JP2015108947A - Image processor, control method thereof, and control program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To extract a specific area by discriminating the specific area, such as an empty area with high accuracy regardless of a composition during photographing.SOLUTION: An image processing part divides an input image into a plurality of areas including a specific area in accordance with a feature amount showing a feature of the input image under control of a control part, determines whether each area other than the specific area is adjacent on condition predetermined as the specific area, and makes an area adjacent on the predetermined condition a specific area candidate. The image processing part performs calculation as a reference pixel value in accordance with a pixel value of the specific area, calculates a representative pixel value in accordance with a pixel value of the specific area candidate, determines whether to integrate the specific area candidate to the specific area in accordance with the reference pixel value and the representative pixel value, and integrates the specific area candidate to the specific area on the basis of the determination result thereof.

Description

  The present invention relates to an image processing apparatus, a control method thereof, and a control program, and more particularly to an image processing apparatus that discriminates and extracts a specific area indicating a predetermined area from an input image.

  In general, when shooting a shooting scene with a sharp contrast in a subject with an imaging device such as a digital camera, for example, a person and a background may be shot in backlight against a bright sky. In such a photographic scene, if the exposure is performed so that the brightness of the person and the background is at an appropriate level, the sky is saturated. On the other hand, if the brightness of the sky is properly exposed, the person and the background are photographed underexposed.

  In order to solve such a situation, for example, the sky and background regions are divided, the optimum exposure condition for each of these regions is determined, two images are taken, and these images are combined. An image processing apparatus is known (see Patent Document 1). Further, as a method for dividing the sky and background regions, for example, there is a method of extracting a region having a feature of being blue and having no edge signal as a sky region.

Japanese Patent Application Laid-Open No. 2004-159211

  However, if an area having the feature of being blue and having no edge signal is extracted as an empty area, the empty area may not be extracted properly depending on the composition.

  FIG. 11 is a diagram for explaining extraction of an empty area in a conventional image processing apparatus. FIG. 11A is a diagram illustrating an example of the composition, and FIG. 11B is a diagram illustrating the extraction result of the empty area in the composition illustrated in FIG.

  In the composition shown in FIG. 11A, the sky is positioned at the top of the screen and the background is positioned at the bottom of the screen. There is a subject 1101 that can be seen through the sky at the top of the screen. When an empty area is extracted from an image obtained in such a composition, the image processing apparatus determines the area indicated by reference numeral B1 as a background area as shown in FIG. 11B, but is indicated by reference numeral B2. If the area has an edge area, it is determined as a background area. Then, the image processing apparatus determines only the area indicated by reference numeral A1 as an empty area.

  As a result, the empty area existing in the area indicated by the symbol B2 is saturated. In order to solve such a problem, if an area is determined only by color information without confirming an edge signal, an object having the same color as the empty area is present in the area indicated by reference numeral B1. It is said that a subject having the same color as the sky area is underexposed in the area indicated by reference numeral B1.

  Accordingly, an object of the present invention is to provide an image processing apparatus, a control method thereof, and a control program capable of accurately identifying and extracting a specific area such as an empty area regardless of the composition at the time of shooting. .

  In order to achieve the above object, an image processing apparatus according to the present invention is an image processing apparatus that discriminates and extracts a specific area as a specific area from an input image according to a feature amount indicating a characteristic of the input image. A dividing unit that divides the input image into a plurality of areas including the specific area, and each of the areas excluding the specific area determines whether or not the specific area is adjacent to the specific area under a predetermined condition. Determining means for setting a region adjacent under the predetermined condition as a specific region candidate, calculating a reference pixel value as a reference pixel value according to a pixel value in the specific region, and further selecting the specific region candidate Calculating means for calculating a representative pixel value representative of the specific area candidate in accordance with the pixel value in, and integrating the specific area candidate in the specific area in accordance with the reference pixel value and the representative pixel value Luke and determines whether to a region integrating means for integrating the specific area of the specific region candidate based on the determination result, characterized in that it has a.

