JP2004080459A

JP2004080459A - Imaging apparatus having camera-shake correcting function

Info

Publication number: JP2004080459A
Application number: JP2002238769A
Authority: JP
Inventors: Kunihisa Yamaguchi; 山口　邦久; Junichi Shinohara; 篠原　純一; Keiichiro Hirahara; 平原　圭一郎; Naoki Koshida; 越田　直紀; Hiroyuki Chiba; 千葉　浩幸
Original assignee: Ricoh Co Ltd
Current assignee: Ricoh Co Ltd
Priority date: 2002-08-20
Filing date: 2002-08-20
Publication date: 2004-03-11
Anticipated expiration: 2022-08-20
Also published as: JP4014970B2

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problems in a digital camera of the possibility of a shutter being released with delay, with respect to the sensitivity of imaging elements and this causing camera-shake, and of the burden on a CPU increases, when a shake-correcting function is built in, to continue shake correction during an exposure operation, resulting in a response delay, in various operations. <P>SOLUTION: The cause of camera-shake lies in a posture or a grip condition of the camera, and the extent of the shake differs in directions, such as horizontal direction or vertical direction, depending on the condition. A direction which must be regarded as important and a direction which may be treated with less importance are determined from the results of detection of a state of the camera or camera-shake, weighting of the time to be allocated to the operation of the CPU is changed with respect to the respective directions. By doing this, correction accuracy is improved as a whole, without increasing the load on the CPU. <P>COPYRIGHT: (C)2004,JPO

Description

【０００１】
【産業上の利用分野】
本発明は、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影装置に関する。詳しくは、これら撮影装置における振れ補正機能に関する。
【０００２】
【従来の技術】
デジタルスチルカメラは、撮像素子の感度の限界により、銀塩写真のような短いシャッタ時間を実現しにくく、露光中の手振れあるいは像振れと呼ばれる振れにより、撮影された画像に「像の流れ」のようなボケが生じやすい。
【０００３】
振れの検出方法としては、たとえば、特開２０００−６９３５３号公報に記載されているように、撮像素子が取り込んだ複数回分の画像データに、何らかの画像処理を施して振れの情報を得る方法や、撮像装置の角加速度を検出し振れ量を算出して用いるものなどがある。
【０００４】
振れの補正方法としては、たとえば、特開２００１−１９４７０１号公報に記載されているように、撮像素子から得られる画像信号の信号レベルを、可変利得増幅手段によって利得を増加させ、シャッタ秒時を振れ限界より短くする電気的補正方法がある。
あるいは、撮像装置のレンズ系の操作や、撮像素子の移動などの機械的補正手段も利用されている。そのような例としては特開平１１−１９６４２８号公報がある。
【０００５】
補正の精度を上げる方法として、特開平５−１０７６１９号公報に、レンズの焦点距離情報に基づいて振れ検出および補正の周期を決定し、レンズ焦点距離が長い場合には補正周期を短くして補正の効果をあげるものが示されている。
【０００６】
いずれの場合も、撮影装置のレンズ光軸に直交する平面内における二次元的な補正を行う場合が多い。ここで、以後の説明の便宜のため、撮影装置における上記二次元の方向の定義をしておく。
図１は、本発明の振れ補正機能付き撮影装置を実施する場合の１つの形態を説明するための図である。図１に示すように、撮影装置のレンズ光軸正面から見た場合の撮影装置の形状が、長方形に近い場合、その長手方向をＸ軸方向と名付ける。Ｘ軸方向と直交する方向をＹ軸方向と名付ける。一般にはＸ軸方向が水平方向になるとは限らない。形状が正方形に近くて長手方向が決められない場合は標準の使用位置における横方向をＸ軸方向とする。
【０００７】
撮影装置１において、図１に示すようにＸ軸方向が水平である場合、Ａで示す矢印はＸ軸方向が変化しないで鉛直方向に向いたＹ軸の方向が変化するのでＸ軸回りの変位と呼ぶ。同様に、Ｂで示す矢印はＹ軸回りの変位と呼ぶ。
Ｘ軸回りの変位による像の移動を補正するためには光学系の光軸、あるいは、撮像素子をＹ軸方向に移動させなければならない。したがって、Ｘ軸回りの変位によって生ずる振れ量の演算結果をΔＹとする。同様にＹ軸回りの変位によって生ずる振れ量の演算結果をΔＸとする。なお、説明の都合上、像面での像の振れ方向で、たとえば、Ｘ方向、あるいは、Ｙ方向と表現することがある。
【０００８】
図５に示す撮影装置１５では、横長方向に構えた場合のシャッタ１６と、縦長方向に構えた場合のシャッタ１７を有している。図５（ａ）に示すように撮影装置１５を横長方向に構えた場合、Ｘ軸回りの変位が鉛直方向の変位、Ｙ軸回りの変位が水平方向の変位となる。また、図５（ｂ）に示すように撮影装置１５を縦長方向に構えた場合、Ｘ軸回りの変位が水平方向の変位、Ｙ軸回りの変位が鉛直方向の変位となる。このように撮影装置１５のＸ軸回りの変位、Ｙ軸回りの変位は、撮影装置の構え方により、鉛直方向変位と水平方向変位のどちらにもなり得る。
【０００９】
【発明が解決しようとする課題】
通常、振れ補正を行う際、たとえばＸ軸回りの振れ量ΔＹを算出し、Ｘ軸回りの振れを補正、次に、Ｙ軸回りの振れ量ΔＸを算出し、Ｙ軸回りの振れを補正するということを一定周期で繰返す。この演算は一般に中央演算装置と呼ばれるＣＰＵで行っている。またＣＰＵではズーム、フォーカスなどの処理も行っており、撮影装置が振れ補正モードに指定されている場合は、ＣＰＵの負担が明らかに増えている。これがそれぞれの処理速度の低下の原因となり、振れ補正における追従性、応答性を低下させることにもつながる。振れ補正モードにあるときには常にこれらの処理を行っているため、消費電力も増えてしまう。
従来技術のように、補正の周期を短くすると、ＣＰＵの処理時間が多くなって、上記問題が顕著になる。逆に、単に補正の周期を長くして、ＣＰＵの負担を小さくしようとすれば、必然的に補正精度の低下になる。
【００１０】
【課題を解決するための手段】
これらの問題を解決するため、請求項１の発明では、撮影光学系と、この撮影光学系を通過した被写体からの光を受光する撮像手段と、前記撮影光学系の光軸に直交する２軸の回りの振れを検出する振れ検出手段と、前記振れ検出手段より検出される検出振れ量を受け入れて所定の処理を行うＣＰＵと、該ＣＰＵの処理結果に基づき、前記２軸の回りの振れ変位を補正する２つの振れ補正手段を有する撮影装置において、前記２軸の回りの振れ補正手段の補正のための処理に要するＣＰＵの演算割り当て時間に重み付けを行うことを特徴とする。
【００１１】
請求項２の発明では、請求項１に記載の撮影装置において、前記ＣＰＵが前記２つの振れ補正手段に与える補正量を演算するための演算式として通常の演算式と簡略演算式を用意し、一方の補正手段に対し通常の演算式を用い、他方の補正手段に対し簡略演算式を用いることでＣＰＵの演算割り当て時間の重み付けを行うことを特徴とする。
請求項３の発明では、請求項１に記載の撮影装置において、前記２つの振れ補正手段の補正周期を異ならせることでＣＰＵの演算割り当て時間の重み付けを行うことを特徴とする。
【００１２】
請求項４の発明では、請求項３に記載の撮影装置において、前記検出振れ量の大きい方の補正周期を短く、小さい方の補正周期を長くすることを特徴とする。
