JP2010191390A - Imaging apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To suppress degradation in the focus detection accuracy, even when correcting a distortion in an imaging element in which normal imaging pixels and AF pixels are arranged. <P>SOLUTION: This imaging apparatus includes the imaging element having a first pixel group that generates a first signal for image formation by photoelectrically converting a subject image formed by an imaging lens, and a second pixel group that generates a second signal for phase difference detection; a first distortion correcting part 201 that corrects an image distortion caused by the imaging lens, for a first signal obtained from the first pixel group; a second distortion correcting part 200 that corrects an image distortion caused by the imaging lens, for a second signal obtained from the second pixel group; a focusing position detecting part 203 that detects the focusing position of the imaging lens from the second signal corrected by the second distortion correcting part; and a control part 109 that causes the first distortion correcting part and second distortion correcting part to perform distortion correcting processes that are different. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a technique for suppressing a decrease in detection accuracy in focus detection using a solid-state imaging device.

  An imaging apparatus such as a digital camera or a digital video camera is generally provided with an autofocus mechanism that automatically performs focus control of an imaging lens so that the subject is always in focus. This autofocus mechanism is classified into a distance measurement method and a focus detection method, based on the principle of the focus method. The distance measuring method measures the distance to the subject and controls the lens position according to the distance. The focus detection method detects the focus on the imaging surface and controls the position of the lens at a focused position. As typical focus detection methods, there are a contrast detection method, a phase difference detection method, and the like. The principle of the phase difference detection method is disclosed in Patent Document 1.

  The technique of Patent Document 1 will be described with reference to FIGS. For example, at the time of focusing, as shown in FIG. 19A, the light a0, b0, c0 that has passed through each part of the imaging lens l converges on the imaging surface m, and as shown in FIG. 19B. An image Z0 in focus on the imaging surface m is obtained. Then, when a so-called front pin state in which the focus position is shifted from the in-focus state shown in FIG. 19, the light a1, b1, c1 that has passed through each part of the imaging lens l is obtained as shown in A of FIG. , It converges after the imaging surface m. And as shown to B of FIG. 20, each light will become another image Za1, Zb1, Zc1 in the imaging surface m, respectively. In a so-called rear pin state, as shown in FIG. 21A, the light a2, b2, c2 that has passed through each part of the imaging lens l converges before the imaging surface m, and is shown in FIG. 21B. In this way, each light on the imaging surface m becomes a different image Za2, Zb2, Zc2. In this case, the image shift direction is reversed between the front pin state and the rear pin state, and the shift direction and the shift amount, so-called defocus amount, can be calculated. Since the relationship between the defocus amount and the amount by which the lens is driven to the in-focus position is determined by the optical system, auto-focus control can be performed by driving the lens to the in-focus position.

  Regarding the defocus amount calculation processing in the phase difference detection method, Patent Document 2 discloses “MIN algorithm” as a known algorithm. FIG. 22 is a diagram showing a general internal structure of a camera for detecting a phase difference correlation by this algorithm. The light incident from the lens is reflected to the lower part of the apparatus by a sub-mirror attached behind the 45-degree mirror, which is the main mirror, and is separated into two images by a lens of a secondary optical system called a spectacle lens, and AF It is incident on the sensor. Then, the output data from these two AF sensors is taken in and the sensor output is correlated. If each AF sensor is sensor 1 and sensor 2, the data of sensor 1 is A [1] to A [n], and the data of sensor 2 is B [1] to B [n], the correlation amount U0 is

(Min (a, b) is the smaller value of a, b)
It is expressed. First, U0 is calculated. Next, as shown in FIG. 23, the correlation amount U1 between the data obtained by shifting the A image by 1 bit of the AF sensor and the data of the B image is calculated. This U1 is

(Min (a, b) is a small value of a, b)
It becomes. Thus, the correlation amount shifted by 1 bit is calculated one after another. If the two images match, the correlation amount is maximized, and the shift amount and direction for obtaining the maximum value are obtained. This value is the defocus amount. From this defocus amount, it is possible to calculate the lens drive amount at which the photographing lens is brought into focus.

