WO2011129387A1 - Image capture device - Google Patents

Image capture device Download PDF

Info

Publication number
WO2011129387A1
WO2011129387A1 PCT/JP2011/059243 JP2011059243W WO2011129387A1 WO 2011129387 A1 WO2011129387 A1 WO 2011129387A1 JP 2011059243 W JP2011059243 W JP 2011059243W WO 2011129387 A1 WO2011129387 A1 WO 2011129387A1
Authority
WO
WIPO (PCT)
Prior art keywords
imaging
image
sampling phase
pixel
distance
Prior art date
Application number
PCT/JP2011/059243
Other languages
French (fr)
Japanese (ja)
Inventor
大森圭祐
若林保孝
Original Assignee
シャープ株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by シャープ株式会社 filed Critical シャープ株式会社
Publication of WO2011129387A1 publication Critical patent/WO2011129387A1/en

Links

Images

Classifications

    • GPHYSICS
    • G03PHOTOGRAPHY; CINEMATOGRAPHY; ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ELECTROGRAPHY; HOLOGRAPHY
    • G03BAPPARATUS OR ARRANGEMENTS FOR TAKING PHOTOGRAPHS OR FOR PROJECTING OR VIEWING THEM; APPARATUS OR ARRANGEMENTS EMPLOYING ANALOGOUS TECHNIQUES USING WAVES OTHER THAN OPTICAL WAVES; ACCESSORIES THEREFOR
    • G03B19/00Cameras
    • G03B19/02Still-picture cameras
    • G03B19/04Roll-film cameras
    • G03B19/07Roll-film cameras having more than one objective
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/50Constructional details
    • H04N23/54Mounting of pick-up tubes, electronic image sensors, deviation or focusing coils
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N25/00Circuitry of solid-state image sensors [SSIS]; Control thereof
    • H04N25/48Increasing resolution by shifting the sensor relative to the scene

