JP5604980B2 - Red-eye determination device and red-eye determination method - Google Patents

Description

  The present invention relates to a red-eye determination device and a red-eye determination method.

  Red eyes may occur in an image captured by an imaging device such as a camera. Conventionally, there is a technique for predicting the possibility of red eyes occurring in a captured image, and imaging without turning on a flashlight when red eyes are predicted to occur.

  For example, Patent Document 1 predicts the possibility of red eyes occurring in a captured image using the luminance of the subject and the distance from the camera to the subject prior to imaging. Using such a technique, it is possible to easily predict the occurrence of red eyes.

JP 2009-157004 A

  However, in reality, the occurrence of red eyes may not be accurately predicted from information such as the brightness of the subject and the distance from the camera to the subject.

  It is an object of the present invention to provide a technique for predicting the occurrence of red eyes with a simple method prior to imaging.

  One embodiment of the present invention that solves the above problems includes a light emitting unit, an image sensor that outputs a captured image obtained by imaging a subject via a lens, a straight line that connects the light emitting unit and the subject, the subject, and the lens. It is predicted that a red eye will be generated in the red-eye occurrence prediction unit, and a red-eye occurrence prediction unit that calculates whether or not red-eye occurs in the captured image based on the angle formed by the straight line connecting A warning unit that issues a warning when the angle is formed based on the coordinates of the eye on the subject surface including the subject, the red-eye occurrence prediction unit calculated based on the position of the eye in the captured image The red-eye occurrence prediction unit calculates a first distance from the lens to the subject surface, a second distance from the image sensor to the lens, and a first position indicating the predetermined position of the lens. A second coordinate which is a coordinate indicating a predetermined position of the light emitting unit, which is determined based on the coordinates, and is determined based on a third coordinate indicating the center position of the captured image. A fourth coordinate indicating the position of the subject is calculated, and is determined from the first distance, the second distance, and a third distance between the third coordinate and the fourth coordinate. Calculating a fourth distance from a point on the optical axis of the lens on the subject surface to the position of the subject, and determining the fourth object distance based on the coordinates of the point on the optical axis of the lens on the subject surface. A fifth coordinate indicating a position is calculated based on the second coordinate, the third distance, and the fourth distance, and the first coordinate, the fifth coordinate, A red-eye determination characterized by calculating the angle formed from the second coordinates. It is the location.
  The red-eye occurrence prediction unit may predict that red-eye occurs in the captured image when the angle formed is narrower than a predetermined angle.
  The red-eye occurrence prediction unit uses the relational expression of the fourth distance = the third distance × (the first distance / the second distance) / the zoom magnification in the image sensor to calculate the fourth distance. The distance may be calculated.
  The warning unit may perform the warning at a timing when the imaging shutter is half-pressed.
  Another aspect of the present invention that solves the above-described problem is a red-eye determination method in an apparatus that includes a light-emitting unit and an image sensor that outputs a captured image obtained by imaging a subject via a lens. A red eye occurrence prediction step of calculating an angle formed by a straight line connecting the subject and a straight line connecting the subject and the lens, and predicting whether or not a red eye is generated in the captured image based on the formed angle; A warning step for giving a warning when red-eye occurrence is predicted in the red-eye occurrence prediction step, and the subject including the subject calculated based on the position of the eye in the captured image in the red-eye occurrence prediction step The angle formed is calculated based on the coordinates of the eye on the surface, and the red eye occurrence prediction step includes a first distance from the lens to the subject surface, and the image A second distance from the disensor to the lens, and a second coordinate that is a coordinate indicating the predetermined position of the light emitting unit, which is determined based on the first coordinate indicating the predetermined position of the lens. Calculating a fourth coordinate indicating the position of the subject in the captured image, which is determined based on the third coordinate indicating the center position of the captured image, and the first distance, the second distance, A fourth distance from a point on the optical axis of the lens on the object plane to the position of the subject is determined from a third distance between the third coordinate and the fourth coordinate. The fifth coordinate indicating the position of the subject, the second coordinate, the third distance, and the fourth distance determined based on the coordinates of the point on the optical axis of the lens on the subject surface. Distance based on the first coordinate and the fifth coordinate Calculating the coordinates, and the second coordinates, the angle from, characterized in that.
  In the red-eye occurrence prediction step, when the formed angle is narrower than a predetermined angle, it is predicted that red-eye is generated in the captured image.
  In the red-eye occurrence predicting step, the fourth distance = the third distance × (the first distance / the second distance) / the zoom magnification ratio in the image sensor is used to calculate the fourth distance. The distance may be calculated.
  In the warning step, the warning may be performed when the imaging shutter is half-pressed.