請求項５の発明では、請求項３または４に記載の撮影装置において、前記補正周期を異ならせた２軸の回りの前記補正量演算において、前記ＣＰＵが前記２つの振れ補正手段に与える補正量を演算するための演算式として通常の演算式と簡略演算式を用意し、一方の軸に関する演算に簡略演算式を用いることを特徴とする。
【００１３】
請求項６の発明では、請求項３、４または５に記載の撮影装置において、前記撮影装置の把持状態を検出する把持状態検出センサを前記撮影装置本体に有し、その検出結果に対応して、予め設定されている第１の補正周期を選定することを特徴とする。
請求項７の発明では、請求項３、４または５に記載の撮影装置において、前記ＣＰＵが撮影装置の姿勢を判定して、姿勢に対応して予め設定されている第１の補正周期を選定することを特徴とする。
【００１４】
請求項８の発明では、請求項２、３、４、または５に記載の撮影装置において、前記検出振れ量から以後の予測振れ量を算出する、振れ量予測演算手段を有することを特徴とする
請求項９の発明では、請求項６，または７に記載の撮影装置において、前記検出振れ量から以後の予測振れ量を算出する振れ量予測演算手段を有し、該予測振れ量に対応して予め設定されている第２の補正周期を選定することを特徴とする。
【００１５】
請求項１０の発明では、請求項９に記載の撮影装置において、前記第１の補正周期と、前記第２の補正周期の相互の、Ｘ、Ｙ軸対応の大小関係が一致しないとき、２軸の補正周期を等しくするか、または、前記第１の補正周期を用いて補正動作を行わせることを特徴とする。
請求項１１の発明では、請求項９に記載の撮影装置において、前記撮影光学系が情報表示機能、もしくは、音、光などによる警告機能を有し、前記第１の補正周期と、前記第２の補正周期の相互の、Ｘ、Ｙ軸対応のの大小関係が一致しないとき、その旨を前記情報表示機能により表示を行うか、または前記警告機能により撮影者に警告を行うことを特徴とする。
請求項１２の発明では、請求項８〜１０のいずれか１つに記載の撮影装置において、前記予測振れ量の、大きい方の補正周期を短く、小さい方の補正周期を長くすることを特徴とする。
【００１６】
請求項１３の発明では、請求項３に記載の撮影装置において、撮影装置の振れの状況を監視する振れ監視手段を前記撮影装置内に設け、該監視手段の検出結果で振れの変化を予測する振れ状況予測演算手段を設け、予測演算の結果で補正周期を所定時間ごとに分けて決めることを特徴とする
請求項１４の発明では、請求項１３に記載の撮影装置において、補正周期は振れの変化の絶対値もしくは振れの変化の比率に対応させて決めることを特徴とする。
【００１７】
請求項１５の発明では、請求項１４において、前記補正周期と振れの変化の絶対値もしくは振れの変化の比率との対応を撮影者が決めるための設定手段を有することを特徴とする振れ補正機能を有する撮影装置。
請求項１６の発明では、請求項３〜１５のいずれか１つに記載の撮影装置において、理論上振れが小さいと見込まれる軸方向、あるいは、検出振れ量もしくは予測振れ量が所定値より小さい軸方向は補正周期を無限大に設定し、実質的に補正動作を省略することを特徴とする。
請求項１７の発明では、請求項１６に記載の撮影装置において、補正動作を省略した軸方向も、所定周期で振れ検出を継続し、検出振れ量が所定量より大きくなったら補正動作を行うことを特徴とする
【００１８】
請求項１８の発明では、請求項３〜１７のいずれか１つに記載の撮影装置において、２軸の補正周期の組合せを複数用意して、どの組合せを採用するかは、撮影者が決められるような選択手段を有することを特徴とすることを特徴とする。
請求項１９の発明では、請求項１〜１８のいずれか１つに記載の撮影装置において、露光中は補正動作の回数を制限することを特徴とする。
請求項２０の発明では、請求項１〜１９のいずれか１つに記載の撮影装置において、補正の状況を表示することを特徴とする。
【００１９】
【目的】
撮影状況、たとえば、カメラの把持状態が片手であるか、両手であるかという把持状態の違い、あるいは、カメラを横長方向に構えているか、縦長方向に構えているかという姿勢の違いによって、Ｘ軸回りあるいはＹ軸回りにおける振れの大きさに差が出てくる。
【００２０】
極端な場合で、仮にＹ軸回りの振れが大きく、Ｘ軸回りの振れは無視できるほど小さければ、Ｙ軸まわりの振れだけを補正することによって、撮影された画像の振れは補正される。
【００２１】
あるいは、振れの大きい方の軸まわりに関しては通常通りの補正演算を行い、振れの小さい方の軸まわりに関しては、補正演算の演算式を簡略化してＣＰＵの負担を軽減することができる。
【００２２】
それほど極端でない場合でも、振れの大きい方の軸の補正頻度を、他方の軸の補正頻度より多くすること、すなわち補正周期を短くすることで、より適正な補正ができる。
【００２３】
本発明は、上記従来の問題点に鑑み、振れ補正手段による振れ補正を状況に応じて、特にＸ軸回りとＹ軸回りで個別に変化させることにより、処理を行っているＣＰＵの負担を軽減し、処理能力、速度の向上を図り、消費電力の無駄を押え、かつ、振れによるボケの少ない快適な画像を撮影することが可能な撮影装置を提供することを目的としている
【００２４】
あるいは、ＣＰＵの負担を増大させることなく、両軸回りの補正精度を最適化することを目的としている。
また、本発明は、露光前の振れ予測情報に基づいて、２軸回りの振れ補正の為の各々の補正周期を決定するものである。それにより、撮影装置のシステムの負荷を上げることなく、全体としての振れ補正の効果を上げることを目的としている。
【００２５】
ところで、振れの大きさの予測については、露光前の振れ検出手段により検出される振れ検出情報に基づき、露光後の振れを予測することが可能である。
よって、露光中の２軸回りの振れの大きさを各々予測し、振れの変化が大きいと予測された軸回りの補正周期を短くして補正精度を高め、また振れの変化が小さいと予測された軸回りの補正周期を長くして無駄な補正精度を抑える。
これにより、システムの負荷を上げることなく、振れの大きい軸回りの補正精度をあげ、また２軸回りの補正精度をほぼ均等にすることができる。
【００２６】
【実施例】
撮影装置１に関して基本的な振れ補正方法を実現するための機能的構成例を図２に示す。符号４、５は振れ検出センサ、符号６はＣＣＤ等の撮像素子を示す。振れ検出センサ４は、撮影装置のＸ軸回りの振れを検出し、振れ検出センサ５は、撮影装置のＹ軸回りの振れを検出する。これらの振れ検出センサ４、５と振れ検出センサ回路７により振れ検出手段８を構成する。振れ検出手段８の出力は演算手段９に送られる。演算手段９はマイクロプロセッサ等により構成され、振れ検出手段８からの出力に応じて、振れ補正量を算出する。この算出された振れ補正量に対応した信号が補正手段駆動回路１０に送られ、それにより補正手段駆動回路１０を通じた振れ補正手段１１、１２が駆動され、撮像素子６を振れを軽減させる方向へ変位させる。振れ補正手段１１はＸ軸回りの振れを補正する方向へ駆動、また振れ補正手段１２はＹ軸回りの振れを補正する方向へ駆動される。
【００２７】
人間の起す手振れは、撮影装置の構え方により変わってくるが、図３に示すようにシャッター押下方向である鉛直方向の振れが大きく、水平方向の振れは少ないのが一般的である。
この場合、振れ補正の繰り返し周期をＸ方向、Ｙ方向とも同じある一定周期で行うとすると、振れが大きい鉛直方向では、図３（ａ）に示すように１周期Δｔ当たりの変化量ΔＸ_ｐが大きいため、補正精度が低くなる。これに対して、振れが小さい水平方向では、図３（ｂ）に示すように単位時間Δｔ当たりの変化量ΔＸ_ｈが少ないため無駄に補正精度が高くなる現象が起こる。
【００２８】
この振れ補正の繰り返し周期は、短いほど補正精度を高くできる。しかし、演算手段９による補正演算は、振れ検出手段８の出力データから振れ補正量を単に算出するといった簡単なものではなく、振れ補正手段１１、１２の機構に起因する線形また非線形の特性を是正する演算の要求もあり、補正演算時間が増大する傾向にある。よって、補正周期を単に短縮することは撮影システムに対する負荷への影響が大きく、限度がある。
そこで、高い精度を必要とする軸回りと、その必要がない軸回りとの間で、ＣＰＵの演算時間の割り当てを変え、ＣＰＵの負荷の増大を招かずに所望の補正精度を得る。
【００２９】
この目的を達成するため、本発明では、Ｘ軸方向またはＹ軸方向の少なくとも一方の補正演算に簡略演算式を用いる。
通常、振れ補正量の演算において、補正手段の応答遅れを補うために振れ量の予測をするための式が用いられており、この式は、現在の振れ量のデータと直前の振れ量のデータの２点から次の点の振れ量を算出する比例式に、振れ速度の変化率すなわち加速度の項を表す微分項を加えている。そこで、高精度を要しないとみなされる軸回りに関しては、この式の微分項を省き、一次式のみで予測をするなど、簡略演算を行うことで、ＣＰＵの負担を軽減し、かつ、より適切な振れ補正を行うことが出来る。
【００３０】
さらに、上記目的を達成するため、本発明では、鉛直方向と水平方向の補正周期を一定周期とするのではなく、振れが大きい軸回りの補正周期を短くして補正精度を高め、さらに振れが小さい軸回りの補正周期を長くして無駄な補正精度を抑え、Ｘ軸回りの１周期当たりの変化量ΔＹと、Ｙ軸回りの１周期当たりの変化量ΔＸとのバランス（ΔＸ≒ΔＹ）をとる。
【００３１】
さらに、上記目的を達成するため、本発明は、露光前の振れ予測情報に基づいて、２軸回りの振れ補正方向各々の補正周期を決定するものである。それにより、撮影装置のシステムの負荷を上げることなく、全体としての手振れ補正の効果を上げるものである。
