, sHc)
When iHc ≦ H <iHb,
(X1, y1) = (iHc, sHc)
(X2, y2) = (iHb, sHb)
When iHb ≦ H <iHm,
(X1, y1) = (iHb, sHb)
(X2, y2) = (iHm, sHm)
When iHm ≦ H,
(X1, y1) = (iHm, sHm)
(X2, y2) = (iHr + 360, sHr + 360)
Next, conversion in the S value and the V value will be described.

The S value and the V value are converted for each area divided by the H value. For example, for the R region:
When S ≦ iSr,
S ′ = S × (sSr ÷ iSr) (Formula 2)
When S> iSr,
S ′ = 1 + (S−1) × {(1-sSr) ÷ (1-iSr)} (Formula 3)
When V ≦ iVr,
V ′ = V × (sVr ÷ iVr) (Formula 4)
When V> iVr,
V ′ = 1 + (V−1) × {(1−sVr) ÷ (1−iVr)} (Formula 5)
It can be calculated by the following formula. Further, other areas can be similarly calculated. In the following, the conversion table defined by the S value conversion formula may be referred to as a saturation correction table, and the conversion table defined by the V value conversion formula may be referred to as a brightness correction table. .

  Thereafter, the converted HSV value is converted into a format suitable for the image printing unit 40 (for example, RGB value). The conversion from the HSV value to the RGB value can be performed according to the above-described known conversion formula.

  In this way, the hue of the original image is converted to the hue of the model image by performing correction processing on the original image so that the second feature amount approaches the first feature amount for each region divided based on the H value. can do.

[1-3-6. Representative value reset process]
Next, the representative value resetting process performed in S112 in the above-described color conversion process (FIG. 4) will be described with reference to the flowchart of FIG.

  When the CPU 71 starts the representative value resetting process, first, in S401, the CPU 71 determines whether one of the six areas divided for each hue in the original image is to be subjected to color conversion processing. Here, whether or not the area is to be subjected to color conversion processing is determined depending on whether or not the area satisfies a conversion target condition described later. If it is determined that the color conversion is to be performed (S401: YES), the process proceeds to S403. On the other hand, if it is determined not to be subject to color conversion (S401: NO), the process proceeds to S402, the representative values of the overall first feature value and the second feature value relating to the region are reset, and then the process proceeds to S403. Transition. A method for resetting the representative value will be described later.

  In S403, it is determined whether or not the determination process for determining whether or not to make all six areas subject to color conversion has been performed. If it is determined that there remains an area for which the process for determining whether or not to perform color conversion remains (S403: NO), the process returns to S401 and is repeated. On the other hand, if it is determined that the determination process has been performed for all regions (S403: YES), the representative value resetting process is terminated.

Here, the conversion target condition in S401 will be described.
(A) Method Using Threshold Thre In S401, the ratio value (the ratio occupied in the model image or the original image) for the target area in the total first feature value and the second feature value is equal to or greater than the threshold value Thre. In this case, it is determined that the conversion target condition is satisfied (the color conversion target). If the ratio value for the target region in at least one of the total first feature amount and the second feature amount is less than the threshold value Thre, in step S402, the total first feature amount and second feature amount relating to that region. The representative value is changed to the same value, and correction processing is performed using the changed representative value. Specifically, the representative value is reset as follows.

When sRateR <Thre or iRateR <Thre,
sHr = 0, sSr = 0.5, sVr = 0.5,
iHr = 0, iSr = 0.5, iVr = 0.5
When sRateG <Thre or iRateG <Thre,
sHg = 120, sSg = 0.5, sVg = 0.5,
iHg = 120, iSg = 0.5, iVg = 0.5
When sRateB <Thre or iRateB <Thre,
sHb = 240, sSb = 0.5, sVb = 0.5,
iHb = 240, iSb = 0.5, iVb = 0.5
When sRateC <Thre or iRateC <Thre,
sHc = 180, sSc = 0.5, sVc = 0.5,
iHc = 180, iSc = 0.5, iVc = 0.5
When sRateM <Thre or iRateM <Thre,
sHm = 300, sSm = 0.5, sVm = 0.5,
iHm = 300, iSm = 0.5, iVm = 0.5
When sRateY <Thre or iRateY <Thre,
sHy = 60, sSy = 0.5, sVy = 0.5,
iHy = 60, iSy = 0.5, iVy = 0.5
In the present embodiment, for the S value and the V value, 0.5, which is an intermediate value of possible values (0 to 1), is adopted, and for the H value, the intermediate value of each region is adopted. These are merely examples, and are not limited to these numerical values.

By changing the representative value in this way, in the correction process, the values of the S value and the V value are not converted as is apparent from the conversion expressions (Expression 2) to (Expression 5) described above. That is, for example, with respect to the R region, when S ≦ iSr,
S ′ = S × (sSr ÷ iSr)
In this equation, sSr = 0.5 and iSr = 0.5, so the above-described equation is
S ′ = S × (0.5 ÷ 0.5) = S (Expression 6)
It becomes. Similarly, when S> iSr, S ′ = S. Similarly, the V value and other areas are not converted.