2 is a block diagram illustrating an example of a hardware configuration of the imaging apparatus 100. FIG. (A) It is a figure which shows the example of an external appearance at the time of seeing the imaging device 100 of this invention from the front. (B) It is a figure which shows the example of an external appearance at the time of seeing the imaging device 100 from the back surface. 2 is a block diagram illustrating an example of a functional configuration of the imaging apparatus 100. FIG. 2 is a diagram illustrating a positional relationship between the imaging apparatus 100 and a subject. FIG. (A) It is a figure which shows the positional relationship of the to-be-photographed object in the captured image and the center position of a captured image. (B) It is a figure which shows the positional relationship of a to-be-photographed object, the lens 173, and the image sensor 174. FIG. It is a flowchart for demonstrating an example of an imaging process. FIG. 11 is a diagram illustrating a modification of the imaging apparatus 100.

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

  The imaging apparatus 100 to which an embodiment of the present invention is applied functions as a general camera such as a digital camera. Further, the imaging apparatus 100 has a function of predicting whether or not red eyes will occur in the captured image prior to imaging, and performing a warning when it is predicted that red eyes will occur.

  FIG. 1 is a block diagram illustrating an example of a hardware configuration of an imaging apparatus 100 to which an embodiment of the present invention is applied. As illustrated, the imaging apparatus 100 includes a CPU (Central Processing Unit) 110, a volatile memory 120, a nonvolatile memory 130, an input device 140, a display 150, an interface 160, and an imaging mechanism 170. Prepare.

  The CPU 110 is an arithmetic device that controls various units to realize various functions of the imaging device 100. The CPU 110 implements various functions by loading a predetermined program stored in a storage medium such as the nonvolatile memory 130 into the volatile memory 120 and executing it.

  The volatile memory 120 is configured by a volatile storage medium such as a RAM (Random Access Memory), for example, temporarily stores various programs and data, and is used as a work memory when the CPU 110 performs calculations.

  The nonvolatile memory 130 is composed of a nonvolatile storage medium such as FlashROM, and stores various programs. For example, the nonvolatile memory 130 stores a program for imaging a subject, various default values used at the time of imaging, and the like. The non-volatile memory 130 also includes an external memory such as an SD card, and stores captured images.

  The input device 140 includes buttons for instructing imaging (hereinafter referred to as “imaging shutter”), a group of buttons for instructing various settings and determinations (hereinafter referred to as “operation buttons”), and the like. The input device 140 receives an instruction from the user via an imaging shutter or operation buttons.

  The display 150 includes a liquid crystal display, an organic EL display, or the like. The display 150 plays a role as a finder (electronic finder) in the imaging mode, and displays an image to be captured (that is, data obtained by converting an analog image signal obtained by the image sensor into a video signal). The display 150 displays a message for warning or announcement to the user, various setting screens, and the like.

  The interface 160 controls data exchange (reception and transmission) with various devices (not shown), and communicates with various devices via a USB cable or a network. Note that various devices include a control device (for example, a PC) that controls the imaging device 100, an information processing device (for example, a PC) that has a function of editing a captured image captured by the imaging device 100, and the like.

  The imaging mechanism 170 images a subject. For example, as illustrated, the imaging mechanism 170 includes a control circuit 171, a light emitting device 172, a lens 173, an image sensor 174, and a distance measuring device 175.

  The control circuit 171 controls each unit provided in the imaging mechanism 170 to control imaging.

  The light emitting device 172 illuminates a subject by turning on a general flash lamp (for example, a xenon lamp).

  The lens 173 is a general camera lens and condenses light from the outside on the image sensor 173.

  The image sensor 174 receives the light collected by the lens 173, accumulates electric charge according to the amount of received light, and outputs it as a captured image (analog signal) (for example, transfers to a control circuit 171 (not shown)). As the image sensor 174, for example, a CCD image sensor or a CMOS image sensor is used.

  The distance measuring device 175 includes a predetermined sensor (so-called distance sensor), and measures the distance from the lens 173 to the subject. For example, the distance measuring device 175 transmits a predetermined radio wave (ultrasonic wave, infrared ray, etc.) toward the subject, receives the reflected radio wave, and measures the distance from the lens 173 to the subject.

  FIG. 2A is an external view when the imaging device 100 is viewed from the front (the surface facing the subject during imaging). As shown in the drawing, the light emitting device 172 and the lens 173 are arranged at predetermined positions with a predetermined distance S therebetween. The predetermined distance S is a distance from the center position O of the lens 173 to the center position A of the light emitting device 172. In addition, the above-described imaging shutter 141 is provided above the imaging device 100.

