JP3934345B2 - Imaging device - Google Patents

Description

で表される。これは、図中、光線Ｏｑの点ｑでの垂線ｑｂと、ミラーＭの点ｑでの接線ｑａとのなす角∠ａｑｂの正接にあたる。
∠ａｑｂは時計回りに９０度回転すれば ∠Ｏｑｑ’に等しい。ｑｑ’はミラーＭの法線なので、∠Ｏｑｑ’は∠ｑ’ｑｒに等しい。また、光線Ｏｑとｚ軸との角度をφとすると、ｚ軸に平行なｑｃと光線との角度∠Ｏｑｃもまたφである。よって、
∠ａｑｂ＝（φ＋θ）／２ ・・・（３）
また図３から、
ｕ＝ｆtan(φ） ・・・（４）
なので、（３）式に（１）、（４）式を代入して、
∠ａｑｂ＝｛φ＋αｆtan(φ)｝／２ ・・・（５）
であれば、このミラーＭは（１）式の条件を満たすことになる。このような形状は（２）式から、方程式
【００２５】
【数２】

の解として得られる。この方程式の解は、
【００２６】
【数３】

と表すことができる。ただしＣは任意の定数である。
図５に、（７）式を 数値的に計算した結果のミラーＭの形状を示す（ただしｘ＞０の部分のみ示す。ｘ＜０の部分はｚ軸に関して対称とする）。 なお、ここでは、Ｃ＝１、αｆ= ６ とした。このミラーＭを水平方向に±１５度の視野角を持つカメラと組み合わせれば、６tan(１５°）×３６０／２π≒９２°なので、約１８４度の範囲の情景を撮影することができる。
双曲面ドームミラーを利用する従来技術特開平０９−５０５４４７号では、「高度利得」という用語で説明されるように、図３のθがφに比例する構成が開示されている。これに対し本発明では、角度θが、角度φではなく距離ｕと比例することを特徴としている。
ここで、通常のビデオカメラは、縦横比３：４の撮影面を持つ。その場合、上記のように水平方向の視野角が±１５度であれば、垂直方向の視野角は
tan^-1(３／４tan(１５°)) ≒１１．４°
となる。
【００２７】
つまり、通常のビデオカメラに図５の曲面ミラーを組み合わせると、縦横比３：４の画面中に、垂直方向±１１．４度、水平方向±９２度の範囲の画像が写ることになり、この画像は横方向にかなり圧縮されていることになる。この画像を横方向に ９２／１１．４×３／４≒６倍に引き延ばして表示すれば、縦横比がほぼ１１．４：９２となり、縦横方向で視野角と画像上の長さとが一致することになり、自然な状態の画像を再現できる。
デジタル画像データに対するこのような横方向の伸長処理は、例えば周知の一次元補間によって簡単に実現し得る。すなわち、注目画素(Ｐ_ｉ)とその直前の画素(Ｐ_ｉ−１)の値を保持しながら、入力画素１画素毎にｊ＝０〜５の計６通りについて下記式（８）によって計算する。
【００２８】
【数４】

その結果、６倍に伸長された画像の画素値Ｑが求められる。上記（８）式の一次元伸長処理は演算量が少ないので、通常のパーソナルコンピュータで、ソフトウエアによって動画データをリアルタイムに処理することが可能である。
図１（ａ）に 本発明の第１の実施の形態のシステム構成を示す。
【００２９】
図示の如く、この撮像装置１において、１１は図５に示す形状を持つ曲面ミラーＭを示す。又、１２は水平方向±１５度の視野角を持ち、平面状のＣＣＤ撮像素子を持つビデオカメラを示す。図１（ｂ）は、ミラー１１とカメラ１２との配置を示す斜視図である。
【００３０】
カメラ１２は、その光学中心が図３のＯに一致するよう配置される。このカメラ１２では上述のように、カメラの周囲、水平視野角±９２度、垂直視野角±１１．４度の範囲の画像が撮影され、これがビデオデータとして出力される。
【００３１】
カメラ１２から出力されたビデオデータは、Ａ／Ｄ変換器１３に入力され、デジタルデータに変換された後、計算機１５に接続された圧縮処理器１４で周知のＭＰＥＧ圧縮アルゴリズムにしたがって圧縮処理される。圧縮されたデータはハードディスク１６に記録される。
【００３２】
このようにして記録されたＭＰＥＧ圧縮データは、計算機１５上で動作するＭＰＥＧデコーダソフトウエアにより、周知のＭＰＥＧ復元アルゴリズムを使用してフレーム毎の画像データに復元される。復元された画像データは、上記（８）式に従って水平方向に６倍に拡大（伸長）された後、計算機１５のフレームメモリに転送され計算機１５の画面に表示される。
図１では動画データの記録と再生表示を一台の計算機１５で処理したが、それぞれの機能毎にネットワークで接続された別々の計算機を利用することも可能である。また、データをハードディスク１６に記録せず、入力された動画データを直接再生することも可能である。
上記第１の実施の形態では、記録された動画データを水平方向に６倍に拡大して表示した。しかしながら、元の動画データが縦横方向に同じ解像度を持つとすれば、このようにして表示される画像は横方向の解像度（帯域）が縦方向に比べて低く、横方向だけにぼけたものになってしまう。
【００３３】