  The control method according to the present invention is a control method for an image processing apparatus that discriminates and extracts a specific region from an input image as a specific region, and the input image is selected according to a feature amount indicating a feature of the input image. A division step of dividing into a plurality of areas including a specific area, and each of the areas excluding the specific area is determined whether it is adjacent to the specific area under a predetermined condition, and the predetermined area is determined. A step of determining an area adjacent to the specific area as a specific area candidate, a reference pixel value is calculated as a reference pixel value according to a pixel value in the specific area, and further according to a pixel value in the specific area candidate A calculation step of calculating a representative pixel value representing the specific area candidate, and whether or not to integrate the specific area candidate into the specific area according to the reference pixel value and the representative pixel value To determine the, and having a, a region integrating step of integrating the specific area of the specific region candidate based on the determination result.

  A control program according to the present invention is a control program used in an image processing apparatus that discriminates and extracts a specific area as a specific area from an input image, and shows the characteristics of the input image on a computer included in the image processing apparatus. A division step of dividing the input image into a plurality of regions including the specific region according to a feature amount, and each of the regions excluding the specific region is adjacent to the specific region under a predetermined condition Determining whether or not to determine an adjacent region under the predetermined condition as a specific region candidate, calculating a reference pixel value as a reference pixel value according to a pixel value in the specific region, and Calculating a representative pixel value representing the specific area candidate according to a pixel value in the specific area candidate; the reference pixel value; and Determining whether to integrate the specific area candidate into the specific area according to a table pixel value, and executing an area integration step of integrating the specific area candidate into the specific area based on the determination result It is characterized by that.

  According to the present invention, it is determined whether or not to integrate the specific area candidate into the specific area according to the reference pixel value in the specific area and the representative pixel value in the specific area candidate. Regardless, the specific area can be accurately identified and extracted.

It is a block diagram which shows an example of an imaging device provided with the image processing apparatus by the 1st Embodiment of this invention. It is a flowchart for demonstrating the area | region discrimination | determination process performed with the camera shown in FIG. It is a figure for demonstrating the area | region discrimination | determination performed with the camera shown in FIG. 1, (a) is a figure which shows an example of an image, (b) is a figure which shows the image after area | region division, (c) is an empty area candidate (D) is a figure which shows the image after area | region discrimination | determination. It is a figure for demonstrating an example of calculation of the reference | standard pixel value when an empty area | region has gradation. FIG. 5 is a diagram for explaining the average luminance for each block in the image shown in FIG. 4, in which (a) shows the average luminance value for each block when the vertical block position is 0, and (b) is horizontal. The figure which shows the average luminance value for every block when the block position of a direction is set to 0, (c) is a figure which shows the average luminance value for every block when the block position of a vertical direction is set to 3. FIG. It is a flowchart for demonstrating the empty area | region integration process shown in FIG. It is a flowchart for demonstrating an example of the sky area | region integration process performed with the camera which concerns on the 2nd Embodiment of this invention. It is a figure for demonstrating the process which makes the pixel of a sky region candidate a pixel of a sky region, (a) is a figure which shows a pixel of a sky region and a pixel of a sky region candidate, (b) is a pixel of a sky region candidate. The figure which shows an example of the state changed into the pixel of the sky area, (c) is a figure which shows an example of the state which changed the pixel of the sky area candidate into the pixel of the sky area. It is a flowchart for abolishing description of an example of the area | region discrimination | determination process performed with the camera which concerns on the 3rd Embodiment of this invention. FIG. 10 is a diagram for explaining calculation of the reliability distribution shown in FIG. 9, where (a) is a diagram showing a sky probability that is a sky region for each pixel, and (b) is a diagram showing a histogram of sky probability. It is a figure for demonstrating extraction of the empty area | region in the conventional image processing apparatus, (a) is a figure which shows an example of a composition, (b) shows the extraction result of the empty area in the composition shown to (a). FIG.