Definitions

  • the present invention relates to an imaging apparatus using a plurality of imaging systems.
  • imaging devices such as digital still cameras and digital video cameras have increased in the number of pixels and the resolution has been increased, and the optical system such as a lens has become large in order to obtain sufficient optical performance. There is a problem of becoming.
  • Patent Document 1 includes a displacement unit that displaces the image sensor as illustrated in FIG.
  • the horizontal direction and the vertical direction mean pixel arrangement directions as shown in FIG.
  • Patent Document 2 proposes a technique for improving the resolution by arranging a plurality of imaging systems so as to realize pixel shifting in a direction perpendicular to the pixel arrangement direction.
  • FIG. 23 shows a situation where a common subject 2301 is imaged by the two imaging systems 2302 and 2303.
  • FIG. 24 shows a region to be imaged by the imaging system 2302 by a grid 2401 and a region to be imaged by the imaging system 2303 by a grid 2402 (broken line), and each of the smallest squares is imaged by one pixel. Indicates the area.
  • the areas imaged by the pixels of the two imaging systems are almost overlapped, and the sampling phases are aligned.
  • FIG. 25 the area captured by each pixel of the imaging elements of the two imaging systems is shifted by 3/2 pixels in the horizontal direction and 1/2 pixel in the vertical direction, and the sampling phase is shifted.
  • the sampling phase here means the relative positional relationship between the image forming point on the image sensor at a certain point of the subject and the pixel in a plurality of imaging systems. For example, in FIG. 25, images are taken with a 3/2 pixel shift in the horizontal direction, and the horizontal sampling phase shift is 1/2 pixel as a decimal part. Pixel shift means a state in which the sampling phase is shifted.
  • an imaging system 2601 includes an optical system 2602 and an imaging element 2603
  • an imaging system 2605 includes an optical system 2606 and an imaging element 2607.
  • the optical system 2602 and the optical system 2606 are arranged on the same plane, and their optical axes are perpendicular to the plane
  • FIG. 26 is a view of the state from above.
  • the subject is located at a point P0 or P1 on the optical axis of the optical system 2602.
  • the subject image of the subject located at the point P0 is formed at the center of the pixel 2604 of the image sensor 2603 and is formed at the center of the pixel 2608 of the image sensor 2607.
  • the sampling phases are in a uniform state.
  • the imaging position on the imaging device 2607 shifts, and the difference in the shift amount becomes u0-u1.
  • u0 ⁇ u1 f ⁇ B ⁇ ((1 / H0) ⁇ (1 / H1)) (1) It can be expressed.
  • f is a focal length
  • B is a baseline length.
  • the base line length B is a distance between the principal point Q of the optical system 2602 and the principal point R of the optical system 2606.
  • the difference in the shift amount of the imaging position is about 3 ⁇ m from the equation (1).
  • the focal length f or the base line length B may be reduced so as not to shift the imaging position from the equation (1).
  • the base line length B is the distance between the principal points of the optical systems of the two image pickup systems.
  • the arrangement direction of the pixels of the image pickup device of the reference image pickup system is the horizontal direction and the vertical direction.
  • the horizontal component of the length B is expressed as a horizontal baseline length BH, and the vertical component is expressed as a vertical baseline length BV.
  • Patent Document 2 by arranging a plurality of imaging systems so that the baseline length in the horizontal direction and the baseline length in the vertical direction are reduced, the change in the sampling phase is small even if the imaging distance changes, and it does not depend on the imaging distance.
  • a technique for shifting pixels is proposed.
  • the sampling phase is uniform depending on the imaging distance, but the base line length in the X direction of the optical system 301a and the optical system 301b is sufficiently small. Even if the distance changes, the change in the sampling phase in the X direction can be kept small.
  • Patent Document 2 proposes a technique that arranges an imaging system so that a change in sampling phase with a change in imaging distance is small, and realizes pixel shifting regardless of the imaging distance.
  • the amount of displacement of the image sensor in the pixel shifting direction is about 1 ⁇ 2 pixel regardless of the imaging distance, but the baseline length is large in the direction perpendicular to the pixel shifting direction, and the distance between the images depends on the imaging distance. Since the amount of deviation changes, it is necessary to know the amount of deviation.
  • stereo matching is based on one image on multiple images taken from the same subject and the corresponding point (projection point of the same spatial point) is assigned to each pixel in the image in the other image. It is to search from.
  • Stereo matching includes various methods such as region-based matching for obtaining corresponding points using template matching and feature-based matching for extracting feature points such as edges and corners.
  • a high-resolution image is generated by arranging each pixel at a corresponding point on the high-resolution image based on the shift amount between the images thus obtained.
  • Patent Document 2 when a high-resolution image is generated by synthesizing images of imaging systems arranged at different heights in the vertical direction, corresponding points are shifted in the vertical direction in the image. Located in place. Therefore, when reading the image from the upper left of the image during processing by hardware, the timing to reach the region where the corresponding point exists in each imaging system is different, so it is necessary to hold the data, and the line memory or frame memory It is necessary to use.
  • Patent Document 2 in order to increase the resolution in all the horizontal and vertical directions, two eyes that are shifted by 1/2 pixel in the horizontal and vertical directions are necessary, and the cameras are juxtaposed in the horizontal or vertical direction. A thin configuration cannot be realized in one direction.
  • the present invention has been made in view of such circumstances, and each camera is arranged in one horizontal or vertical direction so that the amount of memory used is reduced when it is implemented as hardware, depending on the distance of the subject.
  • An object of the present invention is to provide an imaging apparatus capable of increasing the resolution.
  • the present invention has been made to solve the above-described problems.
  • the first technical means of the present invention is:
  • the image pickup system includes at least three image pickup systems each having an optical system and an image pickup element.
  • the image pickup system is arranged in a pixel arrangement direction of the image pickup element and has an image pickup distance within a predetermined range in a picked-up image of the first image pickup system.
  • a subject is imaged with a sampling phase in the arrangement direction shifted from 0.2 to 0.8 pixels with respect to a captured image of the first imaging system by any other imaging system. .
  • the second technical means is In the first technical means, the imaging systems are arranged at substantially equal intervals in the arrangement direction, and the optical axes of the imaging systems are substantially parallel.
  • the third technical means is In the first or second technical means, a subject within an imaging distance within a predetermined range is sampled in a direction perpendicular to the arrangement direction with respect to a captured image of any imaging system by any other imaging system. It is characterized in that the image is taken with a phase shift of 0.2 to 0.8 pixels.
  • the fourth technical means is In the first or second technical means, a subject within an imaging distance within a predetermined range is sampled in a direction perpendicular to the arrangement direction with respect to a captured image of any imaging system by any other imaging system. The image is picked up with a phase shift of approximately 0.5 pixels.
  • the fifth technical means is In the third or fourth technical means, a subject within an imaging distance within a predetermined range is sampled in a direction perpendicular to the arrangement direction with respect to a captured image of any imaging system by any other imaging system. It is characterized in that imaging is performed with almost no phase shift.
  • the sixth technical means is In the first or second technical means, a subject within an imaging distance within a predetermined range is captured in a direction perpendicular to the arrangement direction by any two other imaging systems with respect to a captured image of any imaging system.
  • the sampling phase is imaged with a shift of approximately 1/3 pixels.
  • the seventh technical means includes at least three image pickup systems having an optical system and an image pickup device, the image pickup systems are arranged on a substantially straight line, a plane normal including the straight line, an optical axis of the image pickup system, and Are substantially parallel, the straight line and one arrangement direction of the pixels of the imaging element are substantially parallel, and an object at an imaging distance within a predetermined range is captured by the imaging system of any of the imaging systems. It is characterized by being configured so as to be imaged with the sampling phase of the direction shifted.
  • an object at an imaging distance within a predetermined range is set to have a sampling phase in the arrangement direction of 0.2 to 0.8 by any one of the imaging systems. It is configured to be imaged with a pixel shift.
  • Ninth technical means is characterized in that, in the seventh technical means, the optical axes of the imaging system are arranged at equal intervals in the arrangement direction.
  • an object at an imaging distance within a predetermined range is moved in a direction perpendicular to the arrangement direction by any one of the imaging systems. It is characterized by being configured so as to be imaged with the sampling phase shifted.
  • a subject at an imaging distance within a predetermined range is perpendicular to the arrangement direction by any one of the imaging systems.
  • the sampling phase is configured to be imaged with a deviation of 0.2 to 0.8 pixels.
  • a subject at an imaging distance within a predetermined range is perpendicular to the arrangement direction by any one of the imaging systems.
  • the sampling phase is configured to be imaged with a shift of 0.5 pixels.
  • an object at an imaging distance within a predetermined range is moved in a direction perpendicular to the arrangement direction by any one of the imaging systems.
  • the sampling phase is configured to be imaged without deviation.
  • an object at an imaging distance within a predetermined range is moved in a direction perpendicular to the arrangement direction by any three imaging systems of the imaging system.
  • the sampling phase is configured to be imaged with a shift of 1/3 pixel at a time.
  • the imaging systems are arranged side by side in either one of the horizontal and vertical directions, and a plurality of obtained images are combined to form a pixel.
  • the imaging systems are arranged side by side in either one of the horizontal and vertical directions, and a plurality of obtained images are combined to form a pixel.
  • FIG. 7 is a graph showing a result of overlapping the sampling phase shift of FIG. 5 and the sampling phase shift of FIG.
  • FIG. It is a figure which shows a mode that two imaging systems juxtaposed in the horizontal direction are image
  • difference amount of the sampling phase in the case of QR 11.7mm. It is a graph which shows the deviation
  • shift amount of a sampling phase in case QR 11.0mm. It is a figure which shows the image imaged with two imaging systems juxtaposed in the horizontal direction. It is a figure which shows the image imaged with two imaging systems juxtaposed in the horizontal direction. It is a figure explaining the technique of patent document 1. FIG. It is a figure explaining the horizontal direction and vertical direction with respect to an image sensor.
  • FIG. It is a figure which shows a mode that the common to-be-photographed object is imaged with two imaging systems. It is the figure which showed the field which two imaging systems picturize with a lattice. It is a figure explaining the state of pixel shifting. It is a figure explaining the change of a sampling phase when imaging distance changes in a compound eye system. It is a figure explaining the baseline length BH of a horizontal direction, and the baseline length BV of a perpendicular direction. It is a figure explaining the compound eye system currently indicated by patent documents 2. FIG. It is a figure explaining the compound eye system currently indicated by patent documents 2. FIG. It is a graph which shows the result of having verified high resolution by two imaging systems.
  • FIG. 1 is a diagram showing the configuration of this embodiment.
  • FIG. 1 is a diagram illustrating an arrangement of three imaging systems provided in the imaging apparatus according to the present embodiment.
  • the imaging system 101 includes an optical system 102 and an imaging element 103
  • the imaging system 104 includes an optical system 105 and an imaging element 106
  • the imaging system 107 includes an optical system 108 and an imaging element 109.
  • the three optical systems 101, 104, and 107 are arranged on substantially the same plane, and their optical axes are substantially parallel.
  • the three image sensors 103, 106, and 109 are disposed on substantially the same plane, and the plane on which the optical system is disposed and the plane on which the image sensor is disposed are substantially parallel.
  • the arrangement direction of the pixels of the image sensor 106 is the horizontal direction (X axis) and the vertical direction (Y axis), and the direction parallel to the optical axis is the imaging distance Z axis. Is the principal point of the optical system 105.
  • the arrangement direction of the pixels of the imaging elements 103 and 109 is substantially parallel to the arrangement direction of the pixels of the imaging element 106.
  • the pixel pitches in the horizontal direction of the image sensors 103, 106, and 109 are equal, and the pixel pitch in the vertical direction is also equal.
  • the readout direction of the pixels of each image sensor is assumed to be the X direction.
  • FIG. 3 is an XZ plan view of the imaging device shown in FIG. 1 as viewed from above.
  • the relationship between the sampling phases in the horizontal direction of the imaging systems 104 and 107 will be described with reference to FIG.
  • An image is formed. That is, at the imaging distance H0, the horizontal sampling phases of the imaging systems 104 and 107 are aligned.
  • the sampling phase is changed by changing the relative positional relationship between the optical system and the image sensor, in FIG. 3, when the sampling phases in the horizontal direction of the imaging systems 104 and 107 are aligned at the imaging distance P0. To do. Next, a change in sampling phase when the imaging distance is changed will be described.
  • the focal length f is 5 mm
  • the imaging distance H0 is 500 mm
  • the shift amount of the imaging position when the imaging distance is changed from H0 is calculated from the equation (1).
  • the results are shown.
  • the horizontal axis represents the shooting distance
  • the vertical axis represents the shift amount of the imaging position. The shift amount changes as the imaging distance changes.
  • the sampling phase shift amount is calculated.
  • the sampling phase can be calculated from the value obtained by dividing the shift amount by the pixel pitch of the image sensor and the value after the decimal point of the sum of the deviation amounts of the sampling phase at the imaging distance H0.
  • the shift amount at the imaging distance H1 600 mm
  • the shift amount at the imaging distance H1 600 mm
  • the shift amount is divided by the pixel pitch, 0.02 ⁇ 0.006 ⁇ 3.33 (pixels)
  • there is a shift of 3.33 pixels from the imaging distance H0 ( 500 mm).
  • the sampling phase shift amount is a negative value according to the above calculation, a value obtained by adding 1 is the sampling phase shift amount.
  • Fig. 5 shows the imaging distance dependence of the sampling phase.
  • an imaging area that cannot be increased in resolution by the imaging systems 104 and 107 is increased in resolution by the imaging systems 104 and 101, thereby realizing an increase in resolution over the entire imaging area.
  • the imaging element 103 is arranged so that the sampling phases of the imaging systems 104 and 101 are shifted by 1/2 pixel.
  • the subject at the point P0 forms an image at the center of the pixel 106a of the image sensor 106 in the image pickup system 104, but forms an image between the pixel 103a of the image sensor 103 and the image sensor 103b in the image pickup system 101.
  • the sampling phase is shifted by 1/2 pixel.
  • the shift amount of the imaging position when the imaging distance is changed from H0 when the focal length f is 5 mm, the base length B ( QR) is 12 mm, the imaging distance H0 is 500 mm, is calculated from the equation (1).
  • FIG. 7 shows the result of overlapping the sampling phase shift between the imaging systems 104 and 107 shown in FIG. 5 and the sampling phase shift between the imaging systems 104 and 101 shown in FIG.
  • the sampling phase shift between the imaging systems 104 and 107 is indicated by a thin line, and the sampling phase shift between the imaging systems 104 and 101 is indicated by a bold line.
  • FIG. 8 shows the display range of the vertical axis in FIG. 7 changed from 0.25 pixel to 0.75 pixel. It can be seen that at any imaging distance, either one of the sampling phase shift amounts is always within the range of 0.25 to 0.75 pixels.
  • FIG. 30 shows the result of verification of high resolution using two imaging systems.
  • the chart for resolution measurement was imaged with two imaging systems arranged horizontally, the two images were synthesized, and the horizontal resolution of the synthesized image was evaluated.
  • the horizontal axis of the figure shows the imaging distance from the imaging system to the chart, and the chart was placed at a total of 26 points in order from 100 cm to 125 cm in 1 cm increments.
  • the vertical axis in the figure indicates the horizontal resolution of the composite image.
  • the vertical axis also shows the sampling phase deviation.
  • the sampling phase shift amount was calculated by the above-described method using the focal length, pixel pitch, imaging distance, and the like of the imaging system.
  • the horizontal baseline length QR of the imaging systems 104 and 101 and the horizontal baseline length RS of the imaging systems 104 and 107 are both equal to 12 mm, and the focal length f is also equal to 5 mm.
  • the focal length f is also equal to 5 mm.
  • FIG. 9 is a view of the image sensors 106 and 109 as viewed from the optical axis direction.
  • R ′ and S ′ indicate points where the optical axes of the imaging systems 104 and 107 intersect with the imaging device, and the heights of the optical axes are the same.
  • the image sensor 109 is arranged so that the sampling phase is shifted by 1/2 pixel in the vertical direction with respect to the image sensor 106. Since the heights of the optical axes coincide with each other, the base line length in the vertical direction is zero, and the sampling phase does not change even when the imaging distance is changed according to the equation (1).
  • the sampling phase in the vertical direction is shifted by 1/2 pixel at all imaging distances, and high resolution can be achieved at all imaging distances.
  • the sampling phase of at least two of the three imaging systems only needs to be shifted from 0.20 to 0.80 pixels at which the resolution can be increased. .
  • the sampling phase of at least two imaging systems out of the three imaging systems may be shifted from 0.20 to 0.80 pixels where the resolution can be increased.
  • the arrangement of the two imaging systems is not limited to the above.
  • the configuration in which the sampling phases of the three imaging systems are shifted by 1/3 pixels may be adopted, or as shown in FIG. 11, the sampling phase of one imaging system with respect to the reference imaging system. May be shifted by 1/2 pixel, and the sampling phase of the other imaging system may be aligned.
  • the sampling phases of the two imaging systems are both 1/2 pixels relative to the reference imaging system.
  • a configuration may be adopted in which the sampling phases of the two imaging systems are aligned.
  • FIG. 14 shows a captured image 1401 of the imaging system 1301 and a captured image 1402 of the imaging system 1302 when the image is captured in the state of FIG.
  • the subject 1303 is imaged with a shift in the vertical direction.
  • a corresponding point is searched from the captured image 1401 based on the captured image 1402 by hardware processing, for example, a point corresponding to the pixel 1404 in the captured image 1402 indicated by shading is the pixel 1403 in the captured image 1401.
  • the corresponding point is searched from the upper left pixel 1405 indicated by diagonal lines, it is necessary to hold data until the corresponding point 1403 is reached, and it is necessary to use a line memory or a frame memory. .
  • FIG. 16 shows a captured image 1601 of the imaging system 1501 and a captured image 1602 of the imaging system 1502 when the image is captured in the state of FIG.
  • the subject 1503 is imaged with a shift in the horizontal direction.
  • a corresponding point is searched from the photographed image 1601 based on the photographed image 1602 by hardware processing, for example, a point corresponding to the pixel 1604 in the photographed image 1602 indicated by shading is the pixel 1603 in the photographed image 1601.
  • the three imaging systems are not arranged at equal intervals, the amount of shift accompanying the change in the imaging distance is different, but if the difference in the baseline length is small, the influence on the sampling phase will be small, and the imaging distance will be reduced. Regardless, the resolution can be increased.
  • the focal length f of the three imaging systems is 5 mm
  • the base lengths RS and QR are 12 mm
  • the range of the imaging distance to be increased in resolution is 1000 mm to 3000 mm
  • An allowable error of the base line length QR for imaging so that the resolution can be achieved is obtained.
  • the sampling phase of the imaging system 104 and the imaging system 107 is shifted by 1 ⁇ 2 pixel at an imaging distance of 1500 mm, and the sampling phase of the imaging system 101 and the imaging system 104 is aligned.
  • the two sampling phase shift amounts are not within the range of 0.20 to 0.80 pixels at the imaging distances of 2320 mm and 2980 mm indicated by circles, and the resolution cannot be increased. Yes.
  • either one of the two sampling phase shift amounts is in the range of 0.20 to 0.80 pixels, and the resolution can be increased.
  • the error is suppressed within an allowable range. As a result, it is possible to increase the resolution regardless of the imaging distance within the range of the predetermined imaging distance.
  • the sampling phase varies depending not only on the base line length but also on the relative positional relationship between the optical axis and the image sensor, the focal length, and the pixel pitch, and also depends on the range of the imaging distance for which high resolution is desired.
  • the error range cannot be defined only by the baseline length, but if the relative positional relationship between the optical axis of the optical system and the image sensor, the focal length, the pixel pitch, and the range of the imaging distance to be increased in resolution are determined.
  • the allowable range can be calculated from the shift amount obtained from the equation (1).
  • three imaging systems in which at least one of the two sampling phase shift amounts falls within a range of 0.20 to 0.80 pixels capable of high resolution within a predetermined imaging distance. By arranging, the horizontal resolution can be improved regardless of the imaging distance.
  • the sampling phase shift is almost constant regardless of the imaging distance in the vertical direction where the baseline length is small, but the sampling phase shift varies depending on the imaging distance in the horizontal direction where the baseline length is large.
  • the shift amount (parallax) must be obtained using a method such as stereo matching. When searching for corresponding points in the stereo matching method, if there is a deviation in the sampling phase, there is no perfectly corresponding corresponding point, and a predetermined corresponding point exists between pixels.
  • FIG. 19 shows images 1901 and 1902 captured by two imaging systems juxtaposed in the horizontal direction.
  • a point corresponding to the pixel 1903 in the captured image 1902 is searched from the captured image 1901, the corresponding point exists between the pixel 1904 and the pixel 1905.
  • FIG. 20 shows images 2001 and 2002 captured by two imaging systems juxtaposed in the horizontal direction, and the sampling phase in the vertical direction is also shifted by 1 ⁇ 2 pixel.
  • a point corresponding to the pixel 2003 in the captured image 2002 exists between the four pixels 2004, 2005, 2006, and 2007 in the captured image 2001.
  • the sampling phases in the vertical direction of the two imaging systems are aligned, and when searching for corresponding points, it is possible to search from at least two images that are not shifted in the vertical direction, and the calculation accuracy is high. improves. By improving the calculation accuracy, it is possible to generate an image with higher resolution by reducing the error when generating each pixel on the corresponding point based on the shift amount between images and generating on the high resolution image. It becomes possible.
  • the vertical sampling phases of the three imaging systems are shifted by 1/3 pixels.
  • the optimum condition is that the sampling phase is shifted by 1/2 pixel, which is because the sampling interval is the narrowest.
  • the sampling phase in the vertical direction is shifted by 0.3 pixels
  • one of the vertical directions is shifted by 0.3 pixels
  • the other is shifted by 0.7 pixels
  • the vertical sampling phase is shifted by 1 ⁇ 2 pixel
  • the vertical direction is 1 ⁇ 2 pixel. Therefore, when three imaging systems are provided, a state in which the sampling phase is shifted by 1/3 pixel is a configuration in which the sampling interval is the shortest, which is suitable for high resolution.
  • three imaging systems are arranged in the pixel arrangement direction, and a subject within a predetermined range of imaging distance is subjected to a sampling phase which is a condition for high resolution by any two imaging systems.
  • a sampling phase which is a condition for high resolution by any two imaging systems.
  • the present invention can be used for an imaging apparatus using a plurality of imaging systems.