次に説明する本発明の第２の実施の形態では、このような不自然さを解消するため、横方向に拡大する代わりに、横方向の拡大率に相当する機能を有するローパスフィルタを縦方向にかけ、縦方向に縮小する。これにより再生時の縦横の解像度を揃えることが可能となる。
図６は本発明の第２の実施の形態の撮像装置１０１のブロック図である。
【００３４】
上述の第１の実施の形態のものと同じカメラ１２、ミラー１１を使用し、Ａ／Ｄ変換後、ＭＰＥＧ圧縮の前に縮小処理手段１７で画像データを縦方向に縮小する。縮小率はカメラ１２、ミラー１１の特性によって決定されるため、この場合には、１／６となる。
【００３５】
Ａ／Ｄ変換後の画像データのサイズを６４０×４８０画素とすると、縮小後は６４０×８０画素のサイズの画像となる。ここでフィルタ処理と縮小処理は、縦方向６画素の平均値をひとつの画素の値とすることで同時に実施するものとする。
この場合、画像を表示する際には上述の第１の実施の形態の場合のような横方向の拡大処理は不要であり、ＭＰＥＧ復元処理された画像データをそのまま表示すれば良い。
次に本発明の第３の実施の形態について説明する。
【００３６】
上述の第１、第２の実施の形態では、（曲面ミラー１１を除く）カメラ１２の解像度を縦横同一としたため、ミラー１１を加えた構成で撮影される際の解像度が縦横で異なってしまう。そこで第３の実施の形態では、カメラ１２自体の解像度をミラー１１の特性に応じて縦横で変えることでこの問題を低減する。
この第３の実施の形態では、撮像装置は長方形の受光セルの配列を持つ撮像素子を持ち、横方向に解像度の高いカメラを使用する。
本実施の形態のブロック図は第１の実施の形態と同様(図１)である。
【００３７】
又、ミラー１１とカメラ１２の配置は上記第１，２の実施の形態のものと同じとする。
【００３８】
これまで説明した実施の形態との主な違いはカメラ１２の撮像素子にあり、この構成を図７に示す。
【００３９】
この撮像素子は、縦横比が６：１なる受光素子を縦に８０画素、横に６４０画素配列したモノクロ撮像素子である。これは、第２の実施の形態の撮像素子（６４０×４８０画素）と比べ、縦方向に６倍長い受光素子が、１／６の個数並ぶことになる。よってこの撮像素子によって得られる画像データは、各受光素子で得られるデータを1画素とすれば、ちょうど第２の実施の形態で縦方向に縮小処理された画像データと同等になる。
各受光素子で得られる画素データをそのままＡ／Ｄ変換し、縦横比１：８の状態の画像データとしてＭＰＥＧ圧縮処理を行い、記録する。この場合も第２の実施の形態と同様に、記録されたＭＰＥＧ圧縮データを復元処理し、そのまま表示することが可能である。
この第３の実施の形態では説明の簡略化のためにモノクロ撮像素子を使用した例としたが、色分解フィルタなどを利用し、カラーで撮影することも同様に実施できることは言うまでもない。
次に本発明の第４の実施の形態について説明する。
【００４０】
通常の撮像素子は正方形に近い受光素子を持つため、上述の第３の実施の形態のような縦横６：１の比率を持つ受光素子よりなる撮像素子を使用するためには、それらを特別に作成しなければならない。
一方、一枚の撮像素子上に色分解フィルタを配置してカラー撮像素子とする場合、フィルタの配置により解像度に方向性を持たせることが出来る。本実施の形態では、このように色分解フィルタ配置を利用して解像度を適正化するものである。
この第４の実施の形態は第２の実施の形態と同様のブロック構成（図６）を有する。主な相異はその撮像素子にあり、その構成を図８に示す。
図８に示す第４の実施の形態で使用される撮像素子では、縦方向にはＲＧＢＧという順番の繰り返しとされ、横方向には同じフィルタが連続する構成とされ、配列を構成する各受光素子は正方形状である。
【００４１】
ＲＧＢのうちＧフィルタが人間の視感輝度特性に近い分光透過率を持ち、解像度にもっとも寄与する。簡単のため、Ｇフィルタを持つ受光素子だけを抜き出して考えると、横方向には縦方向に比べて２倍の密度で受光素子が並んでいることになる。したがって、このような色分解フィルタ配列を持つ撮像素子では、縦方向に比べて横方向で２倍近い解像度を持つ画像が得られると考えられる。
そのためこの撮像素子で得られる画像は、水平方向に2倍に拡大した状態で初めて縦横の解像度が同等になる。曲面ミラ１１ーとカメラ１２を上述の第１の実施の形態と同じものを使用すると、撮影された画像は表示前に横方向に６倍拡大する必要があるが、本実施の形態では横に２倍拡大した状態で縦横同等となるので、そこから縦横同じ比率で拡縮すればよいことになる。すなわち縦に１／３倍、横に２倍に変倍することによって、縦横の解像度が同等な状態で画像全体の縦横比を横に６倍することができる。
受光素子の総数が上述の第３の実施例のものと同じならば、本実施の形態で得られる画像は第３の実施例と比べ縦横 ２ 倍の画素数となる。よってより有効に撮像データを利用することが出来ることになる。
