JP6331258B2 - Imaging apparatus and method - Google Patents

Description

  The present invention relates to a false color detection apparatus and method for detecting a false color.

  A digital camera that captures a subject image using a solid-state image sensor is known. The solid-state imaging device has a plurality of pixels, and one color filter is provided for each pixel. Since one pixel constituting the image has three color components, the insufficient color component is interpolated with the color components of the surrounding pixels. Therefore, when a subject image having a frequency higher than the interval between the pixels of the solid-state image sensor is incident, the pixels of the solid-state image sensor cannot accurately sample and interpolate the subject image, and a false color is generated in the image. Patent Documents 1 and 2 are known as documents disclosing means for detecting false colors. In Patent Document 1, edge detection is performed on an image signal output from a solid-state imaging device to detect a high-luminance part. Then, a false color is detected depending on whether or not the value of the color difference component in the region close to the detected part is within a predetermined range. Patent Document 2 detects a false color by comparing an image focused on a subject image with an image not focused.

JP 2002-262299 A JP 2011-109496 A

  However, in the configuration according to Patent Document 1, if the frequency of the subject image is too high, the solid-state imaging device cannot resolve the subject image and cannot capture the edge of the subject image. If the edge of the subject image cannot be captured, the edge cannot be detected, so that the false color cannot be detected. Further, in the configuration in which the false color is detected based on the value of the color difference component, there is a possibility that a part other than the false color is also detected. Further, in the configuration according to Patent Document 2, since there is a time lag at the moment when an image focused on the subject image and an image not focused are captured, it may be difficult to compare these two images. is there. In addition, for comparison, special processing such as extracting feature points from the two images and aligning the position of the subject image is required. This special processing complicates the apparatus and increases the cost.

  The present invention has been made in view of these problems, and an object thereof is to obtain a false color detection apparatus and method for detecting a false color included in an image.

  The false color detection device according to the first invention of the present application acquires image data representing the color of a pixel by a luminance component and a color difference component, determines a first false color region in which a false color is likely to occur in a predetermined color space, An arithmetic unit that determines a second false color area that is complementary to the first false color area, and whether or not a color difference component that forms a pixel of interest included in the image data is included in the first false color area. A first determination result is generated, a second determination result indicating whether or not the color difference component forming the target pixel is included in the second false color region is generated, and the first determination result and the second determination result are generated. And a detection unit that determines whether or not the image data includes a false color.

  The calculation unit rotates the matrix composed of the color difference components of the pixel of interest at a predetermined rotation angle to create a rotation color difference component, and the detection unit determines whether the rotation color difference component is included in the first false color region. It is preferable that a first determination result is generated, and a second determination result indicating whether or not the rotation color difference component is included in the second false color region is generated.

  The color space is preferably a plane represented by a two-dimensional orthogonal coordinate system defined by two orthogonal color difference components.

  The color space is preferably a plane represented by a two-dimensional polar coordinate system defined by the declination corresponding to the color difference and the radius corresponding to the saturation.

  The detection unit indicates whether or not the color difference component of the pixel of interest is included in the first false color region and the color difference component of the peripheral pixels within a predetermined range from the pixel of interest is included in the second false color region. A first final determination result is created, the color difference component of the target pixel is included in the second false color area, and the color difference components of peripheral pixels within a predetermined range from the target pixel are included in the first false color area It is preferable to create a second final determination result indicating whether or not the target pixel is a false color pixel based on the first final determination result and the second final determination result.

  The first final determination result indicates that the color difference component of the pixel of interest is included in the first false color area and the color difference component of the peripheral pixel within the predetermined range from the pixel of interest is included in the second false color area. And the second final color region includes a color difference component of the target pixel included in the second false color region and a color difference component of peripheral pixels within a predetermined range from the target pixel included in the first false color region. When the determination result indicates, the target pixel may be determined to be a false color pixel.

  The predetermined range may be determined according to the number of pixels included in the image data, and may be configured such that the predetermined range becomes wider as the number of pixels increases.

  The detection unit creates a second removal result obtained by removing high-frequency components from the second judgment result, and creates a first removal judgment result based on the second removal result and the first judgment result Creating a first removal result obtained by removing the high frequency component from the first judgment result, creating a second removal judgment result based on the first removal result and the second judgment result, Based on the first removal determination result and the second removal determination result, the target pixel may be determined to be a false color pixel.

  The detection unit is determined to be included in the second false color region in the second determination result and the luminance of the target pixel determined to be included in the first false color region in the first determination result. The first luminance that is an arithmetic average of the luminances of all the pixels is calculated, and the luminance of the pixel of interest that is determined to be included in the second false color region in the second determination result, and the first determination A second luminance that is an arithmetic average of all the pixels determined to be included in the first false color region in the result is calculated, and a difference between the first luminance and the second luminance is a threshold value. If it is smaller, the target pixel may be determined to contain a false color.