  Hereinafter, an example of an image processing apparatus according to an embodiment of the present invention will be described with reference to the drawings.

[First Embodiment]
FIG. 1 is a block diagram illustrating an example of an imaging apparatus including an image processing apparatus according to the first embodiment of the present invention.

  The illustrated imaging apparatus is, for example, a digital camera (hereinafter simply referred to as a camera) and includes an imaging unit 101. The imaging unit 101 includes a photographic lens unit (hereinafter simply referred to as a photographic lens), an imaging element, and a drive circuit for driving the imaging element. The imaging element is, for example, a CCD or CMOS sensor, and outputs an electrical signal (analog image signal) corresponding to the optical image formed through the photographing lens.

  The analog image signal output from the imaging unit 101 is converted into a digital image signal (image data) by the A / D conversion circuit 102. The image data is input to the image processing unit 103. The image processing unit 103 performs known white balance processing, noise suppression processing, gradation conversion processing, edge enhancement correction processing, and the like on the image data to process the luminance signal Y and the color difference signals U and V into processed image data. Output as. Further, the image processing unit 103 performs area determination processing described later.

  The processed image data output from the image processing 103 is recorded in the storage medium 106. Further, the display unit 105 displays an image corresponding to the processed image data. The control unit 104 controls the entire camera. That is, the control unit 10 controls operations of the image processing unit 103, the display unit 105, the recording medium 106, and the like. The image processing unit 103, the control unit 104, the display unit 105, and the storage medium 106 are connected to each other via a bus 107, and various data are transmitted / received via the bus 107.

  FIG. 2 is a flowchart for explaining an area determination process performed by the camera shown in FIG. 2 is performed by the image processing unit 103 under the control of the control unit 104. Here, a case will be described in which the image indicated by the image data includes a sky region and a background region, and the sky region and the background region are divided and divided.

  Upon receiving the image data (input image), the image processing unit 103 divides the image into a plurality of blocks having a predetermined size according to the feature amount of the image data. Then, for example, the image processing unit 103 performs area determination for determining an empty area and a background area for each block (step S201).

  When determining an empty area (that is, a specific area in this case), for example, the image processing unit 103 detects an area having a similar color. At this time, the image processing unit 103 converts the RGB color information indicating the color of the pixel into the HSV color system and uses the hue luminance information, and adjacent areas having the same color and luminance are adjacent to each other. Detect as a region. The image processing unit 103 integrates these adjacent areas to obtain an integrated area. Then, when the area of the integrated region is equal to or larger than a predetermined area (that is, a predetermined threshold), the image processing unit 103 sets the integrated region as an empty region.

  FIG. 3 is a diagram for explaining region discrimination performed by the camera shown in FIG. FIG. 3A shows an example of an image, and FIG. 3B shows an image after region division. FIG. 3C is a diagram showing empty region candidates, and FIG. 3D is a diagram showing an image after region discrimination.

  In the example shown in FIG. 3A, it is assumed that all of the region 1, the region 2, and the region 3 have color and luminance information that is determined as an empty region. Here, since the area of the region 1 is equal to or larger than the threshold (area threshold), the image processing unit 103 determines that the region 1 is an empty region. On the other hand, since the areas of the regions 2 and 3 are less than the threshold, the image processing unit 103 classifies the regions 2 and 3 as background regions without determining that the regions 2 and 3 are empty regions.

  As a result, the image processing unit 103 divides the original image shown in FIG. 3A into a sky area and a background area as shown in FIG.

  Referring to FIG. 2 again, following the process of step S201, the image processing unit 103 determines whether or not there is an empty area in the image (step S202). If there is no empty area (NO in step S202), the image processing unit 103 ends the area determination process. On the other hand, if a sky area exists (YES in step S202), the image processing unit 103 performs sky area candidate determination on the area determined as the background area in step S201, and thus a sky area candidate (that is, a specific area candidate). It is determined whether or not there is an area (step S203).