次に本発明の第５の実施の形態について説明する。
【００４２】
これまで説明してきた実施の形態では、図１（ｂ）に示すように、曲面ミラー１１の正面にカメラ１２を配置した。そのため、画面の中央にはカメラ１２自体が写り、それによってその方向の対象物を隠されてしまう。本発明の第５の実施の形態では、ハーフミラーを使用することによって、カメラ１２自体が写らないようにするものである。
図９は本発明の第５の実施の形態におけるミラー１１とカメラ１２の構成を示す側面図である。上述の第１の実施の形態との違いは、曲面ミラー１１とカメラ１２との間にハーフミラー１８を配置したことにある。
【００４３】
図中上向きに設置されたカメラ１２からは、ハーフミラー１８を通してミラー１１が見える。このハーフミラー１８のために、光学的な配置は（破線で示すように）第１の実施の形態と同等なものになる。
次に本発明の第６の実施の形態について説明する。
【００４４】
本実施の形態も上述の第５の実施の形態同様、カメラ自体の写り込みを避ける方法を実現したものである。ここではミラー１１の形状を変更し、カメラ１２が存在する領域が写らないようにするものである。
第１の実施の形態では、上記（１）式で、θと u が比例関係を有するように構成したが、これに
θ＝αｕ＋β ・・・（９）
の如くオフセットβを加えると、上記（７）式で表現されるミラー１１の形状は、
【００４５】
【数５】

となる。この場合、φ＝０（ｕ＝０）で曲面Ｍの傾きは０ではなく、図１０に示すように中央に角を有する形状となる。したがって図の光学中心Ｏからほぼ真上に向かう光は、ミラーＭの角のどちら側に当たるかに応じて左右に振れ、図に示すように死角が生じることになる。したがってこの領域にカメラ１２を配置すれば、カメラ１２自体を写すことを避け、その分視野の限界を広くとることができる。
【００４６】
例えば、上述の第１の実施の形態と同様にカメラ１２の視野角を±１５度、αｆ＝６として、β＝３度とすれば、カメラ１２に写る視野は全体に３度ずれて３度〜９５度（−３度〜−９５度)となり、中央の６度分を除く１９０度の範囲となる。
なお、本発明の実施の形態は上述のものに限られず、本発明の基本概念の基づき様々な形態が考えられることは言うまでもない。
【発明の効果】
請求項１に記載の発明によれば、図２に示す円柱面撮像系と同様の画像を得ることが出来るため、複雑な修正処理無しに目視のための自然な状態の画像を再現することが可能であり、更に位置による解像度の変動の無い、広い視野角の画像を撮影することが可能となる。
又、１次元的な曲面ミラーと通常の光学系の組み合わせによるため、単純な構成で上記所望の機能を実現することが可能である。
【００４７】
又、曲面ミラーによる画像の異方性を補正し、自然な形状を有する画像を生成することができる。
又、請求項２に記載の発明によれば、曲面ミラーによる画像の異方性を補正し、自然な解像度特性を持つ画像を生成することができる。
又、請求項３，４，５に記載の発明によれば、曲面ミラーによる画像の異方性を補正し、自然な形状と解像度特性を持つ画像を生成することができる。
【００４８】
又、請求項６，７に記載の発明によれば、視野内にカメラ自体が入らないので、視野を有効に利用することができる。
【図面の簡単な説明】
【図１】本発明の第１の実施の形態の概略構成を示すブロック図及びミラーとカメラの配置を示す斜視図である。
【図２】円柱撮像面の機能を説明するための斜視図及び平面図である。
【図３】本発明の基本原理を説明するための図である。
【図４】図３の点ｑ近辺を拡大して示す図である。
【図５】図３，４を使用して説明した本発明の基本原理にしたがって計算されたミラーの形状の一例を示す図である。
【図６】本発明の第２の実施の形態の概略構成を示すブロック図である。
【図７】本発明の第３の実施の形態に使用されるカメラの撮像素子の構成を説明するための図である。
【図８】本発明の第４の実施の形態に使用されるカメラの撮像素子の構成を説明するための図である。
【図９】本発明の第５の実施の形態におけるミラー、カメラ及びハーフミラーの配置を示す平面図である。
【図１０】本発明の第６の実施の形態に使用されるミラーの形状を示す横断面図である。
【符号の説明】
１、１０１ 撮像装置
１１ ミラー(曲面鏡）
１２ カメラ（撮像素子を含む）
１３ Ａ／Ｄ変換器
１４ 圧縮処理器
１５ 計算機
１６ ハードディスク
１７ 縮小処理手段
１８ ハーフミラー[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an imaging apparatus, and more particularly to an imaging apparatus used for a moving image recording / reproducing apparatus such as a video recording system.