  On the other hand, for the H value, the points plotted in FIG. 8 are changed to the representative values, so that the conversion amount in that region can be reduced. That is, even when the above-described conversion formula (Formula 1) is used, the conversion amount is reduced by changing the representative value.

  Next, a method for determining the threshold value Thre will be described. This value can be determined based on sensory evaluation, for example. In sensory evaluation, it was confirmed that if the area occupied about 6% or more, the area was easily perceived. Therefore, 6% can be adopted as the threshold value Thre. However, the threshold value Thre is not limited to 6%.

  For example, the threshold value Thre may be determined so as to extract a region having a relatively large area with respect to other regions. Specifically, if the number of regions to be divided is 6, the reciprocal 1/6 is set as the threshold value Thre.

  Here, the number of divided areas is six. The remaining six vertices excluding white and black, which are achromatic colors, from the RGB space (number of vertices 8) which is one of the color gamuts expressing colors. It is. In order for a person to identify a color, it is sufficient to classify the color gamut into the number of vertices of 6, and if the number is less than 6, the user is more likely to feel that the original image is not converted like a model image. On the other hand, if the data is divided more finely than 6, the conversion accuracy is improved, but the possibility that the person cannot be identified increases. In addition, since the amount of calculation increases as the number of divisions increases, the time until a print result is obtained increases and the possibility of increasing user dissatisfaction increases. Therefore, the number of divided regions is preferably six.

In the present embodiment, the same threshold value Thre is used in all regions. However, the present invention is not limited to this, and the threshold value Thre may be changed for each region.
(B) Method Using Information on Maximum Region In the method (A) above, a threshold value Thre is set, and the representative value is changed based on the threshold value Thre, that is, the color conversion process is stopped, the conversion amount is changed. Reduction control was performed. Here, a method of using information on the maximum area in the image in order to reflect only a specific color of the model image in the original image will be described.

  In this case, in S401, it is determined that the conversion target condition is satisfied (to be color conversion target) when the region has the largest ratio value in both the total first feature amount and the second feature amount. And about the area | region which does not satisfy | fill conversion object conditions, the representative value of the comprehensive 1st feature value and 2nd feature value which concerns on the area | region is reset as follows in S402. Here, the largest ratio value among the ratio values of the overall first feature amount is sMaxRate. Further, the largest ratio value among the ratio values of the second feature amount is assumed to be iMaxRate.

When sRateR ≠ iMaxRate or iRateR ≠ sMaxRate,
sHr = 0, sSr = 0.5, sVr = 0.5,
iHr = 0, iSr = 0.5, iVr = 0.5
When sRateG ≠ iMaxRate or iRateG ≠ sMaxRate,
sHg = 120, sSg = 0.5, sVg = 0.5,
iHg = 120, iSg = 0.5, iVg = 0.5
When sRateB ≠ iMaxRate or iRateB ≠ sMaxRate,
sHb = 240, sSb = 0.5, sVb = 0.5,
iHb = 240, iSb = 0.5, iVb = 0.5
When sRateC ≠ iMaxRate or iRateC ≠ sMaxRate,
sHc = 120, sSc = 0.5, sVc = 0.5,
iHc = 120, iSc = 0.5, iVc = 0.5
When sRateM ≠ iMaxRate or iRateM ≠ sMaxRate,
sHm = 300, sSm = 0.5, sVm = 0.5,
iHm = 300, iSm = 0.5, iVm = 0.5
When sRateY ≠ iMaxRate or iRateY ≠ sMaxRate,
sHy = 60, sSy = 0.5, sVy = 0.5,
iHy = 60, iSy = 0.5, iVy = 0.5
Since the representative value is set in this way, only the region having the largest ratio value in both the overall first feature amount and the second feature amount is to be converted. Conversion is not performed for the V value, and the conversion amount can be reduced for the H value.

  Specifically, for example, when only the B region is to be converted, a hue correction table as shown in FIG. 10 is created. In this hue correction table, the H value representative value (iHc = 180, sHc = 180) in the C region adjacent to the B region in the color space and the H value representative value (iHb, sHb) in the B region are linear. In addition, the representative value of the H value (iHm = 300, sHm = 300) in the M region adjacent to the B region in the color space and the representative value of the H value in the B region (iHb, sHb) are connected by a straight line. It will be.

Therefore, the C region where the H value is 180 <H ≦ 210 and the M region where the H value is 270 <H ≦ 300 are also converted. This conversion amount becomes larger as the value is closer to the B region.
In this way, the conversion target region can be selected, and even if the region is not the conversion target, the H value adjacent in the color space is partially converted. It is possible to prevent a pseudo contour (tone jump) from being generated between a region that is not a conversion target.

  By performing such a representative value resetting process, it is possible to stop a part of the correction process or reduce the conversion amount for each divided area based on the size of the area. The user can reflect only a part of the hue of the model image in the hue of the original image.