  In the determination of the empty area candidate, for example, the image processing unit 103 determines an area surrounded by the empty area on the upper and left sides as an empty area candidate. In the background area shown in FIG. 3B, the background area located in the area indicated by the symbol A is in contact with the sky area on the left and right (the boundary between the background area and the sky area exists: predetermined condition) ). For this reason, there is a high possibility that the background region located in the region indicated by the symbol A is an empty region, and as shown in FIG. 3C, the background region located in the region indicated by the symbol A is determined as an empty region candidate. The

  On the other hand, the left and right sides of the background region located in the region indicated by the symbol B are not in contact with the sky region (the background region and the sky region are not in contact on the left and right). For this reason, since there is a low possibility that it is an empty area, the background area located in the area indicated by the symbol B is left as the background area as shown in FIG.

  Subsequently, the image processing unit 103 obtains reference pixel values (Ysky, Usky, Vsky) in the sky area. Here, for example, the reference pixel value in the sky region is obtained using the average value of the pixel values in the region determined as the sky region in step S201. The reference pixel value may be obtained as follows.

  FIG. 4 is a diagram for explaining an example of calculation of the reference pixel value when the sky region has gradation.

  Assuming that the sky region has a horizontal gradation, the image is divided into a plurality of blocks 1001 and an average luminance value (Y_AVE (xb, yb)) is obtained for each of the blocks 1001. Note that xb represents a horizontal block position, and yb represents a vertical block position. In the illustrated example, the number of blocks in the horizontal direction is 12 and the number of blocks in the vertical direction is 9.

  FIG. 5 is a diagram for explaining the average luminance for each block in the image shown in FIG. FIG. 5A is a diagram showing the average luminance value for each block when the vertical block position is 0, and FIG. 5B is the block when the horizontal block position is 0. It is a figure which shows the average luminance value. FIG. 5C is a diagram showing an average luminance value for each block when the block position in the vertical direction is 3.

  In FIG. 5A, the horizontal axis indicates the position of the block, and the vertical axis indicates the average luminance value for each block. Assuming that the average luminance value Y_AVE (xb, 0) when yb = 0 is set, the gradient GRAD_H (xb, 0) between the blocks in the horizontal direction is expressed by the following equation (1).

GRAD_H (xb, 0) = Y_AVE (xb + 1,0) −Y_AVE (xb, 0) (1)
In the illustrated example, since the average value of inclination (xb = 0 to 10) is equal to or greater than a predetermined value, it is determined that the sky region has a gradation in the horizontal direction.

  In FIG. 5B, assuming that the average luminance value Y_AVE (0, yb) when xb = 0 is satisfied, the gradient GRAD_V (0, yb) between the blocks in the vertical direction is expressed by the following equation (2). .

GRAD_V (0, yb) = Y_AVE (0, yb + 1) −Y_AVE (0, yb) (2)
In the illustrated example, the average value of the slope (yb = 0 to 7) is less than a predetermined value, so that it is determined that the sky region does not have gradation in the vertical direction.

  Since the sky region has a gradation in the horizontal direction, the block of the candidate sky region according to the gradient GRAD_H (xb, 0) between the blocks in the horizontal direction and the average luminance value Y_AVE (xb-1, yb) for each block. A reference luminance value is obtained. The reference luminance value Y_TARGET (xb, yb) is expressed by the following equation (3).

Y_TARGET (xb, yb) = Y_AVE (xb-1, yb) + GRAD_H (xb, 0) (3)
In the case of yb = 3 shown in FIG. 5C, since the blocks up to xb = 0, 1, and 2 are empty regions, the empty regions are obtained for the blocks of xb = 3, 4, 5, 6, and 7. A reference luminance value Y_TARGET of a candidate block is obtained.