[0002]
[Prior art]
As such an imaging apparatus, for example, “omnidirectional visual system” described in Japanese Patent No. 029339087, “panoramic image apparatus” described in Japanese Patent No. 03054146, and Japanese Patent Application Laid-Open No. 09-505447 Panoramic monitoring system ".
[0003]
These all relate to a 360-degree panoramic camera using a hyperbolic dome-shaped mirror. In these configurations, a scene of 360 degrees (hemisphere) can be photographed by photographing an image reflected by a dome-shaped mirror. If such a camera is installed on the ceiling or the like, the entire room can be observed, which is suitable for monitoring applications.
In the above configuration, the captured image is distorted in a circumferential shape. However, the “panoramic image device” described in Japanese Patent No. 03054146 captures such an image with a normal camera. Includes means to correct as if
[Problems to be solved by the invention]
However, the data processing for converting a circumferentially distorted image into a flat shape requires a large amount of calculation, and a high-speed computer is required to process a moving image in real time, which increases the scale of the apparatus. At the same time, there is a problem that the apparatus becomes expensive.
Furthermore, since the scaling factor fluctuates depending on the position depending on such conversion, there is a problem that the substantial resolution of the converted image varies depending on the position.
In addition, in applications other than monitoring, for example, when it is desired to take a panoramic picture of a landscape, it is not always necessary to take a 360-degree all-around scene.
One ideal configuration for capturing a panoramic image is a perspective transformation imaging system using a cylindrical imaging surface as shown in FIG. A plan view of this configuration is shown in FIG.
In this case, the image of the imaging target existing at the point a (the component in the direction perpendicular to the paper surface is arbitrary) is a cylindrical imaging surface (thick curve) centered on O toward the optical center O. ) To point b.
In imaging with such an imaging system, there is no particular correction (by simply developing a cylindrical imaging surface as it is on a plane), locally compared to imaging with a normal planar imaging surface. An image with a wide field of view can be obtained without inferiority.
In this case, there is no problem of variation in resolution depending on the position. However, since it is difficult to arrange photodiode elements and the like in a curved shape, it is difficult to realize such an ideal imaging system as a product as it is.
[Means for Solving the Problems]
An object of the present invention is to solve the above-described problems and to provide an image pickup apparatus that can easily realize an image pickup system having a function equivalent to that of the configuration shown in FIG.
[0004]
The feature of the imaging system using the cylindrical surface as described above is that, as shown in FIG. 2B, the angle of the object projected on the reference horizontal plane is θ, and the position on the imaging surface on which the object is projected. When x is x, θ is proportional to x.
In order to realize such a configuration, the invention according to claim 1 includes an optical system on which a light beam from an object is incident and an imaging element on which the light beam is imaged.
The incident angle of a light beam with respect to the optical system when projected onto a predetermined plane has a linear relationship with one coordinate value representing the position on the image sensor surface where the light beam is imaged.
[0005]
Also, before Symbol optical system, the more the isotropic optical system about the optical axis, a curved mirror cylindrical surface shape to the base curve.
[0007]
Further, the data obtained by the image sensor includes means for performing processing having different characteristics depending on the direction, and the processing having different characteristics depending on the direction includes scaling processing having different magnification depending on the direction.
[0008]
In the invention according to claim 2 , the process having different characteristics depending on the direction further includes a smoothing process.
[0009]
According to a third aspect of the present invention, the image sensor generates image data having different resolution characteristics depending on directions.
[0010]
According to a fourth aspect of the present invention, the image sensor has a configuration in which light receiving elements having different sizes depending on directions are arranged.
[0011]
According to a fifth aspect of the present invention, the image sensor has a striped color separation filter.
[0012]
In the invention described in claim 6 , a half mirror is provided between the curved mirror of the optical system and the image sensor.
[0013]
In the invention of claim 7 , the curved mirror of the optical system comprises a mirror having an angle that generates a blind spot,
The imaging device is arranged in the blind spot.