  FIG. 2B is an external view when the imaging device 100 is viewed from the back surface (the surface facing forward during imaging). As illustrated, the operation button 142 and the display 150 described above are provided on the back surface of the imaging apparatus 100.

  The imaging apparatus 100 to which this embodiment is applied has the hardware configuration described above. However, the hardware configuration of the imaging apparatus 100 is not limited to this. For example, the imaging mechanism 170 of the imaging apparatus 100 may further include a finder (a sight window that is not an electronic finder) that is used to determine the composition by looking through the eyes.

  FIG. 3 is a functional configuration diagram of the imaging apparatus 100. As illustrated, the imaging apparatus 100 includes an input reception unit 101, a display processing unit 102, a red-eye occurrence prediction unit 103, a warning unit 104, and an imaging processing unit 105.

  The input receiving unit 101 receives an instruction from a user. For example, the input receiving unit 101 receives a focus adjustment instruction by half-pressing the imaging shutter 141. Further, the input receiving unit 101 receives an imaging instruction by long pressing the imaging shutter 141. The input receiving unit 101 receives various setting and determination instructions via the operation buttons 142.

  The display processing unit 102 displays image data (more precisely, a video signal converted from an image signal) obtained by the image sensor 174 in the imaging mode, and causes the display 150 to function as a viewfinder. In addition, the display processing unit 102 displays a message for giving a warning or announcement to the user, various setting screens, and the like on the display 150.

  The red-eye occurrence prediction unit 103 predicts whether or not red-eye occurs in the captured image (image data displayed by the display processing unit 102). Here, the red-eye occurrence prediction unit 103 predicts whether or not red-eye will occur at the timing when the input reception unit 101 receives a focus adjustment instruction (timing when the imaging shutter 141 is half-pressed) in the imaging mode. .

  In general, red eyes in a captured image are generated when light emitted from the light emitting device 172 is reflected by the eyes of a subject (for example, a person or an animal) and the reflected light enters the lens 173. However, even when reflected light with low intensity is incident on the lens 173, red eyes are not generated in the captured image, and when reflected light with high intensity is incident on the lens 173, red eyes are generated in the captured image.

  Therefore, the red-eye occurrence prediction unit 103 according to the present invention generates red-eye when the distance between the imaging device (specifically, the lens 173) 100 and the subject (specifically, the subject's eye) is equal to or greater than a predetermined reference value. When the distance between the imaging apparatus 100 and the subject is less than a predetermined reference value, it is predicted that red eyes will not occur. Note that the reference value used here is a highly reliable value obtained through experiments, and is, for example, 20 times the distance from the lens 173 to the light emitting device 172.

  In other words, the red-eye occurrence predicting unit 103 determines that the straight line connecting the light emitting device (specifically, the center position A of the light emitting device 172) 172 and the subject (specifically, the eye position B of the subject) and the subject And the lens (specifically, the center position O of the lens 173) 173 and the angle θ formed between the lens and the lens (specifically, the center position O of the lens 173) are predicted to generate red eyes when the angle θ is smaller than a predetermined reference angle R. When the angle is larger than the reference angle R, it is predicted that red eyes will not occur in the captured image. The predetermined reference angle R used here is a value when the distance between the imaging device 100 and the subject is 20 times the distance from the lens 173 to the light emitting device 172. Specifically, tan (R) = A value satisfying (1/20).

  That is, the red-eye occurrence prediction unit 103 calculates an angle θ formed by a straight line connecting the light-emitting device 172 and the subject and a straight line connecting the subject and the lens 173, and red-eye is generated in the captured image based on the calculated angle θ. Predict whether or not to do.

  Hereinafter, a method for calculating the angle θ formed by the straight line connecting the light emitting device 172 and the subject and the straight line connecting the subject and the lens 173 will be described.

  FIG. 4 is a diagram illustrating the positional relationship between the imaging apparatus 100 and the subject. As shown in the figure, a plane on which two orthogonal vectors (x axis, y axis) parallel to the front surface of the imaging device 100 (the surface facing the subject at the time of imaging) is stretched with the center position O of the lens 173 provided in the imaging device 100 as the origin. Is an “imaging device side plane”. A vector passing through the center position O of the lens 173 and orthogonal to the “imaging device side plane” is defined as the z-axis.

At this time, as described above, the light emitting device 173 is disposed at a position separated from the lens 172 by a predetermined distance S, and the position of the light emitting device 173, that is, the coordinates of the center position A of the light emitting device 173 (relative position with respect to the lens 172). Can be placed as (p, q, r). However, it is assumed that S = {p ² + q ² + r ² } ^(1/2) holds, and p, q, and r are real numbers, respectively. The coordinates of the center position A of the light emitting device 173 are stored in advance in a memory such as the non-volatile memory 130, and the red-eye generation prediction unit 103 reads out the coordinates from the memory (p, q, r). Can be obtained.