  Incidentally, as an apparatus for performing phase difference detection type autofocusing, Patent Document 3 discloses a device as shown in FIG. That is, in a solid-state imaging device (hereinafter referred to as an image sensor) in which photoelectric conversion cells that convert an optical image in an optical system into an electrical signal are arranged in two dimensions, some photoelectric conversion cells (pixels S1 and S2 in FIG. 24). Are used for focus detection in the phase difference detection method. According to Patent Document 3, an image forming lens for a distance measuring sensor as shown in FIG. 22, a spectacle lens for generating a phase difference, and the like are not necessary, and the imaging apparatus can be reduced in size and cost can be reduced. Can do.

JP-A-4-267211 Japanese Patent Laid-Open No. 9-43507 Japanese Unexamined Patent Publication No. 2000-156823 JP-A-6-237413

  Since the optical lens has aberration, optical distortion occurs in the subject optical image formed on the image sensor via the optical lens. As a result, the video signal also becomes an image having distortion. As the optical distortion, there are “thread-wound distortion” as shown in FIG. 25A and “barrel distortion” as shown in FIG. These distortions are distortions such that image information that should originally be at the position indicated by the dotted line in FIG. 25 forms an image at the solid line position. As a correction process for correcting the distortion of the image signal accompanied with the optical distortion, there is a method of reading the image signal temporarily stored in the memory by shifting the read address according to the distortion.

Patent Document 4 describes a correction method for correcting distortion characteristics in which distortion increases as the distance from the lens center of a light beam input through an optical lens increases. FIG. 26 shows an example of an optical distortion characteristic that is a relationship between a relative distance (%) from the optical axis (the center of the lens) and a distortion rate D (%) in the zoom lens. The distortion rate D will be described with reference to FIG. A range surrounded by a rectangle is an image output position when no distortion occurs. A range surrounded by a barrel is an output position of an image in which distortion has actually occurred. For example, when the original image output position at the point A in FIG. 9 is the image output position A ′ where distortion does not occur, the image heights of A and A ′ (distances from the optical axis to A and A ′) are If h and h ′, respectively, the distortion rate at that time is
D (%) = h / h ′ (1)
It can be expressed as This distortion rate D increases with the distance from the optical axis, as shown in the characteristics of FIG. Further, the optical characteristics of a general lens have a distortion rate distributed concentrically with respect to the optical axis.

  By the way, the distortion correction based on the above distortion rate is applied to an imaging element having a configuration in which a normal imaging pixel (hereinafter referred to as a normal pixel) and a focus detection pixel (hereinafter referred to as an AF pixel) as in Patent Document 3 are arranged. In this case, if the AF pixel is subjected to the same processing as that of the normal pixel, there is a concern that the accuracy of focus detection deteriorates.

  The present invention has been made in view of the above-described problems, and an object of the present invention is to achieve focus detection accuracy even when distortion correction is performed in an imaging element having a configuration in which normal imaging pixels and AF pixels are arranged. It is to be able to suppress the decrease in the number of times.

  In order to solve the above-described problems and achieve the object, an imaging apparatus according to the present invention generates a first signal for generating an image by photoelectrically converting a subject image formed by an imaging lens. Imaging having a pixel group and a second pixel group that divides a pupil region of the imaging lens and photoelectrically converts a subject image from the divided pupil region to generate a second signal for phase difference detection A first distortion correction means for correcting distortion of an image generated by the imaging lens with respect to a first signal obtained from the element, the first pixel group, and a first signal obtained from the second pixel group; A second distortion correction unit that corrects distortion of an image generated by the imaging lens with respect to the second signal, and the second signal that has been corrected for distortion by the second distortion correction unit. Focusing position detection means for detecting the focusing position; Said first distortion correction means, in said second distortion correction means, and control means for causing the different distortion correction process, characterized in that it comprises a.