[0014]
Therefore, by combining a planar imaging element that is easy to manufacture and an optical system in which the position on the planar imaging surface corresponding to the angle of the object is proportional, the same function as the configuration shown in FIG. 2 is achieved. It is possible to realize an imaging system having a natural shape for visual observation without complicated correction processing, and it is possible to obtain image data with a wide viewing angle without any change in resolution depending on the position. .
[0015]
Specifically, it is possible to obtain the desired characteristic with a simple structure of adding the one-dimensional curved mirror in the optical system normal.
Further, it is possible to obtain subjected to a treatment of different characteristics depending on the direction relative to the deformed captured data to one direction, the aspect of the shape with the image data equally aligned resolution.
Further, it is possible to reproduce a natural shape images Taking the image vertically and horizontally.
In the invention according to claim 2 , it is possible to make the resolution of the captured image data equal in length and width.
According to the third , fourth, and fifth aspects of the invention, it is possible to obtain image data having the same horizontal and vertical shapes and resolutions.
[0016]
In the invention according to claim 6 , it is possible to obtain an image without concealment by a camera (including an image sensor).
According to the seventh aspect of the present invention, it is possible to photograph a wide range of scenes except for an area hidden by a camera (including an image sensor).
DETAILED DESCRIPTION OF THE INVENTION
First, the basic principle of the present invention will be described.
[0017]
The present invention is an imaging apparatus having a configuration in which θ and x are proportional to each other when θ is an angle from a certain reference on the horizontal plane of the imaging target and x is a position on the imaging surface. This is realized by combining a planar imaging device and a curved mirror.
The operation of the present invention will be described below with reference to FIG.
[0018]
Note that the camera used here is provided with a normal imaging optical system, but for the sake of simplicity of explanation, the camera will be described here as a pinhole model.
[0019]
The figure shows the entire optical system as viewed from above (y-axis).
[0020]
In the following description, for the sake of simplicity, it is assumed that the camera has a wide viewing angle in the horizontal direction. This is generally desirable from human visual characteristics, but it should be understood that other directions can be magnified by rotating the entire device.
In the figure, O represents the optical center, thick line M represents a mirror, and I represents an imaging surface.
[0021]
The mirror M is assumed to have a straight cross section in the y-axis (perpendicular to the paper surface) direction and bends only in the xz direction. The imaging surface I is orthogonal to the z-axis at a distance f from the optical center O.
[0022]
On the imaging surface, the ray reaching the point p at a distance u from the z axis is traced in reverse. That is, if the line connecting the point p and the optical center O is extended and the point intersecting the mirror M is q, the light beam is reflected symmetrically with respect to the normal line qq ′ of the mirror M at the point q and reaches the point r. Here, the angle between the ray qr and the z-axis is θ.
Here, assuming that the object to be imaged is sufficiently far away and ignoring the substantial movement of the optical center, the position u on the imaging surface and the angle θ of the light beam reaching there
θ = αu (1)
Then, an image equivalent to that of the configuration of FIG. 2 is obtained.
The shape of the mirror M for realizing the characteristic of the formula (1) will be described with reference to FIG. This figure is an enlarged view of the vicinity of the point q in FIG.
Here, the shape (xz cross-sectional shape) of the mirror M on the xz plane is expressed as r (φ) in polar coordinate display. Where φ is the angle from the z-axis, and r (φ) is the distance from the optical center O.
[0023]
With this definition, the radial gradient of the surface of the mirror M is
[0024]
[Expression 1]

It is represented by This corresponds to the tangent of the angle ∠aqb formed by the perpendicular qb of the ray Oq at the point q and the tangent qa of the mirror M at the point q.
∠aqb is equal to ∠Oqq ′ if rotated 90 degrees clockwise. Since qq ′ is the normal of the mirror M, ∠Oqq ′ is equal to ∠q′qr. If the angle between the light beam Oq and the z-axis is φ, the angle ∠Oqc between qc and the light beam parallel to the z-axis is also φ. Therefore,
∠aqb = (φ + θ) / 2 (3)
Also from FIG.
u = ftan (φ) (4)
So, substituting (1) and (4) into (3),
∠aqb = {φ + αftan (φ)} / 2 (5)
If so, the mirror M satisfies the condition of the expression (1). Such a shape is obtained from the equation (2) and the equation:
[Expression 2]

Is obtained as a solution of The solution of this equation is
[0026]
[Equation 3]

It can be expressed as. However, C is an arbitrary constant.
FIG. 5 shows the shape of the mirror M as a result of numerically calculating the expression (7) (however, only the part of x> 0 is shown. The part of x <0 is symmetric with respect to the z axis). Here, C = 1 and αf = 6. When this mirror M is combined with a camera having a viewing angle of ± 15 degrees in the horizontal direction, since 6 tan (15 °) × 360 / 2π≈92 °, a scene in a range of about 184 degrees can be photographed.
Japanese Patent Application Laid-Open No. 09-505447 using a hyperboloidal dome mirror discloses a configuration in which θ in FIG. 3 is proportional to φ, as described in terms of “high gain”. On the other hand, the present invention is characterized in that the angle θ is proportional to the distance u instead of the angle φ.
