JP4158344B2 - Image input device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention is an image input in which a large number of optical fiber strands are bundled to form a multi-core structure, and a plurality of image input units configured using the structure are arranged radially to perform omnidirectional imaging. It relates to the device.
[0002]
[Prior art]
Conventionally, there are several means as an image input apparatus for performing omnidirectional photographing. One of them is a reflection mirror type panoramic photographing apparatus shown in FIG. 12A. For example, a reflection mirror part 61 of a convex mirror is attached to the ceiling of a building, and a certain distance is provided directly below the reflection mirror part 61. A camera unit 62 for photographing is arranged at a distance, and the camera unit 62 is configured to capture an image projected on the reflection mirror unit 61. FIG. 12B shows a video image displayed by the camera unit 62. This video image is very distorted.
[0003]
Another one that performs omnidirectional imaging is imaging with a camera unit 62 to which a super-wide-angle lens unit 63 is attached, as in the super-wide-angle lens panoramic imaging device shown in FIG. The captured image in this case is also a very distorted image as shown in FIG.
[0004]
In addition to the above, as shown in FIG. 14A, a camera (not shown) to which a wide-angle lens is attached is housed in a substantially cubic box body 64, and each side surface and top surface of the box body 64 are accommodated. Therefore, there is an image input device configured to be capable of photographing all directions in response to desire for wide-angle lenses 65a to 65e (wide-angle lenses 65d and 65e are not shown). The image input apparatus configured as described above has less image distortion than the above two reflection mirror type and super wide angle lens type, but distortion occurs in the peripheral image due to the use of the wide angle lens. FIG. 14B is a developed image taken by each camera when the horizontal and vertical angles of view are 90 degrees.
[0005]
[Problems to be solved by the invention]
The above-described omnidirectional image input device shown in FIGS. 12 and 13 employs means for taking an omnidirectional image by capturing an omnidirectional image into an image pickup unit in an optically very distorted form. . For this reason, since the omnidirectional video obtained in the imaging unit is very distorted, there is a problem that enormous calculation processing is required to process the video without distortion.
[0006]
In addition, since the image input apparatus shown in FIG. 14 (a) covers all directions with a plurality of cameras, it differs when viewing the whole image or cutting out video in any direction and angle of view. It is necessary to connect the images taken with the camera. At this time, the images of the peripheral images with the largest distortion are connected, and there is a problem that enormous arithmetic processing is still required for image alignment (stitching) and distortion correction.
[0007]
Furthermore, even when a plurality of camera images are combined to form an omnidirectional video, the shooting range covered by each camera is wide. For example, even when eight cameras are used, each camera has a tens of millimeters in terms of 35 mm. The angle of view corresponds to an ultra-wide-angle lens. For this reason, the image by each camera contains the distortion of the image peculiar to a wide angle lens. When synthesizing omnidirectional images obtained by such a system, it is necessary to correct peripheral distortion and connect the adjacent images. For this reason, the process requires a great deal of time. In addition, when an image of an arbitrary narrow rectangular area is taken out from these image videos, it is necessary to perform a coordinate conversion process on the distorted image and output it after removing the distortion.
[0008]
In order to solve these problems, there is a means for increasing the number of photographing cameras to narrow the angle of view covered by one camera. However, there are disadvantages that new problems such as extremely difficult to synchronize when recording and reproducing videos from a plurality of photographing cameras occur, and that the entire system scale is increased.
[0009]
The present invention has been made in view of the above circumstances, and without using a reflection mirror, an ultra-wide angle lens, or the like, an image input unit configured by bundling a large number of optical fiber strands is arranged radially, and the image input unit This eliminates the need for arithmetic processing to correct the image distortion in the image pickup unit so that all distortion-free images taken from the narrow angle of view of each front can be guided to the image pickup unit. It is an object of the present invention to provide an image input device that can constitute a small camera for omnidirectional photographing only by an image input unit configured by bundling a plurality of images and an image pickup unit.
[0010]
[Means for Solving the Problems]
In order to achieve the above object, an image input device according to the present invention includes a multi-core structure formed by bundling a number of optical fiber strands,
An image input unit for capturing an image formed at one end of the multi-core structure;
Means for bundling a plurality of the multi-core structures;
Means for reading an image captured from the image input unit provided at each end of the multi-core structure bundled in plural by this means to the other end by the multi-core structure and reading with an image sensor via optical means When,
Equipped with,
The image input unit is arranged radially on each surface of an image capturing unit made of a regular polyhedron .