  The detection unit is determined to be included in the second false color region in the second determination result and the saturation of the target pixel determined to be included in the first false color region in the first determination result. Calculating the first saturation which is a root mean square with the saturation of all the pixels being present, and the saturation of the pixel of interest determined to be included in the second false color region in the second determination result; A second saturation which is a root mean square with the saturation of all the pixels determined to be included in the first false color region in the first determination result is calculated, and the first saturation and the second saturation are calculated. When the difference from the saturation of is smaller than the threshold value, it may be determined that the target pixel includes a false color.

  The calculation unit preferably determines the rotation angle according to the characteristics of the device that created the image data.

  The image data may be created by an image sensor that captures a subject image via a color filter provided on the optical path of incident light, and the calculation unit may determine the rotation angle according to the characteristics of the color filter.

  The computing unit may determine the rotation angle according to the image processing used when creating the image data.

  The range of the high-frequency component removed from the first determination result and the second determination result is determined according to the number of pixels included in the image data, and the range may be expanded as the number of pixels increases.

  An imaging device according to a second invention of the present application includes the false color detection device, an imaging device that creates image data, and a color filter that is provided on an optical path of light incident on the imaging device, and the arithmetic unit includes a color filter The rotation angle is determined in accordance with the characteristics.

  According to a third aspect of the present invention, there is provided a false color detection method comprising: obtaining image data representing a pixel color from a luminance component and a color difference component; and a first false color region in which a false color is likely to occur in a predetermined color space. Determining a second false color area having a complementary color relationship with the first false color area, and whether or not the first false color area includes a color difference component forming the target pixel. Creating a first judgment result, creating a second judgment result indicating whether or not the color difference component forming the target pixel is included in the second false color area, the first judgment result, and the first judgment result And determining whether or not the image data includes a false color based on the determination result of 2.

  A false color detection program according to a fourth invention of the present application includes a step of acquiring image data representing a pixel color from a luminance component and a color difference component, and a first false color region in which a false color is likely to occur in a predetermined color space. A step of determining, a step of determining a second false color region complementary to the first false color region, and a first indicating whether or not a color difference component forming the target pixel is included in the first false color region A first determination result, a second determination result indicating whether or not the color difference component forming the target pixel is included in the second false color region, the first determination result, and the second determination result And determining whether or not the image data includes a false color based on the determination result.

  According to the present invention, a false color detection apparatus and method for detecting a false color included in an image is obtained.

It is the block diagram which showed schematically the digital camera provided with the false color detection apparatus by 1st Embodiment. It is the figure shown about rotation conversion. It is the figure shown about rotation conversion. It is the figure shown about the 1st and 2nd intermediate determination result. It is the flowchart which showed the 1st false color determination process. It is the figure shown about LPF. It is the flowchart which showed the 3rd false color determination process by 3rd Embodiment.

  A false color detection apparatus according to a first embodiment of the present invention will be described with reference to FIGS. FIG. 1 shows a digital camera 100 equipped with a false color detection device. The configuration of the digital camera 100 will be described with reference to FIG.

  The digital camera 100 mainly includes a DSP 110, a CMOS 121 that is an image sensor, a focusing lens 131, and an LCD 122.

  The focusing lens 131 is composed of a plurality of lenses, is folded and stored in the body of the digital camera 100, and is extended from the body when the digital camera 100 is set to an operation mode for recording a captured image. The focusing lens 131 that has been extended focuses on the subject and forms an object image on the CMOS 121. A focus adjustment ring 132 and a drive motor 133 are provided on the outer peripheral surface of the focusing lens 131. The drive motor 133 drives the focusing lens 131 to focus on the subject. The focus adjustment ring 132 is rotated by the user to focus the focusing lens 131 on the subject.

  A diaphragm blade 123 and a shutter curtain 124 are provided on the optical path between the focusing lens 131 and the CMOS 121. The diaphragm blade 123 adjusts the amount of light reaching the CMOS 121, and the shutter curtain 124 opens and closes the optical path from the focusing lens 131 to the CMOS 121.

  The CMOS 121 includes a plurality of pixels 121a and a color filter 121b provided for each of the pixels 121a. One color filter 121b has one of three colors of red, green, and blue. The red, green, and blue color filters 121b are arranged according to an arrangement method such as a Bayer arrangement. The CMOS 121 receives the synchronous clock signal generated by the DSP 110 and operates in synchronization therewith. The plurality of pixels 121a pick up the formed subject image and output it as charges. The CMOS 121 continuously outputs image data constituted by the magnitude of electric charges to the DSP 110. That is, the image data expresses the color of each pixel 121a by the luminance component and the color difference component.

  The DSP 110 mainly includes a detection unit 111 and a calculation unit 112. The detection unit 111 and the calculation unit 112 detect a false color included in the image data. Details of these members will be described later.