  According to the present invention, it is possible to suppress a decrease in focus detection accuracy even when distortion correction is performed in an imaging device having a configuration in which normal imaging pixels and AF pixels are arranged.

It is a functional block diagram showing the outline of the whole composition of the system concerning a 1st embodiment of the present invention. It is explanatory drawing which shows the detail of the digital signal processing part which performs distortion correction in the 1st Embodiment of this invention. It is a flowchart of the distortion correction process in the 1st Embodiment of this invention. It is explanatory drawing which shows how to arrange the image pick-up element in the 1st Embodiment of this invention. It is explanatory drawing which shows how to obtain | require the coordinate position before distortion correction in the 1st Embodiment of this invention. It is a figure which shows the interpolation method in AF pixel distortion correction of the 1st Embodiment of this invention. It is a figure which shows the interpolation method in the normal pixel distortion correction of the 1st Embodiment of this invention. It is a figure which shows the example of the area | region division of the image after distortion correction. It is explanatory drawing which shows the AF pixel interpolation method in the 1st Embodiment of this invention. It is explanatory drawing which shows the detail of the digital signal processing part which performs distortion correction in the 2nd Embodiment of this invention. It is a flowchart of the distortion correction process in the 2nd Embodiment of this invention. It is a figure which shows the interpolation method in the normal pixel distortion correction of the 2nd Embodiment of this invention. It is explanatory drawing which shows the detail of the digital signal processing part which performs distortion correction in the 3rd Embodiment of this invention. It is a flowchart of the distortion correction process in the 3rd Embodiment of this invention. It is explanatory drawing which shows the change of the to-be-photographed object in a normal pixel distortion correction process. It is a figure which shows how to take the AF frame in the 3rd Embodiment of this invention. It is a figure which shows the method of taking the AF frame in A1 vicinity of FIG. 8 in the 3rd Embodiment of this invention. It is a figure which shows the method of taking the AF frame in A1 vicinity of FIG. 8 in the 3rd Embodiment of this invention. It is explanatory drawing which shows a focusing state. It is explanatory drawing which shows a front pin state. It is explanatory drawing which shows a back pin state. It is a figure which shows schematic structure of a phase difference AF optical system. It is explanatory drawing of a correlation calculation (MIN algorithm). It is a figure which shows the example of the solid-state image sensor containing a ranging pixel. It is a figure which shows the example of an optical distortion. It is a figure which shows the example of an optical distortion characteristic. It is explanatory drawing of a distortion rate.

  DESCRIPTION OF EMBODIMENTS Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings.

(First embodiment)
FIG. 1 is a block diagram showing a system configuration of a digital still camera as an imaging apparatus according to the first embodiment of the present invention.

  In FIG. 1, reference numeral 100 denotes an optical system for forming a subject image, which includes an imaging lens, a shutter, a diaphragm, and the like. In the present embodiment, a zoom lens having a variable focal length is used as the imaging lens. Reference numeral 101 denotes an image sensor (imaging device). For example, it is constituted by a CCD (Charge Coupled Device), a CMOS sensor, or the like. In the present embodiment, the imaging element captures a subject and generates normal image data (first pixel group) (first pixel group) for generating image data (first signal) by JPEG compression or the like, and for focus detection. And an AF pixel (second pixel group for phase difference detection) that generates the above signal (second signal). In such an image sensor, it is assumed that AF pixels are arranged at intervals of 10 pixels with normal pixels in between as shown in FIG. The AF pixel divides the pupil region of the imaging lens, and photoelectrically converts the subject image from the divided pupil region to generate a signal for phase difference detection.

