JP2015154201A - Image processing system, control method for the same, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technology that enables an input image area to be accurately divided into a monochromatic area and a color area.SOLUTION: An image processing system of the present invention, which divides an input image area into a monochromatic area and a color area, includes storage means in which a plurality of division patterns different from one another are pre-recorded, and acquisition means for acquiring a determined pattern in which a monochromatic attribute or a color attribute is allocated to a divided area having the area divided according to one division pattern selected from among the plurality of division patterns, on the basis of the input image.

Description

  The present invention relates to an image processing apparatus, a control method for the image processing apparatus, and a program.

  Conventionally, there has been proposed an image processing apparatus that divides an input image area into a monochrome area (monochrome image area) and a color area (color image area), and separately applies gamma correction to the monochrome area and the color area of the input image. Yes. For example, γ correction (hereinafter referred to as “DICOM γ correction”) defined by DICOM (Digital Imaging and Communication in Medicine) Part 14 is applied to a monochrome area, and γ correction of γ = 2.2 (hereinafter referred to as “DICOM γ correction”). 2. Description of the Related Art An image processing apparatus that performs “2.2γ correction” has been proposed. Such an image processing apparatus may be mounted on, for example, a medical image display apparatus. By using such an image processing apparatus, when displaying a monochrome image such as an X-ray image and a color image such as an endoscopic image, DICOMγ correction can be performed on the monochrome image, and 2.2γ is applied to the color image. Corrections can be made. As a result, each image can be displayed with an appropriate gradation.

Techniques for dividing an input image area into a monochrome area and a color area are disclosed in Patent Documents 1 and 2, for example.
In the technique disclosed in Patent Document 1, it is determined for each pixel of the input image whether the area of the pixel is a monochrome area or a color area.
In the technique disclosed in Patent Document 2, the area of the input image is divided into a plurality of rectangular areas. Then, for each rectangular area, it is determined whether the rectangular area is a monochrome area or a color area.

  In another technical field, a technique has been proposed in which a plurality of candidates for a photographic area in a form are registered in advance, and for each registered candidate, it is determined whether or not the candidate is a photographic area (Patent Document 3).

JP 2001-169101 A JP 2003-244469 A JP-A-5-314312

  However, with the techniques disclosed in Patent Documents 1 and 2, the input image may not be accurately divided into a monochrome area and a color area.

Specifically, the color image may contain some monochrome pixels, or the monochrome image may contain some color pixels, but the color image area should be determined as a color area. The area of the monochrome image should be determined as a monochrome area. However, in the technique disclosed in Patent Document 1, since the determination is performed for each pixel, the monochrome pixel area of the color image is erroneously determined as the monochrome area, or the color pixel area of the monochrome image is erroneously determined as the color area. Sometimes.
In the technique disclosed in Patent Document 1, since the determination is performed for each pixel, a long processing time is required for the determination, and reflection of the determination result for the γ correction may be delayed.

In the technique disclosed in Patent Document 2, since the determination is performed for each rectangular area for a plurality of pixels, the frequency of the above-described erroneous determination can be reduced as compared with the case where the determination is performed for each pixel. However, even if the technique disclosed in Patent Document 2 is used, the above-described erroneous determination cannot be completely suppressed.
In the technique disclosed in Patent Document 2, when the boundary between the color image region and the monochrome image region is located in the rectangular region, an erroneous determination occurs in the rectangular region including the boundary. Specifically, the rectangular area including the boundary includes both a color image area and a monochrome image area, but the entire rectangular area is erroneously determined as a monochrome area or a color area. End up. That is, a part of the color image area is erroneously determined as a monochrome area, or a part of the monochrome image area is erroneously determined as a color area.

  By using the technique disclosed in Patent Document 3, it is possible to obtain a technique for registering a plurality of candidates for a monochrome area in advance and determining whether the candidate is a monochrome area for each registered candidate based on an input image. . However, even in the above technique obtained based on the technique disclosed in Patent Document 3, there are cases where the input image cannot be accurately divided into a monochrome area and a color area. Specifically, in the technique disclosed in Patent Document 3, all candidates that satisfy a predetermined condition are determined as photographic regions. Therefore, when a candidate satisfies a predetermined condition, even if the candidate is an area including a color image area, the area is erroneously determined as a monochrome area.

  An object of the present invention is to provide a technique capable of accurately classifying an input image area into a monochrome area and a color area.

The first aspect of the present invention is:
An image processing apparatus that divides an input image area into a monochrome area and a color area,
Storage means in which a plurality of different division patterns are recorded in advance;
Obtaining means for obtaining a fixed pattern in which a monochrome attribute or a color attribute is assigned to a divided region obtained by dividing the region according to one divided pattern selected from the plurality of divided patterns based on an input image;
An image processing apparatus comprising:

The second aspect of the present invention is:
A control method of an image processing apparatus for dividing an input image area into a monochrome area and a color area,
The image processing apparatus has storage means in which a plurality of different division patterns are recorded in advance,
The control method of the image processing apparatus is:
Obtaining an input image;
Obtaining a fixed pattern in which a monochrome attribute or a color attribute is assigned to a divided region obtained by dividing the region according to one divided pattern selected from the plurality of divided patterns based on the input image;
A control method for an image processing apparatus, comprising:

  According to a third aspect of the present invention, there is provided a program characterized by causing a computer to execute each step of the above-described image processing apparatus control method.

  According to the present invention, an input image area can be accurately divided into a monochrome area and a color area.

1 is a block diagram illustrating an example of a functional configuration of an image processing apparatus according to a first embodiment. Diagram showing an example of a division pattern Diagram showing an example of an input image The figure which shows an example of block attribute information Diagram showing an example of attribute pattern The figure which shows an example of the comparison result of block attribute information and attribute pattern information FIG. 9 is a block diagram illustrating an example of a functional configuration of an image processing apparatus according to a second embodiment. FIG. 9 is a block diagram illustrating an example of a functional configuration of an image processing apparatus according to a third embodiment. Diagram showing an example of an input image The figure which shows an example of a division | segmentation pattern and pattern boundary information FIG. 9 is a block diagram illustrating an example of a functional configuration of an image processing apparatus according to a fourth embodiment. FIG. 10 is a block diagram illustrating an example of a functional configuration of an image processing apparatus according to a fifth embodiment. Diagram showing an example of attribute pattern

<Example 1>
Hereinafter, an image processing apparatus and a control method thereof according to Embodiment 1 of the present invention will be described.

  FIG. 1 is a block diagram illustrating an example of a functional configuration of the image processing apparatus 10 according to the present embodiment. As shown in FIG. 1, the image processing apparatus 10 includes a monochrome count unit 101, a single color determination unit 102, a block attribute assignment unit 103, a divided pattern storage unit 104, an attribute pattern generation unit 105, an evaluation value calculation unit 106, and a confirmed pattern selection. Unit 107, γ correction unit 108, and the like. The image processing apparatus 10 acquires an input image (input image data) S1, divides the area of the input image S1 into a monochrome area and a color area, and individually performs image processing on the monochrome area and the color area of the input image S1, An image (image data) S2 after image processing is output. The image S2 is input to an external device such as a display device, for example. In the present embodiment, the images S1 and S2 have a size of 1920 pixels in the horizontal direction × 1200 pixels in the vertical direction. The pixel values of the images S1 and S2 are RGB values (combinations of R value, G value, and B value), and the R value, G value, and B value are 8-bit values.

  In this embodiment, as a result of dividing the monochrome area and the color area, a fixed pattern in which the monochrome attribute or the color attribute is assigned to the divided area obtained by dividing the area is acquired. An area to which the monochrome attribute is assigned is a monochrome area, and an area to which the color attribute is assigned is a color area.