Abstract

In an image capture device that generates a high-resolution image by employing a plurality of image capture assemblies to composite a capture image, memory usage increases when the process is carried out in hardware. Disclosed is an image capture device comprising at least three image capture assemblies, further comprising an optical assembly and an image capture element. The image capture assemblies are positioned in the pixel array direction of the image capture elements. A photographic subject, which is within an image capture distance of a prescribed range, is captured such that a sampling phase in the array direction is shifted by 0.2-0.8 pixels, relative to an image captured by a first image capture assembly, by any of the other image capture assemblies.

Description

撮像装置Imaging device
 本発明は、撮像系を複数用いた撮像装置に関する。 The present invention relates to an imaging apparatus using a plurality of imaging systems.
 近年、デジタルスチルカメラやデジタルビデオカメラなどの撮像装置の多画素化や高精細化が進み、その光学性能を十分に出すためにレンズなどの光学系が大きくなってしまったり、撮像素子そのものが大きくなってしまうという問題がある。 In recent years, imaging devices such as digital still cameras and digital video cameras have increased in the number of pixels and the resolution has been increased, and the optical system such as a lens has become large in order to obtain sufficient optical performance. There is a problem of becoming.
 光学系や撮像素子の大きさを変えずに画像の解像度を向上させる方法として、光学系と撮像素子との相対的位置関係を時間的にずらしたり、プリズムを用いて光を複数に分離し、複数の撮像素子に入射させたりすることにより、被写体像と撮像素子の画素との位置関係が相互にずれた複数の画像情報を取得し、これらの複数の画像情報を合成することにより高解像度な画像を取得する「画素ずらし」と呼ばれる技術が知られている。例えば、特許文献1では、図21に示すように撮像素子を変位させる変位手段を備え、被写体像と撮像素子との相対的位置関係を撮像素子の画素の配列方向に対して斜め方向に変位させることで、水平方向と垂直方向の解像度を向上させる技術が提案されている。ここで、水平方向、垂直方向とは、図22に示すように、画素の配列方向を意味する。 As a method of improving the resolution of the image without changing the size of the optical system and the image sensor, the relative positional relationship between the optical system and the image sensor is shifted in time, or the light is separated into a plurality using a prism, A plurality of pieces of image information in which the positional relationship between the subject image and the pixels of the image pickup device is shifted from each other is acquired by making the light incident on a plurality of image pickup devices, and the high resolution is obtained by combining these pieces of image information. A technique called “pixel shifting” for acquiring an image is known. For example, Patent Document 1 includes a displacement unit that displaces the image sensor as illustrated in FIG. 21, and displaces the relative positional relationship between the subject image and the image sensor in an oblique direction with respect to the arrangement direction of the pixels of the image sensor. Thus, a technique for improving the resolution in the horizontal direction and the vertical direction has been proposed. Here, the horizontal direction and the vertical direction mean pixel arrangement directions as shown in FIG.
 また、画素ずらしの技術は、複数の撮像系を用いる複眼方式においても提案されている。特許文献2では、複数の撮像系を、画素の配列方向に対して垂直な方向に画素ずらしを実現するよう配置することで、解像度を向上させる技術が提案されている。 Also, a pixel shifting technique has been proposed for a compound eye system using a plurality of imaging systems. Patent Document 2 proposes a technique for improving the resolution by arranging a plurality of imaging systems so as to realize pixel shifting in a direction perpendicular to the pixel arrangement direction.
 複眼方式における画素ずらしの技術の理解を深めるため、図23から図26を用いて画素ずらしについて説明する。 In order to deepen the understanding of the pixel shifting technique in the compound eye system, the pixel shifting will be described with reference to FIGS.
 図23は、2つの撮像系2302、撮像系2303で共通の被写体2301を撮像している様子を示している。図24は、撮像系2302の撮像する領域を格子2401で示し、撮像系2303の撮像する領域を格子2402(破線)で示しており、最小の四角形の1つ1つが1つの画素によって撮像される領域を示している。図24では、2つの撮像系の各画素によって撮像される領域がほぼ重なりあっており、サンプリング位相が揃っている。一方、図25では、2つの撮像系の撮像素子の各画素が撮像する領域が水平方向に3/2画素、垂直方向に1/2画素分ずれており、サンプリング位相がずれている。ここでいうサンプリング位相とは、複数の撮像系における、被写体のある点の撮像素子上の結像点と画素との相対的な位置関係を意味する。例えば、図25においては、水平方向に3/2画素ずれて撮像されており、水平方向のサンプリング位相のずれは、小数部分である1/2画素となる。画素ずらしとは、このサンプリング位相がずれた状態を意味する。 FIG. 23 shows a situation where a common subject 2301 is imaged by the two imaging systems 2302 and 2303. FIG. 24 shows a region to be imaged by the imaging system 2302 by a grid 2401 and a region to be imaged by the imaging system 2303 by a grid 2402 (broken line), and each of the smallest squares is imaged by one pixel. Indicates the area. In FIG. 24, the areas imaged by the pixels of the two imaging systems are almost overlapped, and the sampling phases are aligned. On the other hand, in FIG. 25, the area captured by each pixel of the imaging elements of the two imaging systems is shifted by 3/2 pixels in the horizontal direction and 1/2 pixel in the vertical direction, and the sampling phase is shifted. The sampling phase here means the relative positional relationship between the image forming point on the image sensor at a certain point of the subject and the pixel in a plurality of imaging systems. For example, in FIG. 25, images are taken with a 3/2 pixel shift in the horizontal direction, and the horizontal sampling phase shift is 1/2 pixel as a decimal part. Pixel shift means a state in which the sampling phase is shifted.
 次に、図26を用いて、複眼方式において、撮像距離が変化した場合のサンプリング位相の変化について説明する。図26において、撮像系2601は光学系2602と撮像素子2603とを備え、撮像系2605は光学系2606と撮像素子2607とを備えている。光学系2602と光学系2606は同一平面上に配置されており、それぞれの光軸は上記平面に垂直であり、図26はその様子を上から見た図である。図26において、被写体は、光学系2602の光軸上の点P0もしくは点P1に位置している。点P0に位置する被写体の被写体像は、撮像素子2603の画素2604の中心に結像し、また、撮像素子2607の画素2608の中心に結像している。つまり、サンプリング位相が揃った状態にある。被写体がP0にある撮像距離H0の状態から、P1にある撮像距離H1の状態に変化した場合、撮像素子2607での結像位置がシフトし、そのシフト量の差は、u0-u1となり、
 u0-u1=f×B×((1/H0)-(1/H1))・・・(1)
と表せる。ここで、fは焦点距離、Bは基線長である。ここで、基線長Bは、光学系2602の主点Qと光学系2606の主点Rの距離とする。
Next, a change in sampling phase when the imaging distance changes in the compound eye method will be described with reference to FIG. In FIG. 26, an imaging system 2601 includes an optical system 2602 and an imaging element 2603, and an imaging system 2605 includes an optical system 2606 and an imaging element 2607. The optical system 2602 and the optical system 2606 are arranged on the same plane, and their optical axes are perpendicular to the plane, and FIG. 26 is a view of the state from above. In FIG. 26, the subject is located at a point P0 or P1 on the optical axis of the optical system 2602. The subject image of the subject located at the point P0 is formed at the center of the pixel 2604 of the image sensor 2603 and is formed at the center of the pixel 2608 of the image sensor 2607. That is, the sampling phases are in a uniform state. When the subject changes from the imaging distance H0 at P0 to the imaging distance H1 at P1, the imaging position on the imaging device 2607 shifts, and the difference in the shift amount becomes u0-u1.
u0−u1 = f × B × ((1 / H0) − (1 / H1)) (1)
It can be expressed. Here, f is a focal length, and B is a baseline length. Here, the base line length B is a distance between the principal point Q of the optical system 2602 and the principal point R of the optical system 2606.
 例えば、焦点距離fを5mm、基線長Bを12mm、撮像距離H0を700mm、撮像距離H1を725mmとすると、結像位置のシフト量の差は、(1)式より約3μmとなる。撮像素子の画素ピッチが6μmである場合、撮像距離が700mmから725mmに変化することで、サンプリング位相が約1/2画素シフトすることになる。撮像距離700mmでサンプリング位相が1/2画素ずれた状態にあるとすると、撮像距離が725mmに変化した場合、1/2+1/2=1と整数となり、即ち、サンプリング位相が揃ってしまい、高解像度化できなくなる。撮像距離に依らずサンプリング位相がずれた状態にするには、(1)式より、結像位置がシフトしないよう焦点距離fまたは基線長Bを小さくすればよい。基線長Bとは2つの撮像系の光学系の主点間距離であるが、図27に示すように、基準となる撮像系の撮像素子の画素の配列方向を水平方向、垂直方向とし、基線長Bの水平成分を水平方向の基線長BH、垂直成分を垂直方向の基線長BVと表現する。 For example, if the focal length f is 5 mm, the baseline length B is 12 mm, the imaging distance H0 is 700 mm, and the imaging distance H1 is 725 mm, the difference in the shift amount of the imaging position is about 3 μm from the equation (1). When the pixel pitch of the image sensor is 6 μm, the sampling phase is shifted by about ½ pixel by changing the imaging distance from 700 mm to 725 mm. Assuming that the sampling phase is shifted by ½ pixel at an imaging distance of 700 mm, when the imaging distance is changed to 725 mm, 1/2 + 1/2 = 1, which is an integer, that is, the sampling phases are aligned, resulting in high resolution. It becomes impossible to become. In order to obtain a state in which the sampling phase is shifted regardless of the imaging distance, the focal length f or the base line length B may be reduced so as not to shift the imaging position from the equation (1). The base line length B is the distance between the principal points of the optical systems of the two image pickup systems. As shown in FIG. 27, the arrangement direction of the pixels of the image pickup device of the reference image pickup system is the horizontal direction and the vertical direction. The horizontal component of the length B is expressed as a horizontal baseline length BH, and the vertical component is expressed as a vertical baseline length BV.
 特許文献2では、水平方向の基線長と垂直方向の基線長が小さくなるように複数の撮像系を配置することで、撮像距離が変化してもサンプリング位相の変化が少なく、撮像距離に依らず画素ずらしする技術を提案している。 In Patent Document 2, by arranging a plurality of imaging systems so that the baseline length in the horizontal direction and the baseline length in the vertical direction are reduced, the change in the sampling phase is small even if the imaging distance changes, and it does not depend on the imaging distance. A technique for shifting pixels is proposed.
 図28、29を用いて、特許文献2に開示されている複眼方式を説明する。図28に示すように、特許文献2では、光学系301a、301b、301cを、同一平面内に各光軸300a、300b、300cが平行となるよう配置し、光学系を平面内で回転させることで、被写体像の結像位置をずらし、画素ずらしを実現している。このとき、図29に示すように光軸300bと光軸300cのY方向(図中に記載)のずれが十分に小さいため、すなわち、Y方向の基線長が十分に小さいため、撮像距離が変化してもY方向のサンプリング位相の変化を小さく抑えることが出来る。 28, 29, the compound eye system disclosed in Patent Document 2 will be described. As shown in FIG. 28, in Patent Document 2, the optical systems 301a, 301b, and 301c are arranged in the same plane so that the optical axes 300a, 300b, and 300c are parallel, and the optical system is rotated in the plane. Thus, the imaging position of the subject image is shifted to realize pixel shifting. At this time, as shown in FIG. 29, since the deviation between the optical axis 300b and the optical axis 300c in the Y direction (shown in the figure) is sufficiently small, that is, the baseline length in the Y direction is sufficiently small, the imaging distance changes. Even in this case, the change in the sampling phase in the Y direction can be kept small.
 X方向(図中に記載)に関しては、X方向の基線長が大きいため、撮像距離によってはサンプリング位相が揃うが、光学系301aと光学系301bのX方向の基線長が十分に小さいため、撮像距離が変化してもX方向のサンプリング位相の変化を小さく抑えることが出来る。 Regarding the X direction (described in the figure), since the base line length in the X direction is large, the sampling phase is uniform depending on the imaging distance, but the base line length in the X direction of the optical system 301a and the optical system 301b is sufficiently small. Even if the distance changes, the change in the sampling phase in the X direction can be kept small.
 このように、特許文献2では、撮像距離が変化に伴うサンプリング位相の変化が小さくなるよう撮像系を配置し、撮像距離に依らず画素ずらしを実現する技術を提案している。 As described above, Patent Document 2 proposes a technique that arranges an imaging system so that a change in sampling phase with a change in imaging distance is small, and realizes pixel shifting regardless of the imaging distance.
特開平10-304235号公報JP-A-10-304235 国際公開WO2007/013250公報International Publication WO2007 / 013250
 しかしながら、複眼方式の画素ずらしにより高解像度画像を得るためには、各撮像素子における画像間のずれ量(視差)を計算しなければならない。特許文献2において、画素ずらし方向における撮像素子のずれ量は撮像距離に依らず約1/2画素であるが、画素ずらし方向と垂直な方向に関しては、基線長が大きく、撮像距離によって画像間のずれ量が変化するため、そのずれ量を知る必要がある。その方法として、例えば、ステレオマッチングという方法がある。ステレオマッチングとは、同一の被写体を撮影した視点の異なる複数の画像上において、一方の画像を基準とし、その画像中の各画素に対し対応点(同じ空間点の投影点)を他方の画像中から探索することである。ステレオマッチングには、テンプレートマッチングを利用して対応点を求める領域ベースマッチングやエッジやコーナーなどの特徴点を抽出して求める特徴ベースマッチングなど様々な方法がある。そして、このようにして得られた画像間のずれ量をもとに、各画素を生成する高解像度画像上の対応する点に配置する事により、高解像度画像を生成する。この際、特許文献2に示すように、垂直方向に異なる高さに配置された撮像系の画像を合成して高解像度画像を生成する際には、対応点が画像中の垂直方向にずれた場所に位置する。従って、ハードウェアによる処理時に画像の読み出しを画像左上から行う際、各撮像系で対応点の存在する領域に到達するタイミングが異なるため、データを保持しておく必要があり、ラインメモリあるいはフレームメモリを利用する必要がある。 However, in order to obtain a high-resolution image by compound-eye pixel shift, it is necessary to calculate the shift amount (parallax) between images in each image sensor. In Patent Document 2, the amount of displacement of the image sensor in the pixel shifting direction is about ½ pixel regardless of the imaging distance, but the baseline length is large in the direction perpendicular to the pixel shifting direction, and the distance between the images depends on the imaging distance. Since the amount of deviation changes, it is necessary to know the amount of deviation. For example, there is a method called stereo matching. Stereo matching is based on one image on multiple images taken from the same subject and the corresponding point (projection point of the same spatial point) is assigned to each pixel in the image in the other image. It is to search from. Stereo matching includes various methods such as region-based matching for obtaining corresponding points using template matching and feature-based matching for extracting feature points such as edges and corners. A high-resolution image is generated by arranging each pixel at a corresponding point on the high-resolution image based on the shift amount between the images thus obtained. At this time, as shown in Patent Document 2, when a high-resolution image is generated by synthesizing images of imaging systems arranged at different heights in the vertical direction, corresponding points are shifted in the vertical direction in the image. Located in place. Therefore, when reading the image from the upper left of the image during processing by hardware, the timing to reach the region where the corresponding point exists in each imaging system is different, so it is necessary to hold the data, and the line memory or frame memory It is necessary to use.
 また、特許文献2の構成では水平垂直全ての方向について高解像度化をするためには、水平垂直にそれぞれ1/2画素ずれた2眼が必要となり、水平あるいは垂直方向に各カメラを並置し、一方向に薄い構成を実現する事が出来ない。 In addition, in the configuration of Patent Document 2, in order to increase the resolution in all the horizontal and vertical directions, two eyes that are shifted by 1/2 pixel in the horizontal and vertical directions are necessary, and the cameras are juxtaposed in the horizontal or vertical direction. A thin configuration cannot be realized in one direction.