  Similarly, a reference hue value is obtained for color information, and reference pixel values (Y_TARGET (xb, yb), U_TARGET (xb, yb), and V_TARGET (xb, yb)) for each block are determined according to the reference luminance value and the reference hue value. yb)). Then, the reference pixel values (Ysky (x, y), Usky (x, y), Vsky (x, y)) of the sky region are obtained for each pixel from the reference pixel values for each block by known linear interpolation. In this way, the image processing unit 103 obtains the color luminance distribution of the sky area and determines the target color luminance range (step S204).

  Referring to FIG. 2 again, following the processing in step S204, the image processing unit 103 performs sky region integration processing on the pixels determined as the sky region candidate in the sky region candidate determination in step S203 (step S205).

  FIG. 6 is a flowchart for explaining the empty area integration processing shown in FIG.

  When the sky region integration processing is started, the image processing unit 103 first determines whether or not the target pixel is a sky region candidate pixel (step S301). If the pixel of interest is a sky region candidate pixel (YES in step S301), the image processing unit 103 detects the difference between the reference pixel value and the pixel value of the pixel of interest (here, the representative pixel) (step S302). For example, when the pixel value of the pixel of interest is (Y, U, V), the difference SUB is obtained by the following equation (4).

  Note that a pixel value representing a pixel value of a sky region candidate may be used as a representative pixel value, and the representative pixel value may be compared with a reference pixel value to obtain a difference.

SUB = SQRT ((Y-Ysky (x, y)) ^ 2+ (U-Usky (x, y)) ^ 2+ (V-Vsky (x, y)) ^ 2) (4)
Subsequently, the image processing unit 103 determines whether or not the difference SUB is equal to or less than a predetermined threshold (step S303). If the difference SUB is equal to or smaller than the threshold value (YES in step S303), the image processing unit 103 integrates the target pixel into the sky region (step S304). Then, the image processing unit 103 determines whether or not there is a pixel of the next empty area candidate (step S305).

  If there is a next empty region candidate pixel (NO in step S305), the image processing unit 103 returns to the process of step S301. On the other hand, if there is no next empty area candidate pixel (YES in step S305), the image processing unit 103 ends the empty area integration process.

  If the target pixel is not an empty area candidate pixel (YES in step S301), the image processing unit 103 proceeds to the process of step S305. Similarly, when the difference SUB exceeds the threshold value (NO in step S303), the process proceeds to step S305 in the image processing unit 103.

  After performing the sky region integration processing as described above, the image processing unit 103 ends the region determination processing.

  In this way, in the first embodiment of the present invention, it is determined again only for the pixels determined as the sky region candidates whether or not the pixels of the sky region candidates belong to the sky region. Thus, for example, as shown in FIG. 3D, even in a shooting scene where the sky appears to be missing, the areas where the sky appears to be missing (areas 2 and 3 shown in FIG. 3A) are appropriately set as empty areas. Can be determined.

  In the first embodiment, YUV is used to detect the difference between pixel values. However, the present invention is not limited to YUV as long as the pixel value can be expressed. For example, Lab or XYZ may be used. Here, the difference between the pixel values is detected, but the process may be performed in an average value or a block unit including the peripheral pixels of the target pixel.

  Furthermore, in the first embodiment, the example in which the sky region and the background region are discriminated is described. However, the sky region and the background region are not limited to the region discrimination. For example, in a scene to which a person region is added. The same can be applied. In any case, in the first embodiment, it is possible to accurately extract a sky region that is a specific region (that is, a subject region).

  In addition, in the first embodiment described above, when determining the pixel of the empty region candidate, the region surrounded by the upper and left sides of the empty region is determined as the empty region candidate. However, the empty region has a predetermined area. In the case of the above, an empty area candidate may be obtained.

[Second Embodiment]
Next, an example of a camera provided with an image processing apparatus according to the second embodiment of the present invention will be described. The configuration of the camera according to the second embodiment is the same as that of the camera shown in FIG.

  The camera according to the second embodiment performs the area determination process as described with reference to FIG. 2, but here, the sky area integration process is different from that of the first embodiment.