【００３２】
さらに、上記目的を達成するため、本発明では、Ｘ軸回りとＹ軸回りの補正周期を露光中全期間に渡って一定周期とするのではなく、露光期間を所定の単位時間で幾つかの区間に分割して、区間毎に振れの変化が大きい軸回りの補正周期を短くして補正精度を高め、さらに振れの変化が小さい軸回りの補正周期を長くして無駄な補正精度を抑え、振れの大きい軸回りの１周期当たりの変化量と、振れが小さい軸回りの１周期当たりの変化量とのバランスをとる。
【００３３】
この一連の振れ補正動作はＣＣＤ６への露光中に繰り返し行われる。繰り返しの周期は両軸回りとも同じにするのが一般的である。本発明では両軸のうち、補正の必要度合いの大きい方により多くのＣＰＵ処理時間を与えることにした。
その方法の１つとして、本発明の第１の実施形態では、補正周期は従来通り両軸とも同じにするが、演算時間そのものを異ならせる方法である。より精度の高い補正を必要とする軸回りには、前述の通常の演算式を適用し、そうでない軸回りについては前述の簡略演算式を用いる。その極端な場合として、補正の必要がないと判断された軸回りに関して、補正動作を省略することもできる。
第１の実施形態を採用する場合、図２に示す補正周期決定手段１４は省略し得る。
【００３４】
本発明の第２の実施形態として、両軸回りの演算に、共に通常の演算式を用いる場合でも、両者の補正周期を異ならせることでＣＰＵの演算割り当て時間に重み付けを行う。　図４に示す例のように鉛直方向と水平方向の補正周期を一定周期とするのではなく、振れが大きい鉛直方向の補正周期Δｔ_ｐを短くして補正精度を高め、さらに振れが小さい水平方向の補正周期Δｔ_ｈを長くして無駄な補正精度を抑え、鉛直方向の１周期Δｔ_ｐ当たりの変化量ΔＸ_ｐと、水平方向の１周期Δｔ_ｐ当たりの変化量ΔＸ_ｐとのバランス（ΔＸ_ｐ≒ΔＸ_ｈ）をとる。
【００３５】
どちらの方向が変化率が大きいかは、撮影装置の保持状態でもある程度決まる。つまり、カメラが縦長に構えられているのか、横長に構えられているのかで振れやすさの方向が異なる。撮影装置が図５に示すように横長方向にも縦長方向にも構えられる場合では、横長方向か縦長方向かを撮影者が切り替えるモード選択手段を設けるとよい。また、カメラがどの方向に構えられているのか、つまり、カメラが縦長に構えられているのか、横長に構えられているのかを判別する保持状態検出手段１３を設けてもよい。
【００３６】
図６は、本発明の振れ補正機能付き撮影装置を実施する場合の１つの形態を説明するための図である。撮影装置１５は横型に構えられた状態から、任意の傾き角度θだけ傾いて保持されている。
本実施例では、保持状態検出手段は、単に撮影装置保持方向が縦長方向なのか横長方向なのかを判断するだけでなく、撮影装置１５の傾き角度を検出する。
このような場合でも、補正周期決定手段１４は保持状態検出手段１３から得られた撮影装置の傾き角度により、装置が固有で持っている傾き角度毎に対応した補正周期テーブルからＸ軸回りとＹ軸回りの補正周期を決定する。
【００３７】
両軸回りの補正周期を異ならせる様子を、モデル化した例で以下説明する。
システム負荷と補正周期の考え方を、図７に基づいて説明する。例えば、Ｘ軸回りとＹ軸回りの２つの補正手段の補正周期の比率が９：９と同じにした図７（ａ）の第１例と、２つの補正手段の補正周期の比率が６：１８と振れの変化が大きいＸ軸回りの補正周期を短く、振れの小さいＹ軸回りの補正周期を長くした図７（ｂ）の第２例について考える。
図７の横軸は時間軸で、セルＸ、セルＹ、セルＥは横軸に対してそれぞれＸ軸回りの振れ補正処理時間、Ｙ軸回りの振れ補正処理時間、露光中のシステム側が使用可能な時間である。
【００３８】
Ｘ軸回りの振れ補正処理時間とＹ軸回りの振れ補正処理時間は、簡略演算式を用いない場合、方向が違うだけで振れ補正処理内容は同じであるため、セルＸとセルＹの横軸方向の長さは同じであり、説明の便宜のためセルＥもセルＸ、セルＹと同じ長さとし、この長さを単位時間として表現してある。セル中に記載の番号は各々の振れ補正処理が繰り返された回数である。なお、簡略演算式を用いた場合は、セルＸとセルＹの横幅は等しくならない。
【００３９】
ここで、露光中のある一定時間Ｔ内を考える。Ｘ軸回りとＹ軸回りの振れの２つの補正手段の補正周期の比率が９：９の第１例の場合は、時間Ｔ内の振れ補正処理回数はＸ軸回りの補正が２回、Ｙ軸回りの補正が２回であり、システム側が使用できる時間は１４単位時間である。また、Ｘ軸回りとＹ軸回りの振れの２つの補正手段の補正周期の比率が６：１８の第２例の場合は、時間Ｔ内の振れ補正処理回数はＸ軸回りの補正が３回、Ｙ軸回りの補正が１回であり、第１例に比べて、Ｘ軸回りの振れ補正精度は１．５倍、Ｙ軸回りの振れ補正精度は０．５倍となっているが、システム側が使用できる時間は１４単位時間と変わらないことがわかる。
【００４０】
補正周期の長い方の演算に対して更に簡略演算式を用いることで、両軸回りの重み付けの違いをより強調することができる。逆に、補正周期が非常に短くなると補正動作１回当たりに発生する振れ量はかなり小さくなると考えられるので、１回ごとの演算はさほど高精度を必要としなくなる。したがって、補正周期の短い方に簡略演算式を適用してＣＰＵの負荷を減らすこともできる。
このように、Ｘ軸回りとＹ軸回りの振れの大きさに合わせて、２つの補正手段の補正周期を適切に決定することで、システム側の負荷を上げることなく、必要な補正精度を得ることができる。
【００４１】
図８に把持状態を検出できるスチルカメラの例を示す。図に示すようにカメラ本体１０１のグリップ部に、撮影者がカメラを把持していることを判別するための把持状態検出センサ１０２を有している。また、本体のもう一方にも同じく把持状態検出センサ１０３を有している。これらの把持状態検出センサはタッチセンサやフォトインタラプタの様に遮光されることによって反応するセンサなどが考えられる。この把持状態検出センサに指が触れているかどうかによって、カメラが両手で把持されているのか片手で把持されているのかを判別する。
【００４２】
カメラが両手で把持されている場合、左右から同等の力でカメラをはさみこむように構えているため、Ｙ軸回りの振れはすくなく、Ｘ軸回りの振れが大きくなると思われる。片手で把持している場合には、力の加わる方向が一方向であるため、振れ方向はＸ、Ｙの別なく生じる。そこで、両手持ちと判別された場合は、Ｘ軸回りに対してＣＰＵの演算割り当て時間に高い重み付けを行う。片手の場合は、他の条件を考慮して重み付けを行う。重み付けの手段は前述の各手段のいずれを採用してもよい。
【００４３】
図９に把持状態検出センサ１０４、１０５を有する横薄型カメラを示す。このタイプのカメラの場合、図８のようなタイプのカメラに比べ、Ｘ軸回りの振れは大きくなる傾向にあり、両手で把持した場合、Ｘ軸回りに発生する振れはより大きいと思われる。また両手で把持した場合でも、一眼レフタイプの形態のカメラの場合、構え方にも幾つか考えられ、鏡胴の下に左手を添えるような底面把持の構え方の場合は、両手の力の作用方向から考えて、Ｙ軸回りの振れが大きくなると思われる。そこで、このように一眼レフタイプのカメラ形態の場合は、図１０に示すように、カメラ本体の底面や、鏡胴の下側にも把持状態検出センサ１０６、１０７を設けることで、両手の構え方のバリエーションを判別し、振れ補正方向を決定する。
【００４４】
カメラが振れ補正モードにある時、ＣＰＵは、たとえばＸ軸回りの振れを検出し、補正量を算出し、補正を行ったら、次にＹ軸回りの振れを検出、補正量算出、補正を行うという動作が交互に処理されており、その間、ズームやフォーカスなどの動作が逐次処理されている。したがって、振れ補正を行っている間、ＣＰＵにおける処理は増え、このことはＣＰＵの処理速度の低下や、ひいては振れ補正の応答性や追従性にも影響してくる。そこで、本発明では両手、片手の把持方法や、通常把持か底面把持かの別を示す把持状態検出センサの出力結果を利用して、補正不要と判断した振れ補正方向の補正を省略することで、ＣＰＵにかかる負担を軽減することが出来、さらには消費電力の節約にもつながり、適切な振れ補正を行うことが出来る。
【００４５】
本発明における振れ補正手段では、撮影装置本体に取り付けられた把持状態検出センサによって両手把持、片手把持の別を判断し、把持状態に対応してあらかじめ定められている、振れ補正方向ごとの補正周期の決定を行う。
【００４６】
たとえば、図８において把持状態検出センサ１０２、１０３の両方がｏｎであった時、ＣＰＵはＸ軸回りの補正動作のみを所定の周期で実行し、これにより振れ補正手段１１のみが駆動され、Ｙ軸回り振れ検出センサ５の出力の取り込み後の、振れ補正手段１２へのデータ出力までの一連の動作は省略される。これによりＣＰＵ８の負担は軽減される。ここで述べたセンサの例は一例にすぎず、組み合わせは他にも考えられる。
【００４７】
振れ補正方向が前記例のようにＸ軸回りに決まって、振れ補正手段に対しＣＰＵ８が所定の周期で補正動作を行っている間も、継続してＣＰＵ８がＹ軸回りのチェックを行ってもよい。すなわち、振れ検出とその振れ量の演算までを所定の周期、たとえば、Ｘ軸回りの補正動作と同じ周期で行ってもよい。そうすることによって、先の補正方向の決定によって処理を省略すると決定した方向の振れが、所定の値を超えた場合には、その方向に対する振れ補正を開始することができる。
この場合、処理を省略した方の振れに関しては、以後のチェックの際の振れ補正演算に簡略演算式を用いてもよい。
【００４８】