In this embodiment, an example in which the image processing apparatus 10 outputs the image S2 to the outside will be described, but the present invention is not limited to this. For example, the image processing apparatus 10 may output the result of dividing the monochrome area and the color area to the outside. Then, image processing based on the classification result may be executed in a device separate from the image processing device 10. Further, the image processing apparatus 10 may have a display panel. Then, the image processing apparatus 10 may display the image S2 instead of outputting the image S2 to the outside. That is, the image processing apparatus 10 may be a display device. The display panel is, for example, a liquid crystal display panel, a plasma display panel, an organic EL display panel, or the like.
In this embodiment, an example in which the image processing is γ correction will be described, but the image processing is not limited to γ correction. For example, the image processing may be a brightness adjustment process for a monochrome area and a brightness adjustment process for a color area. The image processing may be a color temperature adjustment process for a monochrome area and a color temperature adjustment process for a color area.
Note that the image size may be larger or smaller than 1920 pixels × 1200 pixels. For example, the image size is 1680 pixels × 1050 pixels, 2560 pixels × 1600 pixels, 2
It may be 880 pixels × 1800 pixels, 3840 pixels × 2400 pixels, and the like. The pixel value may not be an RGB value. For example, the pixel value may be a YCbCr value (a combination of a Y value, a Cb value, and a Cr value). The number of bits may be more or less than eight. For example, the number of bits may be 4, 16, etc.

The monochrome count unit 101 counts the number of monochrome pixels of the input image S1 in each block area for each of a plurality of block areas constituting the area of the input image S1. Then, the monochrome count unit 101 outputs monochrome information cnt indicating the total number of monochrome pixels (number of monochrome pixels) for each block area.
In the present embodiment, a total of 60 block areas, which are 10 in the horizontal direction and 6 in the vertical direction, constituting the area of the input image S1 are set. The size of the block area is 192 (= 1920 ÷ 10) pixels in the horizontal direction × 200 (= 1200/6) pixels in the vertical direction. The monochrome count unit 101 counts the number of monochrome pixels for each of the 60 block areas. Then, the monochrome count unit 101 outputs monochrome information cnt indicating the number of 60 monochrome pixels corresponding to the 60 block areas.

In the present embodiment, an example in which the block area is a rectangular area will be described, but the block area may have any shape. For example, the block area may be an n-gon (n is an integer of 3 or more) area, a circular area, an area surrounded by a line connecting a straight line and a curve, and the like.
In this embodiment, pixels with R value = G value = B value are detected as monochrome pixels, but the detection method of monochrome pixels is not limited to this. For example, RGB values may be converted into color difference values such as CbCr values, and pixels with a color difference value of 0 may be detected as monochrome pixels.
The number of block areas may be more or less than 60. Further, the size of the block area may be larger or smaller than 192 pixels × 200 pixels.

  The monochrome determination unit 102 determines, for each block area, whether or not the block area is a monochrome area in which the input image S1 in the block area is a monochrome image (monochromatic image) (monochrome determination process). Then, the single color determination unit 102 outputs single color information flt indicating the result of the single color determination process. In this embodiment, it is determined that a block area in which all 38400 pixels in the block area have the same pixel value is a monochrome area, and other block areas are not a monochrome area.

Based on the input image S1, the block attribute assignment unit 103 assigns one of a plurality of attributes including a monochrome attribute and a color attribute to each block area. Specifically, the block attribute assignment unit 103 assigns an attribute to each block area using the monochrome information cnt and the single color information flt. Then, the block attribute assignment unit 103 outputs block attribute information mc indicating an attribute for each block area. In this embodiment, a monochrome attribute is assigned to a block area that is determined to be a monochrome area. A monochrome attribute is assigned to a block area that is determined not to be a monochrome area and has a monochrome ratio equal to or greater than a threshold value. Then, a color attribute is assigned to a block area that is determined not to be a monochrome area and that has a monochrome ratio that is less than the threshold value. In this embodiment, the monochrome ratio is the ratio of the total number of monochrome pixels in the block area to the total number of pixels in the block area. In this embodiment, 95% lower than 100% is used as the threshold value. Thereby, when there are some color pixels (for example, color pixels constituting a color annotation character) in a block area which is a monochrome image area, a monochrome attribute can be assigned to the block area. In this embodiment, since the size of the block area is 192 pixels × 200 pixels, it is determined that the block area is not a monochrome area, and monochrome is added to the block area having 3648 (= 192 × 200 × 0.95) or more monochrome pixels. Attributes are assigned. Then, it is determined that the area is not a monochrome area, and a color attribute is assigned to a block area where the number of monochrome pixels is less than 3648.

Note that the monochrome attribute or the color attribute may be assigned to each block area using the monochrome information cnt without using the monochrome information flt.
Note that a color rate may be used instead of the monochrome rate. The color rate is a ratio of the total number of color pixels (pixels other than monochrome pixels) in the block area to the total number of pixels in the block area.
Note that the monochrome ratio threshold value may or may not be a fixed value determined in advance by the manufacturer or the like. For example, the threshold of the monochrome ratio may be a value that can be changed by the user, or may be a value determined according to the type of image.

In the division pattern storage unit 104, a plurality of different division patterns are recorded in advance. In the present embodiment, the division pattern storage unit 104 records area information indicating a plurality of divided areas obtained by dividing an area according to the divided pattern as a divided pattern. Specifically, the start point coordinates and end point coordinates of the divided area are recorded as the area information. The divided areas are color area and monochrome area candidates. In the present embodiment, the size of the input image S1 is equal to the screen size, and the division pattern indicates a plurality of divided areas obtained by dividing the area of the input image S1. Specifically, the division pattern indicates a division area composed of one or more block areas. In this embodiment, the division pattern does not indicate an attribute. That is, in the present embodiment, the division pattern indicates a non-attribute area to which no attribute is assigned as a division area.
The division pattern storage unit 104 outputs division pattern information ptn indicating all recorded division patterns.

An example of the division pattern is shown in FIG. In this embodiment, it is assumed that the division patterns 1 to 4 shown in FIG. Divided pattern 2 shows divided areas 1 and 2. Therefore, for the divided pattern 2, the start point coordinates (horizontal position, vertical position) of the divided area 1 (xs1, ys1) and the end point coordinates (xe1, ye1), and the start point coordinates (xs2, ys2) of the divided area 2 and End point coordinates (xe2, ye2) are recorded. Similarly, for the divided patterns 1, 3, and 4, the start point coordinates and end point coordinates of the divided areas are recorded.
As shown in FIG. 2, in this embodiment, a non-divided pattern that does not divide the area is also recorded as the divided pattern 1.

The non-divided pattern may not be recorded.
The divided pattern recorded in the divided pattern storage unit 104 may be a fixed pattern recorded in advance by a manufacturer or the like, or may be a pattern that can be added or changed by the user. For example, as illustrated in FIG. 1, registered pattern information in_ptn indicating a division pattern may be input to the division pattern storage unit 104 in accordance with a user operation. Then, the division pattern indicated by the registered pattern information in_ptn may be added to the division pattern recorded in the division pattern storage unit 104. Further, one division pattern recorded in the division pattern storage unit 104 may be replaced with a division pattern indicated by the registered pattern information in_ptn.
In the present embodiment, an example in which the divided area is an area composed of one or more block areas will be described. However, the divided area may not be a block area unit area. For example, a plurality of divided regions may be set so that the boundary between the divided regions is located in the block region.
In the present embodiment, an example is described in which the size of a region composed of a plurality of divided regions indicated by the divided pattern is the same as the size of the region of the input image S1, but these sizes may be different. . The size of the region composed of a plurality of divided regions indicated by the divided pattern is the input image S
It may be larger or smaller than the size of one region.
The area information may be any information as long as it is information indicating a divided area. For example, as the area information, the position (start point coordinates, end point coordinates, center coordinates, etc.) and size of the divided area may be recorded.
FIG. 2 shows an example in which the divided area is a rectangular area, but the divided area may have any shape. For example, the divided region may be an n-gon (n is an integer of 3 or more) region, a circular region, a region surrounded by a line connecting a straight line and a curve, and the like.