  FIG. 7 is a flowchart for explaining an example of the sky region integration processing performed by the camera according to the second embodiment of the present invention. In the flowchart shown in the figure, the same steps as those in the flowchart shown in FIG.

  If it is determined in step S303 that the difference SEB is equal to or smaller than the threshold (difference threshold), the image processing unit 103 determines whether the distance between the target pixel and the sky region is equal to or smaller than a predetermined threshold (distance threshold) ( Step S404). If the distance is equal to or smaller than the distance threshold (YES in step S404), the image processing unit 103 proceeds to the process of step S304 and integrates the pixel of interest into the sky region. On the other hand, when the distance exceeds the distance threshold (NO in step S404), the image processing unit 103 proceeds to the process of step S305 described above.

  FIG. 8 is a diagram for explaining the process of setting the sky region candidate pixels as the sky region pixels. FIG. 8A is a diagram illustrating a pixel in the sky region and a pixel in the sky region candidate, and FIG. 8B is a diagram illustrating an example of a state in which the pixel in the sky region candidate is changed to a pixel in the sky region. is there. FIG. 8C is a diagram illustrating an example of a state where the pixels of the sky region candidate are changed to the pixels of the sky region.

  In FIG. 8A, pixels determined to belong to the sky region are denoted by Sxx (x: 0 to 4), and pixels of the sky region candidate are denoted by Hxx (x: 0 to 4). Assume that the sanity distance threshold is 2 pixels, and the difference between the pixel H23 and the pixel H34 and the pixel of interest is equal to or less than the difference threshold.

  Since the pixel in the sky region that has the smallest distance from the pixel H23 is the pixel S03 and the distance is two pixels, the image processing unit 103 determines that the pixel H23 is a sky region. Here, as shown in FIG. 8B, the image processing unit 103 changes the pixel H23 of the sky region candidate to a pixel of the sky region, and labels the pixel with S23. That is, the image processing unit 103 integrates the sky area candidate pixels H23 into the sky area.

  For the pixel H34, the pixel in the sky region with the smallest distance is the pixel S23. Since the distance between the pixel H34 and the pixel S23 is within two pixels, the pixel processing unit 103 determines that the pixel H34 is an empty area. The image processing unit 103 changes the pixel H34 of the sky region candidate to a pixel of the sky region, and labels the pixel as S34 (see FIG. 8C). That is, the image processing unit 103 integrates the sky region candidate pixels H34 into the sky region.

  As described above, in the second embodiment of the present invention, it is determined whether the distance between the sky region candidate pixel whose difference is equal to or smaller than the predetermined difference threshold is equal to or smaller than the predetermined distance threshold, Depending on the determination result, the pixels of the sky region candidate are integrated into the sky region. As a result, even in a shooting scene where the sky appears to be missing, it is possible to accurately determine the area where the sky appears to be missing as an empty area.

  In the second embodiment described above, it is determined whether or not the integration process with the sky area is performed only for the pixels of the sky area candidate, but the integration process with the sky area is performed for all the background area pixels. It may be determined whether or not to perform.

[Third Embodiment]
Next, an example of a camera including an image processing device according to the third embodiment of the present invention will be described. The configuration of the camera according to the third embodiment is the same as that of the camera shown in FIG.

  FIG. 9 is a flowchart for explaining an example of the area determination process performed by the camera according to the third embodiment of the present invention. In the illustrated flowchart, the same steps as those in the flowchart shown in FIG. 2 are denoted by the same reference numerals and description thereof is omitted.

  In step S203, after determining the sky area candidate, the image processing unit 103 checks whether or not there is an area (pixel) that is a sky area candidate (step S503). If there is no empty area candidate area (NO in step S503), the image processing unit 103 ends the area determination process.

  On the other hand, when there is an area that is a candidate for the empty area (YES in step S503), the image processing unit 103 calculates a reliability distribution indicating the reliability of whether or not the candidate area is an empty area (step S503). S504).