図２に示すように撮影装置の保持状態を検出する保持状態検出手段１３と、この保持状態検出手段１３より検出される撮影装置の保持状態に基づいて、Ｘ方向及びＹ方向の補正周期を決定する補正周期決定手段１４を有することにより、撮影装置が横長、縦長のどちらに構えた場合でも、鉛直方向の軸回りの補正周期を短くし、水平方向の軸回りの補正周期を長くすることで、Ｘ軸回り、Ｙ軸回りの適切な補正精度を実現することができる。尚、補正周期決定手段１４は演算手段９に含まれてもなんら問題はない。
【００４９】
本実施形態の場合のＸ軸回り、Ｙ軸回りの補正周期決定手順を図１１に示すフローチャートにて説明する。撮影装置１５は、常に撮影者によるレリーズボタン等の開始指示の操作による撮影指示信号が発生しているか否かをチェックする（Ｓ１）。撮影指示信号が検知されると（Ｓ１のＹｅｓ）、撮影装置の保持状態を取り込む（Ｓ２）。Ｓ２で取り込まれた撮影装置の保持情報により撮影装置が横型に構えられているのか、縦型に構えられているのか判断する（Ｓ３）。撮影装置１５が横長方向に構えたと判断された場合（Ｓ３のＹｅｓ）、Ｘ軸回りが鉛直方向となり、Ｘ軸回りの補正周期にＡが、そしてＹ軸回りの補正周期にＢが設定される（Ｓ３１）。また撮影装置１５が縦長方向に構えたと判断された場合（Ｓ３のＮｏ）、Ｙ軸回りが鉛直方向となり、Ｘ軸回りの補正周期にＢが、そしてＹ軸回りの補正周期にＡが設定される（Ｓ３２）。ここで補正周期Ａは補正周期Ｂより小さくＡ＜Ｂである。このようにして決定されたＸ軸回り及びＹ軸回りの補正周期によって露光と同時に振れ補正動作が開始される（Ｓ４）。
【００５０】
Ｓ２の撮影装置の保持情報取り込みでは、例えば撮影装置が図１に示すようにシャッタ押下方向が一方向であり常に横型に構えられる場合では撮影装置がデフォルトで持っている保持情報を読み込む。また、撮影装置が図５に示すように横長方向にも縦長方向にも構えられる場合では横長方向か縦長方向かを切り替えるモード選択手段により撮影者によって設定された情報を読み込む。
また、カメラがどの方向に構えられているのか、つまり、カメラが縦長に構えられているのか、横長に構えられているのかを判別する保持状態検出手段１３を有する場合では、保持状態検出手段より出力された情報により撮影装置が横長方向に構えられているのか縦長方向に構えられているのかを判断する。
【００５１】
カメラの姿勢判別センサをカメラ内部に装備して、カメラの姿勢とユーザーの把持状態の二つの要素から振れ補正の方向を決定することもできる。図１２に第１の実施形態に対応させたフローチャートを示す。具体的には片手でカメラを構え、横長の状態でカメラを構えている場合は、ユーザーにより触れの程度はかわるものの、力の係り具合から考えて、縦横２方向の振れ補正が必要となる。片手持ちで、カメラを縦長方向に構えている場合、レリーズボタンが上下のどちらになるように構えていても、カメラが下方向への移動を支えているということから、縦方向（この場合はＹ軸回り）に振れやすい。このときはＹ軸回りの振れ補正を決定する。両手でカメラを把持し、縦長方向にカメラを構えている場合には、力の作用方向の関係から、横方向（この場合はＸ軸回り）の振れが大きくなる。このときは、Ｘ軸回りの振れ補正を行うこととする。
なお、このフローチャートでは、ＣＰＵ演算時間割り当ての重み付けとして簡略演算式を使う第１の実施形態の例で示したが、簡略演算と有るところを補正の省略と書き換えても使える。また、同じ考え方で、補正周期を異ならせる第２の実施形態を採用しても構わない。
【００５２】
次に、本発明の第３の実施形態に係わる２軸回りの振れ予測と補正周期の決定、及び補正動作の流れについて図１３に基づいて説明する。露光直前及び露光前の所定期間に振れ検出手段８により検出される２軸回りの振れ検出量に基づき、振れ予測演算手段２０は、露光後の２軸回りの振れ予測情報を算出する。この振れ予測演算手段２０で算出された２軸回りの振れ予測情報に基づき、補正周期決定手段１４は撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの振れ補正手段１１、１２の補正周期を各々決定する。これにより演算手段９のＣＰＵを通じて補正手段駆動回路１０が制御され、振れ補正手段１１、１２は各々適切な補正周期での補正動作が繰り返されることになる。尚、振れ予測演算手段２０及び補正周期決定手段１４は演算手段９に含まれても問題ない。
【００５３】
振れ予測演算手段２０の振れ予測情報算出手法としては、図１４に示すように、露光直前時点の振れ検出手段８により検出される２軸回りの振れ検出量に基づき行うものがある。このように、露光直前の最新の振れ検出量を使用することで、露光後の振れを簡単に精度良く予測することができる。
また、図１５に示すように、露光前の所定期間の前記振れ検出手段８により検出される２軸回りの振れ検出量に基づくものがある。このように、露光直前までの所定期間の複数回のデータに基づく振れ軌跡から振れの性質を的確に掴み取ることで、露光後の振れを非常に精度良く予測することが可能となる。
【００５４】
本実施形態の場合のＸ軸回り、Ｙ軸回りの補正周期決定手順を図１７に示すフローチャートと図１６に基づいて説明する。まず振れ検出手段８では随時撮影装置１５の振れ状態を検出しており（Ｓ０）、記憶手段によってメモリへ書き込み（更新記憶）を行なっている（Ｓ１）。従って、メモリでは所定の時間分、すなわち所定の回数分の振れ検出情報を保持していることになる。
【００５５】
一方、撮影装置１５は、常に撮影者によるレリーズボタン等の開始指示の操作による撮影指示信号が発生しているか否かをチェックする（Ｓ２）。撮影指示信号が検知されると（Ｓ２のＹｅｓ）、振れ検出手段８により検出される振れ検出情報を記憶しているメモリデータ量を確認する（Ｓ３）。このチェックにおいて予測に必要な所定量のデータが記憶されていることを確認すると（Ｓ３のＹｅｓ）、メモリに記憶されているこれらの振れ情報をメモリから演算手段９に取り込み（Ｓ４）、例えば、図１６（ａ）に示すように、最小二乗法や回帰線算出により所定期間の振れデータの傾きを算出し（Ｓ５）、この傾きθ_ｘ、θ_ｙ又はその比率θ_ｘ／θ_ｙに基づき補正周期決定手段１４が撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの前記振れ補正手段の補正周期を決定する。
【００５６】
または、図１６（ｂ）に示すように、所定期間の振れデータの最大変位と最小変位の差分Δｕ_ｘ、Δｕ_ｙ又はその比率Δｕ_ｘ／Δｕ_ｙに基づき補正周期決定手段１４が両軸回りの振れ変位を補正する２つの前記振れ補正手段の補正周期を決定する。あるいは、図１６（ｃ）に示すように、所定期間の振れの周期の短い方が、振れの変化が急傾斜であることなどを利用して振れの周期Δｔ_ｘ、Δｔ_ｙ又はその比率Δｔ_ｘ／Δｔ_ｙに基づき補正周期決定手段１４が両軸回りの振れ変位を補正する２つの前記振れ補正手段の補正周期を決定する。このように２つの前記振れ補正手段の補正周期を決定する際には、振れの変化の大きい軸回りの振れ補正周期を短く（Ｓ７１、Ｓ７２の周期Ａ）、また振れの変化の小さい軸回りの振れ補正周期を長く（Ｓ７１、Ｓ７２の周期Ｂ）設定する。その後、露光開始と同時に２つの前記振れ補正手段は各々設定された周期で振れ補正動作を開始する（Ｓ１０）。
図１８に示すフローチャートは、露光直前の前記振れ検出手段８により検出される２軸回りの１回分の振れ検出情報に基づく振れ予測（Ｓ４１）の場合であり、Ｓ７、Ｓ７１、Ｓ７２は図１７で説明した所定の複数回分の振れ検出情報に基づく振れ予測の場合と同じである。
【００５７】
以上の例に示すような考え方で、振れの変化の大きい軸回りの振れ補正周期を短くし、また振れの変化の小さい軸回りの振れ補正周期を長く設定するわけであるが、２軸回りの振れの変化の絶対値の大きさだけでは、または比率だけでは、２軸回りの振れの変化の絶対値がどちらも同じ程度の場合、振れ補正周期を決定することができない。このように、実際に振れ補正周期を決定するためには、２軸回りの振れの変化の大きさ、及びその比率に加えてシステム負荷の限度を超えないようにバランスを考慮するなどの、幾つかの有効な判断手段を組み合わせて、２軸の振れ補正方向各々の補正周期を適切に決定する。
【００５８】
図１９に示すフローチャートでは、前記振れ予測演算手段２０から算出された２軸回りの振れ予測情報による振れの変化の大きさの違いが、振れ変化差分基準値以下の場合（Ｓ６）、システム負荷の限度を考慮してＸ軸回り及びＹ軸回りの振れ補正周期を同じ周期Ｃ（周期Ａ＜周期Ｃ＜周期Ｂ）としている（Ｓ６１）。
【００５９】
前述のように、人間の起す手振れは、撮影装置の構え方により変わってくるが、撮影装置を図１に示すように横長方向に構えた状態でシャッター２５を押下した場合、Ｘ軸回りの振れの変化が大きく、Ｙ軸回りの振れの変化は少ないのが一般的である。従って、振れ予測演算手段２０で予測された結果も、Ｘ軸回りの振れの変化が大きく、Ｙ軸回りの振れの変化は少なく判断されるはずである。よって、もし撮影装置の構え方により予測される振れの変化の大きさの傾向と、振れ予測演算手段２０で判断された振れの変化の大きさの傾向が異なる場合、すなわち、大小関係が一致しない場合、撮影者の撮影装置の保持形態や撮影姿勢に問題があり、それが原因で振れ予測演算手段２０による露光前の振れによるＸ軸回り及びＹ軸回りの振れの変化の予測が大きく外れる場合がある。この場合には、Ｘ軸回り及びＹ軸回りの振れ補正周期を決定する際、振れ予測演算手段２０による振れ予測情報を考慮しない、または撮影者に撮影装置の保持形態や撮影姿勢について注意を促すようにすることで、振れ補正精度の低下した撮影が行われることを防ぐ。