  The attribute pattern generation unit 105, the evaluation value calculation unit 106, and the confirmation pattern selection unit 107 acquire a confirmed pattern based on the input image S1. In this embodiment, a pattern in which a monochrome attribute or a color attribute is assigned to a divided area obtained by dividing an area according to one divided pattern selected from a plurality of divided patterns is acquired as a deterministic pattern. That is, a pattern in which a monochrome attribute or a color attribute is assigned to a divided area indicated by one divided pattern selected from a plurality of divided patterns is acquired as a definite pattern.

  The attribute pattern generation unit 105 performs processing for generating an attribute pattern by assigning an attribute to a non-attribute area (divided area) indicated by the division pattern for each of the plurality of division patterns recorded in the division pattern storage unit 104. . Then, the attribute pattern generation unit 105 outputs attribute pattern information dptn indicating the plurality of generated attribute patterns. In this embodiment, since four division patterns are recorded in advance, four attribute patterns are generated, and attribute pattern information dptn indicating the four attribute patterns is output. In the attribute pattern generation unit 105, a monochrome attribute or a color attribute is assigned to the non-attribute area. The attribute pattern is a pattern indicating an attribute area to which a monochrome attribute or a color attribute is assigned. In the present embodiment, the attribute pattern generation unit 105 determines a plurality of division patterns recorded in the division pattern storage unit 104 from the division pattern information ptn. Then, the attribute pattern generation unit 105 assigns an attribute to each non-attribute area using the block attribute information mc. Specifically, the color attribute is assigned to the non-attribute area including the block area to which the color attribute is assigned, and the monochrome attribute is assigned to the non-attribute area not including the block area to which the color attribute is assigned. That is, in the present embodiment, attributes are assigned to non-attribute areas in consideration of only block areas to which color attributes are assigned.

In addition, when the size of the area composed of a plurality of divided areas indicated by the divided pattern is different from the size of the area of the input image S1, it is included in the attribute-free area after performing a scaling process for matching the sizes. A block area is determined.
The “block area included in the attribute-free area” may be a block area that is entirely included in the attribute-free area, or a block area that is at least partially included in the attribute-free area. Good.
Note that the method of assigning attributes to the no-attribute area is not limited to the above method. For example, the monochrome attribute may be assigned to the non-attribute area including the block area to which the monochrome attribute is assigned, and the color attribute may be assigned to the non-attribute area not including the block area to which the monochrome attribute is assigned. A monochrome attribute may be assigned to an attributeless area where the total number of monochrome block areas included in the attributeless area is larger than the total number of color block areas included in the attributeless area. A color attribute may be assigned to an attribute-free area where the total number of color block areas included in the attribute-free area is larger than the total number of monochrome block areas included in the attribute-free area. The color block area is a block area to which a color attribute is assigned, and the monochrome block area is a block area to which a monochrome attribute is assigned. Further, the monochrome ratio of the attribute-free area may be calculated using the input image S1. An attribute based on the monochrome ratio may be assigned to the attribute-free area.

The evaluation value calculation unit 106 calculates an evaluation value for each of the plurality of attribute patterns generated by the attribute pattern generation unit 105, and outputs evaluation information dptn_vl indicating the evaluation value for each attribute pattern. In this embodiment, since four attribute patterns are generated, four evaluation values corresponding to the four attribute patterns are calculated, and evaluation information dptn_vl indicating the four evaluation values is output. The evaluation value represents the degree of coincidence between the attribute assigned to the block area and the attribute assigned to the attribute area. In this embodiment, the evaluation value calculation unit 106 determines an attribute pattern from the attribute pattern information dptn. Then, the evaluation value calculation unit 106 calculates an evaluation value for each attribute pattern using the block attribute information mc. Specifically, the evaluation value is calculated using the following formula 1. In Equation 1, “number of matching blocks” is the total number of block areas in which the attribute assigned to the block area matches the attribute assigned to the attribute area including the block area. “Total number of blocks” is the total number of block areas.

Evaluation value = number of matching blocks / total number of blocks × 100 (%) (Formula 1)

  The confirmed pattern selection unit 107 selects one of the plurality of attribute patterns generated by the attribute pattern generation unit 105 as a confirmed pattern. Then, the confirmed pattern selection unit 107 outputs confirmed pattern information dptn_s indicating the confirmed pattern. In this embodiment, the confirmed pattern selection unit 107 determines an attribute pattern from the attribute pattern information dptn, and determines an evaluation value of the attribute pattern from the evaluation information dptn_vl. Then, the confirmed pattern selection unit 107 selects the attribute pattern having the highest evaluation value among the plurality of attribute patterns generated by the attribute pattern generation unit 105 as the confirmed pattern. In other words, in the present embodiment, the attribute pattern having the largest number of block areas in which the attribute assigned to the block area matches the attribute assigned to the attribute area including the block area is selected as the confirmed pattern. When there are a plurality of attribute patterns having the highest evaluation value, one of the plurality of attribute patterns having the highest evaluation value is selected as the confirmed pattern. For example, one of a plurality of attribute patterns having the highest evaluation value is selected at random.

  Note that priority may be added to a plurality of division patterns recorded in the division pattern storage unit 104. If there are a plurality of attribute patterns with the highest evaluation value, the attribute pattern generated from the division pattern with the highest priority among the plurality of attribute patterns with the highest evaluation value is selected as the confirmed pattern. Good.

  The γ correction unit 108 performs γ correction on the input image S1 in accordance with the confirmed pattern indicated by the confirmed pattern information dptn_s. Then, the γ correction unit 108 outputs the image S2 after γ correction. For example, the γ correction unit 108 performs 2.2 γ correction on the color area (attribute area to which the color attribute is assigned) of the input image S1, and the monochrome area (attribute area to which the monochrome attribute is assigned) of the input image S1. DICOMγ correction is applied to The 2.2γ correction is a γ correction of γ = 2.2. The DICOM γ correction is a γ correction defined by DICOM (Digital Imaging and Communication in Medicine) Part14.

  A specific example of the operation of the image processing apparatus 10 will be described.

An example of the input image S1 is shown in FIG. The input image S1 shown in FIG. 3 is an image in which an X-ray picture 301 is arranged on the left side and an endoscopic image 302 is arranged on the right side. The X-ray picture 301 is a monochrome image, and all the pixels of the X-ray picture 301 are monochrome pixels (R value = G value = B value). The endoscopic image 302 is a color image, but includes monochrome pixels. Specifically, the region of the endoscopic image 302 includes an imaging region 303 indicating an image photographed by the endoscope, and background regions 304 and 305 of black single color (R value = G value = B value = 0). , Consisting of. Of the imaging area 303, the upper area is an achromatic area 306, and the lower area is a colored area 307. The achromatic region 306 is a region that includes many monochrome pixels, and the chromatic region is a region that shows a colored image. The achromatic region 306 is, for example, a region where light is not so much irradiated at the time of photographing, a region where light is excessively irradiated, or the like. Since the same image processing should be applied to the achromatic region 306 and the chromatic region 307, the same attribute should be assigned to the achromatic region 306 and the chromatic region 307.