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings.
[0012]
[First Embodiment]
FIGS. 1A to 1C are a schematic configuration diagram and an enlarged view showing a first embodiment of the present invention. In FIGS. 1A to 1C, a number of optical fiber strands 11 are bundled together. The multi-core structure (hereinafter referred to as image fiber) 12a, 12b, 12c,... Is configured so that an image can be taken from the end face. As shown in FIG. 1B, one end of each of the image fibers 12a, 12b, 12c... Is mounted and covered inside the cylindrical body 14, and is configured into a plurality of image input units 15a, 15b, 15c. Is done.
[0013]
The other ends of the image fibers 12a, 12b, 12c... Are formed in a quadrangular shape so that they can be read simultaneously by an image sensor such as an image sensor (CCD image sensor) via optical means as will be described later. Attached to the fiber bundle 16. As shown in the drawing, a plurality of cylindrical holes 17a, 17b, 17c,... These cylindrical holes 17a, 17b, 17c... Are inserted with the ends of the image fibers 12a, 12b, 12c... Constituting the image input units 15a, 15b, 15c. The end faces of 12b, 12c,... Are aligned flat so that they are all the same face.
[0014]
FIG. 1C is an enlarged view of the image input units 15a, 15b, 15c..., And the objective lens 13 or a mechanism having an equivalent function is incorporated in these image input units 15a, 15b, 15c. . As shown in FIG. 1C, the objective lens 13 is supported and provided by the cylindrical body 14.
[0015]
The image input units 15a, 15b, 15c,... Are provided in the omnidirectional video capturing unit 20 formed in a spherical shape shown in FIGS. 2A to 2E so that omnidirectional video can be captured. They are arranged radially at intervals of 30 degrees when viewed in terms of latitude and longitude. In addition, microphones 21 are arranged at intervals of 60 degrees around the equatorial plane of the spherical omnidirectional video capturing unit 20. As will be described later, the microphone 21 is configured to record audio in multiple channels and reproduce it according to the direction of the video being viewed during reproduction. The omnidirectional video capturing unit 20 is supported and fixed to a base (not shown) or the like by a column 22.
[0016]
FIG. 3 is a schematic configuration diagram showing an overall configuration in the first embodiment configured using FIGS. 1 and 2. In FIG. 3, the images captured by the image input units 15a, 15b, 15c... Arranged in the omnidirectional image capturing unit 20 are sent to the fiber bundling unit 16 via the image fibers 12a, 12b, 12c. An image is obtained on the end face of the fiber bundle portion 16.
[0017]
FIG. 4 is a schematic configuration diagram in which the image image obtained on the end face of the fiber bundle portion 16 is formed on the element surface of the image pickup device 24 configured as an image sensor via the optical means 23. is there.
[0018]
FIG. 5 is an explanatory diagram showing a state in which the image image obtained on the end face of the fiber bundle portion 16 is formed on the image sensor. In the image on the image sensor in FIG. 5, the image fibers 12a and 12b are illustrated. , 12c... Are arranged in a lattice shape as shown in the figure, so that the frame portion of the fiber bundle portion 16 is observed as black. Therefore, the angle of view (circle shown in FIG. 6) by the image fibers on the image input units 15a, 15b, 15c... Arranged on the omnidirectional video capturing unit 20 overlap each other as shown in the figure. It is possible to eliminate blind spots by taking a wide area. In FIG. 6, a substantially square portion shown in the drawing is an angle of view to be used.
[0019]
In addition, since an image image is obtained on the end face of the fiber bundling portion 16, the image input portions 15a, 15b, 15c... Arranged radially and the image fibers 12a, 12b, 12c. The subsequent image processing becomes easier if the mutual positional relationship with the end face and the arrangement of the top and bottom are aligned.
[0020]
The omnidirectional images captured by the image input units 15a, 15b, 15c... Configured as described above are imaged on the image sensor 24 via the image fibers 12a, 12b, 12c. The image sensor 24 converts these images into electric signals and inputs them to the signal processing unit 41 shown in FIG.
[0021]
FIG. 7 is a block configuration diagram of a circuit device that converts an image obtained by the image input unit into an electric signal by an image pickup device and processes the electric signal. The electric signal from the image pickup device 24 is processed by the signal processing unit 41, and the processed signal Is recorded in the recording device 43 via the device interface unit 42 and supplied to an external processing apparatus (not shown) via the external interface unit 44.