  Then, the red-eye occurrence prediction unit 103 acquires the distance from the lens 173 to the subject (not the eye position but a point on the optical axis of the lens 173 on the subject surface). Specifically, the red-eye occurrence prediction unit 103 uses the distance measurement device 175 to obtain the distance a from the lens 173 to the subject. At this time, the coordinates of the position O ′ of the point on the optical axis (z axis) of the lens 173 on the subject surface are (0, 0, a). However, a is a real number.

  A plane including the position O ′ and parallel to the “imaging device side plane” is referred to as a “subject side plane (subject plane)”, and the eye position B of the subject exists on the “subject side plane”. At this time, the coordinates of the eye position B of the subject can be set as (m, n, a). However, m and n are each real numbers. Further, the distance from the position O ′ to the eye position B of the subject is set to α.

  Then, the red-eye occurrence prediction unit 103 uses the two-dimensional coordinates (m, n) of the eye position B of the subject using image data (more precisely, a video signal converted from the image signal) obtained by the image sensor 174. To calculate.

  FIG. 5A is a diagram illustrating the positional relationship between the center position of image data (captured image) obtained by the image sensor 174 and the subject in the image data (captured image).

As illustrated, in the imaging mode, image data (captured image) obtained by the image sensor 174 is displayed on the display 150 by the display processing unit 102. Then, the red eye occurrence prediction unit 103 detects the position of the eye position B ′ of the subject in the image data (captured image) by a predetermined method. The present invention is not limited to a method for detecting the eye position in the image data. For example, the eye position is detected by comparison with a template image having eye characteristics. Then, the red-eye occurrence prediction unit 103 sets the center position O ″ of the image data (captured image) as the origin (0, 0), and the subject's eye position (center position O ″ in the image data (captured image)). The coordinates (m ′, n ′) of the relative position B ′ are obtained from the pixel position (number of pixels). Specifically, the red-eye occurrence prediction unit 103 multiplies the number of pixels from the center position O ″ to the eye position B ′ by a predetermined coefficient (the size of the image sensor, the number of color pixels, etc.) to obtain the position B The coordinates (m ', n') of 'are calculated. That is, m ′ = (number of pixels in the x-axis direction from the center position O ″ to the eye position B ′) × 24 × 10 ³ × 2832, n ′ = (from the center position O ″ to the eye position B ′) The coordinates (m ′, n ′) of the position B ′ are calculated using (number of pixels in the y-axis direction) × 24 × 10 ³ × 2832. At the same time, the red-eye occurrence prediction unit 103 obtains a distance β from the center position O ″ in the image data (captured image) to the eye position B ′ of the subject in the image data (captured image). Note that the distance β can be easily obtained by using the following equation: β = {m ′ ² + n ′ ² } ^(1/2) . Of course, a predetermined coefficient (the size of the image sensor, the number of color pixels, etc.) is determined by {(number of pixels in the x-axis direction from the center position O ″ to the eye position B ′) ² + (from the center position O ″) The distance β may be obtained by multiplying the number of pixels in the y-axis direction to the eye position B ′) ² } ^(1/2) .

  Then, the red-eye occurrence prediction unit 103 calculates the two-dimensional coordinates (m, n) of the actual position B of the subject's eyes based on the positional relationship between the subject, the lens 173, and the image sensor 174. The two-dimensional coordinates (m, n) represent the relative position with respect to the center position O of the lens 173.

  FIG. 5B is a diagram illustrating a positional relationship between the subject, the lens 173, and the image sensor 174. As shown in the figure, the light reflected by the subject's eyes (position B) passes through the lens 173 and forms an image at the position B ′ of the image sensor 174. That is, the distance α from the center position O ′ of the “subject side plane (subject plane)” to the eye position B of the subject is the distance β (from the center position O ″ shown to the position B ′ in the image sensor 174). (Distance). The reduction ratio (β / α) is determined by the distance e from the lens (center position O) 173 to the image sensor (center position O ″) 174 and from the lens (center position O) 173 to the subject (center position O ′). The distance a is equal to (or proportional to) the ratio. Therefore, in the simplest system, the relational expression β / α = e / a holds. The distance e from the lens (center position O) 173 to the image sensor (center position O ″) 174 is stored in advance in a memory such as the non-volatile memory 130, and the red-eye occurrence prediction unit 103 reads out from the memory. Thus, the distance e can be obtained.