  When the input image S1 is the image shown in FIG. 3, the block attribute information mc shown in FIG. 4 is acquired. That is, the monochrome attribute is assigned to the block area included in the X-ray photograph 301 area and the block area included in the achromatic area 306. A color attribute is assigned to the block area included in the chromatic area 307, and a single color attribute is assigned to the background areas 304 and 305.

  When the block attribute information mc shown in FIG. 4 is acquired and the division pattern information ptn indicates the division patterns 1 to 4 in FIG. 2, the attribute pattern information dptn indicating the attribute patterns 1 to 4 in FIG. 5 is acquired. Since the divided area 1 of the divided pattern 1 includes the block area to which the color attribute is assigned, the attribute pattern 1 to which the color attribute is assigned to the divided area 1 of the divided pattern 1 is generated. The divided area 1 of the divided pattern 2 does not include the block area to which the color attribute is assigned, and the divided area 2 of the divided pattern 2 includes the block area to which the color attribute is assigned. Therefore, a monochrome attribute is assigned to the divided area 1 of the divided pattern 2, and an attribute pattern 2 is generated in which the color attribute is assigned to the divided area 2 of the divided pattern 2. Similarly, the attribute pattern 3 is divided from the divided patterns 3 and 4. , 4 are generated.

The evaluation value calculation unit 106 compares the block attribute information mc and the attribute pattern information dptn, and calculates an evaluation value of the attribute pattern indicated by the attribute pattern information dptn based on the comparison result. When the block attribute information mc shown in FIG. 4 and the attribute pattern information dptn showing the attribute patterns 1 to 4 in FIG. 5 are acquired, the comparison result shown in FIG. 6 is obtained. In FIG. 6, a block area indicated by a circle is a block area (hereinafter, referred to as a matching block area) to which the same attribute as the attribute area including the block area is assigned. The block area in which the cross mark is described is a block area to which an attribute different from the attribute area including the block area is assigned (hereinafter referred to as a mismatched block area). Then, using the comparison result of FIG. 6 and Expression 1, the following values are calculated as the evaluation values of the attribute patterns 1 to 4. The evaluation value calculation unit 106 outputs evaluation information dptn_vl indicating the evaluation values of the attribute patterns 1 to 4.

Evaluation value of attribute pattern 1 = 15/60 × 100 = 25%
Evaluation value of attribute pattern 2 = 45/60 × 100 = 75%
Evaluation value of attribute pattern 3 = 27/60 × 100 = 45%
Evaluation value of attribute pattern 4 = 45/60 × 100 = 75%

In the example of FIG. 6, since the block area to which the single color attribute is assigned is not used for comparison, no mark is described in the block area. However, a block area to which a single color attribute is assigned may be used for comparison, and the block area may be determined as a mismatched block area.

The evaluation value will be described in detail.
As shown in FIG. 3, generally, the achromatic region 306 and the chromatic region 307, which are regions of the same color image, are present at close positions. Therefore, in the division pattern corresponding to the input image S1, the achromatic region 306 and the chromatic region 307, which are the same color image region, are included in the same division region. In the present embodiment, the achromatic region 306 and the chromatic region 307 are included in the divided region 2 of the divided pattern 2. The attribute pattern generation unit 105 assigns a color attribute to a divided area including one or more block areas to which a color attribute is assigned. Therefore, a color attribute is assigned to the divided area 2 of the divided pattern 2. That is, even if the region of the endoscopic image 302 includes the achromatic region 306, the color attribute is assigned to the divided region 2 (the divided region 2 of the divided pattern 2) corresponding to the region of the endoscopic image 302. It is done. Further, since there is no block area to which the color attribute is assigned in the divided area 1 of the divided pattern 2, the monochrome attribute is assigned. The evaluation value of the division pattern 2 is 75% because only the block area included in the achromatic color area 306 of the color image is determined as the mismatched block area.
On the other hand, in the attribute patterns 1 and 3, the color attribute is assigned to the attribute region including the region of the radiograph 301. Therefore, more block areas than the attribute pattern 2 are determined as mismatched block areas, and a value lower than the evaluation value of the attribute pattern 2 is calculated as an evaluation value. The attribute pattern 4 has the same attribute area distribution as the attribute pattern 2 (a distribution in which a monochrome attribute is assigned to the left half area and a color attribute is assigned to the right half area). Therefore, the same value as the evaluation value of attribute pattern 2 is calculated as the evaluation value.

The confirmed pattern selection unit 107 selects the attribute pattern having the highest evaluation value indicated by the evaluation information dptn_vl as the confirmed pattern. Then, the confirmed pattern selection unit 107 outputs confirmed pattern information dptn_s indicating the selected confirmed pattern. When the block attribute information mc shown in FIG. 4 and the attribute pattern information dptn indicating the attribute patterns 1 to 4 in FIG. 5 are acquired, the evaluation value of the attribute pattern 2 (pattern having a registration number of 2) and the attribute pattern 4 ( The evaluation value of the pattern whose registration number is 4) is the highest. Therefore, the attribute pattern 2 having a small registration number is selected as the confirmed pattern. The registration number is a number corresponding to the priority, and is smaller as the priority is higher.
Note that the attribute pattern 4 may be selected as the final pattern without considering the priority.

  As described above, according to the present embodiment, a division pattern indicating a candidate for arrangement of a monochrome image or a color image is prepared in advance. Then, based on the input image, a fixed pattern in which a monochrome attribute or a color attribute is assigned to the divided area obtained by dividing the area according to one divided pattern selected from the divided patterns prepared in advance is acquired. Thereby, the area of the input image can be accurately divided into a monochrome area and a color area. Specifically, it is possible to prevent a color image area from being erroneously determined as a monochrome area or a monochrome image area from being erroneously determined as a color area. In particular, in the medical field, medical images are often arranged at fixed positions, and it is easy to assume candidates for arrangement of monochrome images or color images, so that it is suitable for use with the image processing apparatus according to the present embodiment.

Further, according to the present embodiment, since determination for each pixel (determination of whether or not it is a monochrome area) is not performed, the area of the input image can be changed to a monochrome area and a color in a shorter time than when determination is performed for each pixel. It can be divided into areas.
Further, according to the present embodiment, a pattern indicating an attribute-free area is prepared in advance as a division pattern. According to such a configuration, since it is not necessary to prepare patterns having only different attributes, the number of division patterns to be prepared can be reduced, and the storage capacity of the division pattern storage unit 104 can be reduced.

  Further, according to the present embodiment, the attribute pattern having the largest number of block areas in which the attribute assigned to the block area matches the attribute assigned to the attribute area including the block area is selected as the confirmed pattern. Thereby, a definite pattern in which appropriate attributes are assigned to the monochrome pixel area of the color image and the color pixel area of the monochrome image can be obtained more reliably. Specifically, a definite pattern in which a color attribute is assigned to a monochrome pixel area of a color image and a monochrome attribute is assigned to a color pixel area of a monochrome image can be obtained more reliably.

Further, according to the present embodiment, since a division pattern prepared in advance is used, it is possible to suppress erroneous detection of the boundary between the color image and the monochrome image. Specifically, in the configuration in which only the determination for each block area is performed, the boundary between the color image and the monochrome image is erroneously detected when the boundary between the color image and the monochrome image exists in the block area. According to the present embodiment, it is possible to prepare in advance a division pattern indicating a division area that is not a block area unit area. For example, it is possible to prepare in advance a division pattern indicating a plurality of divided regions such that the boundary between the divided regions is located in the block region. For this reason, the boundary between the color image and the monochrome image can be detected with higher accuracy than the configuration in which only the determination for each block region is performed, and the input image region can be accurately divided into the monochrome region and the color region.
When assigning attributes to attribute-free areas, consider attributes of only block areas that are entirely included in the attribute-free area without considering attributes of block areas that are at least partially included in the attribute-free area. Thus, the erroneous detection of the boundary can be further suppressed.