Here, a normal video camera has a shooting surface with an aspect ratio of 3: 4. In that case, if the horizontal viewing angle is ± 15 degrees as described above, the vertical viewing angle is
tan ^-1 (3/4 tan (15 °)) ≒ 11.4 °
It becomes.
[0027]
In other words, when the curved mirror shown in FIG. 5 is combined with a normal video camera, an image in the range of ± 11.4 degrees in the vertical direction and ± 92 degrees in the horizontal direction is captured in the screen with an aspect ratio of 3: 4. The image is considerably compressed in the horizontal direction. If this image is displayed by stretching it 92 / 11.4 × 3 / 4≈6 times in the horizontal direction, the aspect ratio becomes approximately 11.4: 92, and the viewing angle and the length on the image match in the vertical and horizontal directions. As a result, a natural image can be reproduced.
Such a lateral expansion process for digital image data can be easily realized by, for example, well-known one-dimensional interpolation. That is, while maintaining the values of the pixel of interest (P _i ) and the immediately preceding pixel (P _i-1 ), a total of 6 patterns of j = 0 to 5 are calculated for each input pixel by the following equation (8). .
[0028]
[Expression 4]

As a result, the pixel value Q of the image expanded by 6 times is obtained. Since the one-dimensional decompression processing of the above equation (8) has a small amount of calculation, it is possible to process moving image data in real time by software with a normal personal computer.
FIG. 1A shows the system configuration of the first embodiment of the present invention.
[0029]
As shown in the figure, in this imaging apparatus 1, reference numeral 11 denotes a curved mirror M having the shape shown in FIG. Reference numeral 12 denotes a video camera having a viewing angle of ± 15 degrees in the horizontal direction and having a planar CCD image sensor. FIG. 1B is a perspective view showing the arrangement of the mirror 11 and the camera 12.
[0030]
The camera 12 is arranged such that its optical center coincides with O in FIG. As described above, the camera 12 captures an image in the range around the camera, the horizontal viewing angle ± 92 degrees, and the vertical viewing angle ± 11.4 degrees, and outputs this as video data.
[0031]
Video data output from the camera 12 is input to an A / D converter 13 and converted into digital data, and then compressed by a compression processor 14 connected to a computer 15 in accordance with a well-known MPEG compression algorithm. . The compressed data is recorded on the hard disk 16.
[0032]
The MPEG compressed data recorded in this way is restored to image data for each frame using a known MPEG restoration algorithm by MPEG decoder software operating on the computer 15. The restored image data is expanded (expanded) six times in the horizontal direction according to the above equation (8), then transferred to the frame memory of the computer 15 and displayed on the screen of the computer 15.
In FIG. 1, recording and playback display of moving image data is processed by one computer 15, but it is also possible to use separate computers connected by a network for each function. It is also possible to directly play the input moving image data without recording the data on the hard disk 16.
In the first embodiment, the recorded moving image data is enlarged and displayed 6 times in the horizontal direction. However, if the original video data has the same vertical and horizontal resolution, the image displayed in this way has a lower horizontal resolution (bandwidth) than the vertical direction, and is blurred only in the horizontal direction. turn into.
[0033]
In a second embodiment of the present invention to be described next, in order to eliminate such unnaturalness, a low-pass filter having a function corresponding to a horizontal enlargement factor is used in the vertical direction instead of expanding in the horizontal direction. To reduce in the vertical direction. This makes it possible to align the vertical and horizontal resolutions during playback.
FIG. 6 is a block diagram of the imaging apparatus 101 according to the second embodiment of the present invention.
[0034]
The same camera 12 and mirror 11 as those in the first embodiment are used, and after A / D conversion, the image data is reduced in the vertical direction by the reduction processing means 17 before MPEG compression. Since the reduction ratio is determined by the characteristics of the camera 12 and the mirror 11, it is 1/6 in this case.
[0035]
If the size of the image data after A / D conversion is 640 × 480 pixels, the image has a size of 640 × 80 pixels after reduction. Here, it is assumed that the filtering process and the reduction process are simultaneously performed by setting the average value of the six pixels in the vertical direction as the value of one pixel.
In this case, when the image is displayed, the horizontal enlargement process as in the case of the first embodiment described above is unnecessary, and the image data subjected to the MPEG restoration process may be displayed as it is.
Next, a third embodiment of the present invention will be described.
[0036]
In the first and second embodiments described above, the resolution of the camera 12 (excluding the curved mirror 11) is the same in both the vertical and horizontal directions. Therefore, the resolution when shooting with the configuration including the mirror 11 is different in the vertical and horizontal directions. Therefore, in the third embodiment, this problem is reduced by changing the resolution of the camera 12 itself vertically and horizontally in accordance with the characteristics of the mirror 11.