  Then, the red-eye occurrence prediction unit 103 uses the above relational expression (β / α = e / a), and the distance from the center position O ′ of the “subject side plane (subject plane)” to the eye position B of the subject. α is calculated. Specifically, the red-eye occurrence prediction unit 103 substitutes β, e, a, and f previously acquired (or calculated) into the relational expression β / α = (e / a) × f. Then, the distance α from the center position O ′ of the “subject side plane (subject plane)” to the eye position B of the subject is calculated. Here, f indicates a set zoom magnification.

  Then, the red-eye occurrence prediction unit 103 sets the reciprocal (α / β) of the reduction ratio (β / α) to the coordinates (m ′, n ′) of the eye position B ′ of the subject in the image data (captured image). Multiply to obtain the two-dimensional coordinates (m, n) of the actual position B of the subject's eyes. That is, the two-dimensional coordinates (m, n) of the position B are obtained by the calculation formula of (m, n) = (m ′ × (α / β), n ′ × (α / β)). Here, since the z coordinate of the position B is a as described above, the three-dimensional coordinate of the position B is (m, n, a).

In this way, the red-eye occurrence prediction unit 103 can obtain the positional relationship (each coordinate) between the center position O of the lens 173, the center position A of the light emitting device 173, and the eye position B of the subject. Then, the red-eye occurrence prediction unit 103 determines a straight line (AB) connecting the light-emitting device 172 and the subject and a straight line (B-) connecting the subject and the lens 173 from the positional relationship (three-point coordinates) of O, A, and B. The angle θ formed by (O) can be calculated. For example, the red-eye occurrence prediction unit 103 calculates the inner product of the vector BA from the subject eye position B to the center position A of the light emitting device 173 and the vector BO from the subject eye position B to the center position O of the lens 173. Is calculated, and the angle θ formed is calculated. That is, cos (θ) is calculated using a relational expression of BA · BO = | BA | × | BO | × cos (θ). However, “·” represents an inner product, and | BA | = {(p−m) ² + (q−n) ² + (r−a) ² } ^(1/2) , | BO | = {(− m ) ² + (− n) ² + (− a) ² } ^(1/2) . Note that the red-eye occurrence prediction unit 103 converts the calculated cos (θ) into tan (θ).

  As described above, the red-eye occurrence prediction unit 103 can calculate the angle θ formed by the straight line connecting the light emitting device 172 and the subject and the straight line connecting the subject and the lens 173. Then, the red-eye occurrence prediction unit 103 determines whether or not red-eye occurs in the captured image based on the calculated angle θ (specifically, tan (θ)). Specifically, the red-eye occurrence prediction unit 103 determines that tan (θ) ≧ (1/20) when the calculated tan (θ) is equal to or greater than the predetermined reference value (1/20) described above. In this case, since the formed angle θ is equal to or larger than the predetermined reference angle R, it is determined that red-eye does not occur in the captured image. On the other hand, when the calculated tan (θ) is less than the predetermined reference value (1/20) described above, that is, when tan (θ) <(1/20), Since the formed angle θ is narrower than the predetermined reference angle R, it is determined that red-eye occurs in the captured image.

  In addition, the red-eye occurrence prediction unit 103 notifies the warning unit 104 of a prediction result as to whether or not red-eye occurs in the captured image.

  Returning to FIG. 3, the warning unit 104 issues a warning when it is predicted that red eyes will occur in the captured image. Specifically, the warning unit 104 sends a message to the effect that red eyes are generated via the display processing unit 102 when the red-eye generation prediction unit 103 is notified that red-eye occurs in the captured image (prediction result). Is displayed. Note that the warning method performed by the warning unit 104 is not limited to this. For example, the imaging shutter 141 may be locked so that the user cannot press it, or the user is notified that red eyes are generated by voice. May be.

  The imaging processing unit 105 controls the imaging mechanism 170 to image a subject. Specifically, in the imaging mode, the imaging processing unit 105 captures an image of a subject at a timing at which the input receiving unit 101 receives an imaging instruction (timing at which the imaging shutter 141 is held down). That is, the imaging processing unit 105 stores image data (captured image) obtained by the image sensor 174 in the imaging mode in a predetermined memory (for example, the non-volatile memory 130) at a timing when the input receiving unit 101 receives an imaging instruction. Store.

  The imaging apparatus 100 to which this embodiment is applied has the above-described configuration. However, the configuration of the imaging apparatus 100 is not limited to this.

  In addition, each of the above-described constituent elements is classified according to main processing contents in order to facilitate understanding of the configuration of the imaging apparatus 100. The present invention is not limited by the way of classification and names of the constituent elements. The configuration of the imaging apparatus 100 can be further classified into more components depending on the processing content. Moreover, it can also classify | categorize so that one component may perform more processes. Further, the processing of each component may be executed by one hardware or may be executed by a plurality of hardware.