<Example 2>
Hereinafter, an image processing apparatus and a control method thereof according to Embodiment 2 of the present invention will be described. In the first embodiment, an example has been described in which a monochrome attribute is erroneously assigned to a block area including monochrome pixels of a color image, or a color attribute is erroneously assigned to a block area including color pixels of a monochrome image. In this embodiment, when such an erroneous assignment is not performed, the boundary between the color image and the monochrome image can be detected with high accuracy, and the area of the input image is accurately divided into the monochrome area and the color area. A configuration that can be performed will be described.

  FIG. 7 is a block diagram illustrating an example of a functional configuration of the image processing apparatus 70 according to the present embodiment. In addition, the same code | symbol is attached | subjected to the same function part as Example 1 (FIG. 1), and the description is abbreviate | omitted.

The division pattern selection unit 701 selects one of a plurality of division patterns indicated by the division pattern information ptn. Then, the division pattern selection unit 701 outputs selection pattern information ptn_s indicating the selected division pattern.
In this embodiment, the division pattern is selected based on the division pattern information ptn and the block attribute information mc. Specifically, a division pattern is selected in which the attributes of all the block areas that are entirely included in the non-attribute area (division area) are the same. Then, a division pattern indicating at least one attributeless area including the entire color block area and the entire monochrome block area is not selected. When there are a plurality of division patterns that satisfy the condition that “the attributes of all block areas that are entirely included in the non-attribute area are equal to each other”, one of the plurality of division patterns that satisfy the condition is selected. . For example, a division pattern with the highest priority is selected from among a plurality of division patterns that satisfy the above conditions. Further, when there is no division pattern that satisfies the above conditions, a non-division pattern is selected.

  The confirmed pattern generation unit 702 determines the division pattern selected by the division pattern selection unit 701 from the selection pattern information ptn_s. Then, the fixed pattern generation unit 702 generates a fixed pattern by assigning a monochrome attribute or a color attribute to the no-attribute area indicated by the division pattern selected by the division pattern selection unit 701 based on the input image S1. Thereafter, the confirmed pattern generation unit 702 outputs confirmed pattern information dptn_s indicating the generated confirmed pattern. In the present embodiment, similarly to the attribute pattern generation unit 105 of the first embodiment, attributes are assigned to non-attribute areas using the block attribute information mc.

As described above, according to the present embodiment, by assigning attributes to the attribute-free area indicated by the division pattern that satisfies the condition that “the attributes of all block areas that are entirely included in the attribute-free area are equal to each other” In addition, a definite pattern is generated. Thereby, the area of the input image can be accurately divided into a monochrome area and a color area. Specifically, the boundary between the color image and the monochrome image can be detected with high accuracy, and the area of the input image can be accurately divided into the monochrome area and the color area.
According to the present embodiment, the non-divided pattern is selected when there is no divided pattern that satisfies the above conditions. Thereby, one of the monochrome attribute and the color attribute is assigned to the entire area of the input image, and deterioration of the image quality of the color image or the monochrome image can be suppressed. Specifically, image quality degradation due to the assignment of both monochrome and color attributes to a color image area and image quality degradation due to the assignment of both monochrome and color attributes to a monochrome image area are suppressed. can do.

  Note that the pattern selection method in this embodiment may be applied to the first embodiment. Specifically, in the confirmed pattern selection unit 107 according to the first embodiment, attribute patterns having the same attributes in all block regions that are entirely included in the attribute region may be selected as the confirmed pattern. Then, when there is no attribute pattern in which all the block areas that are entirely included in the attribute area have the same attributes, the attribute pattern generated from the non-divided pattern may be selected as the confirmed pattern.

<Example 3>
Hereinafter, an image processing apparatus and a control method thereof according to Embodiment 3 of the present invention will be described. In the present embodiment, the division pattern selection method is different from that in the second embodiment. In this embodiment, the attribute assignment method for the non-attribute area (divided area) is different from those in the first and second embodiments.

  FIG. 8 is a block diagram illustrating an example of a functional configuration of the image processing apparatus 80 according to the present embodiment. In addition, the same code | symbol is attached | subjected to the same function part as Example 1, 2 (FIG. 1, 7), and the description is abbreviate | omitted.

  The boundary detection unit 801 detects a boundary between a plurality of images when a plurality of images are arranged in the input image S1. For example, an image in which a plurality of medical images are arranged may include a boundary image arranged between the medical images. Specifically, as shown in FIG. 9, the input image may include a boundary image 901 between the X-ray photograph 301 and the endoscopic image 302. For example, the boundary detection unit 801 detects the boundary image. The boundary detection unit 801 outputs detection boundary information frm indicating the detection result of the boundary. In this embodiment, a bitmap image (bitmap image data) having a 1-bit pixel value is output as the detection boundary information frm. Specifically, as the detection boundary information frm, an image in which the pixel value of the boundary portion between the boundary image and another image is 1 and the pixel value of the other portion is 0 is output. The detection boundary information frm is generated, for example, by calculating the luminance value of each pixel of the input image S1 and differentiating the luminance value of each pixel in units of pixels. Specifically, a pixel value 1 is assigned to a pixel in which at least one of the absolute value of the vertical differential result and the absolute value of the horizontal differential result is greater than or equal to a threshold value, and both of the two absolute values are threshold values. A pixel value of 0 is assigned to pixels that are less than. Thereby, the detection boundary information frm is generated.

In the divided pattern / boundary storage unit 802, a plurality of divided patterns are recorded in advance as in the divided pattern storage unit 104 of the first and second embodiments. Furthermore, a plurality of pattern boundary information corresponding to a plurality of divided patterns is recorded in advance in the divided pattern / boundary storage unit 802. The pattern boundary information is information indicating a boundary between divided regions indicated by the divided pattern. For example, the pattern boundary information includes position information indicating the position of the boundary and direction information indicating the direction. The position information is, for example, start point coordinates and end point coordinates of a rectangular area including a boundary. The direction information is information indicating which direction the boundary is along. In this embodiment, the direction information is “horizontal” or “vertical”.
The division pattern / boundary storage unit 802 outputs division pattern information ptn indicating all recorded division patterns and a pattern boundary information group ptn_frm indicating all recorded pattern boundary information.

An example of the division pattern according to the present embodiment is shown in FIG. In the present embodiment, four pattern boundary information corresponding to the four divided patterns 1 to 4 shown in FIG. 10 are recorded in advance.
Since there is no boundary for the divided pattern 1, “no boundary” is recorded as the pattern boundary information.
As for the divided pattern 2, since there is one boundary along the vertical direction, pattern boundary information indicating the boundary is recorded in advance. Specifically, the start point coordinates (fxs1, fys1) and end point coordinates (fxe1, fy1) of the rectangular area including the boundary are recorded as position information, and “vertical” is recorded as direction information.
For the divided pattern 3, since there are two boundaries along the vertical direction, pattern boundary information indicating the two boundaries is recorded in advance. Specifically, for the first boundary, start point coordinates (fxs1, fys1) and end point coordinates (fxe1, fy1) are recorded as position information, and “vertical direction” is recorded as direction information. For the second boundary, start point coordinates (fxs2, fys2) and end point coordinates (fxe2, fy2) are recorded as position information, and “vertical direction” is recorded as direction information.
As for the divided pattern 4, since there are two boundaries, a boundary along the horizontal direction and a boundary along the vertical direction, pattern boundary information indicating the two boundaries is recorded in advance. Regarding the boundary along the horizontal direction, the start point coordinates (fxs1, fys1) and the end point coordinates (fxe1, fye1) are recorded as position information, and the “horizontal direction” is recorded as direction information. For the boundary along the vertical direction, start point coordinates (fxs2, fys2) and end point coordinates (fxe2, fye2) are recorded as position information, and “vertical direction” is recorded as direction information.