【００６０】
デジタル式の撮影装置の多くが、撮影状態その他の情報を表示するための画像情報もしくは文字情報の表示手段を持っている。本発明の撮影装置においても、撮影装置本体に画像表示手段を備えていると振れ補正量や振れ補正方向などが表示できるほか、上記のような注意や警告も表示できることになるので便利である。このほかに発光もしくは発音により、警告を行えるように構成するのもよい。
【００６１】
図５に示す撮影装置１５では、横長方向に構えた場合のレリーズボタン１６と、縦長方向に構えた場合のレリーズボタン１７を有している。図５（ａ）に示すように撮影装置１５を横長方向に構えた場合は、前に説明したがＸ軸回りの振れの変化が大きく、Ｙ軸回りの振れの変化は少ないのが一般的である。また、図５（ｂ）に示すように撮影装置１５を縦長方向に構えた場合は、Ｘ軸回りの振れの変化が小さく、Ｙ軸回りの振れの変化が大きいのが一般的となる。このように撮影装置１５のＸ軸回り、Ｙ軸回りの振れの変化の大きさの傾向は、撮影装置の構え方により変化する。
【００６２】
よって、撮影装置は、撮影装置保持方向が縦長方向か横長方向かを判断するために、撮影者が撮影装置保持状態を設定することができるモード設定手段を有する、または、図１３に示すように撮影装置の保持状態を検出する保持状態検出手段１３を有し、常に撮影保持状態を監視することで、撮影装置の構え方により予測される振れの変化の大きさの傾向と、振れ予測演算手段２０で判断された振れの変化の大きさの傾向の一致・不一致を判断することができる。
【００６３】
図２０に示すフローチャートでは、振れ予測演算手段２０で判断された振れの変化の大きさにより決定した補正方向別周期判断と、撮影装置保持情報から決定した補正方向別周期判断が異なる場合、すなわち、大小関係が逆になっている場合（Ｓ９）、システム負荷の限度を考慮してＸ軸回り及びＹ軸回りの振れ補正周期を同じ周期Ｃ（周期Ａ＜周期Ｃ＜周期Ｂ）としている（Ｓ９１）。ただし、Ｓ９１の処理はこれに限らず、たとえば、撮影装置保持情報から決定した周期を採用してもよい。
【００６４】
また、図２１に示すフローチャートでは、振れ予測演算手段２０で判断された振れの変化の大きさにより決定した補正方向別周期判断と、撮影装置保持情報から決定した補正方向別周期判断が異なる場合（Ｓ９）、撮影者に撮影装置の保持状態または撮影姿勢について警告する（Ｓ９２）。
Ｓ８の撮影装置の保持情報取り込は、前述の図８の場合と同様である。
【００６５】
本発明の第４の実施形態の機能的構成例を図２２にて説明する。振れ監視センサ２２は、撮影装置のＸ軸回りの振れを検出し、振れ監視センサ２３は、撮影装置のＹ軸回りの振れを検出する。これらの振れ監視センサ２２、２３と振れ監視センサ回路２１により振れ監視手段２４を構成する。
露光中の所定期間に振れ監視手段２４により検出される２軸回りの振れ検出量に基づき、振れ予測演算手段２０は、振れ予測情報を算出する。この振れ予測演算手段２０で算出された２軸回りの振れ予測情報に基づき、補正周期決定手段１４は撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの振れ補正手段１１、１２の補正周期を各々決定する。これにより演算手段９のＣＰＵを通じて補正手段駆動回路１０が制御され、振れ補正手段１１、１２は各々適切な補正周期での補正動作が繰り返される。
【００６６】
本実施形態では露光期間は所定の単位時間でいくつかの区間に分割されており、補正周期の変更はこの区間毎に行われる。尚、振れ監視手段２４は、振れ検出手段８と兼用であっても問題ない。また、予測演算手段２０及び補正周期決定手段は演算手段９に含まれても問題ない。
【００６７】
Ｘ軸回りとＹ軸回りの補正周期を露光中全期間に渡って一定周期とするのではなく、図２３に示すように、露光期間を所定の単位時間で幾つかの区間に分割して、区間毎に振れの変化が大きい軸回りの補正周期Δｔ_１を短くして補正精度を高め、さらに振れの変化が小さい軸回りの補正周期Δｔ_２を長くして無駄な補正精度を抑え、振れの大きい軸回りの１周期Δｔ_１当たりの変化量Δｕ_１と、振れが小さい軸回りの１周期Δｔ_２当たりの変化量Δｕ_２とのバランス（Δｕ_１≒Δｕ_２）をとる。
【００６８】
補正周期決定手段１４により決定される撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの振れ補正手段１１、１２の補正周期と、振れ監視手段２４により検出される２軸回りの振れの関係を図２４に基づいて説明する。
尚、振れ検出手段８による振れの検出は、振れ補正動作毎に実施されるため、補正周期の変更区間毎に振れ検出周期が変化するが、振れ予測の精度を考慮すれば、振れ監視手段２４による振れの監視は一定周期であることが望ましい。勿論、振れ検出手段８による振れの検出と振れ監視手段２４による振れの監視は同じ振れを検出しているものであり、振れ検出結果の振れ遷移を全体としてみれば同じものであり、振れ検出手段８と振れ監視手段２４を同一として、振れ監視を振れ補正動作毎の振れ検出データを用いて行っても問題ない。むしろ振れ監視による若干のシステム負荷の増加を考えれば、振れ検出手段８と振れ監視手段２４を同一として、振れ監視用データとして振れ補正動作毎の振れ検出データを用いるのが望ましい。
【００６９】
図２４は露光中の任意の連続する４つの区間ａ、区間ｂ、区間ｃ、区間ｄのＸ軸回りとＹ軸回りの振れの遷移を示している。区間ａはＸ軸回りとＹ軸回りの振れの変化がともに小さい区間であり、Ｘ軸回りとＹ軸回りともに補正周期は長く設定される。区間ｂはＸ軸回りの振れの変化は小さく、Ｙ軸回りの振れ変化が大きい区間であり、Ｘ軸回りの補正周期は長く、Ｙ軸回りの補正周期は短く設定される。区間ｃはＸ軸回りの振れの変化が大きく、Ｙ軸回りの振れ変化は小さい区間であり、Ｘ軸回りの補正周期は短く、Ｙ軸回りの補正周期は長く設定されている。区間ｄはＸ軸回りとＹ軸回りの振れの変化がともに大きい区間であり、Ｘ軸回りとＹ軸回りともに補正周期は短く設定される。
【００７０】
これらの区間ａ〜区間ｄのＸ軸回りとＹ軸回りの補正周期の決定とシステム負荷の関係を図２５に基づいて説明する。なお、横軸とそれぞれの記号の定義は前述の図７と同じである。また、この例では、システム負荷の限度として、システム側で使用できる時間が一定時間Ｔ内で１４単位時間以上必要な場合を仮定している。
【００７１】
区間ａの一定時間Ｔ内を考える。Ｘ軸回りとＹ軸回りの振れの２つの補正手段の補正周期の比率が１８：１８であり、時間Ｔ内の振れ補正処理回数はＸ軸回りの振れ補正が１回、Ｙ軸回りの補正が１回であり、システム側が使用できる時間は１６単位時間である。
区間ｂの一定時間Ｔ内を考える。Ｘ軸回りとＹ軸回りの振れの２つの補正手段の補正周期の比率が１８：６であり、時間Ｔ内の振れ補正処理回数はＸ軸回りの振れ補正が１回、Ｙ軸回りの補正が３回であり、システム側が使用できる時間は１４単位時間である。
区間ｃの一定時間Ｔ内を考える。Ｘ軸回りとＹ軸回りの振れの２つの補正手段の補正周期の比率が６：１８であり、時間Ｔ内の振れ補正処理回数はＸ軸回りの振れ補正が３回、Ｙ軸回りの補正が１回であり、システム側が使用できる時間は１４単位時間である。
区間ｄの一定時間Ｔ内を考える。Ｘ軸回りとＹ軸回りの振れの２つの補正手段の補正周期の比率が９：９であり、時間Ｔ内の振れ補正処理回数はＸ軸回りの振れ補正が２回、Ｙ軸回りの補正が２回であり、システム側が使用できる時間は１４単位時間である。
【００７２】
区間ｄのＸ軸回りとＹ軸回りの振れの変化がともに大きい区間では、一定時間Ｔ内の振れ補正処理回数はＸ軸回りの振れ補正が２回、Ｙ軸回りの補正が２回となり、システム側が使用できる時間は１４単位時間と、システム負荷限界で設定されている。
区間ｂでは一定時間Ｔ内の振れ補正処理回数はＸ軸回りの振れ補正が１回、Ｙ軸回りの補正が３回であり、区間ｄに比較するとＸ軸回りの振れ補正精度は０．５倍、Ｙ軸回りの振れ補正精度は１．５倍となる。また、区間ｃでは一定時間Ｔ内の振れ補正処理回数はＸ軸回りの振れ補正が３回、Ｙ軸回りの補正が１回であり、区間ｄに比較するとＸ軸回りの振れ補正精度は１．５倍、Ｙ軸回りの振れ補正精度は０．５倍となる。このように、区間ｂ及び区間ｃではＸ軸回り及びＹ軸回りの振れ補正精度が区間ｄとは異なるが、この場合でも、システム側が使用できる時間は１４単位時間となっており、区間ｄと同じシステム負荷限界での設定となっている。
【００７３】
さらに、図２４、図２５で説明した制御を実現するために、２軸回りの振れ補正を実行する回数に制限を設けることで、システム負荷についても考慮して補正周期を決定することになる。このように、各区間におけるＸ軸回りとＹ軸回りの振れの変化の大きさに合わせて、さらに２軸回りの振れ補正を実行する回数に制限を設けることで区間毎にシステム負荷がシステム負荷限界以下となるように、Ｘ軸回りとＹ軸回りの振れの２つの補正手段の補正周期を適切に決定する。このようにすることで、露光期間中のシステム側の負荷を上げることなく、Ｘ軸回りとＹ軸回りで必要な補正精度を得ることができる。
【００７４】
ところで、予測に必要な所定期間の振れ監視手段２４から検出される振れ検出量データは、例えばメモリ等に記憶される。よって、メモリ等に記憶された予測該当区間の直前までの所定期間の振れ検出量データを用いて予測演算手段２０では予測演算が実施される。この予測演算方法としては、例えば、一連の所定期間の振れ検出量データを一次微分、二次微分して、該当区間直前の振れ速度、及び振れ加速度を算出し、該当区間の振れの変化の大きさを算出するなどの方法がある。