  The DSP 110 determines whether or not the focusing lens 131 is focused on the subject based on the image data, and transmits a signal corresponding to the determination to the drive circuit 125. The drive circuit 125 operates the drive motor 133 according to the received signal. Thereby, the focusing lens 131 focuses on the subject.

  The DSP 110 performs predetermined processing on the image data while using the memory 127 as a temporary storage device, creates an image file and transmits the image file to the memory card 128, and continuously transmits the processed image data to the LCD 122. . The memory card 128 is an SD card, for example, provided so as to be detachable from the digital camera 100, and stores the received image file. The LCD 122 displays the received image data.

  An operation switch 126 is connected to the DSP 110. The operation switch 126 includes a shutter release button. The shutter release button is a two-stage switch that transmits a half-press signal to the DSP 110 when the user half-presses it, and transmits a full-press signal to the DSP 110 when the user fully presses it.

  When the DSP 110 receives the half-press signal, the aperture value and shutter speed are calculated according to the received image data, and the LCD 122 displays the corrected image. When the DSP 110 receives the full-press signal, the aperture blade 123 and the shutter curtain 124 are driven with the calculated aperture value and shutter speed, and the CMOS 121 causes the subject image to be captured. The image data obtained from the CMOS 121 is subjected to predetermined image processing and transmitted to the memory card 128. The memory card 128 stores the received image file.

  Next, the calculation unit 112 will be described with reference to FIGS. FIG. 2A shows a region 20 where false colors are likely to occur in a coordinate system defined by two color difference components. This coordinate system is a two-dimensional orthogonal coordinate system defined by two orthogonal color difference components and a two-dimensional polar coordinate system defined by a declination corresponding to the color difference and a radius corresponding to the saturation. Hereinafter, the color difference Cb is described as the horizontal axis and the color difference Cr as the vertical axis.

  The region 20 where false colors are likely to occur is appropriately determined according to the spectral sensitivity of the color filter 121b, color correction processing performed in image processing executed by the DSP 110, and the like. Spectral sensitivity and color correction processing vary depending on the device that created the image. Therefore, you may determine according to the apparatus which produced the image. A means for determining the region 20 will be described. First, a CZP (Circular Zone Plate) chart is imaged using a device whose rotation angle is to be determined. The image data obtained by this has a false color. Therefore, by using the luminance component and the color difference component included in the image data, a region where a false color is generated in the coordinate system shown in FIG. The size of the region 20 is changed according to the degree to which the false color is generated.

  Next, a rectangular first unrotated false color region 21b and second unrotated false color region 22b are set in the region 20 (see FIG. 2B). By approximating the region 20 to a rectangular region, it becomes easy to determine whether or not the color difference component of the image data is included. However, since the longitudinal directions of the first unrotated false color area 21b and the second unrotated false color area 22b are not parallel to the vertical axis but are inclined in parallel, the color difference components of the image data are included in these areas. If it is determined whether or not, the calculation becomes complicated and the calculation load increases. Therefore, the first rotated false color area 21c and the second rotated false color area 22c are created by rotationally converting the first unrotated false color area 21b and the second unrotated false color area 22b (FIG. 2 (c)). Since the longitudinal direction of the first rotation false color area 21c and the second rotation false color area 22c is parallel to the vertical axis, it is determined whether or not the color difference component of the image data is included in these areas. The calculation becomes easy and the calculation load can be suppressed. The first rotation false color region 21c and the second rotation false color region 22c are in a complementary color relationship with each other, and the length in the vertical direction matches the saturation, and the length in the horizontal direction is the hue and the hue. Match.

Similar to the means for determining the region 20, the rotation conversion angle is appropriately determined according to the spectral sensitivity of the color filter 121b, the color correction performed in the image processing executed by the DSP 110, and the like. That is, it is determined according to the device that created the image. A means for determining the rotation angle will be described. Similar to the means for determining the region 20, the CZP chart is imaged using the device whose rotation angle is to be determined, and the region where the false color is generated is determined. Then, when this region is rotationally converted, a rotation angle is determined so that the longitudinal direction of the region is parallel to the vertical axis. The determined rotation angle is a rotation angle unique to the device. When the color difference component before rotation is (Cb, Cr), the color difference component after rotation is (Cb ′, Cr ′), and the rotation angle is θ, the color difference component after rotation (Cb ′, Cr ′) is obtained by the following equation. expressed.
Cb ′ = Cb · cos θ−Cr · sin θ
Cr ′ = Cb · sin θ + Cr · cos θ
Using this equation, the calculation unit 112 rotationally converts the color difference component of the acquired image data to create a rotated color difference component. In other words, the calculation unit 112 rotationally converts the hue of the image data. This process is called a rotation conversion process.

  FIG. 3 is a diagram showing a region 30a where a false color of a device having characteristics different from that shown in FIG. A rectangular first unrotated false color region 31a and second unrotated false color region 32a are set in the region 30a (see FIG. 3A).