In the third embodiment, the imaging apparatus has an imaging element having an array of rectangular light receiving cells and uses a camera with high resolution in the lateral direction.
The block diagram of this embodiment is the same as that of the first embodiment (FIG. 1).
[0037]
The arrangement of the mirror 11 and the camera 12 is the same as that in the first and second embodiments.
[0038]
The main difference from the embodiment described so far lies in the image sensor of the camera 12, and this configuration is shown in FIG.
[0039]
This image pickup device is a monochrome image pickup device in which light receiving elements having an aspect ratio of 6: 1 are arranged vertically in 80 pixels and horizontally in 640 pixels. This means that the number of light receiving elements that are six times longer in the vertical direction than the imaging element (640 × 480 pixels) of the second embodiment is arranged in 1/6. Therefore, the image data obtained by this image sensor is equivalent to the image data reduced in the vertical direction in the second embodiment if the data obtained by each light receiving element is one pixel.
The pixel data obtained by each light receiving element is A / D converted as it is, and is subjected to MPEG compression processing and recorded as image data having an aspect ratio of 1: 8. In this case, similarly to the second embodiment, the recorded MPEG compressed data can be restored and displayed as it is.
In the third embodiment, for the sake of simplicity, an example in which a monochrome imaging device is used has been described. Needless to say, it is also possible to perform color imaging using a color separation filter or the like.
Next, a fourth embodiment of the present invention will be described.
[0040]
Since an ordinary image sensor has a light receiving element close to a square, in order to use an image sensor made up of light receiving elements having a ratio of 6: 1 in the vertical direction as in the third embodiment, they are specially used. Must be created.
On the other hand, when a color separation filter is arranged on a single image pickup device to form a color image pickup device, the direction of the resolution can be given by the arrangement of the filters. In the present embodiment, the resolution is optimized by using the color separation filter arrangement as described above.
The fourth embodiment has the same block configuration (FIG. 6) as that of the second embodiment. The main difference is in the image sensor, and its configuration is shown in FIG.
In the imaging device used in the fourth embodiment shown in FIG. 8, the RGBG is repeated in the order of the vertical direction, and the same filter is continuous in the horizontal direction. Is square.
[0041]
Among RGB, the G filter has a spectral transmittance close to that of human visual luminance and contributes most to the resolution. For simplicity, if only light receiving elements having a G filter are extracted, the light receiving elements are arranged in the horizontal direction at a density twice that of the vertical direction. Therefore, it is considered that an image having such a color separation filter array can obtain an image having a resolution close to twice in the horizontal direction compared to the vertical direction.
For this reason, the images obtained with this image sensor have the same vertical and horizontal resolutions when they are enlarged twice in the horizontal direction. When the same curved mirror 11- and camera 12 as those in the first embodiment are used, the photographed image needs to be magnified 6 times in the horizontal direction before display. Since it becomes the same in the vertical and horizontal directions in the state of being doubled, it can be enlarged and reduced at the same ratio in the vertical and horizontal directions. That is, by changing the magnification by 1/3 times vertically and 2 times horizontally, the aspect ratio of the entire image can be increased by 6 times horizontally with the same vertical and horizontal resolutions.
If the total number of light receiving elements is the same as that of the third embodiment, the image obtained in this embodiment has twice as many pixels as the third embodiment. Therefore, imaging data can be used more effectively.
Next, a fifth embodiment of the present invention will be described.
[0042]
In the embodiment described so far, the camera 12 is disposed in front of the curved mirror 11 as shown in FIG. Therefore, the camera 12 itself appears in the center of the screen, thereby hiding the object in that direction. In the fifth embodiment of the present invention, the camera 12 itself is prevented from being photographed by using a half mirror.
FIG. 9 is a side view showing the configuration of the mirror 11 and the camera 12 in the fifth embodiment of the present invention. The difference from the first embodiment described above is that a half mirror 18 is disposed between the curved mirror 11 and the camera 12.
[0043]
From the camera 12 installed upward in the figure, the mirror 11 can be seen through the half mirror 18. Because of this half mirror 18, the optical arrangement is equivalent to that of the first embodiment (as indicated by the broken line).
Next, a sixth embodiment of the present invention will be described.
[0044]
This embodiment also realizes a method for avoiding the reflection of the camera itself, as in the fifth embodiment. Here, the shape of the mirror 11 is changed so that the area where the camera 12 exists is not captured.
In the first embodiment, θ and u have a proportional relationship in the above equation (1), but θ = αu + β (9)
When the offset β is added as shown below, the shape of the mirror 11 expressed by the above equation (7) is
[0045]
[Equation 5]

It becomes. In this case, φ = 0 (u = 0) and the slope of the curved surface M is not 0, but has a shape having a corner at the center as shown in FIG. Accordingly, the light traveling almost directly from the optical center O in the figure swings left and right depending on which side of the corner of the mirror M hits, and a blind spot is generated as shown in the figure. Therefore, if the camera 12 is arranged in this region, the camera 12 itself can be avoided from being copied, and the field of view can be widened accordingly.