[0022]
The electric circuit after the signal processing unit 41 can use almost the same device as a normal digital video recording device, but it is possible to reduce the amount of data by using compression processing or the like if necessary. It is.
[0023]
For audio recording, omnidirectional recording using omnidirectional microphones and general stereo recording, as well as multi-channel recording using a number of microphones, can be played back according to the direction of the video being viewed. An audio signal for switching a sound channel to be processed may be processed by the signal processing unit 41. A power supply unit 46 is provided to supply power to the signal processing unit 41 and other units.
[0024]
In the above description, the case where the omnidirectional video capturing unit 20 is spherical has been described, but FIGS. 8A, 8B, and 8C are planes when the omnidirectional video capturing unit 20 is configured as a regular octahedron. It is a figure, a front view, and a side view. The image input units 15a to 15h are arranged at the centers of equilateral triangles constituting the surface portions 20a to 20h (the surface portion 20h is not shown) of the regular octahedron so that images of all directions can be taken. In addition, when this regular octahedron omnidirectional video capturing unit 20 is used, the eight image input units 15a to 15h cover all directions, so the range covered by one image input unit is the horizontal plane. The angle of view is 90 degrees, which is 1/4 the angle of view of the hemisphere.
[0025]
FIG. 9 is a schematic configuration diagram of the fiber bundling portion 16a in the case where the regular octahedron omnidirectional image capturing portion 20 is used. In this fiber bundling portion 16a, cylindrical holes 17a, 17b,. Is formed.
[0026]
The cylindrical hole formed in the fiber bundling portion 16a can be freely selected depending on how the image fibers are bundled. However, the later image processing becomes easier when the fiber is configured in a lattice shape.
[0027]
In the first embodiment configured as described above, an example has been described in which the whole sky is divided at an interval of 30 degrees by latitude and longitude, but this interval can be arbitrarily set. When the shooting angle of view is set narrow as described above, there is an advantage that the angle of view of the standard lens (near 50 mm) in a normal 35 mm camera is slightly closer to the telephoto side, and there is little distortion or the like of the shot image. This is because there is almost no distortion of each image because a plurality of image fibers are used and only images captured from a narrow angle of view of each front are used.
[0028]
In addition, since all image fibers are bundled and imaged by one image sensor, there is an advantage that it is not necessary to synchronize a plurality of camera images as in the case of shooting using a plurality of cameras. . Further, on the image sensor, the images of the image fibers arranged horizontally are arranged so as to be imaged on the same horizontal line. Therefore, when an image on the image pickup device is taken out, a necessary range of video can be synthesized by reading out data for necessary pixels (pixels) from the horizontal direction and the vertical direction based on the cut-out position. That is, the composition of the video is very simple compared to the conventional method.
[0029]
For this reason, when images from a plurality of image fibers are connected and viewed, it is very easy to connect the images. This provides a great advantage that the following processing is simple and fast when a video having an arbitrary angle of view is cut out from an omnidirectional video taken.
[0030]
If the processing becomes simple, distortion correction of each image becomes unnecessary, and it is not necessary to synchronize a plurality of images, and furthermore, the vertical / horizontal relationship of the shooting target is maintained vertically and horizontally on the image sensor, Even when a rectangular area is cut out, the processing time is shortened because complicated coordinate conversion is not required.
[0031]
Further, among the image input units 15a, 15b, 15c..., When the image fibers 12a, 12b, 12c... Of the image input units arranged at the same latitude on the horizontal plane are arranged in a lattice shape by the fiber bundle unit 16, Place on the same horizontal line.
[0032]
Next, when the image input units on the same longitude are arranged in the same vertical column and the image is captured in the image sensor with the spherical surface expanded, the omnidirectional image can be viewed like a world map. . Accordingly, it is clear from the positional relationship between the videos that the video having an arbitrary angle of view is cut out from the video, and there is an advantage that it can be performed very easily.
[0033]
As described above, in the image input device according to the first embodiment of the present invention, an omnidirectional video can be captured easily and without distortion, and the omnidirectional video can be captured even when monitoring the captured video. In addition to the high visibility, there is an advantage that the labor of image processing can be greatly simplified even when a partial video is cut out.
[0034]
In addition, when using a computer with very high computing capability and the signal processing capability such as coordinate conversion is sufficient, the arrangement of the image input unit is set to a direction in which the resolution of the image sensor is more effectively utilized. This increases the flexibility of the entire system. An embodiment of such an omnidirectional video capturing unit will be described with reference to FIG.