  While the region 20 in FIG. 2 extends in a straight line in the longitudinal direction, the region 30a is bent in a “<” shape. Therefore, even if the entire region 30a is rotated at a constant rotation angle, the longitudinal directions of the two rotation false color regions are not parallel to the vertical axis. Therefore, the first rotation angle θ1 such that the longitudinal direction of the first unrotated false color region 31a is parallel to the longitudinal axis, and the longitudinal direction of the second unrotated false color region 32a is parallel to the longitudinal axis. Such a second rotation angle θ2 is obtained. The obtained first rotation angle θ1 and second rotation angle θ2 are rotation angles unique to the device. The computing unit 112 rotationally converts the color difference component of the acquired image data using the first rotation angle θ1 and the second rotation angle θ2, and creates a rotation color difference component. In other words, the calculation unit 112 rotationally converts the hue of the image data. This process is called a rotation conversion process.

  Next, the detection unit 111 will be described with reference to FIGS. The detection unit 111 determines whether or not the color difference component of each pixel 121a included in the image data is included in the first rotation false color region 21c and the second rotation false color region 22c, and the image is determined according to the result. Determine whether the data contains false colors. That is, for a pixel 121a (a false color possibility pixel) that may have a false color, a false color possibility that is complementary to the false color possibility pixel within a predetermined range from the false color possibility pixel. If a pixel is present, it is determined that the false color possibility pixel has a false color. Hereinafter, this process will be described in detail.

  First, the detecting unit 111 selects one target pixel to be inspected from the pixels 121a constituting the image data. Then, it is determined whether or not the color difference component forming the target pixel is included in the first rotation false color region 21c. If it is included in the first rotation false color region 21c, the value 1 is output as the determination result, and if it is not included, the value 0 is output as the determination result. This process is executed for all the pixels 121a. A matrix obtained by arranging the output determination results so as to correspond to the arrangement of the pixels 121a is the first intermediate determination result. Similarly, it is determined whether or not the color difference component forming the target pixel is included in the second rotation false color region 22c, and a second intermediate determination result is created.

  Next, the detection unit 111 detects a target pixel having a value 1 in the first intermediate determination result, and sets the target pixel in the second intermediate determination result and the value of the pixel 121a within a predetermined range from the target pixel to 1. It is determined whether or not is included. FIG. 4A shows the target pixel 41 having a value 1 in the first intermediate determination result and the values of the pixel 121a included in the range of two pixels from the target pixel. FIG. 4B and FIG. 4C are a part of the second intermediate determination result, which is a matrix centered on the target pixel, and the target pixel having the value 1 in the first intermediate determination result. The corresponding pixel 42 and the value of the pixel 121a included in the range of two pixels from the target pixel are shown. 4B and 4C are 5 × 5 square matrices, and a range of two pixels from the target pixel corresponds to a predetermined range. The predetermined range is determined according to the number of pixels of the image data, and the predetermined range becomes wider as the number of pixels increases. FIG. 4B includes a pixel 121a having a value of 1. However, FIG. 4C does not include the pixel 121a that takes the value 1, and the values of all the pixels 121a take 0. As shown in FIG. 4B, among the target pixels having the value 1 in the first intermediate determination result, the target pixel in the second intermediate determination result and the value of the pixel 121a within a predetermined range from the target pixel. When 1 is included, it is determined that the target pixel includes a false color, and the value 1 is output as a determination result. On the other hand, as shown in FIG. 4C, among the target pixels having the value 1 in the first intermediate determination result, the target pixel in the second intermediate determination result and the pixels 121a within a predetermined range from the target pixel. When 1 is not included in the value, it is determined that the target pixel does not include a false color, and the value 0 is output as a determination result. This process is executed for all the pixels 121a. A matrix obtained by arranging the output determination results so as to correspond to the arrangement of the pixels 121a is the first final determination result. Similarly, the detection unit 111 detects a target pixel having a value of 1 in the second intermediate determination result, and sets the target pixel in the first intermediate determination result and the value of the pixel 121a within a predetermined range from the target pixel. It is determined whether 1 is included. A matrix obtained by arranging the determination results so as to correspond to the arrangement of the pixels 121a is the second final determination result.

  Then, the detection unit 111 adds the first final determination result and the second final determination result for each corresponding pixel to create a false color determination result. The false color determination result is the same matrix as the matrix created by the pixel 121a of the image data. When the value of the corresponding element in the matrix is 1, the pixel 121a is determined to be a false color. When the value of the corresponding element is 0, it indicates that the pixel 121a is determined not to be a false color.

  Next, the first false color determination process will be described with reference to FIG. The first false color determination process is a process executed by the calculation unit 112 and the detection unit 111.

  In the first step S51, the calculation unit 112 reads image data. The image data is image data composed of a luminance component and a color difference component as in the JPEG format, for example.

  In the next step S52, the calculation unit 112 performs rotation conversion processing on the hue of the image data to create a rotated color difference component.