【００７５】
以上の例に示すような考え方で予測された該当区間の２軸回りの振れの変化の大きさにより、２軸回りの振れの変化の大きさの絶対値に基づいて２軸回りの振れ変位を補正する２つの振れ補正手段の補正周期を決定したり、２軸回りの振れの変化の大きさの比率に基づいて２軸回りの振れ変位を補正する２つの振れ補正手段の補正周期を決定することができる。さらにこれに加えてシステム負荷の限度を超えないようなバランスを考慮するなどの、幾つかの有効な判断手段を組み合わせて、２軸回りの振れ補正方向各々の補正周期を各区間毎に適切に決定する。
振れの変化の大きさの絶対値、あるいは、振れの変化の大きさの比率に対応すべき補正周期の組合せは、固定的であってもよいが、複数の組合せを用意しておいて、撮影者が選択できるように、選択手段を設けてもよい。
【００７６】
本実施形態の場合のＸ軸回り、Ｙ軸回りの補正周期決定手順を図２６、図２７に示すフローチャートに基づいて説明する。
まず図２６は露光前動作を説明するものであり、振れ監視手段２４では随時撮影装置１５の振れ状態を検出しており（Ｓ０）、記憶手段によってメモリへ書き込み（更新記憶）を行なっている（Ｓ１）。従って、メモリでは所定の時間分すなわち、所定の複数回分の振れ監視情報を保持していることになる。
【００７７】
一方、撮影装置１５は、常に撮影者によるレリーズボタン等の開始指示の操作による撮影指示信号が発生しているか否かをチェックする（Ｓ２）。撮影指示信号が検知されると（Ｓ２のＹｅｓ）、振れ監視手段２４により検出される振れ監視データが記憶されていることを確認すると（Ｓ３のＹｅｓ）、メモリに記憶されているこれらの振れ情報をメモリから演算手段９に取り込み（Ｓ４）、例えば、一連の所定期間の振れ検出量データを一次微分、二次微分して、振れ速度、及び振れ加速度を算出し、露光直後の振れの変化の大きさを予測する（Ｓ５）。このステップ５の処理により算出された予測振れ情報により、補正周期決定手段１４は、露光直後の区間における撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの前記振れ補正手段の補正周期を決定し、露光開始と同時に、Ｘ軸回りとＹ軸回り各々決定された補正周期で振れ補正が開始される（Ｓ１０）。
【００７８】
図２７は露光中の振れ補正動作を説明するものであり、振れ検出手段８からの出力を取り込み（Ｓ１１）、この出力に応じて、振れ補正量を算出する（Ｓ１２）。そして算出された振れ補正量に対応した信号が補正手段駆動回路１０に送られ、それにより補正手段駆動回路１０を通じた振れ補正手段１１、１２が駆動される（Ｓ１３）。この一連の振れ補正動作はＸ軸回り、Ｙ軸回り別々の周期で露光中に繰り返し行われる（Ｓ１９）。
【００７９】
ところで、振れ監視手段２４から検出される振れ検出量データ（Ｓ０）は露光前動作と同じく随時メモリ等の記憶手段に記憶（更新記憶）されている（Ｓ１）。さらに演算手段９のマイクロプロセッサは図２４に例示する区間ａから区間ｄに相当する区間の変わり目を判断し（Ｓ１５）、区間の変わり目である場合には、メモリに記憶された次区間の直前までの所定期間の振れ検出量データを取り込み（Ｓ４）そのデータを用いて予測演算手段１８により予測演算が実施される（Ｓ５）。このＳ５の処理により算出された予測振れ情報により、補正周期決定手段１４は、露光直後の区間における撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの前記振れ補正手段の補正周期を決定し、次区間開始と同時にＸ軸回りとＹ軸回り各々決定された補正周期で振れ補正が開始される（Ｓ１９のＹｅｓ）。尚、区間毎のＸ軸回りとＹ軸回りの補正周期決定では、振れ予測情報だけではなく、システム負荷についても考慮して補正周期を決定することより、区間の変わり目は図２５のＰ時点及びＱ時点のようにＸ軸回りとＹ軸回りの振れ補正動作開始が同じタイミングである時点であることが望ましい。
【００８０】
以上、本実施形態では振れ補正手段が撮像素子を振れとは反対方向に変位駆動させる振れ補正方式の場合について説明したが、振れ補正手段が撮影光学系に含まれる補正レンズによる場合や、撮影光学系の光軸上に設けた可変頂角プリズムによるものであっても同様のことが言える。
【００８１】
【発明の効果】
請求項１記載の発明は、振れ補正機能を有する撮影装置において、撮影装置の光軸に直交する２軸の回りの振れ補正手段の、補正のための処理に要するＣＰＵの演算割り当て時間に重み付けを行うことにしたので、ＣＰＵの負荷の増加を招かずに、各々の軸周りの振れ特性に合わせた適切な補正精度での振れ補正が行なえる。よって、全体として撮影画像の振れ補正効果を向上させることができる。
【００８２】
請求項２記載の発明は、請求項１に記載の撮影装置において、ＣＰＵが２つの振れ補正手段に与える補正量を演算するための演算式として通常の演算式と簡略演算式を用意し、一方の補正手段に対し通常の演算式を用い、他方の補正手段に対し簡略演算式を用いることでＣＰＵの演算割り当て時間の重み付けを行うことにしたので、ＣＰＵの負荷の増加を招かずに、各々の軸周りの振れ特性に合わせた適切な補正精度での振れ補正が行なえる。よって、全体として撮影画像の振れ補正効果を向上させることができる。
【００８３】
請求項３記載の発明は、請求項１記載の撮影装置において、撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの前記振れ補正手段の補正周期を各々個別に設定することにより、ＣＰＵの演算割り当て時間の重み付けを行うことにしたので、ＣＰＵの負荷の増加を招かずに、各々の軸周りの振れ特性に合わせた適切な補正精度での振れ補正が行なえる。よって、全体として撮影画像の振れ補正効果を向上させることができる。
【００８４】
請求項４記載の発明は、請求項３に記載の撮影装置のおいて、撮影装置の光軸に直交する２軸回りの振れ方向のうち、振れが大きく振れる方向の前記補正手段の補正周期を短くし、また振れが小さく振れる方向の前記補正手段の補正周期を長くして無駄な補正精度を抑える。このことはたとえば、水平方向での補正精度を抑えることにより得られた余り時間を、鉛直方向の補正時間に活用することであり、撮影システムの負荷を増大させることなく、全体として撮影画像の手振れ補正効果を向上させることができる。
【００８５】
請求項５記載の発明は、請求項３または４に記載の撮影装置のおいて、補正周期を異ならせた２軸の回りの補正量演算において、ＣＰＵが２つの振れ補正手段に与える補正量を演算するための演算式として通常の演算式と簡略演算式を用意し、一方の軸に関する演算に簡略演算式を用いることにしたので、ＣＰＵの負荷の増加を招かずに、各々の軸周りの振れ特性に合わせた適切な補正精度での振れ補正が行なえる。よって、全体として撮影画像の振れ補正効果を向上させることができる。
【００８６】
請求項６記載の発明は、請求項３、４、または５に記載の撮影装置において、振れ補正機能を有する撮影装置において、撮影装置の把持状態検出センサを有し、その検出結果に基づいて、撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの振れ補正手段の第１の補正周期を各々決定する補正周期決定手段を有することにより、各々の軸周りの振れ特性に合わせた適切な補正精度での振れ補正が行なえる。よって、全体として撮影画像の振れ補正効果を向上させることができる。
【００８７】
請求項７記載の発明は、請求項３、４または５に記載の撮影装置のおいて、撮影装置の姿勢すなわち、保持状態が縦長方向なのか横長方向を判断し、撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの前記振れ補正手段の第１の補正周期を各々決定する補正周期決定手段を有することにより、各々の軸周りの振れ特性に合わせた適切な補正精度での振れ補正が行なえる。よって、全体として撮影画像の振れ補正効果を向上させることができる。
【００８８】
請求項８記載の発明は、請求項２、３、４または５記載の撮影装置において、露光以前の振れ検出手段により検出される２軸回りの振れ検出量に基づき、露光後の２軸回りの予測振れ量を算出する振れ予測演算手段と、振れ予測演算手段で算出された２軸回りの予測振れ量に基づき、撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの振れ補正手段の補正周期を各々決定する補正周期決定手段とを有することにより、露光前の振れ検出手段により検出される予測振れ量に基づき、露光後の２軸回りの振れ方向の振れの変化の大きさを各々予測し、振れの変化が大きい軸回りの補正周期を短くして補正精度を高め、振れの変化が小さい軸回りの補正周期を長くし無駄な補正精度を抑え、撮影システムの負荷を増大させることなく、全体として撮影画像の手振れ補正効果を向上させることができる。
【００８９】
請求項９記載の発明は、請求項６または７に記載の撮影装置において、露光以前の振れ検出手段により検出される２軸回りの振れ検出量に基づき、露光後の２軸回りの予測振れ量を算出する振れ予測演算手段と、振れ予測演算手段で算出された２軸回りの予測振れ量に基づき、撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの振れ補正手段の第２の補正周期を各々決定する補正周期決定手段とを有することにより、露光前の振れ検出手段により検出される予測振れ量に基づき、露光後の２軸回りの振れ方向の振れの変化の大きさを各々予測し、振れの変化が大きい軸回りの補正周期を短くして補正精度を高め、振れの変化が小さい軸回りの補正周期を長くし無駄な補正精度を抑え、撮影システムの負荷を増大させることなく、全体として撮影画像の手振れ補正効果を向上させることができる。