 本発明は、このような事情に鑑みてなされたもので、水平、あるいは垂直の一方向に各カメラを配置し、ハードウェア化した際にメモリの使用量が少ない構成で、被写体の距離に依らず高解像度化することができる撮像装置を提供することを目的とする。 The present invention has been made in view of such circumstances, and each camera is arranged in one horizontal or vertical direction so that the amount of memory used is reduced when it is implemented as hardware, depending on the distance of the subject. An object of the present invention is to provide an imaging apparatus capable of increasing the resolution.
 本発明は上述した課題を解決するためになされたもので、
 本発明の第一の技術手段は、
光学系と撮像素子とを有する撮像系を少なくとも3つ備え、前記撮像系は、前記撮像素子の画素の配列方向に配置され、第一の撮像系の撮像画像中の所定範囲内の撮像距離にある被写体が、前記第一の撮像系の撮像画像に対し、他のいずれかの撮像系により、前記配列方向のサンプリング位相が0.2から0.8画素ずれて撮像されることを特徴とする。
The present invention has been made to solve the above-described problems.
The first technical means of the present invention is:
The image pickup system includes at least three image pickup systems each having an optical system and an image pickup element. The image pickup system is arranged in a pixel arrangement direction of the image pickup element and has an image pickup distance within a predetermined range in a picked-up image of the first image pickup system. A subject is imaged with a sampling phase in the arrangement direction shifted from 0.2 to 0.8 pixels with respect to a captured image of the first imaging system by any other imaging system. .
 第二の技術手段は、
第一の技術手段において、前記撮像系は前記配列方向に略等間隔に配置され、前記撮像系の光軸は略平行であることを特徴とする。
The second technical means is
In the first technical means, the imaging systems are arranged at substantially equal intervals in the arrangement direction, and the optical axes of the imaging systems are substantially parallel.
 第三の技術手段は、
 第一または第二の技術手段において、所定範囲の撮像距離内にある被写体が、いずれかの撮像系の撮像画像に対し、他のいずれかの撮像系により、前記配列方向と垂直な方向のサンプリング位相が0.2から0.8画素ずれて撮像されることを特徴とする。
The third technical means is
In the first or second technical means, a subject within an imaging distance within a predetermined range is sampled in a direction perpendicular to the arrangement direction with respect to a captured image of any imaging system by any other imaging system. It is characterized in that the image is taken with a phase shift of 0.2 to 0.8 pixels.
 第四の技術手段は、
第一または第二の技術手段において、所定範囲の撮像距離内にある被写体が、いずれかの撮像系の撮像画像に対し、他のいずれかの撮像系により、前記配列方向と垂直な方向のサンプリング位相が略0.5画素ずれて撮像されることを特徴とする。
The fourth technical means is
In the first or second technical means, a subject within an imaging distance within a predetermined range is sampled in a direction perpendicular to the arrangement direction with respect to a captured image of any imaging system by any other imaging system. The image is picked up with a phase shift of approximately 0.5 pixels.
 第五の技術手段は、
第三または第四の技術手段において、所定範囲の撮像距離内にある被写体が、いずれかの撮像系の撮像画像に対し、他のいずれかの撮像系により、前記配列方向と垂直な方向のサンプリング位相がほぼずれなく撮像されることを特徴とする。
The fifth technical means is
In the third or fourth technical means, a subject within an imaging distance within a predetermined range is sampled in a direction perpendicular to the arrangement direction with respect to a captured image of any imaging system by any other imaging system. It is characterized in that imaging is performed with almost no phase shift.
 第六の技術手段は、
第一または第二の技術手段において、所定範囲の撮像距離内にある被写体が、いずれかの撮像系の撮像画像に対し、他のいずれか2つの撮像系により、前記配列方向と垂直な方向のサンプリング位相が略1/3画素ずつずれて撮像されることを特徴とする。
The sixth technical means is
In the first or second technical means, a subject within an imaging distance within a predetermined range is captured in a direction perpendicular to the arrangement direction by any two other imaging systems with respect to a captured image of any imaging system. The sampling phase is imaged with a shift of approximately 1/3 pixels.
 第七の技術手段は、光学系と撮像素子とを有する撮像系を少なくとも3つ備え、前記撮像系は略直線上に配置され、前記直線を含む平面の法線と前記撮像系の光軸とが略平行をなし、前記直線と前記撮像素子の画素の一方の配列方向とが略平行をなし、前記撮像系のいずれかの撮像系により、所定範囲内の撮像距離にある被写体が、前記配列方向のサンプリング位相がずれて撮像されるように構成されることを特徴とする。 The seventh technical means includes at least three image pickup systems having an optical system and an image pickup device, the image pickup systems are arranged on a substantially straight line, a plane normal including the straight line, an optical axis of the image pickup system, and Are substantially parallel, the straight line and one arrangement direction of the pixels of the imaging element are substantially parallel, and an object at an imaging distance within a predetermined range is captured by the imaging system of any of the imaging systems. It is characterized by being configured so as to be imaged with the sampling phase of the direction shifted.
 第八の技術手段は、第七の技術手段において、前記撮像系のいずれかの撮像系により、所定範囲内の撮像距離にある被写体が、前記配列方向のサンプリング位相が0.2から0.8画素ずれて撮像されるように構成されることを特徴とする。 According to an eighth technical means, in the seventh technical means, an object at an imaging distance within a predetermined range is set to have a sampling phase in the arrangement direction of 0.2 to 0.8 by any one of the imaging systems. It is configured to be imaged with a pixel shift.
 第九の技術手段は、第七の技術手段において、前記撮像系の光軸は、前記配列方向に等間隔に配置されることを特徴とする。 Ninth technical means is characterized in that, in the seventh technical means, the optical axes of the imaging system are arranged at equal intervals in the arrangement direction.
 第十の技術手段は、第七から第九のいずれかの技術手段において、前記撮像系のいずれかの撮像系により、所定範囲内の撮像距離にある被写体が、前記配列方向と垂直な方向のサンプリング位相がずれて撮像されるように構成されることを特徴とする。 According to a tenth technical means, in any one of the seventh to ninth technical means, an object at an imaging distance within a predetermined range is moved in a direction perpendicular to the arrangement direction by any one of the imaging systems. It is characterized by being configured so as to be imaged with the sampling phase shifted.
 第十一の技術手段は、第七から第九のいずれかの技術手段において、前記撮像系のいずれかの撮像系により、所定範囲内の撮像距離にある被写体が、前記配列方向と垂直な方向のサンプリング位相が0.2から0.8画素ずれて撮像されるように構成されることを特徴とする。 In an eleventh technical means according to any one of the seventh to ninth technical means, a subject at an imaging distance within a predetermined range is perpendicular to the arrangement direction by any one of the imaging systems. The sampling phase is configured to be imaged with a deviation of 0.2 to 0.8 pixels.
 第十二の技術手段は、第七から第九のいずれかの技術手段において、前記撮像系のいずれかの撮像系により、所定範囲内の撮像距離にある被写体が、前記配列方向と垂直な方向のサンプリング位相が0.5画素ずれて撮像されるように構成されることを特徴とする。 In a twelfth technical means according to any one of the seventh to ninth technical means, a subject at an imaging distance within a predetermined range is perpendicular to the arrangement direction by any one of the imaging systems. The sampling phase is configured to be imaged with a shift of 0.5 pixels.
 第十三の技術手段は、第十一または第十二の技術手段において、前記撮像系のいずれかの撮像系により、所定範囲内の撮像距離にある被写体が、前記配列方向と垂直な方向のサンプリング位相がずれなく撮像されるように構成されることを特徴とする。 According to a thirteenth technical means, in the eleventh technical means or the twelfth technical means, an object at an imaging distance within a predetermined range is moved in a direction perpendicular to the arrangement direction by any one of the imaging systems. The sampling phase is configured to be imaged without deviation.
 第十四の技術手段は、第十または第十一の技術手段において、前記撮像系のいずれか3つの撮像系により、所定範囲内の撮像距離にある被写体が、前記配列方向と垂直な方向のサンプリング位相が1/3画素ずつずれて撮像されるように構成されることを特徴とする。 According to a fourteenth technical means, in any one of the tenth and eleventh technical means, an object at an imaging distance within a predetermined range is moved in a direction perpendicular to the arrangement direction by any three imaging systems of the imaging system. The sampling phase is configured to be imaged with a shift of 1/3 pixel at a time.
 本発明によれば、複数の撮像系を用いて共通の被写体を撮像する撮像装置において、撮像系を水平あるいは垂直のどちらか一方向に並置することにより、得られた複数の画像を合成し画素数が多く解像度の高い画像を生成する際に、メモリの使用量を低減することが可能となる。また、一方向に並置することにより薄型化することが可能となる。 According to the present invention, in an imaging device that captures a common subject using a plurality of imaging systems, the imaging systems are arranged side by side in either one of the horizontal and vertical directions, and a plurality of obtained images are combined to form a pixel. When generating a large number of images with high resolution, it is possible to reduce the amount of memory used. Moreover, it becomes possible to reduce in thickness by juxtaposing in one direction.
本発明の第一の実施形態の構成を示す図である。It is a figure which shows the structure of 1st embodiment of this invention. 撮像素子の画素の配列方向とX軸、Y軸、Z軸との関係を説明する図である。It is a figure explaining the relationship between the arrangement direction of the pixel of an image sensor, and an X-axis, a Y-axis, and a Z-axis. 3つの撮像素子の配置を説明する図である。It is a figure explaining arrangement of three image sensors. 撮像装置を上から見たXZ平面図である。It is XZ top view which looked at an imaging device from the top. 撮像距離を変化させた場合の結像位置のシフト量を示すグラフである。It is a graph which shows the shift amount of the image formation position at the time of changing imaging distance. サンプリング位相の撮像距離依存性を示すグラフである。It is a graph which shows the imaging distance dependence of a sampling phase. サンプリング位相のずれを計算した結果を示すグラフである。It is a graph which shows the result of having calculated the shift of sampling phase. 図5のサンプリング位相のずれと図6のサンプリング位相のずれを重ねた結果を示すグラフである。FIG. 7 is a graph showing a result of overlapping the sampling phase shift of FIG. 5 and the sampling phase shift of FIG. 6; 図7の縦軸の表示範囲を0.25画素から0.75画素にした結果を示すグラフである。It is a graph which shows the result of having changed the display range of the vertical axis | shaft of FIG. 7 from 0.25 pixel to 0.75 pixel. 撮像素子106と109を光軸方向から見た図である。It is the figure which looked at the image sensors 106 and 109 from the optical axis direction. 3つの撮像系の配置の他の例を示す図である。It is a figure which shows the other example of arrangement | positioning of three imaging systems. 3つの撮像系の配置のさらに他の例を示す図である。It is a figure which shows the further another example of arrangement | positioning of three imaging systems. 3つの撮像系の配置のさらに他の例を示す図である。It is a figure which shows the further another example of arrangement | positioning of three imaging systems. 垂直方向に並置された2つの撮像系が被写体を撮像している様子を示す図である。It is a figure which shows a mode that the two imaging systems juxtaposed in the perpendicular direction are image | photographing a to-be-photographed object. 図13の状態で撮像した場合の撮像画像を示す図である。It is a figure which shows the captured image at the time of imaging in the state of FIG. 水平方向に並置された2つの撮像系が被写体を撮像している様子を示す図である。It is a figure which shows a mode that two imaging systems juxtaposed in the horizontal direction are image | photographing a to-be-photographed object. 図15の状態で撮像した場合の撮像画像を示す図である。It is a figure which shows the captured image at the time of imaging in the state of FIG. QR=11.7mmの場合のサンプリング位相のずれ量を示すグラフである。It is a graph which shows the shift | offset | difference amount of the sampling phase in the case of QR = 11.7mm. QR=11.0mmの場合のサンプリング位相のずれ量を示すグラフである。It is a graph which shows the deviation | shift amount of a sampling phase in case QR = 11.0mm. 水平方向に並置した2つの撮像系で撮像した画像を示す図である。It is a figure which shows the image imaged with two imaging systems juxtaposed in the horizontal direction. 水平方向に並置した2つの撮像系で撮像した画像を示す図である。It is a figure which shows the image imaged with two imaging systems juxtaposed in the horizontal direction. 特許文献1に記載の技術を説明する図である。It is a figure explaining the technique of patent document 1. FIG. 撮像素子に対する水平方向および垂直方向を説明する図である。It is a figure explaining the horizontal direction and vertical direction with respect to an image sensor. 2つの撮像系で共通の被写体を撮像している様子を示す図である。It is a figure which shows a mode that the common to-be-photographed object is imaged with two imaging systems. 2つの撮像系の撮像する領域を格子で示した図である。It is the figure which showed the field which two imaging systems picturize with a lattice. 画素ずらしの状態を説明する図である。It is a figure explaining the state of pixel shifting. 複眼方式において撮像距離が変化した場合のサンプリング位相の変化を説明する図である。It is a figure explaining the change of a sampling phase when imaging distance changes in a compound eye system. 水平方向の基線長BHおよび垂直方向の基線長BVを説明する図である。It is a figure explaining the baseline length BH of a horizontal direction, and the baseline length BV of a perpendicular direction. 特許文献2に開示されている複眼方式を説明する図である。It is a figure explaining the compound eye system currently indicated by patent documents 2. FIG. 特許文献2に開示されている複眼方式を説明する図である。It is a figure explaining the compound eye system currently indicated by patent documents 2. FIG. 2つの撮像系による高解像度化を検証した結果を示すグラフである。It is a graph which shows the result of having verified high resolution by two imaging systems.
<第一の実施形態>
 以下、本発明の第一の実施形態について、図面を参照して説明する。図1は本実施形態の構成を示す図である。図1は、本実施形態における撮像装置内に備えられる3つの撮像系の配置を示す図である。撮像系101は光学系102と撮像素子103を備え、撮像系104は光学系105と撮像素子106を備え、撮像系107は光学系108と撮像素子109を備えている。3つの光学系101、104、107は略同一平面上に配置されており、それぞれの光軸は略平行である。また、3つの撮像素子103、106、109は略同一平面上に配置されており、光学系が配置されている平面と撮像素子が配置されている平面は略平行である。
<First embodiment>
Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing the configuration of this embodiment. FIG. 1 is a diagram illustrating an arrangement of three imaging systems provided in the imaging apparatus according to the present embodiment. The imaging system 101 includes an optical system 102 and an imaging element 103, the imaging system 104 includes an optical system 105 and an imaging element 106, and the imaging system 107 includes an optical system 108 and an imaging element 109. The three optical systems 101, 104, and 107 are arranged on substantially the same plane, and their optical axes are substantially parallel. The three image sensors 103, 106, and 109 are disposed on substantially the same plane, and the plane on which the optical system is disposed and the plane on which the image sensor is disposed are substantially parallel.
 本実施形態では、図2Aに示すように、撮像素子106の画素の配列方向を水平方向(X軸)、垂直方向(Y軸)とし、光軸と平行な方向を撮像距離Z軸とし、原点を光学系105の主点とする。また、図2Bに示すように、撮像素子103と109の画素の配列方向は撮像素子106の画素の配列方向と略平行である。また、撮像素子103、106、109の水平方向の画素ピッチは等しく、垂直方向の画素ピッチも等しい。また、各撮像素子の画素の読み出し方向はX方向であるものとする。 In the present embodiment, as shown in FIG. 2A, the arrangement direction of the pixels of the image sensor 106 is the horizontal direction (X axis) and the vertical direction (Y axis), and the direction parallel to the optical axis is the imaging distance Z axis. Is the principal point of the optical system 105. In addition, as illustrated in FIG. 2B, the arrangement direction of the pixels of the imaging elements 103 and 109 is substantially parallel to the arrangement direction of the pixels of the imaging element 106. The pixel pitches in the horizontal direction of the image sensors 103, 106, and 109 are equal, and the pixel pitch in the vertical direction is also equal. In addition, the readout direction of the pixels of each image sensor is assumed to be the X direction.