  In the next step S53, the detection unit 111 creates a first intermediate determination result and a second intermediate determination result.

  In the next step S54, after the detection unit 111 creates the first final determination result and the second final determination result based on the first intermediate determination result and the second intermediate determination result, the first final determination is performed. The result and the second final determination result are added to create a false color determination result. Thereby, the pixel 121a having a false color is detected.

  According to the present embodiment, since the false color determination is performed using the first rotation false color area 21c and the second rotation false color area 22c that is complementary to the first rotation false color area 21c, It is possible to prevent a false color from being erroneously determined as a false color and to detect the false color with high accuracy.

  In addition, since the calculation unit 112 performs rotation conversion, it is possible to prevent the calculation from becoming complicated and to reduce the calculation load.

  Further, as the number of pixels of the image data increases, the predetermined range described above is expanded, that is, by increasing the size of the square matrix shown in FIGS. 4B and 4C, the number of pixels of the image data is increased. By setting a suitable range, it is possible to accurately detect a region including a false color.

  Next, a second embodiment will be described with reference to FIGS. In the second embodiment, the first final determination result and the second final determination result are calculated after the first intermediate determination result or the second intermediate determination result is subjected to LPF (low pass filter). Different from the embodiment. The same components as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted.

  Next, the detection unit 111 will be described with reference to FIG. The detection unit 111 determines whether or not the color difference component of each pixel 121a included in the image data is included in the first rotation false color region 21c and the second rotation false color region 22c, and multiplies the result by LPF. In accordance with the result, it is determined whether the image data includes a false color. Hereinafter, these processes will be described in detail. Since the procedure for creating the first intermediate determination result and the second intermediate determination result is the same as that in the first embodiment, the description thereof is omitted.

FIG. 6A shows an LPF having a size of 5 × 5, and FIG. 6B shows a second intermediate determination result. Here, assuming that the target pixel is A13, a value LPF_A13 obtained by multiplying A13 by LPF is obtained by the following equation.
LPF_A13 = L1 ・ A1 + L2 ・ A2 + L3 ・ A3 + L4 ・ A4 + L5 ・ A5 + L6 ・ A6 + L7 ・ A7 + L8 ・ A8 + L9 ・ A9 + L10 ・ A10 + L11 ・ A11 + L12 ・ A12 + L13 ・ A18 + L15 ・ A14 + L15 A25
Similar processing is performed on all the pixels 121a included in the second intermediate determination result to obtain a second removal result. Then, the first intermediate determination result and the second removal result are multiplied for each element at the same position. The result of placing the multiplication result at the same position as the multiplied element is the first removal determination result. As a result, only the pixel 121a that takes a non-zero value in both the first intermediate determination result and the second removal result takes a non-zero value in the first removal determination result. The pixel 121a that is 0 takes the value 0 in the first removal determination result. Similarly, the detection unit 111 multiplies the first intermediate determination result by LPF to obtain a first removal result. Then, the second intermediate determination result and the first removal result are multiplied for each element at the same position to obtain a second removal determination result. When LPF is applied to the first intermediate determination result and the second intermediate determination result, the region that takes a non-zero value in the first removal result and the second removal result is expanded. That is, if high frequency components are removed from the first intermediate determination result and the second intermediate determination result, the pixels included in the first false color region increase in the second determination result, and the pixels included in the second false color region Increases in the first determination result. That is, the area that may contain a false color is enlarged. Thereby, the area | region containing a false color can be detected more reliably.

  Then, the detection unit 111 adds the first removal determination result and the second removal determination result for each corresponding element to create a false color determination result. The false color determination result is the same matrix as the matrix created by the pixel 121a of the image data. When the value of the corresponding element in the matrix is 1, the pixel 121a is determined to be a false color, and the corresponding element When the value of is 0, it is determined that the pixel 121a is not a false color.

  FIG. 6C is an example of an LPF in which all elements have the same value. The first intermediate determination result and the second intermediate determination result created by determining whether or not they are included in the first rotation false color region 21c and the second rotation false color region 22c having a complementary color relationship are usually The pixel 121a included in the first rotation false color region 21c or the second rotation false color region 22c is not included at the same position. Here, when the second intermediate determination result is multiplied by LPF, a region that takes a non-zero value in the second removal result, that is, a region that may contain a false color, is larger than the second intermediate determination result. To do. As a result, the second removal result and the first intermediate determination result have a complementary color relationship.

  Further, as the number of times of applying the LPF to the first intermediate determination result and the second intermediate determination result is increased, a region that takes a non-zero value in the first removal result and the second removal result is expanded. Therefore, if the number of times of applying the LPF is increased as the number of pixels of the image data is increased, the area that may contain the false color can be expanded and the area containing the false color can be detected more reliably.

  Alternatively, as the size of the LPF is increased, a region that takes a non-zero value in the first removal result and the second removal result can be expanded. Therefore, if the size of the LPF is increased as the number of pixels of the image data is increased, the area that may contain the false color can be expanded, and the area containing the false color can be detected more reliably.