  FIG. 10 is a diagram for explaining calculation of the reliability distribution shown in FIG. FIG. 10A is a diagram showing a sky probability that is a sky region for each pixel, and FIG. 10B is a diagram showing a histogram of sky probability.

  In FIG. 10A, the horizontal axis indicates the hue, and the vertical axis indicates the sky probability. Here, the image processing unit 103 obtains a sky probability that is a sky region for each pixel of the sky region candidate, but the hue Hue of the pixel of interest (that is, the sky region candidate pixel) is less than or equal to the hue Hue1 or the hue Hue4. The image processing unit 103 sets the sky probability to 0. On the other hand, when the hue Hue of the target pixel is greater than or equal to the hue Hue2 and less than or equal to the hue Hue3, the image processing unit 103 sets the sky probability to 1.

  When the hue Hue of the pixel of interest is between the hue Hue1 and less than the hue Hue2, the image processing unit 103 obtains the sky probability according to the following equation (5).

Empty probability = (Hue−Hue1) × (Hue2−Hue1) (5)
Similarly, when the hue Hue of the pixel of interest is between the hue Hue3 and less than the hue Hue4, the image processing unit 103 obtains the sky probability according to the following equation (6).

Empty probability = (Hue-Hue3) × (Hue4-Hue3) +1 (6)
When obtaining the hue Hue, an operation such as Hue = tan-1 (b * / a *) is used after conversion to the Lab space.

  After obtaining the sky probability for each of the pixels of the sky region candidate, the image processing unit 103 obtains a histogram relating to the sky probability (see FIG. 10B). Then, the image processing unit 103 calculates the reliability that is an empty region according to the average value, peak value, or variance value of the histogram. For example, as shown in FIG. 10B, when the variance of the histogram is large, there is a high possibility that subjects in the background area as well as the sky are mixed, and the image processing unit 103 has a reliability of being an empty area. Judge as low. In this case, the image processing unit 103 reduces the reliability.

  Subsequently, the image processing unit 103 determines whether or not the reliability of the empty region candidate is equal to or higher than a predetermined threshold (reliability threshold) (step S505). If the reliability is equal to or higher than the reliability threshold (YES in step S505), the image processing unit 103 determines that the reliability related to the pixels of the sky region candidate is high, and performs region division related to the sky region and the background region. No (step S506). Then, the image processing unit 103 ends the area determination process.

  On the other hand, if the reliability is less than the reliability threshold (NO in step S505), the image processing unit 103 determines that the reliability related to the sky region candidate pixel is low, and uses the sky region candidate pixel as the background region. The area is divided (step S507). Then, the image processing unit 103 ends the area determination process.

  As described above, in the third embodiment of the present invention, since the region division related to the sky region and the background region is not performed according to the reliability indicating the reliability of the sky region with respect to the sky region candidate, the sky region In addition, false detection can be reduced for the background region.

  As is clear from the above description, in the example shown in FIG. 1, the control unit 104 and the image processing unit 103 function as a dividing unit, a determining unit, a calculating unit, and a region integrating unit. Here, at least the image processing unit 103 and the control unit 104 constitute an image processing apparatus.

  As mentioned above, although this invention was demonstrated based on embodiment, this invention is not limited to these embodiment, Various forms of the range which does not deviate from the summary of this invention are also contained in this invention. .

  For example, the function of the above embodiment may be used as a control method, and this control method may be executed by the image processing apparatus. In addition, a program having the functions of the above-described embodiments may be used as a control program, and the control program may be executed by a computer included in the image processing apparatus. The control program is recorded on a computer-readable recording medium, for example.

  Each of the control method and the control program has at least a division step, a determination step, a calculation step, and a region integration step.

  The present invention can also be realized by executing the following processing. That is, software (program) that realizes the functions of the above-described embodiments is supplied to a system or apparatus via a network or various recording media, and the computer (or CPU, MPU, etc.) of the system or apparatus reads the program. To be executed.