 図3は、図1に示す撮像装置を上から見たXZ平面図である。図3を用いて、撮像系104と107の水平方向のサンプリング位相の関係について述べる。図3に示すように、Z=H0の点P0にある被写体は、撮像系104では撮像素子106の画素106aの中央に結像しており、撮像系107では撮像素子109の画素109aの中央に結像している。即ち、撮像距離H0では、撮像系104と107の水平方向のサンプリング位相は揃っている。サンプリング位相は光学系と撮像素子の相対的な位置関係を変化させることで変わるが、図3においては、撮像距離P0において撮像系104と107の水平方向のサンプリング位相が揃っている状態にあるとする。次に、撮像距離を変化させた場合のサンプリング位相の変化について述べる。 FIG. 3 is an XZ plan view of the imaging device shown in FIG. 1 as viewed from above. The relationship between the sampling phases in the horizontal direction of the imaging systems 104 and 107 will be described with reference to FIG. As shown in FIG. 3, the subject at the point P0 where Z = H0 forms an image at the center of the pixel 106a of the image sensor 106 in the imaging system 104, and in the center of the pixel 109a of the image sensor 109 in the imaging system 107. An image is formed. That is, at the imaging distance H0, the horizontal sampling phases of the imaging systems 104 and 107 are aligned. Although the sampling phase is changed by changing the relative positional relationship between the optical system and the image sensor, in FIG. 3, when the sampling phases in the horizontal direction of the imaging systems 104 and 107 are aligned at the imaging distance P0. To do. Next, a change in sampling phase when the imaging distance is changed will be described.
 図4に、焦点距離fを5mm、基線長B(=RS)を12mm、撮像距離H0を500mmとして、撮像距離をH0から変化させた場合の結像位置のシフト量を(1)式より計算した結果を示す。横軸は撮影距離、縦軸が結像位置のシフト量である。撮像距離の変化とともに、シフト量が変化している。 In FIG. 4, the focal length f is 5 mm, the base length B (= RS) is 12 mm, the imaging distance H0 is 500 mm, and the shift amount of the imaging position when the imaging distance is changed from H0 is calculated from the equation (1). The results are shown. The horizontal axis represents the shooting distance, and the vertical axis represents the shift amount of the imaging position. The shift amount changes as the imaging distance changes.
 次に、サンプリング位相のずれ量を計算する。サンプリング位相は、前記シフト量を撮像素子の画素ピッチで除した値と撮像距離H0でのサンプリング位相のずれ量の和の小数点以下の値から算出出来る。例えば、図4において、撮像距離H1(=600mm)でのシフト量は(1)式より、
u0-u1=5×12×((1/500)-(1/600))=0.02 (mm)
となり、画素ピッチを6μmとして、シフト量を画素ピッチで除すると、
0.02÷0.006≒3.33 (画素)
となり、撮像距離H0(=500mm)から3.33画素シフトしている。撮像距離H0でサンプリング位相が0.5画素ずれているとすると、撮像距離H1でのサンプリング位相のずれ量は、
3.33+0.5=3.83 (画素)
となり、小数点以下の0.83画素が撮像距離H1でのサンプリング位相のずれ量となる。上記計算によりサンプリング位相のずれ量が負の値となる場合は、1を足した値がサンプリング位相のずれ量となる。
Next, the sampling phase shift amount is calculated. The sampling phase can be calculated from the value obtained by dividing the shift amount by the pixel pitch of the image sensor and the value after the decimal point of the sum of the deviation amounts of the sampling phase at the imaging distance H0. For example, in FIG. 4, the shift amount at the imaging distance H1 (= 600 mm) is calculated from the equation (1):
u0−u1 = 5 × 12 × ((1/500) − (1/600)) = 0.02 (mm)
When the pixel pitch is 6 μm and the shift amount is divided by the pixel pitch,
0.02 ÷ 0.006 ≒ 3.33 (pixels)
Thus, there is a shift of 3.33 pixels from the imaging distance H0 (= 500 mm). Assuming that the sampling phase is shifted by 0.5 pixels at the imaging distance H0, the amount of sampling phase shift at the imaging distance H1 is
3.33 + 0.5 = 3.83 (pixels)
Thus, 0.83 pixels below the decimal point is the sampling phase shift amount at the imaging distance H1. When the sampling phase shift amount is a negative value according to the above calculation, a value obtained by adding 1 is the sampling phase shift amount.
 図5にサンプリング位相の撮像距離依存性を示す。水平方向の画素ピッチは6μmとする。基準とした撮像距離H0(=500mm)ではサンプリング位相が揃っているとする。撮像距離を変えるとサンプリング位相が変化することがわかる。従って、撮像系104と107は撮像距離によってサンプリング位相が揃う距離があり、撮像距離に依っては高解像度化出来ない距離がある。 Fig. 5 shows the imaging distance dependence of the sampling phase. The pixel pitch in the horizontal direction is 6 μm. It is assumed that the sampling phases are aligned at the reference imaging distance H0 (= 500 mm). It can be seen that the sampling phase changes when the imaging distance is changed. Therefore, the imaging systems 104 and 107 have a distance at which the sampling phases are aligned according to the imaging distance, and there are distances at which the resolution cannot be increased depending on the imaging distance.
 そこで、本実施形態では、撮像系104と107では高解像度化出来ない撮像領域を、撮像系104と101により高解像度化し、撮像領域全体に渡る高解像度化を実現する。 Therefore, in the present embodiment, an imaging area that cannot be increased in resolution by the imaging systems 104 and 107 is increased in resolution by the imaging systems 104 and 101, thereby realizing an increase in resolution over the entire imaging area.
前述のように、基準とした撮像距離H0で撮像系104と107のサンプリング位相が揃っているため、撮像系104と101のサンプリング位相が1/2画素ずれるよう、撮像素子103を配置する。図3において、点P0にある被写体は、撮像系104では撮像素子106の画素106aの中央に結像しているが、撮像系101では撮像素子103の画素103aと撮像素子103bの間に結像しており、サンプリング位相が1/2画素ずれている。 As described above, since the sampling phases of the imaging systems 104 and 107 are aligned at the reference imaging distance H0, the imaging element 103 is arranged so that the sampling phases of the imaging systems 104 and 101 are shifted by 1/2 pixel. In FIG. 3, the subject at the point P0 forms an image at the center of the pixel 106a of the image sensor 106 in the image pickup system 104, but forms an image between the pixel 103a of the image sensor 103 and the image sensor 103b in the image pickup system 101. The sampling phase is shifted by 1/2 pixel.
 次に、撮像距離を変化させた場合のサンプリング位相の変化について述べる。 Next, the change in sampling phase when the imaging distance is changed will be described.
 図6に、焦点距離fを5mm、基線長B(=QR)を12mm、撮像距離H0を500mmとして、撮像距離をH0から変化させた場合の結像位置のシフト量を(1)式より計算しサンプリング位相のずれを計算した結果を示す。基準とした撮像距離H0(=500mm)ではサンプリング位相が1/2が画素ずれており、撮像距離を変えるとサンプリング位相が変化することがわかる。 In FIG. 6, the shift amount of the imaging position when the imaging distance is changed from H0 when the focal length f is 5 mm, the base length B (= QR) is 12 mm, the imaging distance H0 is 500 mm, is calculated from the equation (1). The results of calculating the sampling phase shift are shown below. It can be seen that at the reference imaging distance H0 (= 500 mm), the sampling phase is ½ pixel shifted, and the sampling phase changes when the imaging distance is changed.
 次に、図5に示した撮像系104と107のサンプリング位相のずれと、図6に示した撮像系104と101のサンプリング位相のずれを重ねた結果を図7に示す。撮像系104と107のサンプリング位相のずれを細線で、撮像系104と101のサンプリング位相のずれを太線で示している。図7からわかるように、一方のサンプリング位相が揃う撮像距離において、他方のサンプリング位相がずれており、すべての撮像距離において少なくとも一方は画素ずらしの状態となっている。図7の縦軸の表示範囲を0.25画素から0.75画素にしたものを、図8に示す。すべての撮像距離において、サンプリング位相のずれ量のどちらか一方が必ず0.25画素から0.75画素の範囲内に入っていることがわかる。 Next, FIG. 7 shows the result of overlapping the sampling phase shift between the imaging systems 104 and 107 shown in FIG. 5 and the sampling phase shift between the imaging systems 104 and 101 shown in FIG. The sampling phase shift between the imaging systems 104 and 107 is indicated by a thin line, and the sampling phase shift between the imaging systems 104 and 101 is indicated by a bold line. As can be seen from FIG. 7, at the imaging distance where one sampling phase is aligned, the other sampling phase is shifted, and at least one of the imaging distances is in a pixel shifted state. FIG. 8 shows the display range of the vertical axis in FIG. 7 changed from 0.25 pixel to 0.75 pixel. It can be seen that at any imaging distance, either one of the sampling phase shift amounts is always within the range of 0.25 to 0.75 pixels.
 次に、高解像度化するために必要なサンプリング位相のずれ量について述べる。図30に、2つの撮像系を用いて高解像度化の検証を行った結果を示す。水平に配置した2つの撮像系で解像度測定用のチャートを撮像し、2つの画像を合成し、合成画像の水平方向の解像度を評価した。図の横軸は撮像系からチャートまでの撮像距離を示しており、チャートを100cmから1cm刻みで125cmまで、順に計26点に配置して撮像した。図の縦軸は、合成画像の水平方向の解像度を示している。また、縦軸には、サンプリング位相のずれ量も合わせて示している。サンプリング位相のずれ量は、撮像系の焦点距離や画素ピッチ、撮像距離等を用いて、上述の方法で算出した。図より、十分に解像度を向上させるには、0.20画素から0.80画素のサンプリング位相のずれが必要であることがわかる。したがって、3つの撮像系を用い、すべての撮像距離において、サンプリング位相のずれ量のどちらか一方が必ず0.25画素から0.75画素の範囲内に入っている図8の条件においては、すべての撮像距離において高解像度化出来る。 Next, we will describe the amount of sampling phase shift required for higher resolution. FIG. 30 shows the result of verification of high resolution using two imaging systems. The chart for resolution measurement was imaged with two imaging systems arranged horizontally, the two images were synthesized, and the horizontal resolution of the synthesized image was evaluated. The horizontal axis of the figure shows the imaging distance from the imaging system to the chart, and the chart was placed at a total of 26 points in order from 100 cm to 125 cm in 1 cm increments. The vertical axis in the figure indicates the horizontal resolution of the composite image. The vertical axis also shows the sampling phase deviation. The sampling phase shift amount was calculated by the above-described method using the focal length, pixel pitch, imaging distance, and the like of the imaging system. From the figure, it can be seen that a sampling phase shift of 0.20 pixels to 0.80 pixels is necessary to sufficiently improve the resolution. Therefore, all three imaging systems are used, and at any imaging distance, either one of the sampling phase shift amounts is always within the range of 0.25 pixel to 0.75 pixel. The resolution can be increased at an imaging distance of.
 本実施形態では、撮像系104と101の水平方向の基線長QRと、撮像系104と107の水平方向の基線長RSをいずれも12mmと等しい構成とし、焦点距離fも5mmと等しくした。基線長、焦点距離を等しくすることで、撮像距離に変化に伴うシフト量の変化量が等しくなり、基準とする距離において、ある1つの撮像系(ここでは撮像系104)とサンプリング位相の揃う撮像系(ここでは撮像系107)と、サンプリング位相の1/2画素ずれる撮像系(ここでは撮像系101)を備えることで、すべての撮像距離にある被写体を、高解像度化可能な画素ずらしの状態で撮像することが可能となる。 In this embodiment, the horizontal baseline length QR of the imaging systems 104 and 101 and the horizontal baseline length RS of the imaging systems 104 and 107 are both equal to 12 mm, and the focal length f is also equal to 5 mm. By making the base line length and the focal length equal, the amount of change in the shift amount accompanying the change in the imaging distance becomes equal, and imaging with the same sampling phase as that of a certain imaging system (imaging system 104 in this case) at the reference distance. By providing a system (here, the imaging system 107) and an imaging system (here, the imaging system 101) that is shifted by 1/2 pixel of the sampling phase, it is possible to shift the pixels at all imaging distances so that the resolution can be increased. It becomes possible to pick up an image.
 次に、垂直方向の高解像度化について述べる。図9に、撮像素子106と109を光軸方向から見た図を示す。R’とS’は撮像系104と107のそれぞれの光軸と撮像素子が交わる点を示しており、光軸の高さは一致している。撮像素子109は、撮像素子106に対し、垂直方向にサンプリング位相が1/2画素ずれるよう配置されている。光軸の高さが一致していることから、垂直方向の基線長は零であり、式(1)より、撮像距離が変化してもサンプリング位相が変化しない。従って、すべての撮像距離において垂直方向のサンプリング位相が1/2画素ずれており、すべての撮像距離において高解像度化出来る。また、水平方向と同様、垂直方向の高解像度化についても、3つの撮像系のうち、少なくとも2つの撮像系のサンプリング位相が高解像度化可能な0.20から0.80画素ずれていればよい。 Next, we will discuss the high resolution in the vertical direction. FIG. 9 is a view of the image sensors 106 and 109 as viewed from the optical axis direction. R ′ and S ′ indicate points where the optical axes of the imaging systems 104 and 107 intersect with the imaging device, and the heights of the optical axes are the same. The image sensor 109 is arranged so that the sampling phase is shifted by 1/2 pixel in the vertical direction with respect to the image sensor 106. Since the heights of the optical axes coincide with each other, the base line length in the vertical direction is zero, and the sampling phase does not change even when the imaging distance is changed according to the equation (1). Accordingly, the sampling phase in the vertical direction is shifted by 1/2 pixel at all imaging distances, and high resolution can be achieved at all imaging distances. Similarly to the horizontal direction, in order to increase the resolution in the vertical direction, the sampling phase of at least two of the three imaging systems only needs to be shifted from 0.20 to 0.80 pixels at which the resolution can be increased. .
 以上のような構成により、撮像系を水平方向に3つ配置することで、撮像距離に依らず、垂直方向にも水平方向にも高解像度化可能となる。 With the configuration as described above, by arranging three imaging systems in the horizontal direction, high resolution can be achieved both in the vertical direction and in the horizontal direction regardless of the imaging distance.
 なお、垂直方向の高解像度化については、3つの撮像系のうち、少なくとも2つの撮像系のサンプリング位相が高解像度化可能な0.20から0.80画素ずれた状態であればよいため、3つの撮像系の配置は、上記に限らない。例えば、図10に示すように、3つの撮像系のサンプリング位相が1/3画素ずつずれた構成でもよいし、図11に示すように、基準の撮像系に対し、一方の撮像系のサンプリング位相が1/2画素ずれており、他方の撮像系のサンプリング位相が揃う構成でもよいし、図12に示すように、基準の撮像系に対し、2つの撮像系のサンプリング位相が共に1/2画素ずれており、その2つの撮像系のサンプリング位相が揃う構成にしてもよい。 Note that, in order to increase the resolution in the vertical direction, the sampling phase of at least two imaging systems out of the three imaging systems may be shifted from 0.20 to 0.80 pixels where the resolution can be increased. The arrangement of the two imaging systems is not limited to the above. For example, as shown in FIG. 10, the configuration in which the sampling phases of the three imaging systems are shifted by 1/3 pixels may be adopted, or as shown in FIG. 11, the sampling phase of one imaging system with respect to the reference imaging system. May be shifted by 1/2 pixel, and the sampling phase of the other imaging system may be aligned. As shown in FIG. 12, the sampling phases of the two imaging systems are both 1/2 pixels relative to the reference imaging system. A configuration may be adopted in which the sampling phases of the two imaging systems are aligned.
 次に、複眼方式の画素ずらしにより高解像度画像を得るために必要な、複数の画像のずれ量(視差)の計算について述べる。図13は、2つの撮像系1301と1302が垂直方向に並置され、被写体1303を撮像している。図14に、図13の状態で撮像した場合の撮像系1301の撮像画像1401と、撮像系1302の撮像画像1402を示す。被写体1303は垂直方向にずれて撮像される。ハードウェア処理により、撮影画像1402を基準とし撮影画像1401中から対応点を探索する場合、例えば、網掛けで示す撮影画像1402中の画素1404に対応する点は、撮影画像1401中の画素1403となるが、対応点の探索を斜線で示す左上の画素1405から行うと、対応点である画素1403に到達するまでデータを保持しておく必要があり、ラインメモリあるいはフレームメモリを利用する必要が生じる。 Next, calculation of shift amounts (parallax) of a plurality of images necessary for obtaining a high-resolution image by compound-eye pixel shifting will be described. In FIG. 13, two imaging systems 1301 and 1302 are juxtaposed in the vertical direction to image the subject 1303. FIG. 14 shows a captured image 1401 of the imaging system 1301 and a captured image 1402 of the imaging system 1302 when the image is captured in the state of FIG. The subject 1303 is imaged with a shift in the vertical direction. When a corresponding point is searched from the captured image 1401 based on the captured image 1402 by hardware processing, for example, a point corresponding to the pixel 1404 in the captured image 1402 indicated by shading is the pixel 1403 in the captured image 1401. However, if the corresponding point is searched from the upper left pixel 1405 indicated by diagonal lines, it is necessary to hold data until the corresponding point 1403 is reached, and it is necessary to use a line memory or a frame memory. .