[0035]
FIGS. 10A and 10B are schematic configuration diagrams of the omnidirectional video capturing unit 20 configured as a regular icosahedron, in which optical fiber strands are bundled in a regular triangle at the center of each surface part of the regular icosahedron. Input units 15a, 15b, 15c... Are arranged.
[0036]
On the other hand, on the image sensor side, as shown in FIG. 11, the image sensor is not limited to the spatial arrangement of the image input units 15a, 15b, 15c... So that the resolution of the image sensor (CCD image sensor) can be used effectively. Lined up. In this case, the image input unit and the lens are appropriately selected so that the shooting angle of view can cover the space without any gap. In such a configuration, the utilization efficiency of the image sensor is improved. In the case of the regular icosahedron, the image input unit is configured as an equilateral triangle, but the same effect can be obtained with other shapes.
[0037]
【The invention's effect】
As described above, according to the present invention, the distortion of the omnidirectional video is very small, and at the same time, the visibility of the monitoring video is very high. In addition, it is difficult to create shadows due to lighting, so the visibility of images for simultaneous monitoring is further improved.
[0038]
The image taken in through a multi-core structure composed of many optical fiber strands is less distorted, and the positional relationship between each other is simple and easy to understand, making it easy to cut out images at any angle of view and direction. . In addition, by changing the multi-core structure layout of the video capturing unit, it is possible to flexibly cope with changes in the pattern and range of the captured video.
[Brief description of the drawings]
FIGS. 1A to 1C are a schematic configuration diagram and an enlarged view showing a first embodiment of the present invention.
FIGS. 2A to 2E are a front view, a right side view, a plan view, a cross-sectional view, and a left side view of an omnidirectional video capturing unit;
FIG. 3 is a schematic overall configuration diagram of the first embodiment.
FIG. 4 is a schematic configuration diagram when an image image obtained on an end face of a fiber bundle portion is formed on an image sensor.
FIG. 5 is an explanatory diagram showing a state in which an image image obtained on the end face of the fiber bundle portion is formed on the image sensor.
FIG. 6 is an explanatory diagram of an angle of view of an image fiber and an angle of view to be used.
FIG. 7 is a block configuration diagram of a circuit device that converts an image obtained by an image input unit into an electrical signal by an image sensor and processes it.
FIGS. 8A, 8B, and 8C are a plan view, a front view, and a side view when the omnidirectional video capturing unit is configured as a regular octahedron. FIGS.
FIG. 9 is a perspective view of a fiber bundle portion.
FIGS. 10A and 10B are schematic configuration diagrams of an omnidirectional video capturing unit configured as a regular icosahedron. FIGS.
FIG. 11 is a schematic configuration diagram showing an image arrangement on an image sensor (CCD image sensor).
FIG. 12 is a schematic configuration diagram of a reflection mirror type panoramic photographing apparatus which is an image input apparatus that performs conventional omnidirectional photographing.
FIG. 13 is a schematic configuration diagram of a conventional super wide-angle lens panoramic photographing device that is an image input device that performs omnidirectional photographing.
FIG. 14 is a schematic configuration diagram of an image input apparatus that performs omnidirectional imaging using a plurality of wide-angle lenses.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 11 ... Optical fiber 12a, 12b, 12c ... Multi-core structure 13 ... Objective lens 14 ... Cylindrical body 15a, 15b, 15c ... Image input part 16 ... Fiber bundle part 17a, 17b, 17c ... Cylindrical hole DESCRIPTION OF SYMBOLS 20 ... Omnidirectional image capture part 21 ... Microphone 22 ... Support | pillar 23 ... Optical means 24 ... Image sensor 41 ... Signal processing part 42 ... Device interface part 43 ... Recording device 44 ... External interface part 45 ... Multi microphone part 46 ... Power supply part

Claims (3)

A multi-core structure formed by bundling a number of optical fiber strands;
An image input unit for capturing an image formed at one end of the multi-core structure;
Means for bundling a plurality of the multi-core structures;
Means for reading an image captured from the image input unit provided at each end of the multi-core structure bundled in plural by this means to the other end by the multi-core structure and reading with an image sensor via optical means When,
Equipped with,
The image input device, wherein the image input unit is arranged radially on each surface of an image capturing unit made of a regular polyhedron . The image input device according to claim 1, wherein the image input unit disposed in the image capturing unit made of the regular polyhedron is configured by bundling optical fiber strands into regular triangles . 2. The image input device according to claim 1 , wherein the other end of the multi-core structure matches the positional relationship with the image input units arranged radially .

Images