  Next, a characteristic operation of the imaging apparatus 100 having the above configuration will be described.

<Imaging processing>
FIG. 6 is a flowchart for explaining an example of an imaging process executed by the imaging apparatus 100 of the present embodiment.

  The imaging apparatus 100 starts this flow at a predetermined timing. For example, this flow is started at a timing when an instruction to shift to the imaging mode is given.

  When this flow is started, the input receiving unit 101 of the imaging apparatus 100 determines whether there is a request for imaging processing (step S101). Specifically, the input receiving unit 101 waits until the imaging shutter 141 is half-pressed (step S101; No). When the imaging shutter 141 is half-pressed, a request for imaging processing is made. It determines with having been carried out (step S101; Yes), and transfers a process to step S102. When the imaging shutter 141 is pressed halfway, the input receiving unit 101 determines that an instruction for focus adjustment has been given, and the imaging processing unit 105 performs control such as focus adjustment.

  When the process proceeds to step S102, the red-eye occurrence prediction unit 103 determines the distance S between the lens 173 and the light emitting device 172, that is, the coordinates (p, q, r) of the relative position A of the light emitting device 172 with respect to the lens 173. Reading from a predetermined memory (step S102).

  Next, the red-eye occurrence prediction unit 103 acquires the distance a from the imaging device 100 to the subject (step S103). Specifically, the red-eye occurrence prediction unit 103 transmits predetermined radio waves (ultrasonic waves, infrared rays, etc.) from the distance measuring device 175 to the subject, receives the reflected radio waves, and receives the reflected radio waves from the lens 173 (that is, The distance a to the point on the optical axis of the lens 173 in the “subject side plane (subject plane)” is acquired.

  Then, the red-eye occurrence prediction unit 103 calculates an angle θ formed by a straight line (AB) connecting the light-emitting device 172 and the subject and a straight line (BO) connecting the subject and the lens 173 (step S104).

  Specifically, the red-eye occurrence prediction unit 103 uses the positional relationship (coordinates) between the center position O of the lens 173, the center position A of the light emitting device 173, and the eye position B of the subject to make an angle θ Is calculated.

  For this purpose, the red-eye occurrence prediction unit 103 sets the coordinates of the center position O of the lens 173 as the origin (0, 0, 0), and reads the coordinates of the center position A of the light emitting device 173 in step S102 (p , Q, r).

The red-eye occurrence prediction unit 103 reads a distance e from the lens (center position O) 173 to the image sensor (center position O ″) 174 from the memory in order to calculate the coordinates of the eye position B of the subject. The zoom magnification f specified by the user is specified. Further, the red-eye occurrence prediction unit 103 uses the captured image obtained by the image sensor 174 to calculate the number of pixels (from the x-axis direction and the position B ′ of the subject's eye in the captured image from the center position O ″ of the captured image). The number of pixels in the y-axis direction) is counted. Then, as described above, the red-eye occurrence prediction unit 103 determines that “m ′ = (number of pixels in the x-axis direction from the center position O ″ to the eye position B ′) × 24 × 10 ³ × 2832”, “n ′. = (Number of pixels in the y-axis direction from the center position O ″ to the eye position B ′) × 24 × 10 ³ × 2832 ”,“ β / α = (e / a) × f ”,“ m = m ′ The coordinates (m, n, a) of the actual position B are calculated for the subject's eyes using the calculation formulas of “× (α / β)” and “n = n ′ × (α / β)”. For reference, the calculation formulas used to calculate the coordinates (m, n, a) of the eye position B of the subject are summarized as follows: “m = (x-axis from the center position O ″ to the eye position B ′) Number of pixels in the direction) × 24 × 10 ³ × 2832 × (a / e) / f ”,“ n = (number of pixels in the y-axis direction from the center position O ″ to the eye position B ′) × 24 × 10 ³ × 2832 × (a / e) / f) ”.

  Then, the red-eye occurrence prediction unit 103 calculates the straight line (AB) connecting the light emitting device 172 and the subject and the straight line (BO) connecting the subject and the lens 173 from the calculated coordinates of the three points O, A, and B. And the angle θ formed by That is, as described above, the red-eye occurrence prediction unit 103 calculates cos (θ) and tan (θ) using the relational expression BA · BO = | BA | × | BO | × cos (θ). .