【００９０】
請求項１０〜１１記載の発明は、請求項９記載の撮影装置において、第１の補正周期と第２の補正周期の相互のＸ、Ｙ軸対応の大小関係が一致しないとき、２軸の補正周期を等しくするか、または、第１の補正周期を用いる。またそのような状況を撮影装置に表示を行うか、あるいは撮影者に音または光によって警告を行うので、振れ補正精度の低下した状態での撮影を防ぐことができる。
【００９１】
請求項１２記載の発明は、請求項８〜１０のいずれか１つに記載の撮影装置において、前記補正周期決定手段は、前記振れ予測演算手段から算出された２軸回りの予測振れ量より、撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの前記振れ補正手段の補正周期のうち、振れの変化の大きい軸回りの補正周期を短くし、振れの変化の小さい軸回りの補正周期を長く変化させることにより、振れの変化が大きいと予測された軸回りの補正精度を高め、振れの変化が小さいと予測された軸回りの補正周期を長くして無駄な補正精度を抑えことで、システムの負荷を上げることなく、振れの大きい方向の補正精度をあげ、また２軸回りの補正精度を均一にすることができるため、全体として撮影画像の手振れ補正効果を向上させることができる。
【００９２】
請求項１３記載の発明は、請求項３に記載の撮影装置において、露光中の当該撮影装置の光軸に直交する２軸回りの振れ状態を監視する振れ監視手段と、前記振れ監視手段により検出される２軸回りの振れ検出量に基づき、露光後の２軸回りの予測振れ量を算出する振れ予測演算手段と、前記振れ予測演算手段で算出された２軸回りの予測振れ量に基づき、撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの前記振れ補正手段の補正周期を所定区間毎に各々変化させる補正周期決定手段とを有することにより、Ｘ軸回りとＹ軸周りの振れの変化の大きさに合わせて、Ｘ軸回りとＹ軸周りの２つの補正手段の補正周期を適切に決定することで、Ｘ軸回りとＹ軸周りで必要な補正精度を得ることができる撮影装置を提供できる。
【００９３】
請求項１４に記載の発明は、請求項１３に記載の撮影装置において、前記補正周期決定手段により、前記振れ予測演算手段から算出された２軸回りの振れ予測情報より決定される撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの前記振れ補正手段の補正周期を、２軸回りの振れ予測情報による２軸回りの振れの変化の大きさの絶対値に基づいて変化させることにより、もしくは２軸回りの振れ予測情報による２軸回りの振れの変化の大きさの比率に基づいて変化させることにより、２軸方向の振れの変化の大きさの判断を容易に、また正確に実施することができ、振れの変化が大きいと予測された軸回りの補正精度を高め、振れの変化が小さいと予測された軸回りの補正周期を長くして無駄な補正精度を抑えことで、システムの負荷を上げることなく、振れの大きい方向の補正精度をあげ、また２軸回りの補正精度を均一にすることができるため、全体として撮影画像の手振れ補正効果を向上させることができる。
【００９４】
請求項１５、に記載の発明は、請求項１４に記載の撮影装置において、撮影装置が、前記振れ予測演算手段から算出された２軸回りの予測振れ量による振れの変化の大きさの絶対値、また２軸回りの予測振れ量による振れの変化の大きさの比率と、前記補正周期決定手段により決定される２軸回りの振れ変位を補正する２つの前記振れ補正手段の補正周期との関係を、撮影者が適宜設定することが可能である設定手段を有することで、より細やかな２軸方向の振れの変化の大きさの判断が可能となる。
【００９５】
請求項１６記載の発明は請求項３〜１５のいずれか１つに記載の撮影装置において、理論上振れが小さいと見込まれる軸方向、あるいは、検出振れ量もしくは予測振れ量が所定値より小さい軸方向の振れ補正を省略するので、ＣＰＵの処理量が軽減され、処理速度が向上し、さらに消費電力の軽減になる
請求項１７に記載の発明は、請求項１６に記載の撮影装置において、振れ補正動作を省略した軸方向の振れ量も継続的に測定し、必要に応じて補正を行うので、より振れによるボケのすくない鮮明な撮影を行うことが出来る
【００９６】
請求項１８記載の発明は、請求項３〜１７のいずれか１つに記載の撮影装置において、２軸の補正周期の組合せを複数用意して、どの組合せを採用するかは、撮影者が決められるような選択手段を設けたので、実際の撮影状況に合わせて撮影者が最適の組合せを選ぶことができ、より精度の高い補正が可能となる。
【００９７】
請求項１９記載の発明は、請求項１〜１８のいずれか１つに記載の撮影装置において、露光中に撮影装置の光軸に直交する２軸回りの振れ変位を補正する２つの振れ補正手段による２軸回りの振れ補正を実行する回数に制限を設けることで、システム負荷がシステム負荷限界以下となるように、Ｘ軸回りとＹ軸周りの振れの２つの補正手段の補正周期を適切に決定していくことができ、露光期間中のシステム側の負荷を上げることなく、Ｘ軸回りとＹ軸周りで必要な補正精度を得ることがで、全体として撮影画像の手振れ補正効果を向上させることができる。
【００９８】
請求項２０記載の発明は請求項１〜１９のいずれか１つに記載の撮影装置において、補正の状況を表示することにしたので、撮影者は自分の撮影した画像がどのような状態で撮影されたのかを知ることができ、以後の撮影の参考に利用できる。
【図面の簡単な説明】
【図１】本発明の実施形態を説明するための撮影装置の概略斜視図である。
【図２】本発明の第２の実施形態の機能的構成例を示すブロック図である。
【図３】従来の２方向の手振れ変位と補正周期の関係を示す図である。
【図４】本発明の２方向の手振れ変位と補正周期の関係を示す図である。
【図５】撮影装置を横長方向及び縦長方向に保持した状態を示す概略斜視図である。
【図６】本発明の実施形態を説明するための撮影装置の概略斜視図である。
【図７】本発明のシステム負荷と補正周期の関係を説明するための図である。
【図８】把持状態検出センサを有する一般的な形状のスチルカメラを示す。
【図９】把持状態検出センサを有する横薄型カメラを示す
【図１０】底面や、鏡胴の下側にも把持状態検出センサを設けたカメラを示す
【図１１】本発明の第２の実施形態の補正周期決定手順を示す概略フローチャートである。
【図１２】本発明の第１の実施形態において撮影装置の姿勢と把持状態を考慮したフローチャートである。
【図１３】本発明の第３の実施形態の機能的構成例を示すブロック図である。
【図１４】本発明の第３の実施形態を説明するための図である。
【図１５】本発明の第３の実施形態を説明するための図である。
【図１６】本発明の第３の実施形態を説明するための図である。
【図１７】本発明の第３の実施形態の補正周期決定手順を示す概略フローチャートである。
【図１８】本発明の第３の実施形態の補正周期決定手順を示す概略フローチャートである。
【図１９】本発明の第３の実施形態の補正周期決定手順を示す概略フローチャートである。
【図２０】本発明の第３の実施形態の補正周期決定手順を示す概略フローチャートである。
【図２１】本発明の第３の実施形態の補正周期決定手順を示す概略フローチャートである。
【図２２】本発明の第４の実施形態の機能的構成例を示すブロック図である。
【図２３】本発明における振れ量の時間的変化と補正周期の関係を示す図である。
【図２４】本発明の第４の実施形態を説明するための図である。
【図２５】本発明のシステム負荷と補正周期の関係を説明するための図である。
【図２６】本発明の台４の実施形態の補正周期決定手順を示す概略フローチャートである。
【図２７】本発明の台４の実施形態の補正周期決定手順を示す概略フローチャートである。
【符号の説明】
１　撮影装置
２　撮影レンズ
３　撮影光学系
４　振れ検出センサ（Ｘ軸回り）
５　振れ検出センサ（Ｙ軸回り）
６　撮像素子
７　振れ検出センサ回路
８　振れ検出手段
９　演算手段
１０　補正手段駆動回路
１１　振れ補正手段（Ｘ軸回り）
１２　振れ補正手段（Ｙ軸回り）
１３　保持状態検出手段
１４　補正周期決定手段
１５　撮影装置
１６　レリーズボタン（横型に構えた場合）
１７　レリーズボタン（縦型に構えた場合用）
２０　振れ予測演算手段
２１　振れ監視センサ回路
２２　振れ監視センサ（Ｘ軸回り）
２３　振れ監視センサ（Ｙ軸回り）
２４　振れ監視手段
２５　レリーズボタン
１０１　カメラ
１０２、１０３、１０４、１０５、１０６、１０７　把持状態検出センサ[0001]
[Industrial applications]
The present invention relates to an imaging device such as a digital still camera and a digital video camera. More specifically, the present invention relates to a shake correction function in these photographing apparatuses.