[1-4. effect]
As described above, the multifunction machine 10 of the present embodiment inputs a plurality of types of model images from the image reading unit 20 (S101, S105), and represents the first characteristic representing the overall characteristics of the input types of model images. An overall first feature quantity that is a feature quantity is specified (S102 to S104, S106, S107). In addition, an original image to be subjected to color conversion is input from the memory card (S108), and a second feature amount representing the characteristics of the input original image is specified (S110, S111). Then, color conversion processing is performed on the original image so that the second feature amount approaches the overall first feature amount (S112, S120).

  According to such a multifunction device 10, the user can cause the multifunction device 10 to perform color conversion processing based on these features by using a plurality of types of model images. For this reason, the user can intuitively create colors that do not exist in the model image at hand by using an image that mixes the colors of multiple types of model images, compared to the case where only one model image is used. Therefore, it is possible to perform color conversion processing that reflects the user's intention more accurately.

  Specifically, when a model image is input (S101), a first feature amount representing the feature of the model image is specified and stored (S102 to S104), and the first comprehensive amount based on the first feature amount of each model image is stored. The feature amount is specified (S106, S107). For this reason, it is not necessary to store all the input model images at the same time, and it is possible to save the storage capacity necessary for storing the model images.

  Further, as shown in FIG. 11, the total first feature amount is specified based on only the first feature amount of each model image without taking into account the relative size difference between the input multiple types of model images. Therefore, even when model images having different sizes are used, color conversion can be performed without being affected by the difference in size.

  On the other hand, in this multifunction device 10, since the corrected image image representing the color conversion processing result is displayed on the display unit 32 before the image is printed (S114), can the user obtain the desired color conversion processing result? It is possible to easily determine whether or not based on the corrected image. Then, if the corrected image does not represent a desired color conversion processing result, the user can newly add a model image used for the color conversion process or exclude an unnecessary model image (S115). S119). For this reason, it is possible to finely adjust the color conversion process without wasteful printing, and an appropriate color conversion process result intended by the user can be obtained.

  In addition, since a corrected image is generated by performing color conversion processing on the reduced image of the original image, a corrected image image is generated as compared with the case where color conversion processing is performed on the original image itself. The processing to be performed can be simplified.

[1-5. Correspondence with Claims]
In the MFP 10 of the first embodiment, the CPU 71 that executes the processing of S101 and S105 in the color conversion processing (FIG. 4) corresponds to the first image input means, and the CPU 71 that executes the processing of S102 and S103. The CPU 71 that executes the processes of S104, S106, and S107 corresponds to the first feature amount specifying means. The CPU 71 that executes the process of S108 corresponds to the second image input unit, and the CPU 71 that executes the processes of S109, S113, and S114 corresponds to the output unit. Further, the CPU 71 that executes the processes of S110 and S111 corresponds to a second feature amount specifying unit, and the CPU 71 that executes the processes of S112 and S115 to S120 corresponds to an image correcting unit.

[2. Second Embodiment]
Next, the multifunction machine 10 of the second embodiment will be described.
The MFP 10 of the second embodiment is different from the MFP 10 of the first embodiment in that the color conversion process shown in the flowchart of FIG. 12 is performed instead of the color conversion process (FIG. 4) of the first embodiment described above. Different. Description of other common contents will be omitted.

  Compared with the color conversion process of FIG. 4, the color conversion process of FIG. 12 is partially different from the process contents of S104, S106, and S107, and the remaining S501 to S503 and S505. , S508 to S521 have the same processing contents as S101 to S103, S105, and S108 to S121. Therefore, the processing related to this difference will be mainly described, and description of common portions will be omitted.

  In S504, the first feature value specified in S503 and the range used to specify the first feature value are stored in the RAM 73. Here, the “range” is the number of constituent pixels used for calculating each representative value (that is, the number of constituent pixels belonging to each region). If the first feature amount and range for another model image are already stored in the RAM 73, the first feature amount and range are stored in a newly added form while maintaining the storage state of the first feature amount and range. Specifically, the first feature value and range stored in the RAM 73 are erased at the start or end of the main color conversion process, and from the start to the end of the main color conversion process. The memory state is maintained. Therefore, by repeating the processing of S501 to S504, a plurality of first feature amounts and ranges for a plurality of model images are stored in the RAM 73. Note that the storage destination of the first feature amount and the range is not limited to the RAM 73. For example, when the multifunction machine 10 includes a hard disk, it may be stored in the hard disk.

In S506, all the first feature values (hereinafter also referred to as “first feature value group”) and ranges (hereinafter also referred to as “range groups”) stored in the RAM 73 in S504 are read out.
Subsequently, in S507, based on the first feature amount group and the range group read in S506, an overall first feature amount that is a first feature amount representing the overall features of a plurality of types of model images is specified.