  Next, the second false color determination process will be described with reference to FIG. The second false color determination process is a process executed by the calculation unit and detection unit 111.

  Steps S51 to S53 are the same as the first false color determination process, and thus description thereof is omitted.

  In step S54, after the detection unit 111 creates the first removal determination result and the second removal determination result based on the first intermediate determination result and the second intermediate determination result, The false removal determination result is created by adding the second removal determination result. Thereby, the pixel 121a having a false color is detected.

  According to the present embodiment, the same effects as those of the first embodiment can be provided, and a false color can be detected using the LPF included in the conventional imaging apparatus. Therefore, conventional image processing can be performed without providing a new component. An IC can detect false colors with high accuracy.

  Further, by increasing the number of times of applying the LPF as the number of pixels of the image data increases, it is possible to expand the region that may contain the false color and more reliably detect the region that includes the false color.

  Furthermore, by increasing the size of the LPF as the number of pixels of the image data increases, it is possible to enlarge the area that may contain false colors and more reliably detect areas that contain false colors.

  Next, a third embodiment will be described with reference to FIGS. In the third embodiment, brightness and saturation are determined between the creation of the first intermediate judgment result and the second intermediate judgment result and the creation of the first removal judgment result and the second removal judgment result. This is different from the first and second embodiments. The same components as those in the first and second embodiments are denoted by the same reference numerals and description thereof is omitted.

  Next, the third false color determination process will be described with reference to FIGS. The third false color determination process is a process executed by the calculation unit 112 and the detection unit 111.

  Steps S51 to S53 are the same as the first and second false color determination processes, and a description thereof will be omitted.

  In the next step S71, the false color is determined using the luminance. More specifically, the luminance of the pixel of interest that takes the value 1 in the first intermediate determination result in FIG. 4A and the second intermediate determination in FIG. 4B that is complementary to the first intermediate determination result. In the result, the arithmetic average with the luminance of all the pixels 121a taking the value 1 is calculated. The obtained value is set as the first luminance. Similarly, the luminance of the pixel of interest that takes the value 1 in the second intermediate determination result and the luminance of all the pixels 121a that take the value 1 in the first intermediate determination result that is complementary to the second intermediate determination result. Calculate the arithmetic mean. The obtained value is set as the second luminance. Next, a difference between the first luminance and the second luminance is obtained, and when the difference is smaller than a predetermined threshold, it is determined that there is a possibility of including a false color, and in the first intermediate determination result, The value of the corresponding pixel 121a (target pixel) is maintained as 1. On the other hand, if the difference is greater than or equal to the threshold value, it is determined that there is no possibility of including a false color, and the value of the corresponding pixel 121a (target pixel) is changed from 1 to 0 in the first intermediate determination result. The pixel 121a in which the false color is generated uses a phenomenon in which the color difference components having a complementary color relationship are greatly different from each other, but the luminance components have a small difference. That is, it is determined that the pixel 121a having a large difference in color difference components and a small difference in luminance components may contain a false color. Accordingly, it is possible to prevent a color having a complementary color relationship from being erroneously determined as a false color when an object having an object having a complementary color relationship is imaged.

  In the next step S72, the false color is determined using the saturation. More specifically, the saturation of the pixel of interest that takes the value 1 in the first intermediate determination result of FIG. 4A and the second intermediate of FIG. 4B that is complementary to the first intermediate determination result. An arithmetic average with the saturation of all the pixels 121a having a value of 1 in the determination result is calculated. Saturation is the root mean square of the color difference Cb and the color difference Cr. The obtained value is set as the first saturation. Similarly, the saturation of the pixel of interest that takes the value 1 in the second intermediate determination result and the saturation of all the pixels 121a that take the value 1 in the first intermediate determination result that is complementary to the second intermediate determination result. The arithmetic average is calculated. The obtained value is set as the second saturation. Next, a difference between the first saturation and the second saturation is obtained, and if the difference is smaller than a predetermined threshold, it is determined that there is a possibility of including a false color, and the first intermediate determination In the result, the value of the corresponding pixel 121a (target pixel) is maintained as 1. On the other hand, if the difference is greater than or equal to the threshold value, it is determined that there is no possibility of including a false color, and the value of the corresponding pixel 121a (target pixel) is changed from 1 to 0 in the first intermediate determination result. The pixel 121a in which the false color is generated uses a phenomenon in which the color difference components having a complementary color relationship are greatly different from each other, but the saturation component is not much different. That is, it is determined that the pixel 121a having a large difference in color difference components and a small difference in saturation components may include a false color. Accordingly, it is possible to prevent a color having a complementary color relationship from being erroneously determined as a false color when an object having an object having a complementary color relationship is imaged.