  Reference numeral 102 denotes a motor drive circuit, which generates a drive signal for driving the motor 114 by a control signal from the CPU 109. Reference numeral 114 denotes a motor that drives the optical system 100 by a drive signal from the motor drive circuit 102. Reference numeral 103 denotes an image sensor driving unit. When the image sensor 101 is constituted by a CCD, it generates horizontal and vertical transfer drive signals for reading out the photoelectrically converted electric signal. Reference numeral 104 denotes an analog processing unit. Correlated double sampling for removing reset noise contained in the input image pickup signal, variable gain amplification for amplifying the image pickup signal level with variable gain, and the like are performed. Reference numeral 105 denotes an A / D conversion unit that quantizes the analog signal output from the analog processing unit 104 into digital data. Reference numeral 107 denotes a DRAM which is a storage device for temporarily storing the image data output from the A / D conversion unit 105. The DRAM 107 also functions as a storage device for temporarily storing image data that has been signal-processed by a digital signal processing unit 108 described later. Reference numeral 108 denotes a digital signal processing unit that performs image processing on the digital image data output from the A / D conversion unit 105. In general, the image data is subjected to white balance adjustment, gain adjustment, filtering processing, interpolation processing, and the like, resized to an appropriate image size, compressed, and recorded on the media 112. Alternatively, a process for displaying on the LCD or the like of the display unit 111 is performed. A timing generator 106 generates and transmits timing signals to the image sensor driving unit 103, the analog processing unit 104, the A / D conversion unit 105, and the digital signal processing unit 108. Reference numeral 109 denotes a CPU that controls the entire system. Reference numeral 110 denotes an operation unit such as a switch operated by the user from the outside. A ROM 113 stores CPU instructions and the like. In the present embodiment, a distortion rate used for distortion correction is stored in the ROM.

  FIG. 2 is a diagram showing in detail a block for performing distortion correction for the inside of the digital signal processing unit 108 in FIG. 1, and is a diagram for explaining the operation of the distortion correction processing. FIG. 3 illustrates a processing flow in the present embodiment in which AF pixels are subjected to distortion correction, focus detection processing (focus position detection processing) is performed, and distortion is corrected for captured normal pixels after focusing. FIG. The function of the distortion correction block will be described with reference to FIG.

  In FIG. 2, reference numeral 200 denotes an AF pixel distortion correction address generation unit (second distortion correction unit) that generates an address used when distortion correction is performed on the AF pixel. Reference numeral 201 denotes a normal pixel distortion correction address generation unit (first distortion correction unit) that generates an address used when distortion correction is performed on a normal pixel. Reference numeral 209 denotes an AF pixel interpolation address generation unit that generates an address used when AF pixel interpolation is performed. Note that the interpolation of the AF pixel means an operation of interpolating the AF pixel using a signal of a surrounding normal pixel because the AF pixel is treated as a defective pixel in normal image shooting.

  A selector 202 receives an instruction from the CPU 109 and selects an address to be transmitted to the DRAM 107. Reference numeral 203 denotes a defocus amount calculation unit that accumulates AF pixel data, divides the AF pixel data into a plurality of regions, and calculates a defocus amount corresponding to each region. Reference numeral 204 denotes an AF pixel interpolation value generation unit that calculates an interpolation value of the AF pixel from the normal pixel data designated by the AF pixel interpolation address generation unit 209. A signal processing unit 205 performs signal processing on data written in the DRAM 107 after AF pixel interpolation processing. The data recorded in the DRAM after the AF pixel interpolation process is pixel data (hereinafter referred to as RGB data) in a dot sequential Bayer array that is generally used in a CCD. The signal processing unit 205 converts the RGB data into YUV data. And write back to DRAM again.

  Reference numeral 206 denotes a normal pixel distortion correction value generation unit that calculates a distortion correction value for a normal pixel read based on the address calculated by the normal pixel distortion correction address generation unit 201. Reference numeral 207 denotes a memory controller that selects data and addresses exchanged between the DRAM and each processing unit based on the processing flow. DRAM addresses other than distortion correction and AF pixel interpolation processing can also be created. Reference numeral 208 denotes a media control unit for writing normal pixel data whose distortion has been corrected to the media 112.