Here, a case where the image reading unit 20 reads three example images A to C will be described as an example. Here, the first feature amount of each of the model images A to C is defined as follows.
<First feature amount and range of sample image A>
-Representative values of R region: sHr (A), sSr (A), sVr (A)
-Representative values of G region: sHg (A), sSg (A), sVg (A)
-Representative value of B region: sHb (A), sSb (A), sVb (A)
-Representative value of C region: sHc (A), sSc (A), sVc (A)
-Typical values of M region: sHm (A), sSm (A), sVm (A)
-Representative values of the Y region: sHy (A), sSy (A), sSy (A)
The ratio of the R region in the model image: sRateR (A)
-Ratio of the G area in the model image: sRateG (A)
-Ratio of area B in the model image: sRateB (A)
-Ratio of the C area in the model image: sRateC (A)
-Ratio of M area in model image: sRateM (A)
-Proportion of the Y area in the model image: sRateY (A)
A range (Pixel) used to calculate the first feature amount of the R region: sAreaR (A)
A range (Pixel) used to calculate the first feature amount of the G region: sAreaG (A)
A range (Pixel) used to calculate the first feature amount of the B area: sAreaB (A)
A range (Pixel) used to calculate the first feature amount of the C area: sAreaC (A)
A range (Pixel) used to calculate the first feature amount of the M region: sAreaM (A)
A range (Pixel) used to calculate the first feature amount of the Y region: sAreaY (A)
<First feature amount and range of model image B>
・ Representative value of R region: sHr (B), sSr (B), sVr (B)
-Typical values in the G region: sHg (B), sSg (B), sVg (B)
-Representative value of B region: sHb (B), sSb (B), sVb (B)
-Representative value of C region: sHc (B), sSc (B), sVc (B)
-Typical values in the M region: sHm (B), sSm (B), sVm (B)
-Representative values of the Y region: sHy (B), sSy (B), sSy (B)
The ratio of the R area in the model image: sRateR (B)
-Ratio of the G area in the model image: sRateG (B)
-Ratio of area B in the model image: sRateB (B)
-Ratio of the C area in the model image: sRateC (B)
The ratio of the M area in the model image: sRateM (B)
-Proportion of the Y area in the model image: sRateY (B)
A range (Pixel) used to calculate the first feature amount of the R region: sAreaR (B)
A range (Pixel) used to calculate the first feature amount of the G region: sAreaG (B)
A range (Pixel) used to calculate the first feature amount of the B area: sAreaB (B)
A range (Pixel) used to calculate the first feature amount of the C area: sAreaC (B)
A range (Pixel) used to calculate the first feature amount of the M region: sAreaM (B)
A range (Pixel) used to calculate the first feature amount of the Y region: sAreaY (B)
<First feature amount and range of model image C>
・ Representative values of R region: sHr (C), sSr (C), sVr (C)
-Representative value of G region: sHg (C), sSg (C), sVg (C)
-Representative value of B region: sHb (C), sSb (C), sVb (C)
-Representative value of C region: sHc (C), sSc (C), sVc (C)
-Representative values of M region: sHm (C), sSm (C), sVm (C)
-Representative values of the Y region: sHy (C), sSy (C), sSy (C)
The ratio of the R region in the model image: sRateR (C)
-Ratio of the G area in the model image: sRateG (C)
-Ratio of area B in the model image: sRateB (C)
-Ratio of the C area in the model image: sRateC (C)
-Ratio of the M area in the model image: sRateM (C)
The ratio of the Y area in the model image: sRateY (C)
A range (Pixel) used to calculate the first feature amount of the R region: sAreaR (C)
A range (Pixel) used to calculate the first feature amount of the G region: sAreaG (C)
A range (Pixel) used to calculate the first feature amount of the B area: sAreaB (C)
A range (Pixel) used to calculate the first feature amount of the C area: sAreaC (C)
A range (Pixel) used to calculate the first feature amount of the M region: sAreaM (C)
A range (Pixel) used to calculate the first feature amount of the Y region: sAreaY (C)
Further, the representative value (HSV value) of each area of the total first feature amount specified in S507 and the ratio of each area in the model image are defined as follows.
-Typical values of R region: sHr, sSr, sVr
-Typical values in the G region: sHg, sSg, sVg
-Typical values of the B region: sHb, sSb, sVb
-Typical value of C region: sHc, sSc, sVc
-Typical values in the M region: sHm, sSm, sVm
-Typical values of the Y region: sHy, sSy, sSy
The ratio of the R region in the model image: sRateR
-Ratio of the G area in the model image: sRateG
-Ratio of area B in the model image: sRateB
-Ratio of the C area in the model image: sRateC
-Ratio of M area in the model image: sRateM
The ratio of the Y area in the model image: sRateY
In step S507, the average value (average value weighted based on the range) of the first feature amounts of the model images A to C is specified as the comprehensive first feature amount.

Specifically, for example, the representative value of the R region is obtained as follows.
sHr = ((sHr (A) × sAreaR (A))
+ (SHr (B) × sAreaR (B))
+ (SHr (C) × sAreaR (C)))
/ (SAreaR (A) + sAreaR (B) + sAreaR (C))
sSr = ((sSr (A) × sAreaR (A))
+ (SSr (B) × sAreaR (B))
+ (SSr (C) × sAreaR (C)))
/ (SAreaR (A) + sAreaR (B) + sAreaR (C))
sVr = ((sVr (A) × sAreaR (A))
+ (SVr (B) × sAreaR (B))
+ (SVr (C) × sAreaR (C)))
/ (SAreaR (A) + sAreaR (B) + sAreaR (C))
In this way, the average value of the representative values is calculated in consideration of the relative size difference of the model images. It should be noted that the representative values of other areas are calculated in the same manner.