 一方、図15は、2つの撮像系1501と1502が水平方向に並置され、被写体1503を撮像している。図16に、図15の状態で撮像した場合の撮像系1501の撮像画像1601と、撮像系1502の撮像画像1602を示す。被写体1503は水平方向にずれて撮像される。ハードウェア処理により、撮影画像1602を基準とし撮影画像1601中から対応点を探索する場合、例えば、網掛けで示す撮影画像1602中の画素1604に対応する点は、撮影画像1601中の画素1603となるが、垂直方向のずれがないため、対応点の探索を斜線で示す左上の画素1605から行う必要はなく、対応点である画素1603と同じ高さのライン上の左端の画素である横線で示す画素1606から探索すればよく、対応点に到達するまで保持しておくデータ量を、撮像系を垂直に並置する場合に比べて大幅に低減することが可能となる。 On the other hand, in FIG. 15, two imaging systems 1501 and 1502 are juxtaposed in the horizontal direction to image the subject 1503. FIG. 16 shows a captured image 1601 of the imaging system 1501 and a captured image 1602 of the imaging system 1502 when the image is captured in the state of FIG. The subject 1503 is imaged with a shift in the horizontal direction. When a corresponding point is searched from the photographed image 1601 based on the photographed image 1602 by hardware processing, for example, a point corresponding to the pixel 1604 in the photographed image 1602 indicated by shading is the pixel 1603 in the photographed image 1601. However, since there is no deviation in the vertical direction, it is not necessary to search for the corresponding point from the upper left pixel 1605 indicated by the diagonal line, and the horizontal line which is the leftmost pixel on the line having the same height as the pixel 1603 which is the corresponding point. It is only necessary to search from the pixel 1606 shown, and the amount of data held until reaching the corresponding point can be significantly reduced as compared with the case where the imaging systems are juxtaposed vertically.
 また、3つの撮像系を水平方向に並置することにより、従来の垂直方向と水平方向に並置する構成と比較し、垂直方向に薄型化することが可能となる。 Further, by juxtaposing the three image pickup systems in the horizontal direction, it is possible to reduce the thickness in the vertical direction as compared with the conventional arrangement in which the vertical and horizontal directions are juxtaposed.
<第二の実施形態>
 次に、本発明の第二の実施形態について述べる。第一の実施形態では、焦点距離の等しい3つの撮像系を水平方向の基線長が等しくなるよう配置することで、撮像距離に依らず垂直方向、水平方向ともに高解像度化可能な構成を実現した。
<Second Embodiment>
Next, a second embodiment of the present invention will be described. In the first embodiment, by arranging three imaging systems having the same focal length so that the baseline lengths in the horizontal direction are equal, a configuration capable of increasing the resolution in both the vertical and horizontal directions regardless of the imaging distance is realized. .
 第二の実施形態では、撮像装置の実際の製造過程において位置精度に誤差が生じた場合においても撮像距離に依らず高解像度化可能な構成について説明する。 In the second embodiment, a configuration capable of increasing the resolution regardless of the imaging distance even when an error occurs in the position accuracy in the actual manufacturing process of the imaging device will be described.
 3つの撮像系が等間隔に並んでいない場合、撮像距離の変化に伴うシフト量の大きさが異なるが、基線長の違いが微小であれば、サンプリング位相に与える影響も微小となり、撮像距離に依らず高解像度化可能となる。 If the three imaging systems are not arranged at equal intervals, the amount of shift accompanying the change in the imaging distance is different, but if the difference in the baseline length is small, the influence on the sampling phase will be small, and the imaging distance will be reduced. Regardless, the resolution can be increased.
 例えば、図3において、3つの撮像系の焦点距離fを5mm、基線長RSとQRを12mmとして、高解像度化したい撮像距離の範囲を1000mmから3000mmとした場合において、上記撮像距離範囲内において高解像度化出来るよう撮像するための基線長QRの許容誤差を求める。撮像距離1500mmで撮像系104撮像系107のサンプリング位相が1/2画素ずれる状態とし、撮像系101と撮像系104のサンプリング位相が揃った状態とする。QRをQR=12mmの状態から1mmずつ変化させると、QR=11.7mmからQR=12.4mmの範囲においては、撮像距離1000mmから3000mmの間における撮像系104と撮像系107のサンプリング位相と、撮像系101と撮像系104のサンプリング位相の少なくとも一方が0.20から0.80画素ずれており、撮像距離に依らず高解像度化出来ると言える。図17にQR=11.7mmの場合のサンプリング位相のずれ量を示し、図18にQR=11.0mmの場合のサンプリング位相のずれ量を示す。図18では、丸で示した撮像距離2320mm付近と2980mm付近において2つのサンプリング位相のずれ量が共に0.20から0.80画素の範囲に入っておらず、高解像度化出来ない状態となっている。一方、図17では、すべての距離において2つのサンプリング位相のずれ量のいずれか一方が0.20から0.80画素の範囲に入っており、高解像度化出来る状態となっている。 For example, in FIG. 3, when the focal length f of the three imaging systems is 5 mm, the base lengths RS and QR are 12 mm, and the range of the imaging distance to be increased in resolution is 1000 mm to 3000 mm, An allowable error of the base line length QR for imaging so that the resolution can be achieved is obtained. The sampling phase of the imaging system 104 and the imaging system 107 is shifted by ½ pixel at an imaging distance of 1500 mm, and the sampling phase of the imaging system 101 and the imaging system 104 is aligned. When QR is changed by 1 mm from the state of QR = 12 mm, in the range of QR = 11.7 mm to QR = 12.4 mm, the sampling phases of the imaging system 104 and the imaging system 107 between the imaging distances of 1000 mm and 3000 mm, At least one of the sampling phases of the imaging system 101 and the imaging system 104 is shifted by 0.20 to 0.80 pixels, and it can be said that the resolution can be increased regardless of the imaging distance. FIG. 17 shows the amount of sampling phase shift when QR = 11.7 mm, and FIG. 18 shows the amount of sampling phase shift when QR = 11.0 mm. In FIG. 18, the two sampling phase shift amounts are not within the range of 0.20 to 0.80 pixels at the imaging distances of 2320 mm and 2980 mm indicated by circles, and the resolution cannot be increased. Yes. On the other hand, in FIG. 17, at any distance, either one of the two sampling phase shift amounts is in the range of 0.20 to 0.80 pixels, and the resolution can be increased.
 以上のように、第二の実施形態では、撮像系101と104の基線長QRと撮像系104と107の基線長RS基線長に誤差がある場合においても、許容範囲内にその誤差を抑えることにより、所定の撮像距離の範囲内において、撮像距離に依らず高解像度化することが可能となる。 As described above, in the second embodiment, even when there is an error in the baseline length QR of the imaging systems 101 and 104 and the baseline length RS of the imaging systems 104 and 107, the error is suppressed within an allowable range. As a result, it is possible to increase the resolution regardless of the imaging distance within the range of the predetermined imaging distance.
 なお、サンプリング位相は基線長だけでなく、光軸と撮像素子の相対的な位置関係、焦点距離、画素ピッチによっても変化し、また、高解像度化したい撮像距離の範囲にも依存するため、許容される誤差の範囲を基線長だけで定義することは出来ないが、光学系の光軸と撮像素子の相対的な位置関係、焦点距離、画素ピッチ、高解像度化したい撮像距離の範囲が決まれば、許容範囲は式(1)から求まるシフト量より計算することが可能となる。 Note that the sampling phase varies depending not only on the base line length but also on the relative positional relationship between the optical axis and the image sensor, the focal length, and the pixel pitch, and also depends on the range of the imaging distance for which high resolution is desired. The error range cannot be defined only by the baseline length, but if the relative positional relationship between the optical axis of the optical system and the image sensor, the focal length, the pixel pitch, and the range of the imaging distance to be increased in resolution are determined. The allowable range can be calculated from the shift amount obtained from the equation (1).
 本実施形態では、所定の撮像距離の範囲内において、2つのサンプリング位相のずれ量のうち、少なくとも一方が、高解像度化可能な0.20から0.80画素の範囲に入るよう3つの撮像系を配置することで、撮像距離に依らず、水平方向の解像度を向上させることが可能となる。 In the present embodiment, three imaging systems in which at least one of the two sampling phase shift amounts falls within a range of 0.20 to 0.80 pixels capable of high resolution within a predetermined imaging distance. By arranging, the horizontal resolution can be improved regardless of the imaging distance.
 また、垂直方向に関しては、第一の実施形態と同様の構成にすることで、撮像距離に依らず高解像度化可能となる。 Further, with respect to the vertical direction, by adopting the same configuration as that of the first embodiment, high resolution can be achieved regardless of the imaging distance.
<第三の実施形態>
 次に、本発明の第三の実施形態について述べる。第三の実施形態では、図11や図12に示すように、水平方向に配置した3つの撮像系のうち、2つの撮像系の垂直方向のサンプリング位相が揃い、もう1つの撮像系の垂直方向のサンプリング位相が他の2つの撮像系に対し1/2画素ずれる構成とする。
<Third embodiment>
Next, a third embodiment of the present invention will be described. In the third embodiment, as shown in FIG. 11 and FIG. 12, among the three imaging systems arranged in the horizontal direction, the vertical sampling phases of the two imaging systems are aligned, and the vertical direction of the other imaging system is aligned. The sampling phase is shifted by 1/2 pixel with respect to the other two imaging systems.
 本発明では、基線長の小さな垂直方向については撮像距離に依らずサンプリング位相のずれがほぼ一定であるが、基線長の大きな水平方向については撮像距離によってサンプリング位相のずれが変化するため、画像間のずれ量(視差)をステレオマッチングなどの方法を用いて求める必要がある。ステレオマッチング法において対応点を探索する際、サンプリング位相にずれがある場合には、完全に一致する対応点は存在せず、所定の対応点は画素と画素の間に存在することになる。 In the present invention, the sampling phase shift is almost constant regardless of the imaging distance in the vertical direction where the baseline length is small, but the sampling phase shift varies depending on the imaging distance in the horizontal direction where the baseline length is large. The shift amount (parallax) must be obtained using a method such as stereo matching. When searching for corresponding points in the stereo matching method, if there is a deviation in the sampling phase, there is no perfectly corresponding corresponding point, and a predetermined corresponding point exists between pixels.
 図19を用いて説明する。図19は、水平方向に並置した2つの撮像系で撮像した画像1901と1902を示している。撮像画像1902中の画素1903に対応する点を撮像画像1901中から探索すると、対応する点は画素1904と画素1905の間に存在する。 This will be described with reference to FIG. FIG. 19 shows images 1901 and 1902 captured by two imaging systems juxtaposed in the horizontal direction. When a point corresponding to the pixel 1903 in the captured image 1902 is searched from the captured image 1901, the corresponding point exists between the pixel 1904 and the pixel 1905.
 更に、垂直方向にもサンプリング位相にずれがある場合について、図20を用いて説明する。図20は、水平方向に並置した2つの撮像系で撮像した画像2001と2002を示しており、垂直方向のサンプリング位相も1/2画素ずれている。図20において、撮像画像2002中の画素2003に対応する点は、撮像画像2001中の4つの画素2004、2005、2006、2007の間に存在する。 Furthermore, the case where there is a deviation in the sampling phase in the vertical direction will be described with reference to FIG. FIG. 20 shows images 2001 and 2002 captured by two imaging systems juxtaposed in the horizontal direction, and the sampling phase in the vertical direction is also shifted by ½ pixel. In FIG. 20, a point corresponding to the pixel 2003 in the captured image 2002 exists between the four pixels 2004, 2005, 2006, and 2007 in the captured image 2001.
 このように、対応点探索の際、サンプリング位相にずれがある場合は完全に一致する対応点が存在せず、計算精度が低下する恐れがある。第三の実施形態では、2つの撮像系の垂直方向のサンプリング位相が揃っており、対応点探索をする際、少なくとも垂直方向にはずれのない2つの画像から検索することが可能となり、計算精度が向上する。計算精度が向上することにより、画像間のずれ量をもとに各画素を対応する点に配置し高解像度画像上を生成する際の誤差が小さくなり、より解像度の高い画像を生成することが可能となる。 Thus, when searching for corresponding points, if there is a shift in the sampling phase, there is no corresponding matching point, and the calculation accuracy may be reduced. In the third embodiment, the sampling phases in the vertical direction of the two imaging systems are aligned, and when searching for corresponding points, it is possible to search from at least two images that are not shifted in the vertical direction, and the calculation accuracy is high. improves. By improving the calculation accuracy, it is possible to generate an image with higher resolution by reducing the error when generating each pixel on the corresponding point based on the shift amount between images and generating on the high resolution image. It becomes possible.
<第四の実施形態>
 次に、本発明の第四の実施形態について述べる。第四の実施形態では、図10に示すように、3つの撮像系の垂直方向のサンプリング位相が1/3画素ずつずれる構成とする。上述のように、2つの撮像系で高解像度化する場合、サンプリング位相が1/2画素ずれている状態が最適な条件であるが、これは、サンプリングの間隔が最も狭くなるためである。例えば、垂直方向のサンプリング位相が0.3画素ずれている状態であるとすると、上下方向のうち、一方は0.3画素ずれているが、他方は0.7画素ずれており、サンプリングの間隔が広い。垂直方向のサンプリング位相が1/2画素ずれている状態であれば、上下方向ともに1/2画素となる。したがって、撮像系を3つ備える場合は、サンプリング位相が1/3画素ずつずれている状態が、サンプリングの間隔が最も狭くなる構成であり、高解像度化に好適である。
<Fourth embodiment>
Next, a fourth embodiment of the present invention will be described. In the fourth embodiment, as shown in FIG. 10, the vertical sampling phases of the three imaging systems are shifted by 1/3 pixels. As described above, when the resolution is increased by the two imaging systems, the optimum condition is that the sampling phase is shifted by 1/2 pixel, which is because the sampling interval is the narrowest. For example, if the sampling phase in the vertical direction is shifted by 0.3 pixels, one of the vertical directions is shifted by 0.3 pixels, while the other is shifted by 0.7 pixels, and the sampling interval Is wide. If the vertical sampling phase is shifted by ½ pixel, the vertical direction is ½ pixel. Therefore, when three imaging systems are provided, a state in which the sampling phase is shifted by 1/3 pixel is a configuration in which the sampling interval is the shortest, which is suitable for high resolution.
 上述のように、本発明は、3つの撮像系を画素の配列方向に配置し、所定範囲の撮像距離内にある被写体が、いずれか2つの撮像系により、高解像度化の条件であるサンプリング位相のずれが0.2から0.8画素ずれた状態で撮像される構成にすることで、水平・垂直両方向を高解像度化するとともに、3つの撮像系を一方向に配置することで、ハードウェアで視差算出などの画像処理を行うために必要な処理量やメモリ量を削減することが出来る。 As described above, according to the present invention, three imaging systems are arranged in the pixel arrangement direction, and a subject within a predetermined range of imaging distance is subjected to a sampling phase which is a condition for high resolution by any two imaging systems. By adopting a configuration in which an image is captured with a shift of 0.2 to 0.8 pixels, the resolution is improved in both the horizontal and vertical directions, and the hardware is provided by arranging the three imaging systems in one direction. Thus, it is possible to reduce a processing amount and a memory amount necessary for performing image processing such as parallax calculation.
 本発明は撮像系を複数用いた撮像装置に利用可能である。 The present invention can be used for an imaging apparatus using a plurality of imaging systems.
101、104、107、1301、1302、1501、1502 撮像系
102、105、108 光学系
103、106、109 撮像素子
1303、1503 被写体
1401、1402、1601、1602、1901、1902、2001、2002 撮像画像
1403、1404、1405、1603、1604、1605、1606、1903、1904、1905、2003、2004、2005、2006、2007 画素
101, 104, 107, 1301, 1302, 1501, 1502 Imaging system 102, 105, 108 Optical system 103, 106, 109 Imaging device 1303, 1503 Subject 1401, 1402, 1601, 1602, 1901, 1902, 2001, 2002 Captured image 1403, 1404, 1405, 1603, 1604, 1605, 1606, 1903, 1904, 1905, 2003, 2004, 2005, 2006, 2007 pixels