A method for generating distortion correction data will be described in detail with reference to FIG. A is an output position of an image where distortion has actually occurred, and A ′ is an image output position corresponding to A where distortion does not occur. Assuming that an image in which distortion does not actually occur is a coordinate A ′ (x2, y2) that is separated from the image center O by a distance r ′, the image in which distortion has occurred is expressed by D * r from the image center O using Equation (1). The image is formed at the coordinate A (x1, y1) of the distance '. Therefore, the coordinates of A ′ use the distortion rate D,
(X2, y2) = (x1 / D, y1 / D) (2)
It can be expressed as. Therefore, when the pixel at the coordinate position (x2, y2) is obtained by distortion correction, the pixel (x1, y1) before distortion correction calculated backward from Equation (2) is read, and the pixel at (x2, y2) is read. Replace it. However, the calculation result of the coordinates of A (x1, y1) calculated backward from Equation (2) is rarely an integer value coordinate, and there are often no pixels at the calculated coordinate position.

  Therefore, it is necessary to read out the peripheral pixel of the calculated coordinates (x1, y1) and obtain the pixel value by interpolation. As an interpolation method used in this case, there are methods such as a nearest neighbor method, a linear interpolation method, and a cubic interpolation method. In the case of normal pixels, interpolation from many pixels using a polynomial often improves image quality. However, when such polynomial interpolation is performed on AF pixels, there is a concern that the accuracy of focus detection deteriorates.

  Therefore, in this embodiment, pixel data after distortion correction is obtained by interpolation calculation of four neighboring pixels for normal pixels, and pixel data after distortion correction is obtained by replacing the AF pixels with the nearest AF pixel data. Hereinafter, how to obtain the pixel values of the AF pixel and the normal pixel will be described.

As for the AF pixel, the AF pixel at the nearest coordinates of the coordinate position (x1, y1) before distortion correction obtained by calculation is read from the DRAM 107. In FIG. 6, if the AF pixels are arranged as B, C, D, and E, and the coordinates before correction of the AF pixels are A, A is replaced with the nearest D value. Therefore, the corrected value is A '= D
Can be expressed as

For normal pixels, pixel data is obtained by interpolation of four neighboring pixels. In FIG. 7, it is assumed that normal pixels are arranged as W, X, Y, and Z, and the coordinates before correction of the normal pixels are A. W, X, Y, and Z, which are four normal pixels in the horizontal and vertical vicinity of A, are read out and obtained from these four pixel data by interpolation calculation. When the distance between A and each pixel is represented as k, l, m, and n in FIG.
A ′ = (lnW + lmX + knY + kmZ) / (k + 1) (m + n) (3)
It can be expressed as.

  The distortion rate D is stored in the ROM 113 in advance. However, since the distortion rate varies depending on the zoom magnification, a plurality of distortion rates are recorded in the ROM 113 according to the zoom magnification.

  The operation of the present embodiment will be described using the flowchart of FIG. First, the AF pixel distortion correction flow will be described.

  In step S300, distortion correction processing is started. Since the distortion of the AF pixel is corrected, the CPU 109 instructs the selector 202 to select the address of the AF pixel distortion correction address generation unit 200 and designates the coordinate position for obtaining the AF pixel value. In step S301, the AF pixel distortion correction address generation unit 200 reads the AF pixel distortion rate D corresponding to the designated coordinate position and zoom magnification from the ROM 113, and calculates the coordinate position of the AF pixel before distortion correction using the formula (2). ) To calculate backward. In step S302, as shown in FIG. 6, the nearest AF pixel of the coordinate position inversely calculated by equation (2) is read from the DRAM 107 and replaced with the AF pixel after distortion correction. The AF pixels after distortion correction are accumulated in the defocus amount calculation unit 203. In step S303, it is determined whether the AF pixel whose distortion is to be corrected is the final pixel. If the AF pixel is not the final pixel, the coordinate position is incremented and the process proceeds to step S301. If it is determined that the pixel is the last pixel, the process proceeds to step S304. In step S304, the defocus amount calculation unit 203 divides the image into a plurality of regions, for example, as shown in FIG. 8 for the accumulated AF pixel data after distortion correction. Then, the defocus amount is calculated for each region by the MIN algorithm described above. Based on the defocus amount, the CPU 109 drives the lens so as to focus on the region to be focused. After the lens is driven and brought into the in-focus state, the image is picked up again in step S305, the image is acquired in the DRAM, and the process proceeds to the interpolation process flow of the AF pixel A.