The ratio of each area in the model image is obtained as follows.
First, the total value of each color range is calculated.
sumAreaR = sAreaR (A) + sAreaR (B) + sAreaR (C)
sumAreaG = sAreaG (A) + sAreaG (B) + sAreaG (C)
sumAreaB = sAreaB (A) + sAreaB (B) + sAreaB (C)
sumAreaC = sAreaC (A) + sAreaC (B) + sAreaC (C)
sumAreaM = sAreaM (A) + sAreaM (B) + sAreaM (C)
sumAreaY = sAreaY (A) + sAreaY (B) + sAreaY (C)
Next, the total value of all ranges is calculated.

sumArea = sumAreaR + sumAreaG + sumAreaB
+ SumAreaC + sumAreaM + sumAreaY
Using this total value, the existence ratio of each region is calculated.
The ratio of the R region in the model image:
sRateR = sumAreaR / sumArea
-Ratio of G area in the model image:
sRateG = sumAreaG / sumArea
-Ratio of area B in the model image:
sRateB = sumAreaB / sumArea
-Ratio of the C area in the model image:
sRateC = sumAreaC / sumArea
-Ratio of the M area in the model image:
sRateM = sumAreaM / sumArea
The ratio of the Y area in the model image:
sRateY = sumAreaY / sumArea
As described above, in the multifunction machine 10 according to the second embodiment, as shown in FIG. 13, the first feature amount of each model image and the first feature amount thereof are specified for a plurality of types of model images input. The total first feature value is specified from the size of the range of the model image used in the above. Therefore, it is possible to mix the colors of a plurality of types of model images by reflecting the difference in relative sizes of the model images. As a result, for example, when an image having a size larger than the document placement surface of the image reading unit 20 is to be used as a model image, the original image is temporarily reproduced once by dividing the image into a plurality of images. The color conversion process equivalent to the case where the image data is read by can be performed.

  In the MFP 10 of the second embodiment, the CPU 71 that executes the processes of S501 and S505 in the color conversion process (FIG. 12) corresponds to the first image input means, and the CPU 71 that executes the processes of S502 and S503. The CPU 71 that corresponds to the first feature quantity specifying means and executes the processes of S504, S506, and S507 corresponds to the comprehensive first feature quantity specifying means. The CPU 71 that executes the process of S508 corresponds to the second image input unit, and the CPU 71 that executes the processes of S509, S513, and S514 corresponds to the output unit. Further, the CPU 71 that executes the processes of S510 and S511 corresponds to the second feature amount specifying means, and the CPU 71 that executes the processes of S512, S515 to S520 corresponds to the image correcting means.

[3. Third Embodiment]
Next, the multifunction machine 10 of the third embodiment will be described.
The MFP 10 of the third embodiment is different from the MFP 10 of the first embodiment in that the color conversion process shown in the flowchart of FIG. 14 is performed instead of the color conversion process (FIG. 4) of the first embodiment described above. Different. Description of other common contents will be omitted.

  The color conversion process of FIG. 14 is compared with the color conversion process of FIG. 4 in that the process of S601 to S606 is performed instead of the process of S101 to S107, and the process of S616 to S618 is replaced with the process of S117 to S119. The difference is in the processing. In addition, the processing contents of S607 to S615, S619, and S620 are the same as the processing contents of S108 to S116, S120, and S121. Therefore, the processing related to this difference will be mainly described, and description of common portions will be omitted.

  When the CPU 71 starts the color conversion process, first, in step S601, the entire document placement surface of the image reading unit 20 is read as one whole image. Although the format of the read image is not particularly limited, the present embodiment will be described on the assumption of the RGB format.

  In step S602, a model image is extracted from the entire image read in step S601. In this embodiment, an image read in each reading area obtained by dividing the document placement surface into a plurality of parts is recognized as an independent model image. Specifically, for example, as shown in FIG. 15 (b), the document placement surface can be divided into four reading areas in the vertical and horizontal directions. In this case, FIG. As shown in FIG. 15, if a plurality of (three in the example in the figure) photos are set, the images of the photos are recognized as independent images as shown in FIG. Therefore, at this stage, the reading area (Image 1 in FIG. 15C) where no photograph is set is also recognized as one image (white image). Note that the number of divisions of the reading area may be set in advance or may be set by the user.

  Subsequently, in S603, it is determined whether or not there is a sample image extracted in S602 having an average luminance value meanV of the entire image equal to or higher than a preset threshold value ThreV (for example, 0.97). judge. Here, the threshold value ThreV is set as a reference value for determining whether or not a document is set on the document placement surface of the image reading unit 20, and the average luminance value meanV is equal to or greater than the threshold value ThreV ( If it is determined that the image is white only), there is a high possibility that the document does not exist.