Claims (6)

  1.  光学系と撮像素子とを有する撮像系を少なくとも3つ備え、前記撮像系は、前記撮像素子の画素の配列方向に配置され、第一の撮像系の撮像画像中の所定範囲内の撮像距離にある被写体が、前記第一の撮像系の撮像画像に対し、他のいずれかの撮像系により、前記配列方向のサンプリング位相が0.2から0.8画素ずれて撮像されることを特徴とする撮像装置。 The image pickup system includes at least three image pickup systems each having an optical system and an image pickup element. The image pickup system is arranged in a pixel arrangement direction of the image pickup element and has an image pickup distance within a predetermined range in a picked-up image of the first image pickup system. A subject is imaged with a sampling phase in the arrangement direction shifted from 0.2 to 0.8 pixels with respect to a captured image of the first imaging system by any other imaging system. Imaging device.
  2.  前記撮像系は前記配列方向に略等間隔に配置され、前記撮像系の光軸は略平行であることを特徴とする請求項1に記載の撮像装置。 2. The imaging apparatus according to claim 1, wherein the imaging systems are arranged at substantially equal intervals in the arrangement direction, and the optical axes of the imaging systems are substantially parallel.
  3.  所定範囲の撮像距離内にある被写体が、いずれかの撮像系の撮像画像に対し、他のいずれかの撮像系により、前記配列方向と垂直な方向のサンプリング位相が0.2から0.8画素ずれて撮像されることを特徴とする請求項1または2に記載の撮像装置。 A subject within an imaging distance of a predetermined range has a sampling phase of 0.2 to 0.8 pixels in a direction perpendicular to the arrangement direction by any of the other imaging systems with respect to a captured image of any imaging system. The imaging apparatus according to claim 1, wherein the imaging is performed with a shift.
  4.  所定範囲の撮像距離内にある被写体が、いずれかの撮像系の撮像画像に対し、他のいずれかの撮像系により、前記配列方向と垂直な方向のサンプリング位相が略0.5画素ずれて撮像されることを特徴とする請求項1または2に記載の撮像装置。 A subject within an imaging distance within a predetermined range is imaged with a sampling phase in a direction perpendicular to the arrangement direction shifted by approximately 0.5 pixels by any other imaging system with respect to a captured image of any imaging system. The imaging apparatus according to claim 1 or 2, wherein
  5.  所定範囲の撮像距離内にある被写体が、いずれかの撮像系の撮像画像に対し、他のいずれかの撮像系により、前記配列方向と垂直な方向のサンプリング位相がほぼずれなく撮像されることを特徴とする請求項3または4に記載の撮像装置。 A subject within an imaging distance of a predetermined range is imaged with a sampling phase in a direction perpendicular to the arrangement direction with almost no deviation from one of the other imaging systems with respect to a captured image of any imaging system. The imaging apparatus according to claim 3 or 4, wherein the imaging apparatus is characterized.
  6.  所定範囲の撮像距離内にある被写体が、いずれかの撮像系の撮像画像に対し、他のいずれか2つの撮像系により、前記配列方向と垂直な方向のサンプリング位相が略1/3画素ずつずれて撮像されることを特徴とする請求項1または2に記載の撮像装置。 A subject within a predetermined imaging distance is shifted by approximately 1/3 of the sampling phase in the direction perpendicular to the arrangement direction with respect to the captured image of any imaging system by any two other imaging systems. The imaging apparatus according to claim 1, wherein the imaging apparatus captures an image.
PCT/JP2011/059243 2010-04-14 2011-04-14 Image capture device WO2011129387A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2010092879 2010-04-14
JP2010-092879 2010-04-14