  Next, an AF pixel interpolation processing flow will be described. As described above, the AF pixel interpolation processing means that an AF pixel is treated as a defective pixel in normal image shooting, and therefore, an operation of interpolating it using a signal of a surrounding normal pixel is meant. . The reason why the AF pixel interpolation processing is performed here is that the pixel value after the AF pixel interpolation may be used in the subsequent normal pixel distortion correction.

  The CPU 109 sets the selector 202 to select the address of the AF pixel interpolation address generation unit 209, and designates the coordinate position of the image data acquired by re-imaging. In step S310, it is determined whether the pixel corresponding to the coordinate position is an AF pixel or a normal pixel. If it is determined that the pixel is an AF pixel, the process proceeds to step S311. If it is not an AF pixel, the same process is performed again in step S310. In step S311, the AF pixel interpolation address generation unit 209 calculates the addresses of normal pixels around the AF pixel used for AF pixel interpolation. In step S312, the normal pixel calculated in step S311 is read from the DRAM. In step S313, the AF pixel interpolation value generation unit 204 creates an interpolation value from normal pixels around the AF pixel. FIG. 9 shows the positions of normal pixels used for creating an interpolation value. The interpolation method differs depending on which RGB position the AF pixel S to be interpolated corresponds to. The left side (G pixel interpolation method), the middle (R pixel interpolation method), and the right side (B pixel interpolation method) in FIG. ) Is a method of selecting four normal pixels. When four normal pixels are selected, an average value is taken. In step S314, the AF pixel interpolation address generation unit 209 calculates the address of the AF pixel corresponding portion and writes the calculated average value there. In step S315, it is determined whether the AF pixel to be interpolated is the final pixel. If the AF pixel is not the final pixel, the coordinate position is incremented and the process proceeds to step S310. In the case of the final pixel, the process proceeds to the distortion correction flow for the B normal pixel.

  Finally, a normal pixel distortion correction flow will be described.

  In step S320, the signal processing unit 205 performs signal processing on the RGB data that has undergone the AF pixel interpolation processing to produce YUV data. In step S321, YUV data is written into the DRAM 107. A series of addresses used in these processes is generated from the memory controller 207. In step S322, it is determined whether the current coordinate position is the last pixel. If the current pixel position is not the last pixel, the coordinate position is incremented and the process proceeds to step S320. The CPU 109 sets the selector 202 to select the address of the normal pixel distortion correction address generation unit 201. In step S323, the normal pixel distortion correction address generation unit 201 reads the distortion rate D of the normal pixel from the ROM 113 based on the coordinate position of the image after distortion correction, and reverses the coordinate position before distortion correction using Equation (2). calculate. Since the calculated coordinate position is not an integer value in most cases, as described with reference to FIG. 7, the addresses of four nearby coordinate points are calculated. In step S324, YUV data is read from the DRAM 107 from this address. In step S325, the normal pixel distortion correction value generation unit 206 generates YUV data after distortion correction from YUV data after distortion correction at four neighboring points by interpolation calculation. In step S326, the YUV interpolation data generated in step S324 is written into the DRAM. In step S327, it is determined whether the current coordinate position is the last pixel. If the current pixel position is not the last pixel, the coordinate position is incremented and the process proceeds to step S323. If it is the last pixel, the process proceeds to step S328. In step S328, the distortion-corrected YUV data is written to the medium 112 through the media control unit 208. At this time, display on the display unit 111 may be performed. Alternatively, compression processing (not shown) may be performed, the image data may be subjected to compression processing, converted into a file, and written to the medium 112. After writing the medium, the process ends in step S329.