The pattern boundary information is not limited to the above information. For example, information indicating a point or a line may be used as position information instead of information indicating a region.
It is preferable that the recorded pattern boundary information is also updated in accordance with the update of the recorded divided pattern. For example, as illustrated in FIG. 8, registered pattern boundary information in_ptn_frm, which is pattern boundary information corresponding to the registered pattern information in_ptn, may be input to the divided pattern / boundary storage unit 802 in accordance with a user operation. Then, the recorded division pattern and pattern boundary information may be updated using the registered pattern information in_ptn and the registered pattern boundary information in_ptn_frm.

The division pattern selection unit 803 selects one of a plurality of division patterns indicated by the division pattern information ptn. Then, the division pattern selection unit 803 outputs selection pattern information ptn_s indicating the selected division pattern.
In this embodiment, a plurality of boundary patterns are determined from the division pattern ptn. Then, a division pattern is selected based on the detection boundary information frm and the pattern boundary information group ptn_frm. Specifically, when a plurality of images are arranged in the input image, a division pattern that indicates a plurality of divided areas and whose boundaries coincide with the boundaries of the plurality of images is selected.

A specific example of the processing of the division pattern selection unit 803 will be described. Below, the process with respect to the division | segmentation pattern 2 of FIG. 10 is demonstrated.
First, the position information of the division pattern 2 in FIG. 10 is the start point coordinates (fxs1, fys1) and the end point coordinates (fxe1, fy1), and the direction information is “vertical direction”. Therefore, it is determined whether or not one or more pixels (boundary pixels) having a pixel value of the detection boundary information frm of 1 exist in the horizontal line region from the coordinates (fxs1, fys1) to the coordinates (fxe1, fys1). . The horizontal line region is a region made up of a plurality of pixels having the same vertical position.
Next, the horizontal line area to be determined is shifted by one pixel in the vertical direction. That is, coordinates (
The determination target is switched from the horizontal line area of coordinates (fxs1, fys1) to the horizontal line area of coordinates (fxe1, fys1 + 1) from fxs1, fys1). Then, it is determined whether or not one or more boundary pixels exist in the horizontal line area from the coordinates (fxs1, fys1 + 1) to the coordinates (fxe1, fys1 + 1).
Then, the horizontal line area to be determined is shifted by one pixel in the vertical direction from the coordinates (fxs1, fye1) to the horizontal line area of the coordinates (fxe1, fye1), and the determination is repeated. Then, when boundary pixels are included in all horizontal line regions, it is determined that “the boundaries of the plurality of divided regions match the boundaries of the plurality of images”. If there is a horizontal line that does not include the boundary pixel, it is determined that “the boundaries of the plurality of divided regions do not coincide with the boundaries of the plurality of images”.

When the direction information is “horizontal direction”, it is determined whether or not a boundary pixel exists for each vertical line area of the area indicated by the position information. Specifically, the vertical line area to be determined is shifted by one pixel in the horizontal direction from the leftmost vertical line area to the rightmost vertical line area, and the determination is repeated. Then, when boundary pixels are included in all the vertical line regions, it is determined that “the boundaries of the plurality of divided regions match the boundaries of the plurality of images”. If there is a vertical line that does not include the boundary pixel, it is determined that “the boundaries of the plurality of divided regions do not coincide with the boundaries of the plurality of images”. The vertical line region is a region made up of a plurality of pixels having the same horizontal position.
When the pattern boundary information indicates a plurality of boundaries as in the divided patterns 3 and 4 in FIG. 10, it is determined whether or not “the boundaries of the plurality of divided areas match the boundaries of the plurality of images” as follows. Is done.
That is, only when it is determined that “all boundary regions are included in all line regions (horizontal line region or vertical line region)” for all the boundaries indicated by the pattern boundary information, It is determined that it matches the boundary of the image. When it is determined that “a line region that does not include a boundary pixel exists”, it is determined that “the boundaries of the plurality of divided regions do not coincide with the boundaries of the plurality of images”.
In the non-divided pattern, it is determined that “the boundaries of the plurality of divided regions do not coincide with the boundaries of the plurality of images” without determining whether or not the boundary pixel exists.

  When there are a plurality of division patterns determined that “the boundaries of the plurality of divided areas match the boundaries of the plurality of images”, it is determined that “the boundaries of the plurality of divided areas match the boundaries of the plurality of images”. One of the plurality of divided patterns is selected. For example, the division pattern with the highest priority is selected from among the plurality of division patterns determined that “the boundaries of the plurality of divided regions match the boundaries of the plurality of images”. In addition, when there is no divided pattern determined that “the boundaries of the plurality of divided regions match the boundaries of the plurality of images”, the non-divided pattern is selected.

The monochrome count unit 804 counts the number of monochrome pixels of the input image S1 in the divided area indicated by the divided pattern selected by the divided pattern selection unit 803. When the division pattern selected by the division pattern selection unit 803 indicates a plurality of division regions, the number of monochrome pixels is counted for each of the plurality of division patterns. Then, the monochrome count unit 804 outputs monochrome information cnt_ptn indicating the number of monochrome pixels in the divided area. The division area indicated by the division pattern selected by the division pattern selection unit 803 is determined from the selection pattern information ptn_s.
In addition, when the size of the area composed of a plurality of divided areas indicated by the divided pattern is different from the size of the area of the input image S1, the pixels included in the divided areas are subjected to scaling processing for matching the sizes. Is judged.

The pattern attribute assignment unit 805 assigns a monochrome attribute or a color attribute to the divided area indicated by the division pattern selected by the division pattern selection unit 803. Division pattern selection unit 803
When the division pattern selected in (1) indicates a plurality of division areas, a monochrome attribute or a color attribute is assigned to each of the plurality of division patterns. Then, the pattern attribute assignment unit 805 outputs pattern attribute information mc_ptn indicating the attribute of the divided area.
In this embodiment, attributes are assigned based on the selection pattern information ptn_s and the monochrome information cnt_ptn. Specifically, a color attribute is assigned to an attribute-free area (divided area) where the monochrome ratio is less than a threshold, and a monochrome attribute is assigned to an attribute-free area where the monochrome ratio is greater than or equal to the threshold. In the present embodiment, the monochrome ratio is a ratio of the total number of monochrome pixels in the attribute-free area (second total number) to the total number of pixels in the attribute-free area (first total number). The division area and the first total number indicated by the division pattern selected by the division pattern selection unit 803 are determined from the selection pattern information ptn_s. The second total number is determined from the monochrome information cnt_ptn. In the present embodiment, as in the first embodiment, 95% is used as the monochrome ratio threshold.

The confirmed pattern generation unit 806 generates a confirmed pattern based on the selection pattern information ptn_s and the pattern attribute information mc_ptn. Then, the confirmed pattern generation unit 806 outputs confirmed pattern information dptn_s indicating the generated confirmed pattern.
In the present embodiment, an attribute pattern indicating an attribute area in which the attribute indicated by the pattern attribute information mc_ptn is assigned to the divided area indicated by the selection pattern information ptn_s is generated as a confirmed pattern.

As described above, according to the present embodiment, a division pattern in which the boundary of the divided area coincides with the boundary of the image is selected. And a fixed pattern is produced | generated by assigning an attribute to the no-attribute area | region which the selected division | segmentation pattern shows. When a plurality of images are arranged in the input image, a boundary usually exists between the plurality of images. Therefore, by such a method, the area of the input image can be accurately divided into a monochrome area and a color area.
In the present embodiment, attribute determination for each block area and generation of a plurality of attribute patterns are not performed, so that a reduction in processing load is expected compared to the first and second embodiments.