  In the next step S54, after the detection unit 111 creates the first final determination result and the second final determination result based on the first intermediate determination result and the second intermediate determination result, the first final determination is performed. The result and the second final determination result are added to create a false color determination result. Thereby, the pixel 121a having a false color is detected.

  According to the present embodiment, it is possible to prevent a color having a complementary color relationship from being erroneously determined as a false color when an object having an object having a complementary color relationship is imaged. In addition, it is possible to prevent a subject having different luminance and saturation from being determined as a false color while having a complementary color relationship.

  In the third embodiment, only one of the false color determination using luminance and the false color determination using saturation may be executed in the third false color determination process.

  In the third embodiment, the root mean square of the color difference Cb and the color difference Cr is described as saturation, but the root mean square of the color difference Cb and the color difference Cr may be used as saturation.

  In any of the embodiments, the detection unit 111 does not use the first rotation false color region 21c and the second rotation false color region 22c, but the first non-rotation false color regions 21b and 31a and the second non-rotation color region 21c. The false color may be detected by determining whether or not the rotation false color regions 22b and 32a include the color difference component of the pixel 121a.

  Further, in step S51 of the first false color determination process and step S71 of the second false color determination process, a file of a format other than the JPEG format composed of the luminance component and the color difference component, for example, the RAW format may be read. At this time, for image data that is not composed of a luminance component and a color difference component, the image data is converted into a luminance component and a color difference component.

21b First unrotated false color area 21c First rotated false color area 22b Second unrotated false color area 22c Second rotated false color area 31a First unrotated false color area 32a Second unrotated false Color area 41 Pixel of interest 42 Pixel 100 Digital camera 110 DSP
111 detector 112 arithmetic unit 121 CMOS
121a pixel 121b color filter 122 LCD
123 Aperture blade 124 Shutter curtain 125 Drive circuit 126 Operation switch 127 Memory 128 Memory card 131 Focusing lens 132 Focus adjustment ring 133 Drive motor

Claims (18)