  If it is determined in S603 that there is a model image whose average luminance value meanV is equal to or greater than the threshold value ThreV, the process proceeds to S604, and after the model image is excluded, the process proceeds to S605. That is, in the example of FIG. 15 described above, an image (white image) read by Image1 is excluded from the four reading regions.

On the other hand, if it is determined in S603 that there is no model image whose average luminance value meanV is equal to or greater than the threshold ThreV, the process proceeds to S605 as it is.
In S605, a plurality of types of model images (group of model images) extracted as a result of the processing in S601 to S604 are regarded as one model image (an image connected to one), and each pixel constituting the model image Is converted into HSV parameters (H value: 0 to 360, S value and V value: 0 to 1). Note that the conversion from RGB to HSV parameters and the conversion from HSV to RGB can be performed according to the above-described known conversion formulas.

  Subsequently, in S606, based on the HSV parameter obtained by the conversion process in S605, a first feature amount that is a feature amount representing the feature of the model image (corresponding to the comprehensive first feature amount in each of the embodiments described above). A first feature amount specifying process for specifying is performed. The specific processing content of the first feature amount specifying processing is as described above (FIG. 6).

In S616, a thumbnail image of each model image extracted in S602 is generated and displayed on the display unit 32.
Subsequently, in S617, the model image of the thumbnail image selected by the operation on the operation unit 31 is set to be invalid. That is, the model image from which the thumbnail image is selected is excluded from the plurality of types of model images extracted in S602.

  Subsequently, in S618, the first feature amount is specified by the same method as in S605 and S606 based on a model image other than the model image set to invalid among a plurality of types of model images extracted in S602. Thereafter, the process returns to S611.

  As described above, in the multifunction machine 10 according to the third embodiment, the feature amount representing the feature of one entire image composed of the input plural types of model images is specified as the first feature amount. Therefore, it is possible to mix the colors of a plurality of types of model images by reflecting the difference in relative sizes of the model images. As a result, for example, when an image having a size larger than the document placement surface of the image reading unit 20 is to be used as a model image, the original image is temporarily reproduced once by dividing the image into a plurality of images. The color conversion process equivalent to the case where the image data is read by can be performed.

  Further, in this multifunction machine 10, since an image read in each reading area obtained by dividing the document placement surface into a plurality of parts is input as an independent model image, the user can easily read a plurality of model images. And the working time can be shortened.

  In addition, if the average luminance value of the image read in the reading area is greater than or equal to the threshold value ThreV, it is determined that no image exists in the reading area. Inputting as an image can be prevented.

  In the multifunction machine 10 of the third embodiment, the CPU 71 that executes the processes of S601 to S604 in the color conversion process (FIG. 14) corresponds to the first image input unit, and the CPU 71 that executes the processes of S605 and S606. This corresponds to the comprehensive first feature value specifying means. The CPU 71 that executes the process of S607 corresponds to a second image input unit, and the CPU 71 that executes the processes of S608, S612, and S613 corresponds to an output unit. Further, the CPU 71 that executes the processes of S609 and S610 corresponds to the second feature amount specifying means, and the CPU 71 that executes the processes of S611, S614 to S619 corresponds to the image correcting means.

[4. Other forms]
As mentioned above, although embodiment of this invention was described, it cannot be overemphasized that this invention can take a various form.

  (1) In the third embodiment, the example in which the multifunction device 10 inputs an image read in each reading area obtained by dividing the document placement surface into a plurality of reading images as an independent model image has been described. However, the present invention is not limited to this. is not. For example, the entire document placement surface may be read as one whole image, and the contents of the whole image may be analyzed to extract the image. In this way, a plurality of images can be extracted as independent images regardless of the position on the document placement surface. Further, one model image is not necessarily required for each document, and each image may be extracted as an independent model image from a document on which a plurality of images are printed.

  (2) In the third embodiment, an example in which a plurality of model images are read at once has been described. However, the present invention is not limited to this, and the model image reading operation is repeatedly performed as in the first embodiment. Thus, the model images may be read one by one. In this way, it is possible to reliably prevent one image from being misrecognized as a plurality of images, or conversely, misrecognizing a plurality of images as one image.

(3) In the first and second embodiments, the example in which the model images are read one by one by repeatedly performing the model image reading operation has been described. However, the present invention is not limited to this. Similarly to the embodiment, a plurality of model images may be read at once. (4) In the above-described embodiment, the example in which the multifunction machine 10 reads the model images with the image reading unit 20 has been described. However, the present invention is not limited to this. is not. For example, an image read from a memory card or an image received from the outside via the communication unit 60 may be used as a model image.

  (5) In the above embodiment, an example in which an image (original image) to be subjected to color conversion is input from a memory card has been described. However, the present invention is not limited to this. For example, an image read by the image reading unit 20 or an image received from the outside via the communication unit 60 may be the target of color conversion.

(6) In the above embodiment, in the color conversion process, before performing the process for specifying the first feature value and the second feature value, the process of converting the constituent pixels of the model image and the original image into HSV parameters. However, the present invention is not limited to this. For example, instead of the HSV parameter, it may be converted into another parameter such as an L ^* c ^* h ^* parameter or an RGB parameter.