  Thereafter, the red-eye occurrence prediction unit 103 determines whether or not red-eye occurs in the captured image based on the angle θ (specifically tan (θ)) formed in step S104 (step S105). Specifically, the red-eye occurrence prediction unit 103 determines that tan (θ) calculated in step S104 is equal to or greater than the predetermined reference value (1/20) described above, that is, tan (θ) ≧ (1 / 20), it is determined that the formed angle θ is equal to or greater than the predetermined reference angle R, and no red-eye is generated in the captured image. Further, the red-eye occurrence prediction unit 103 determines that no red-eye is generated in the captured image even when the red-eye cannot be detected from the captured image in step S104 (including a case where the subject is sufficiently far away). On the other hand, the red-eye occurrence prediction unit 103 determines that tan (θ) calculated in step S104 is less than the predetermined reference value (1/20) described above, that is, tan (θ) <(1/20). In some cases, it is determined that the formed angle θ is narrower than a predetermined reference angle R, and it is determined that red-eye occurs in the captured image.

  Here, when it is determined that red eyes are generated in the captured image (step S105; Yes), the red-eye occurrence prediction unit 103 notifies the warning unit 104 of the determination result (prediction result), and the process is performed in step S106. Migrate to

  When the process proceeds to step S <b> 106, the warning unit 104 displays a message indicating that red eyes are generated via the display processing unit 102. At this time, for example, the display processing unit 102 displays a message such as “There is a possibility that red eyes may occur. Do you want to continue shooting?” On the display 150.

  Then, the warning unit 104 moves the process to step S107.

  On the other hand, when it is determined in step S105 that red-eye does not occur in the captured image (step S105; No), the process proceeds to step S107 without warning in step S106. .

  When the process proceeds to step S107, the imaging processing unit 105 determines whether an imaging instruction has been given via the input receiving unit 101 (step S107). Specifically, the imaging processing unit 105 determines that an imaging instruction has been given when the imaging shutter 141 is in a long-pressed state (step S107; Yes), and the process proceeds to step S108.

  When the process proceeds to step S108, the imaging processing unit 105 performs imaging (step S108). Specifically, the imaging processing unit 105 stores image data (captured image) obtained by the image sensor 174 in a predetermined memory (for example, the nonvolatile memory 130). Then, after imaging, the imaging processing unit 105 ends this flow.

  On the other hand, in step s107, when the imaging processing unit 105 returns to the state where the imaging shutter 141 is not pressed, the imaging processing unit 105 determines that an instruction to stop imaging is given (step S107; No), and ends this flow.

  By performing the above-described imaging processing, the imaging apparatus 100 according to the present embodiment performs data based on the intensity (the angle θ formed) of the light (flash) that is reflected by the eyes of the subject (for example, a person or an animal) and enters the lens 173. ) To predict the occurrence of red eyes. Therefore, although it is a simple method, the occurrence of red eyes can be predicted with higher accuracy than in the past.

  Note that each processing unit of each flow described above is divided according to main processing contents in order to make the imaging apparatus 100 easy to understand. The invention of the present application is not limited by the method of classification of the processing steps and the names thereof. The processing performed by the imaging apparatus 100 can be divided into more processing steps. One processing step may execute more processes.

  Moreover, said embodiment intends to illustrate the summary of this invention, and does not limit this invention. Many alternatives, modifications, and variations will be apparent to those skilled in the art.

  For example, in the above embodiment, the red-eye occurrence prediction unit 103 uses the three-dimensional coordinates (positional relationship) of the center position O of the lens 173, the center position A of the light emitting device 173, and the eye position B of the subject. Thus, an angle θ (an angle formed by a straight line connecting the light emitting device 172 and the subject and a straight line connecting the subject and the lens 173) is calculated. However, the present invention is not limited to this, and the angle θ formed may be calculated by another method. For example, the red-eye occurrence prediction unit 103 may calculate the angle θ formed using the two-dimensional coordinates of the center position O of the lens 173, the center position A of the light emitting device 173, and the eye position B of the subject. Good. Of course, in this case, a predetermined approximation or the like is required in the process of calculating the angle θ formed.

  FIG. 7 is a diagram illustrating a modified example of the imaging apparatus 100. The position at which the light emitting device 172 is disposed is not limited to the position illustrated in FIG. 2A, and may be disposed in the upper direction of the lens 173 as illustrated in FIG. That is, the present invention does not limit the position where the light emitting device 172 is disposed. The imaging apparatus 100 may store the distance S from the lens 173 to the light emitting device 172, that is, the relative position A (coordinates) of the light emitting device 172 with respect to the lens 173 in advance in the memory.