Note that the pattern selection method in the present embodiment may be applied to the first and second embodiments. For example, in the fixed pattern selection unit 107 according to the first embodiment, when a plurality of images are arranged in the input image, the plurality of attribute areas are indicated and the boundaries of the plurality of attribute areas coincide with the boundaries of the plurality of images. The attribute pattern to be selected may be selected as the confirmed pattern. An attribute pattern generated from an undivided pattern is selected as a definite pattern when there is no attribute pattern that indicates a plurality of attribute areas and the boundaries of the plurality of attribute areas match the boundaries of the plurality of images. Good. And when one image is arrange | positioned at an input image, the attribute pattern produced | generated from the non-division pattern may be selected as a definite pattern.
Note that the attribute assignment method in the present embodiment may be applied to the first and second embodiments. For example, in the attribute pattern generation unit 105 of the first embodiment and the confirmed pattern generation unit 702 of the second embodiment, attributes may be assigned to the divided areas based on the monochrome ratio of the divided areas.

<Example 4>
Hereinafter, an image processing apparatus and a control method thereof according to Embodiment 4 of the present invention will be described. In the present embodiment, the division pattern selection method is different from that in the third embodiment.

  FIG. 11 is a block diagram illustrating an example of a functional configuration of the image processing apparatus 110 according to the present embodiment. In addition, the same code | symbol is attached | subjected to the same function part as Examples 1-3 (FIG. 1, 7, 8), and the description is abbreviate | omitted.

The division pattern selection unit 1101 selects one of a plurality of division patterns indicated by the division pattern information ptn according to a user operation. Then, the division pattern selection unit 1101
Selection pattern information ptn_s indicating the selected division pattern is output.
For example, numbers (pattern numbers) are assigned to the plurality of division patterns indicated by the division pattern information ptn. A pattern number (user selection number) ptn_no corresponding to a user operation on the menu image displayed on the switch or display device of the image processing apparatus is input to the division pattern selection unit 1101. Then, the division pattern assigned the user selection number ptn_no is selected.

  As described above, according to the present embodiment, a division pattern is selected according to a user operation. The user can select the division pattern in consideration of only the arrangement of the image (monochrome image or color image) in the input image without considering the attribute. Therefore, the user can easily select a division pattern suitable for the input image. As a result, the image processing apparatus 110 can select a division pattern suitable for the input image with higher accuracy than in the third embodiment.

  Note that the pattern selection method in the present embodiment may be applied to the first and second embodiments. For example, in the confirmed pattern selection unit 107 according to the first embodiment, one of a plurality of attribute patterns may be selected as a confirmed pattern in accordance with a user operation. Even in this case, since the user only needs to consider the arrangement of the image (monochrome image or color image) in the input image, the user can easily select an attribute pattern suitable for the input image.

<Example 5>
Hereinafter, an image processing apparatus and a control method thereof according to Embodiment 5 of the present invention will be described. In the present embodiment, an example in which a plurality of attribute patterns are prepared in advance as a plurality of division patterns will be described.

  FIG. 12 is a block diagram illustrating an example of a functional configuration of the image processing apparatus 120 according to the present embodiment. In addition, the same code | symbol is attached | subjected to the same function part as Examples 1-4 (FIGS. 1, 7, 8, 11), and the description is abbreviate | omitted.

A plurality of attribute patterns are recorded in the attribute pattern storage unit 1201 in advance. The attribute pattern storage unit 1201 outputs attribute pattern information dptn indicating a plurality of recorded attribute patterns.
FIG. 13 shows an example of an attribute pattern recorded in the attribute pattern storage unit 1201. Each of the eight attribute patterns 1 to 8 in FIG. 13 indicates one or more attribute areas. Each attribute area is assigned a color attribute or a monochrome attribute.

  The attribute pattern selection unit 1202 sets the first condition that “all the block areas existing in the attribute area to which the monochrome attribute is assigned are monochrome block areas” from among the plurality of attribute patterns indicated by the attribute pattern information dptn. Select an attribute pattern that meets the requirements. In other words, the first condition is a condition that “the color block area does not exist in the attribute area to which the monochrome attribute is assigned”. The monochrome block area is a block area to which a monochrome attribute is assigned, and the color block area is a block area to which a color attribute is assigned. The attribute of the block area is determined from the block attribute information mc. Then, the attribute pattern selection unit 1202 outputs selection attribute pattern information dptn_pck indicating the selected attribute pattern. If there are a plurality of attribute patterns that satisfy the first condition, all attribute patterns that satisfy the first condition are selected. When the image shown in FIG. 3 is input as the input image S1, the attribute patterns 4 and 5 in FIG. 10 are selected.

By narrowing down such attribute patterns, it is possible to suppress erroneous calculation of a high evaluation value due to the presence of a monochrome block region in the achromatic region of the color image. As described in the first embodiment, the achromatic region and the chromatic region of the color image are present at close positions. Therefore, by selecting an attribute pattern that includes at least one color block area in an attribute area to which a monochrome attribute is assigned, attribute patterns that are likely to be miscalculated with high evaluation values are excluded from evaluation value calculation targets. can do. Specifically, it is possible to exclude an attribute pattern from which an evaluation value is highly likely to be erroneously calculated due to the presence of a monochrome block region in an achromatic region of a color image.

If no attribute pattern that satisfies the first condition exists, an attribute pattern indicating one attribute area is selected. In that case, both of the attribute patterns 1 and 2 in FIG. 13 may be selected, or one of the attribute patterns 1 and 2 in FIG. 13 may be selected.
Note that an attribute pattern that satisfies the second condition that “all block areas existing in an attribute area to which a color attribute is assigned is a color block area” may be selected. Only an attribute pattern that satisfies the first condition may be selected, only an attribute pattern that satisfies the second condition may be selected, or an attribute pattern that satisfies at least one of the first condition and the second condition is selected. Also good.

  The evaluation value calculation unit 106 and the confirmed pattern selection unit 107 perform the same processing as in the first embodiment. However, the evaluation value calculation unit 106 calculates the evaluation value of each attribute pattern indicated by the selection attribute pattern information dptn_pck when the selection attribute pattern information dptn_pck indicates a plurality of attribute patterns. Then, when the selected attribute pattern information dptn_pck indicates a plurality of attribute patterns, the determined pattern selection unit 107 selects the attribute pattern having the highest evaluation value from among the plurality of attribute patterns indicated by the selected attribute pattern information dptn_pck as the determined pattern. Is done. When the selection attribute pattern information dptn_pck indicates one attribute pattern, the attribute pattern indicated by the selection attribute pattern information dptn_pck is selected as a confirmed pattern (the evaluation value may or may not be calculated).

As described above, according to the present embodiment, in a configuration in which a plurality of attribute patterns are prepared in advance, as in the first embodiment, an input image area can be accurately divided into a monochrome area and a color area. it can.
Note that the processing of the attribute pattern selection unit 1202 may be omitted, and the attribute pattern information dptn may be input to the evaluation value calculation unit 106 and the confirmed pattern selection unit 107.

<Other examples>
The present invention can also be implemented by a system (or a device such as a CPU or MPU) of a system or apparatus that implements the functions of the above-described embodiments by reading and executing a program recorded in a storage device. The present invention can also be implemented by a method comprising steps executed by a computer of a system or apparatus that implements the functions of the above-described embodiments by reading and executing a program recorded in a storage device, for example. . For this purpose, the program is stored in the computer from, for example, various types of recording media that can serve as the storage device (ie, computer-readable recording media that holds data non-temporarily). Provided to. Therefore, the computer (including devices such as CPU and MPU), the method, the program (including program code and program product), and the computer-readable recording medium that holds the program non-temporarily are all present. It is included in the category of the invention.