Publications (1)

Publication Number Publication Date
WO2011129387A1 true WO2011129387A1 (en) 2011-10-20

Family

ID=44798756

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2011/059243 WO2011129387A1 (en) 2010-04-14 2011-04-14 Image capture device

Country Status (2)

Country Link
JP (1) JP2011239372A (en)
WO (1) WO2011129387A1 (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2013118305A1 (en) * 2012-02-10 2013-08-15 シャープ株式会社 Image generation device, image generation method, and computer-readable recording medium having image generation program recorded therein
US10090608B2 (en) 2016-09-29 2018-10-02 Delphi Technologies, Inc. Electrical connection system having a terminal with contact ridges

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH11355794A (en) * 1998-06-08 1999-12-24 Hitachi Denshi Ltd Television camera device
JP2001078217A (en) * 1999-06-30 2001-03-23 Canon Inc Image pickup device
WO2007060847A1 (en) * 2005-11-22 2007-05-31 Matsushita Electric Industrial Co., Ltd. Imaging device

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH11355794A (en) * 1998-06-08 1999-12-24 Hitachi Denshi Ltd Television camera device
JP2001078217A (en) * 1999-06-30 2001-03-23 Canon Inc Image pickup device
WO2007060847A1 (en) * 2005-11-22 2007-05-31 Matsushita Electric Industrial Co., Ltd. Imaging device

Also Published As

Publication number Publication date
JP2011239372A (en) 2011-11-24

Similar Documents

Publication Publication Date Title
EP3033733B1 (en) Stereo yaw correction using autofocus feedback
TWI628955B (en) Device and method for detecting an object region
EP3514598B1 (en) Image pickup apparatus including lens elements having different diameters
US20190373246A1 (en) Multi-lens camera with a single image sensor
US20120293633A1 (en) Stereo camera
JP2010078768A (en) Stereoscopic image capturing apparatus and stereoscopic image capturing system
TW201622403A (en) Digital refocusing method
US8537256B2 (en) Image pickup device and solid-state image pickup element
JP2009188973A (en) Imaging apparatus, and optical axis control method
JP2009284188A (en) Color imaging apparatus
JP2012189551A5 (en)
TW201043017A (en) Spatial-pixel-offset imaging device
JP2010130628A (en) Imaging apparatus, image compositing device and image compositing method
WO2011129387A1 (en) Image capture device
CN104301590A (en) Three-lens detector array video acquisition device
JP4729011B2 (en) Element image group conversion device, stereoscopic image display device, and element image group conversion program
JP2011095431A (en) Stereoscopic video pickup device and stereoscopic video pickup method
EP2851748B1 (en) Optoelectronic photodetector (variants)
CN102171533B (en) Distance measuring device and method for manufacturing same
TWI794571B (en) Device comprising a multi-aperture imaging device for accumulating image information
JP6022428B2 (en) Image photographing apparatus and image photographing method
WO2016051871A1 (en) Multi-lens imaging device, method for manufacturing same, and information terminal device
JP5474651B2 (en) Imaging device
WO2018139250A1 (en) Entire celestial-sphere image capture device
WO2018167999A1 (en) Projector and projector system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 11768908

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 11768908

Country of ref document: EP

Kind code of ref document: A1