(Second Embodiment)
FIG. 10 is a diagram illustrating in detail the digital signal processing unit 108 according to the second embodiment of the present invention. Since the content of FIG. 10 is substantially the same as FIG. 2, only different parts will be described.

  In the first embodiment, the normal pixel is subjected to signal processing once, and distortion correction is performed on the signal-processed YUV data. However, in this embodiment, distortion correction is performed on RGB data before signal processing. I do.

  Reference numeral 1006 denotes a normal pixel distortion correction value generation unit for reading the RGB image data written back to the DRAM 107 after the AF pixel interpolation process and performing the distortion correction process on the read RGB data.

  A normal pixel distortion correction address generation unit 1001 generates an address of a pixel used for normal pixel distortion correction. Also in the present embodiment, pixel data is obtained by interpolation processing from four neighboring pixels, as in FIG. 7 of the first embodiment. However, in the present embodiment, the way of taking pixel data differs depending on which of RGB is the normal pixel to be interpolated. That is, when the R pixel is interpolated, the left side of FIG. 12 is taken, when the G pixel is interpolated, the middle of the figure, and when the B pixel is interpolated, the neighboring four pixels of the same color are taken as shown by the black dot on the right side of FIG. . Then, using these four points, an interpolation value is calculated from Equation (3).

  Reference numeral 1005 denotes a signal processing unit for performing signal processing on data after normal pixel distortion correction. RGB data is converted to YUV data and written to DRAM. Other configurations are the same as those in the first embodiment.

  Next, the operation of this embodiment will be described using the flowchart of FIG.

  The flow from Step S1100 to Step S1115 is processing from AF pixel distortion correction to AF pixel interpolation, and is the same as Step S300 to Step S305 in FIG. 3 in the first embodiment. In the present embodiment, in order to perform distortion correction on RGB data, after performing AF pixel interpolation processing from step S1100 to step S1115, values before normal pixel distortion correction are calculated from step S1120 to step S1122. Thereafter, signal processing is performed in step S1123, RGB is converted into YUV, and YUV data subjected to signal processing in step S1126 is written to the medium 112.

(Third embodiment)
In the present embodiment, AF processing in which optical distortion is corrected is performed by changing the focus evaluation area in accordance with lens distortion.

  FIG. 13 is a diagram illustrating in detail the digital signal processing unit 108 in the third embodiment. Since the content of FIG. 13 is substantially the same as FIG. 2 of the first embodiment, only different parts will be described.

  In FIG. 2, the distortion correction address is calculated for each AF pixel. However, in this embodiment, the AF pixel does not calculate the distortion correction address. For this reason, reference numeral 1300 in FIG. 13 is an AF pixel address generation unit that only generates an AF pixel address. By not generating the AF pixel distortion correction address, the calculation amount can be reduced and the processing time can be shortened. Other configurations are the same as those in FIG. 2 of the first embodiment, but differ from the first embodiment in that the focus evaluation area in the defocus amount calculation unit 1303 changes due to lens distortion. For example, if the image after distortion correction is divided into regions as shown in FIG. 8, it is determined which AF pixel evaluation value is used for each area according to the distortion, and this mapping information is stored in the ROM 113. Keep it. The defocus amount calculation unit reads this mapping information from the ROM according to the distortion, evaluates the AF pixels in each area of FIG. 8, and calculates the defocus amount from these AF pixels.

  In the following, let us consider focusing on the area A1 of the image after normal pixel distortion correction in FIG.