10, 70, 80, 110, 120 Image processing device 103 Block attribute assignment unit 104 Division pattern storage unit 105 Attribute pattern generation unit 106 Evaluation value calculation unit 107 Determination pattern selection unit 701, 803, 1101 Division pattern selection unit 702, 806 Determination Pattern generation unit 801 Boundary detection unit 802 Division pattern / boundary storage unit 805 Pattern attribute allocation unit 806 Deterministic pattern generation unit 1201 Attribute pattern storage unit 1202 Attribute pattern selection unit

Claims (23)

An image processing apparatus that divides an input image area into a monochrome area and a color area,
Storage means in which a plurality of different division patterns are recorded in advance;
Obtaining means for obtaining a fixed pattern in which a monochrome attribute or a color attribute is assigned to a divided region obtained by dividing the region according to one divided pattern selected from the plurality of divided patterns based on an input image;
An image processing apparatus comprising: The division pattern indicates a non-attribute area to which no attribute is assigned as a division area,
The acquisition means includes
A process of generating an attribute pattern indicating an attribute area to which a monochrome attribute or a color attribute is assigned by assigning a monochrome attribute or a color attribute to an attribute-free area indicated by the division pattern based on the input image. For each of the patterns,
The image processing apparatus according to claim 1, wherein one of a plurality of attribute patterns is selected as the confirmed pattern. An allocation unit that allocates one of a plurality of attributes including a monochrome attribute and a color attribute to each of a plurality of block areas constituting the area of the input image based on the input image;
The acquisition unit selects an attribute pattern having the largest number of block areas in which the attribute assigned to the block area matches the attribute assigned to the attribute area including the block area as the fixed pattern. The image processing apparatus according to claim 2. An allocation unit that allocates one of a plurality of attributes including a monochrome attribute and a color attribute to each of a plurality of block areas constituting the area of the input image based on the input image;
The image processing apparatus according to claim 2, wherein the acquisition unit selects an attribute pattern having the same attributes in a block area whose entire attribute area is included as the definite pattern. The plurality of division patterns include a non-division pattern that does not divide a region,
The acquisition means selects an attribute pattern generated from the non-divided pattern as the definite pattern when there is no attribute pattern in which all the block areas that are entirely included in the attribute area have the same attributes. The image processing apparatus according to claim 4. The acquisition means determines an attribute pattern that indicates a plurality of attribute areas and a boundary of the plurality of attribute areas coincides with a boundary of the plurality of images when a plurality of images are arranged in the input image. The image processing apparatus according to claim 2, wherein the image processing apparatus is selected as a pattern. The plurality of division patterns include a non-division pattern that does not divide a region,
The acquisition means indicates an attribute pattern generated from the non-divided pattern when there is no attribute pattern that indicates a plurality of attribute areas and a boundary of the plurality of attribute areas matches a boundary of the plurality of images. The image processing apparatus according to claim 6, wherein the image processing apparatus is selected as a confirmed pattern. The plurality of division patterns include a non-division pattern that does not divide a region,
The said acquisition means selects the attribute pattern produced | generated from the said non-division pattern as the said definite pattern, when one image is arrange | positioned at the said input image. Image processing apparatus. The image processing apparatus according to claim 2, wherein the acquisition unit selects one of the plurality of attribute patterns as the confirmed pattern in response to a user operation. The division pattern indicates a non-attribute area to which no attribute is assigned as a division area,
The acquisition means includes
Selecting one of the plurality of division patterns;
The image processing apparatus according to claim 1, wherein the fixed pattern is generated by assigning a monochrome attribute or a color attribute to an attribute-free area indicated by the selected divided pattern based on the input image. An allocation unit that allocates one of a plurality of attributes including a monochrome attribute and a color attribute to each of a plurality of block areas constituting the area of the input image based on the input image;
The image processing apparatus according to claim 10, wherein the acquisition unit selects a division pattern in which attributes of all block areas that are entirely included in the non-attribute area are equal to each other. The plurality of division patterns include a non-division pattern that does not divide a region,
12. The image according to claim 11, wherein the acquisition unit selects the non-divided pattern when there is no divided pattern having the same attributes of all the block areas that are entirely included in the non-attribute area. Processing equipment. The acquisition unit selects a division pattern that indicates a plurality of attributeless areas and a boundary of the attributeless areas matches a boundary of the plurality of images when a plurality of images are arranged in the input image. The image processing apparatus according to claim 10. The plurality of division patterns include a non-division pattern that does not divide a region,
The obtaining means selects the non-divided pattern when there is no divided pattern that indicates a plurality of non-attribute areas and the boundaries of the plurality of non-attribute areas match the boundaries of the plurality of images. The image processing apparatus according to claim 13. The plurality of division patterns include a non-division pattern that does not divide a region,
The image processing apparatus according to claim 13 or 14, wherein the acquisition unit selects the non-divided pattern when one image is arranged in the input image.   The image processing apparatus according to claim 10, wherein the acquisition unit selects one of the plurality of division patterns according to a user operation. An allocation unit that allocates one of a plurality of attributes including a monochrome attribute and a color attribute to each of a plurality of block areas constituting the area of the input image based on the input image;
The acquisition means assigns a color attribute to an attribute-free area including a block area to which a color attribute is assigned, and assigns a monochrome attribute to an attribute-free area not including the block area to which the color attribute is assigned. The image processing apparatus according to any one of 2 to 16. The acquisition unit assigns a color attribute to an attributeless area in which a ratio of a total number of monochrome pixels in the attributeless area to a total number of pixels in the attributeless area is less than a threshold, and the attribute is equal to or greater than the threshold The image processing apparatus according to claim 2, wherein a monochrome attribute is assigned to the area. An allocation unit that allocates one of a plurality of attributes including a monochrome attribute and a color attribute to each of a plurality of block areas constituting the area of the input image based on the input image;
The division pattern indicates an attribute area to which an attribute is assigned as a division area,
The acquisition means determines the division pattern having the largest number of block areas in which the attribute assigned to the block area matches the attribute assigned to the attribute area including the block area from the plurality of division patterns. The image processing apparatus according to claim 1, wherein the image processing apparatus is selected as a pattern. An allocation unit that allocates one of a plurality of attributes including a monochrome attribute and a color attribute to each of a plurality of block areas constituting the area of the input image based on the input image;
The division pattern indicates an attribute area to which an attribute is assigned as a division area,
The acquisition means selects, as the definite pattern, an attribute pattern that does not include a block area to which a color attribute is assigned to an attribute area to which a monochrome attribute is assigned from the plurality of division patterns. The image processing apparatus according to claim 1. When there are a plurality of division patterns that do not include the block area to which the color attribute is assigned in the attribute area to which the monochrome attribute is assigned, the acquisition unit assigns the color attribute to the attribute area to which the monochrome attribute is assigned. Among the plurality of division patterns not including the block area, the division pattern having the largest number of block areas in which the attribute assigned to the block area matches the attribute assigned to the attribute area including the block area, The image processing apparatus according to claim 20, wherein the image processing apparatus selects the fixed pattern. A control method of an image processing apparatus for dividing an input image area into a monochrome area and a color area,
The image processing apparatus has storage means in which a plurality of different division patterns are recorded in advance,
The control method of the image processing apparatus is:
Obtaining an input image;
Obtaining a fixed pattern in which a monochrome attribute or a color attribute is assigned to a divided region obtained by dividing the region according to one divided pattern selected from the plurality of divided patterns based on the input image;
A control method for an image processing apparatus, comprising: 23. A program for causing a computer to execute each step of the control method for an image processing apparatus according to claim 22.

Images