  (7) In the above embodiment, in the color conversion process, the second feature value is specified after specifying the first feature value. However, the present invention is not limited to this, and the second feature value is specified after specifying the second feature value. One feature quantity may be specified.

  (8) In the above embodiment, the algorithm for specifying the first feature quantity and the algorithm for specifying the second feature quantity have been described as being the same. However, the present invention is not limited to this. Also good.

  (9) In the above embodiment, the process of stopping a part of the correction process or reducing the conversion amount according to the size of the region by the representative value resetting process is exemplified, but the present invention is not limited to this. However, the representative value resetting process may not be performed.

  (10) In the above embodiment, the corrected image image representing the color conversion processing result based on the model image is displayed on the display unit 32. However, the present invention is not limited to this. For example, instead of display, Simple printing (printing at a low resolution) may be performed.

  (11) In the above embodiment, the color conversion process is exemplified as the image correction process. However, the present invention is not limited to this, and an image correction process other than that exemplified in the above embodiment may be performed.

  (12) In the above embodiment, the H value of the region that is not subject to conversion can reduce the conversion amount, but the conversion amount cannot be zero. This is because, as shown in FIG. 8, linear interpolation is performed between the representative values of the regions that are not to be converted, and therefore, it is affected by the representative values of other regions.

  Therefore, a hue correction table as shown in FIG. 16 can be employed. FIG. 16 is a hue correction table in the case where only the B region is to be converted. In this figure, the number of areas to be converted is one, but the present invention can be similarly applied to cases where a plurality of areas are to be corrected.

  In FIG. 16, since the H values other than the B region are H ′ = H, color conversion is not performed. The H ′ value in the B region can be obtained by the following equation, where the minimum value in the B region is Hmin and the maximum value in the B region is Hmax.

When H <iH,
H ′ = Hmin + (sHb−Hmin) × (H−Hmin) ÷ (iHb−Hmin)
When H> iH,
H ′ = sHb + (Hmax−sHb) × (H−iHb) ÷ (Hmax−iHb)
By using this equation, only the region to be converted can be converted.

In this way, since only the H value to be converted can be converted, the effect of color conversion can be increased.
(13) In the above embodiment, since a correction curve (conversion formula) is independently used for each region with respect to the S value and the V value, a pseudo contour (tone jump) may be generated. That is, as shown in FIG. 17, each region has a table indicating the relationship between S and S ′, and no consideration is given to the nature of the table in the adjacent region.

On the other hand, as shown in FIG. 18, gradation skipping can be prevented by smoothing the correction curve in each color region.
Here, specific processing will be described below. The color conversion process for a part of the C area and a part of the B area will be described with reference to FIGS. 19 and 20, but the contents of the process are basically the same for the other areas.

  The corrected S value (Sb ″) includes the H value (H) of the conversion target area, the intermediate value (Hbmid) of the H value of the conversion target area, and the hue coordinate position of the H value of the pixel to be converted. The hue coordinate position of the intermediate value of the H value of the area to be converted is compared, the hue coordinate position of the H value of the pixel to be converted is close, and the intermediate value of the H value of the area to be converted The H value representative value (Hcmid) of the adjacent region far from the hue coordinate position is converted by a conversion formula corresponding to the conversion formula (Formula 2) (that is, using the saturation correction table of the B region). The S value (Sb ′) of the conversion target area (calculated) and adjacent converted by the conversion expression corresponding to the conversion expression (Expression 3) (that is, calculated using the saturation correction table of the C area) Using the S value (Sc ′) of the region, it can be obtained by the following equation.

Sb ″ = {(H−Hcmid) × Sb ′ + (Hbmid−H) × Sc ′}
÷ {(Hbmid−Hcmid)} (Expression 7)
The Hbmid and Hcmid are the “representative values” that have been reset.

  Further, the corrected V value (Vb ″) in this example includes the H value (H) of the conversion target region, the representative value (Hbmid) of the H value of the conversion target region, and the representative value of the H value of the adjacent region (Hbmid). Hcmid), the V value (Vb ′) of the conversion target area converted by the conversion expression corresponding to the conversion expression (Expression 4) (that is, calculated using the brightness correction table of the B area), and the conversion expression ( Using the S value (Vc ′) of the adjacent region converted by the conversion equation corresponding to Equation 5 (that is, calculated using the brightness correction table of the C region), the following equation can be used. .

Vb ″ = {(H−Hcmid) × Vb ′ + (Hbmid−H) × Vc ′}
÷ {(Hbmid−Hcmid)} (Equation 8)
The above-described processing is performed on part of the B region (H value range: 210 <H ≦ 240) and part of the C region (H value range 180 <H ≦ 210) shown in FIG. As a result, the output saturation value (S ″) and lightness value (V ″) are obtained by weighting calculation according to the input hue value (H), thereby smoothing the correction effect between the hues. Can do.

  (14) In the above embodiment, the multifunction machine 10 is exemplified as the image processing apparatus. However, the present invention is not limited to this, and may be an information processing apparatus such as a personal computer.