  FIG. 15 is an enlarged image near the area A1 in FIG. Solid lines are images before distortion correction, and dotted lines are images after distortion correction. When focusing on the A1 area, if the four gray points in the figure, which are AF pixels corresponding to the A1 area before the AF pixel distortion correction, are used for AF, the image after the normal pixel distortion correction is focused outside the frame of the A1 area. They are aligned and the A1 area is not in focus. Therefore, in the present embodiment, the AF pixel distortion is corrected so that the A1 area is in focus by changing the AF frame. FIG. 16 is a diagram for explaining an example of how to set an AF frame according to the present embodiment. A dotted line indicates an AF evaluation frame when there is no distortion, a solid line indicates an AF evaluation frame when there is distortion, and a round dot indicates an AF pixel. FIG. 17 is an enlarged view of the upper left of FIG. In the present embodiment, when there is no distortion, the upper left area surrounded by the dotted line in FIG. 17 corresponds to the area A1 in FIG. Therefore, when there is no distortion, four AF pixels 00, 01, 10, and 11 within the dotted line are used for the focus evaluation of A1. When there is distortion, the upper left area surrounded by the solid line corresponds to the area A1 in FIG. Therefore, when there is distortion, the two AF pixels 11 and 20 in the solid line frame are used for the focus evaluation of A1. If more AF pixels are to be used for focus detection, AF pixels within a certain range outside the solid line frame may be taken. For example, like the one-dot chain line in FIG. 18, the outside of the AF frame when there is distortion may be enlarged, and two points of AF pixels 10 and 21 within the enlarged frame may be added and used for focus detection. AF pixel mapping information used for AF evaluation according to such optical distortion is stored in the ROM, and the value of this ROM is read out during the focus evaluation of the A1 area, and the AF pixel corresponding to the A1 area Are used for AF evaluation.

  In this way, mapping information corresponding to all areas is provided in the ROM, and AF pixels corresponding to the in-focus area are used for in-focus evaluation, so that the AF area of the image after normal pixel distortion correction is actually The AF area used for evaluation can be matched. In the present embodiment, the AF evaluation is performed using four and two AF pixels, but the AF evaluation may be performed using more AF pixels.

  The operation of the present embodiment will be described using the flowchart of FIG.

  The content of FIG. 14 is almost the same as FIG. However, in step S1401, the AF pixel is read without calculating the AF pixel distortion correction address, and before performing AF processing in step S1404, the defocus amount calculation unit 1303 performs distortion correction AF in step S1403. The only difference is that an evaluation frame is calculated.

Claims (3)

A first pixel group that photoelectrically converts an object image formed by the imaging lens to generate a first signal for generating an image, and a pupil region of the imaging lens are divided, and from the divided pupil region An image sensor having a second pixel group that photoelectrically converts a subject image to generate a second signal for phase difference detection;
First distortion correction means for correcting distortion of an image generated by the imaging lens with respect to the first signal obtained from the first pixel group;
Second distortion correction means for correcting image distortion caused by the imaging lens with respect to the second signal obtained from the second pixel group;
A focus position detecting means for detecting a focus position of the imaging lens from the second signal whose distortion is corrected by the second distortion correcting means;
Control means for causing the first distortion correction means and the second distortion correction means to perform different distortion correction processes;
An imaging apparatus comprising:   The first distortion correction means calculates a coordinate position of the first pixel group when the image is distorted from a coordinate position of the first pixel group when the image is not distorted; 2. The pixel value obtained by interpolation from pixel values in the vicinity of the calculated coordinate position is set as a pixel value at the coordinate position of the first pixel group when there is no distortion of the image. The imaging device described in 1.   The second distortion correction means calculates a coordinate position of the second pixel group when the image is distorted from a coordinate position of the second pixel group when the image is not distorted, The imaging apparatus according to claim 1, wherein a pixel value in the vicinity of the calculated coordinate position is set as a pixel value at the coordinate position of the second pixel group when there is no distortion of the image.

Images