[0046]
For example, if the viewing angle of the camera 12 is ± 15 degrees, αf = 6, and β = 3 degrees as in the first embodiment, the field of view reflected on the camera 12 is shifted by 3 degrees and 3 degrees. It is ˜95 degrees (−3 degrees to −95 degrees), and is in a range of 190 degrees excluding the central 6 degrees.
Needless to say, embodiments of the present invention are not limited to those described above, and various forms are conceivable based on the basic concept of the present invention.
【The invention's effect】
According to the first aspect of the present invention, an image similar to that of the cylindrical surface imaging system shown in FIG. 2 can be obtained, so that an image in a natural state for viewing can be reproduced without complicated correction processing. In addition, it is possible to capture an image with a wide viewing angle without any change in resolution depending on the position.
In addition , since the one- dimensional curved mirror is combined with a normal optical system, the desired function can be realized with a simple configuration.
[0047]
Further, by correcting the anisotropy of the image by the song plane mirror, it is possible to generate an image having a natural shape.
According to the second aspect of the present invention, it is possible to correct the anisotropy of the image by the curved mirror and generate an image having natural resolution characteristics.
According to the third , fourth, and fifth aspects of the invention, the anisotropy of the image by the curved mirror can be corrected, and an image having a natural shape and resolution characteristics can be generated.
[0048]
According to the sixth and seventh aspects of the invention, since the camera itself does not enter the field of view, the field of view can be used effectively.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a schematic configuration of a first embodiment of the present invention and a perspective view showing an arrangement of a mirror and a camera.
FIG. 2 is a perspective view and a plan view for explaining the function of a cylindrical imaging surface.
FIG. 3 is a diagram for explaining a basic principle of the present invention.
4 is an enlarged view showing the vicinity of a point q in FIG.
FIG. 5 is a diagram showing an example of a mirror shape calculated according to the basic principle of the present invention described with reference to FIGS.
FIG. 6 is a block diagram showing a schematic configuration of a second embodiment of the present invention.
FIG. 7 is a diagram for explaining a configuration of an image sensor of a camera used in a third embodiment of the present invention.
FIG. 8 is a diagram for explaining a configuration of an image sensor of a camera used in a fourth embodiment of the present invention.
FIG. 9 is a plan view showing an arrangement of a mirror, a camera, and a half mirror according to a fifth embodiment of the present invention.
FIG. 10 is a cross-sectional view showing the shape of a mirror used in a sixth embodiment of the present invention.
[Explanation of symbols]
1, 101 Imaging device 11 Mirror (curved surface mirror)
12 Camera (including image sensor)
13 A / D converter 14 Compression processor 15 Computer 16 Hard disk 17 Reduction processing means 18 Half mirror

Claims (7)

It consists of an optical system on which light rays from an object are incident and an image sensor on which the light rays are imaged.
There is a linear relationship between the incident angle of a light beam with respect to the optical system when projected onto a predetermined plane and one coordinate value representing the position on the image sensor surface where the light beam is imaged,
The optical system is composed of an isotropic optical system around the optical axis and a columnar curved mirror having a curved bottom,
Furthermore, the data obtained by the imaging device comprises means for performing processing having different characteristics depending on the direction,
The imaging apparatus having a configuration in which the processing having different characteristics depending on the direction includes scaling processing with different magnification depending on the direction .   The imaging apparatus according to claim 1, wherein the process having different characteristics depending on the direction further includes a smoothing process. It consists of an optical system on which light rays from an object are incident and an image sensor on which the light rays are imaged.
There is a linear relationship between the incident angle of a light beam with respect to the optical system when projected onto a predetermined plane and one coordinate value representing the position on the image sensor surface where the light beam is imaged,
The imaging device is configured to generate image data having different resolution characteristics depending on directions.   The image pickup apparatus according to claim 3, wherein the image pickup element has a configuration in which light receiving elements having different sizes depending on directions are arranged.   The imaging device according to claim 3, wherein the imaging element includes a striped color separation filter. It consists of an optical system on which light rays from an object are incident and an image sensor on which the light rays are imaged.
There is a linear relationship between the incident angle of a light beam with respect to the optical system when projected onto a predetermined plane and one coordinate value representing the position on the image sensor surface where the light beam is imaged,
The optical system is composed of an isotropic optical system around the optical axis and a columnar curved mirror having a curved bottom,
An imaging apparatus having a configuration in which a half mirror is provided between the curved mirror of the optical system and an imaging element. The curved mirror of the optical system comprises a mirror having an angle that generates a blind spot,
The imaging device according to claim 1, wherein the imaging element is arranged in a blind spot.

Images