Image data representing pixel colors by luminance components and color difference components is acquired, a first false color area in which false colors are likely to be generated in a predetermined color space is determined, and a complementary color relationship with the first false color area is obtained. An arithmetic unit for determining a second false color region;
Setting a first rectangular area in the first false color area, setting a second rectangular area in the second false color area;
A first determination result indicating whether or not a color difference component forming a target pixel included in the image data is included in the first rectangular area is generated, and the color difference component forming the target pixel is the second rectangular area Create a second decision result that indicates whether or not
A false color detection device including a detection unit that determines whether or not a false color is included in the image data based on the first determination result and the second determination result. Image data representing pixel colors by luminance components and color difference components is acquired, a first false color area in which false colors are likely to be generated in a predetermined color space is determined, and a complementary color relationship with the first false color area is obtained. An arithmetic unit for determining a second false color region;
A first determination result indicating whether or not a color difference component that forms a target pixel included in the image data is included in the first false color region is created, and a color difference component that forms the target pixel is the second false color A second determination result indicating whether or not the image data is included in the color area is created, and it is determined whether or not the image data includes a false color based on the first determination result and the second determination result. ,further,
Indicates whether or not the color difference component of the target pixel is included in the first false color region and the color difference component of surrounding pixels within a predetermined range from the target pixel is included in the second false color region. A first final determination result is created, and a color difference component of the pixel of interest is included in the second false color region, and a color difference component of a peripheral pixel within a predetermined range from the pixel of interest is the first false color Whether or not the pixel of interest is a false color pixel is generated based on the first final determination result and the second final determination result, and a second final determination result indicating whether or not the pixel is included in the color region is created. A false color detection device including a detection unit for determining. The calculation unit rotates a matrix composed of color difference components of the target pixel at a predetermined rotation angle to create a rotation color difference component,
The detection unit creates a first determination result indicating whether or not the rotation color difference component is included in the first false color region, and whether or not the rotation color difference component is included in the second false color region. The false color detection apparatus according to claim 1, wherein a second determination result indicating whether or not is generated.   4. The false color detection device according to claim 1, wherein the color space is a plane represented by a two-dimensional orthogonal coordinate system defined by two orthogonal color difference components. 5.   5. The false color according to claim 1, wherein the color space is a plane represented by a two-dimensional polar coordinate system defined by a declination corresponding to a color difference and a radius corresponding to saturation. Detection device. The first false color region includes a color difference component of the target pixel included in the first false color region, and a color difference component of peripheral pixels within a predetermined range from the target pixel includes the first false color region. final decision result is shown, or the color difference component of the pixel of interest is included in the second false color region, and the false color region color difference component of the peripheral pixels is the first in from the pixel of interest within a predetermined range of 3. The false color detection device according to claim 2 , wherein the target pixel is determined to be a false color pixel when the second final determination result indicates that the pixel is included in the false color pixel.   The false color detection device according to claim 6, wherein the predetermined range is determined according to the number of pixels included in the image data, and the predetermined range becomes wider as the number of pixels increases.   The detection unit creates a second removal result obtained by removing high-frequency components from the second determination result, and performs first removal based on the second removal result and the first determination result. A determination result is generated, a first removal result obtained by removing high frequency components from the first determination result is generated, and a second result is obtained based on the first removal result and the second determination result. The removal determination result is created, and the target pixel is determined to be a false color pixel based on the first removal determination result and the second removal determination result. The false color detection device described.   The detection unit creates a second removal result obtained by increasing the pixels included in the first false color region in the second determination result, and the second removal result and the first determination result Based on the above, a first removal determination result is created, a first removal result obtained by increasing the number of pixels included in the second false color region in the first determination result is created, and the first removal result is created. A second removal judgment result is created based on the removal result and the second judgment result, and the pixel of interest is a false color pixel based on the first removal judgment result and the second removal judgment result. The false color detection device according to claim 1, which determines that   The detection unit includes the luminance of the pixel of interest determined to be included in the first false color region in the first determination result, and included in the second false color region in the second determination result. A first luminance that is an arithmetic mean of the luminances of all the pixels that are determined to be calculated, and the pixel of interest that is determined to be included in the second false color region in the second determination result is calculated. Calculating a second luminance that is an arithmetic average of the luminance and the luminance of all the pixels determined to be included in the first false color region in the first determination result; 6. The false color detection device according to claim 1, wherein when the difference between the second luminance and the second luminance is smaller than a threshold value, the target pixel is determined to include a false color.   The detection unit includes a saturation of a target pixel determined to be included in the first false color region in the first determination result, and is included in the second false color region in the second determination result. The first saturation, which is the root mean square with the saturation of all the pixels that are determined to be determined, is calculated, and the second determination result determines that the first saturation is included in the second false color region A second saturation that is a root mean square of the saturation of the pixel of interest and the saturation of all the pixels that are determined to be included in the first false color region in the first determination result is calculated. The false according to any one of claims 1 to 5 and 9, wherein when the difference between the first saturation and the second saturation is smaller than a threshold, it is determined that the target pixel includes a false color. Color detection device.   The false color detection device according to claim 3, wherein the calculation unit determines the rotation angle according to characteristics of a device that has created the image data.   The image data is created by an image sensor that captures a subject image via a color filter provided on an optical path of incident light, and the calculation unit determines the rotation angle according to characteristics of the color filter. 12. The false color detection device according to 12.   The false color detection apparatus according to claim 12, wherein the calculation unit determines the rotation angle according to image processing used when creating the image data.   The range of the high-frequency component to be removed from the first determination result and the second determination result is determined according to the number of pixels included in the image data, and the range is expanded as the number of pixels increases. The false color detection device described. The false color detection device according to claim 3,
An image sensor for creating the image data;
A color filter provided on an optical path of light incident on the image sensor,
The said calculating part is an imaging device which determines the said rotation angle according to the characteristic of the said color filter. Obtaining image data representing the color of a pixel from a luminance component and a color difference component;
Determining a first false color region in which a false color is likely to occur in a predetermined color space;
Determining a second false color area that is complementary to the first false color area;
Creating a first determination result indicating whether or not a color difference component forming a target pixel is included in the first false color region;
Creating a second determination result indicating whether or not a color difference component forming the target pixel is included in the second false color region;
Determining whether the image data includes a false color based on the first determination result and the second determination result;
Indicates whether or not the color difference component of the target pixel is included in the first false color region and the color difference component of surrounding pixels within a predetermined range from the target pixel is included in the second false color region. A first final determination result is created, and a color difference component of the pixel of interest is included in the second false color region, and a color difference component of a peripheral pixel within a predetermined range from the pixel of interest is the first false color Whether or not the pixel of interest is a false color pixel is generated based on the first final determination result and the second final determination result, and a second final determination result indicating whether or not the pixel is included in the color region is created. A false color detection method comprising: determining. Obtaining image data representing the color of a pixel from a luminance component and a color difference component;
Determining a first false color region in which a false color is likely to occur in a predetermined color space;
Determining a second false color area that is complementary to the first false color area;
Creating a first determination result indicating whether or not a color difference component forming a target pixel is included in the first false color region;
Creating a second determination result indicating whether or not a color difference component forming the target pixel is included in the second false color region;
Determining whether the image data includes a false color based on the first determination result and the second determination result;
Indicates whether or not the color difference component of the target pixel is included in the first false color region and the color difference component of surrounding pixels within a predetermined range from the target pixel is included in the second false color region. A first final determination result is created, and a color difference component of the pixel of interest is included in the second false color region, and a color difference component of a peripheral pixel within a predetermined range from the pixel of interest is the first false color Whether or not the pixel of interest is a false color pixel is generated based on the first final determination result and the second final determination result, and a second final determination result indicating whether or not the pixel is included in the color region is created. A false color detection program comprising the step of determining.

Images