DESCRIPTION OF SYMBOLS 101 Image pick-up part 102 A / D conversion part 103 Image processing part 104 Control part 105 Display part 106 Storage medium

Claims (10)

An image processing apparatus for determining and extracting a specific area as a specific area from an input image,
A dividing unit that divides the input image into a plurality of regions including the specific region according to a feature amount indicating a feature of the input image;
A determination unit that determines whether each of the areas excluding the specific area is adjacent to the specific area under a predetermined condition, and determines the area adjacent to the predetermined area as a specific area candidate When,
Calculating means for calculating a reference pixel value as a reference pixel value according to a pixel value in the specific area, and further calculating a representative pixel value representing the specific area candidate according to a pixel value in the specific area candidate; ,
Region integration for determining whether or not to integrate the specific region candidate into the specific region according to the reference pixel value and the representative pixel value and integrating the specific region candidate into the specific region based on the determination result Means,
An image processing apparatus comprising: The feature amount is hue luminance information indicating hue and luminance,
The image processing apparatus according to claim 1, wherein the dividing unit sets a region having a predetermined area or more among the plurality of regions as the specific region.   The area integration unit integrates the specific area candidates into the specific area when a difference between the representative pixel value and the reference pixel value is equal to or less than a predetermined threshold. The image processing apparatus described.   The area integration unit is configured such that a difference between the representative pixel value and the reference pixel value is equal to or less than a predetermined difference threshold value, and a distance between the specific area candidate and the specific area is equal to or less than a predetermined distance threshold value. The image processing apparatus according to claim 1, wherein if there is, the specific area candidate is integrated into the specific area.   The area integration unit obtains a reliability indicating that the specific area candidate is the specific area according to the reference pixel value and the representative pixel value, and the reliability is equal to or higher than a predetermined reliability threshold. The image processing apparatus according to claim 1, wherein the specific area candidate is integrated into the specific area.   The image processing apparatus according to claim 5, wherein the region integration unit obtains, as the reliability, a probability that the specific region candidate is the specific region according to a hue indicated by the representative pixel value.   The image processing apparatus according to claim 1, wherein the reference pixel value and the representative pixel value are average values of pixel values in the specific region and the specific region candidate, respectively.   The image processing apparatus according to claim 1, wherein the specific area is an empty area indicating sky. A control method for an image processing apparatus for determining and extracting a specific area from an input image as a specific area,
A division step of dividing the input image into a plurality of regions including the specific region according to a feature amount indicating a feature of the input image;
A step of determining whether or not each of the areas excluding the specific area is adjacent to the specific area under a predetermined condition, and setting the area adjacent to the predetermined condition as a specific area candidate When,
Calculating a reference pixel value as a reference pixel value according to a pixel value in the specific area, and calculating a representative pixel value representing the specific area candidate according to a pixel value in the specific area candidate; ,
Region integration for determining whether or not to integrate the specific region candidate into the specific region according to the reference pixel value and the representative pixel value and integrating the specific region candidate into the specific region based on the determination result Steps,
A control method characterized by comprising: A control program used in an image processing apparatus that identifies and extracts a specific area as a specific area from an input image,
In the computer provided in the image processing apparatus,
A division step of dividing the input image into a plurality of regions including the specific region according to a feature amount indicating a feature of the input image;
A step of determining whether or not each of the areas excluding the specific area is adjacent to the specific area under a predetermined condition, and setting the area adjacent to the predetermined condition as a specific area candidate When,
Calculating a reference pixel value as a reference pixel value according to a pixel value in the specific area, and calculating a representative pixel value representing the specific area candidate according to a pixel value in the specific area candidate; ,
Region integration for determining whether or not to integrate the specific region candidate into the specific region according to the reference pixel value and the representative pixel value and integrating the specific region candidate into the specific region based on the determination result Steps,
A control program characterized by causing

Images