DESCRIPTION OF SYMBOLS 100 ... Imaging device, 101 ... Input reception part, 102 ... Display processing part, 103 ... Red eye generation | occurrence | production prediction part, 104 ... Warning part, 105 ... Imaging processing part, 110 ... CPU, 120 ... volatile memory, 130 ... non-volatile memory, 140 ... input device, 141 ... imaging shutter, 142 ... operation buttons, 150 ... display, 160 ... Interface, 170 ... Imaging mechanism, 171 ... Control circuit, 172 ... Light emitting device, 173 ... Lens, 174 ... Image sensor, 175 ... Distance measuring device

Claims (8)

A light emitting unit;
An image sensor that outputs a captured image obtained by imaging a subject via a lens;
A red-eye occurrence prediction that calculates an angle formed by a straight line connecting the light emitting unit and the subject and a straight line connecting the subject and the lens, and predicts whether or not red-eye occurs in the captured image based on the formed angle. And
A warning unit that issues a warning when red-eye occurrence is predicted to occur in the red-eye occurrence prediction unit,
The red-eye occurrence prediction unit calculates the angle formed based on the coordinates of the eye on the subject surface including the subject, calculated based on the position of the eye in the captured image ,
The red-eye occurrence prediction unit
A predetermined distance of the light emitting unit determined based on a first distance from the lens to the subject surface, a second distance from the image sensor to the lens, and a first coordinate indicating a predetermined position of the lens; A second coordinate which is a coordinate indicating the position of
Calculating a fourth coordinate indicating the position of the subject in the captured image, determined based on the third coordinate indicating the center position of the captured image;
A point on the optical axis of the lens on the object plane determined from the first distance, the second distance, and a third distance between the third coordinate and the fourth coordinate. A fourth distance from the subject to the subject position,
The fifth coordinate indicating the position of the subject, the second coordinate, the third distance, and the fourth distance determined based on the coordinates of a point on the optical axis of the lens on the subject surface. And calculated based on
Calculating the angle formed from the first coordinate, the fifth coordinate, and the second coordinate;
A red-eye determination device characterized by that. The red-eye determination device according to claim 1,
The red-eye occurrence prediction unit
Predicting that red-eye occurs in the captured image when the angle formed is narrower than a predetermined angle;
A red-eye determination device characterized by that. The red-eye determination device according to claim 1 or 2 ,
The red-eye occurrence prediction unit
The fourth distance is calculated using a relational expression of the fourth distance = the third distance × (the first distance / the second distance) / the zoom magnification in the image sensor.
A red-eye determination device characterized by that. The red-eye determination device according to any one of claims 1 to 3 ,
The warning part is
The warning is given when the imaging shutter is half pressed.
A red-eye determination device characterized by that. A red-eye determination method in an apparatus comprising: a light emitting unit; and an image sensor that outputs a captured image obtained by imaging a subject via a lens,
A red-eye occurrence prediction that calculates an angle formed by a straight line connecting the light emitting unit and the subject and a straight line connecting the subject and the lens, and predicts whether or not red-eye occurs in the captured image based on the formed angle. Steps,
A warning step for giving a warning when red-eye is predicted to occur in the red-eye occurrence prediction step;
And
In the red-eye occurrence prediction step, the angle formed based on the coordinates of the eye on the subject surface including the subject, calculated based on the position of the eye in the captured image ,
In the red-eye occurrence prediction step,
A predetermined distance of the light emitting unit determined based on a first distance from the lens to the subject surface, a second distance from the image sensor to the lens, and a first coordinate indicating a predetermined position of the lens; A second coordinate which is a coordinate indicating the position of
Calculating a fourth coordinate indicating the position of the subject in the captured image, determined based on the third coordinate indicating the center position of the captured image;
A point on the optical axis of the lens on the object plane determined from the first distance, the second distance, and a third distance between the third coordinate and the fourth coordinate. A fourth distance from the subject to the subject position,
The fifth coordinate indicating the position of the subject, the second coordinate, the third distance, and the fourth distance determined based on the coordinates of a point on the optical axis of the lens on the subject surface. And calculated based on
Calculating the angle formed from the first coordinate, the fifth coordinate, and the second coordinate;
A red-eye determination method characterized by the above. The red-eye determination method according to claim 5 ,
In the red-eye occurrence prediction step,
Predicting that red-eye occurs in the captured image when the angle formed is narrower than a predetermined angle;
A red-eye determination method characterized by the above. The red-eye determination method according to claim 5 or 6 ,
In the red-eye occurrence prediction step,
The fourth distance is calculated using a relational expression of the fourth distance = the third distance × (the first distance / the second distance) / the zoom magnification in the image sensor.
A red-eye determination method characterized by the above. The red-eye determination method according to any one of claims 5 to 7 ,
In the warning step,
The warning is given when the imaging shutter is half pressed.
A red-eye determination method characterized by the above.

Images