1 is a perspective view illustrating an appearance of a multifunction machine according to a first embodiment. FIG. 2 is a block diagram illustrating a schematic configuration of a control system of the multifunction machine. FIG. 6 is an explanatory diagram illustrating an outline of a user operation and a process of a multifunction peripheral in color conversion processing. It is a flowchart of the color conversion process of 1st Embodiment. It is explanatory drawing of the display screen of a correction | amendment image image. It is a flowchart of a 1st feature-value specific process. It is a flowchart of the 2nd feature-value specific process. It is explanatory drawing of a hue correction table. It is a flowchart of a representative value reset process. It is explanatory drawing of the hue correction table produced when only B area | region is made into conversion object. It is explanatory drawing of the identification method of the comprehensive feature-value in 1st Embodiment. It is a flowchart of the color conversion process of 2nd Embodiment. It is explanatory drawing of the identification method of the comprehensive feature-value in 2nd Embodiment. It is a flowchart of the color conversion process of 3rd Embodiment. It is explanatory drawing for demonstrating the recognition method of several example images. It is the figure which showed the hue correction table in a modification. It is the figure which showed the saturation correction table. It is the figure which showed the change of the curve of saturation correction | amendment. It is the figure which showed the saturation correction table in B area | region and C area | region. It is the figure which showed the corrected S value in B area | region and C area | region. It is the figure which showed that a part of B area | region and C area | region became conversion object.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... MFP, 11 ... Main body casing, 12 ... Opening, 20 ... Image reading part, 21 ... Document cover, 30 ... Operation panel, 31 ... Operation part, 32 ... Display part, 40 ... Image printing part, 50 ... Card slot , 60 ... communication section, 70 ... control section, 71 ... CPU, 72 ... ROM, 73 ... RAM, 80 ... signal line

Claims (10)

First image input means capable of inputting a plurality of types of first images;
First feature amount specifying means for specifying a first feature amount representing a feature of each first image input by the first image input means;
Based on each first feature amount specified by the first feature amount specifying means, a comprehensive first feature amount that includes a plurality of representative values representing features of the plurality of types of first images as a whole is specified. 1 feature quantity specifying means;
Second image input means for inputting a second image which is an image to be subjected to image correction;
Second feature amount specifying means for specifying a second feature amount including a plurality of representative values representing features of the second image input by the second image input means;
Image correction means for performing image correction processing on the second image based on a correspondence relationship between the plurality of representative values of the total first feature quantity and the plurality of representative values of the second feature quantity;
An image processing apparatus comprising: The first feature amount specifying unit specifies the first feature amount based on a pixel value of a constituent pixel of each first image input by the first image input unit,
The comprehensive first feature quantity specifying unit is a pixel of a constituent pixel of the first image used for specifying the first feature quantity based on each first feature quantity specified by the first feature quantity specifying unit. in weighting according to the number, the image processing apparatus according to claim 1, characterized in that identifying the overall first feature amount. Output means for outputting a corrected image representing an image correction processing result by the image correction means;
Print control means for printing the second image on which image correction processing has been performed by the image correction means;
The image processing apparatus according to claim 1 or claim 2, characterized in that it comprises a. On condition that the corrected image is output by the output means, the first image is input by the first image input means, and the image correction means is output by the input first image and the output means. The image processing apparatus according to claim 3 , wherein an image correction process is performed on the second image based on an overall first feature amount in the first image used for the image correction process of the corrected image. The image correction means is a predetermined one of a plurality of types of first images used for image correction processing of the correction image image output by the output means, on condition that the correction image image is output by the output means. from General first feature data based on the first image other than the first image, the image processing apparatus according to claim 3 or claim 4, characterized in that performing the image correction process for the second image. The correction images, the image processing apparatus according to any one of claims 3, characterized in that the reduced image of the second image is obtained performing the image correction process to claim 5 . The first image input means inputs an image optically read from a print medium set at a predetermined reading position, and when a plurality of images are read from the print medium set at the reading position The image processing apparatus according to any one of claims 1 to 6 , wherein each image can be input as an independent first image. The image processing apparatus according to claim 7 , wherein the first image input unit inputs an image read in each reading area obtained by dividing the reading position into a plurality of parts as an independent first image. The first image input unit, when the average luminance value of the image read by the reading area is above a predetermined value, to claim 8, wherein determining that the image is not present in the reading area The image processing apparatus described. First image input means capable of inputting a plurality of types of first images;
First feature amount specifying means for specifying a first feature amount representing a feature of each first image input by the first image input means;
Based on each first feature amount specified by the first feature amount specifying means, a comprehensive first feature amount that includes a plurality of representative values representing features of the plurality of types of first images as a whole is specified. 1 feature quantity specifying means;
Second image input means for inputting a second image which is an image to be subjected to image correction;
Second feature amount specifying means for specifying a second feature amount including a plurality of representative values representing features of the second image input by the second image input means;
A computer is caused to function as image correction means for performing image correction processing on the second image based on a correspondence relationship between the plurality of representative values of the total first feature amount and the plurality of representative values of the second feature amount. An image processing program.