[0002]
[Prior art]
Digital still cameras are difficult to achieve a short shutter time like silver halide photography due to the limit of the sensitivity of the image sensor, and the camera shake during image exposure or image shake causes "image flow" to occur in the captured image. Such blur is likely to occur.
[0003]
As a method of detecting a shake, for example, as described in Japanese Patent Application Laid-Open No. 2000-69353, a method of performing image processing on a plurality of times of image data captured by an image sensor to obtain shake information, There is an image pickup apparatus that detects angular acceleration of an imaging apparatus, calculates a shake amount, and uses the calculated shake amount.
[0004]
As a method of correcting the shake, for example, as described in Japanese Patent Application Laid-Open No. 2001-194701, the signal level of an image signal obtained from an image sensor is increased by a variable gain amplifying means to increase the shutter time. There is an electrical correction method that makes the length shorter than the runout limit.
Alternatively, mechanical correction means such as operation of a lens system of an imaging apparatus and movement of an imaging element are also used. Japanese Patent Application Laid-Open No. 11-196428 discloses such an example.
[0005]
As a method for improving the accuracy of correction, Japanese Unexamined Patent Publication No. Hei 5-107719 discloses a method of determining a shake detection and correction cycle based on focal length information of a lens, and shortening the correction cycle when the lens focal length is long. Are shown to be effective.
[0006]
In either case, two-dimensional correction is often performed in a plane orthogonal to the lens optical axis of the imaging device. Here, for convenience of the following description, the two-dimensional direction in the imaging device is defined.
FIG. 1 is a diagram for explaining an embodiment in which a photographing apparatus with a shake correction function according to the present invention is implemented. As shown in FIG. 1, when the shape of the photographing device when viewed from the front of the optical axis of the lens of the photographing device is close to a rectangle, the longitudinal direction is referred to as the X-axis direction. A direction orthogonal to the X-axis direction is referred to as a Y-axis direction. In general, the X-axis direction is not always horizontal. If the shape is close to a square and the longitudinal direction cannot be determined, the horizontal direction at the standard use position is defined as the X-axis direction.
[0007]
In the imaging device 1, when the X-axis direction is horizontal as shown in FIG. 1, the arrow indicated by A indicates the displacement around the X-axis because the direction of the Y-axis that is oriented vertically without changing the X-axis direction changes. Call. Similarly, the arrow indicated by B is referred to as a displacement about the Y axis.
In order to correct the movement of the image due to the displacement about the X axis, the optical axis of the optical system or the image sensor must be moved in the Y axis direction. Therefore, the calculation result of the shake amount caused by the displacement around the X axis is defined as ΔY. Similarly, the calculation result of the shake amount caused by the displacement about the Y axis is defined as ΔX. For convenience of description, the direction of image shake on the image plane may be expressed as, for example, the X direction or the Y direction.
[0008]
The photographing device 15 shown in FIG. 5 has a shutter 16 when held in a horizontally long direction and a shutter 17 when held in a vertically long direction. As shown in FIG. 5A, when the photographing device 15 is held in the horizontal direction, the displacement around the X axis is the displacement in the vertical direction, and the displacement around the Y axis is the displacement in the horizontal direction. When the photographing device 15 is held in the vertically long direction as shown in FIG. 5B, the displacement around the X axis is the displacement in the horizontal direction, and the displacement around the Y axis is the displacement in the vertical direction. As described above, the displacement of the photographing device 15 around the X axis and the displacement around the Y axis can be either vertical displacement or horizontal displacement depending on how the photographing device is held.
[0009]
[Problems to be solved by the invention]
Normally, when performing shake correction, for example, a shake amount ΔY around the X axis is calculated to correct shake around the X axis, and then a shake amount ΔX around the Y axis is calculated to correct shake around the Y axis. Is repeated at regular intervals. This calculation is generally performed by a CPU called a central processing unit. The CPU also performs processes such as zooming and focusing, and when the image capturing apparatus is designated in the shake correction mode, the burden on the CPU is clearly increased. This causes a reduction in the respective processing speeds, and also leads to a reduction in follow-up and response in shake correction. Since these processes are always performed when in the shake correction mode, the power consumption also increases.
As in the related art, when the correction cycle is shortened, the processing time of the CPU is increased, and the above-described problem becomes conspicuous. Conversely, if the correction cycle is simply lengthened to reduce the load on the CPU, the correction accuracy necessarily decreases.
[0010]
[Means for Solving the Problems]
In order to solve these problems, according to the first aspect of the present invention, an imaging optical system, imaging means for receiving light from a subject passing through the imaging optical system, and two axes orthogonal to the optical axis of the imaging optical system A shake detecting means for detecting a shake around, a CPU for performing a predetermined process by receiving a detected shake amount detected by the shake detecting means, and a shake displacement around the two axes based on a processing result of the CPU. In a photographing apparatus having two shake correcting means for correcting the above, weighting is performed on the computation allocation time required by the CPU for the processing for correction by the shake correcting means around the two axes.
[0011]
According to a second aspect of the present invention, in the image capturing apparatus according to the first aspect, a normal arithmetic expression and a simplified arithmetic expression are prepared as arithmetic expressions for calculating the correction amount given to the two shake correction units by the CPU. It is characterized in that the CPU assigns a calculation time by using a normal calculation formula for one correction means and a simplified calculation formula for the other correction means.
According to a third aspect of the present invention, in the photographing apparatus according to the first aspect, the calculation allocation time of the CPU is weighted by making the correction periods of the two shake correction units different.
[0012]
According to a fourth aspect of the present invention, in the imaging device according to the third aspect, the correction cycle of the larger detection shake amount is shortened, and the correction cycle of the smaller one is lengthened.
According to a fifth aspect of the present invention, in the photographing apparatus according to the third or fourth aspect, the correction amount given to the two shake correction units by the CPU in the correction amount calculation around two axes having different correction periods. A normal operation expression and a simplified operation expression are prepared as the operation expression for calculating, and the simplified operation expression is used for the operation on one axis.
[0013]
According to a sixth aspect of the present invention, in the imaging device according to the third, fourth, or fifth aspect, the imaging device main body includes a gripping state detection sensor that detects a gripping state of the imaging device. , A preset first correction cycle is selected.
According to a seventh aspect of the present invention, in the photographing device according to the third, fourth or fifth aspect, the CPU determines a posture of the photographing device and selects a first correction cycle set in advance corresponding to the posture. It is characterized by doing.
[0014]
According to an eighth aspect of the present invention, there is provided the photographing apparatus according to the second, third, fourth, or fifth aspect, further comprising a shake amount prediction calculating means for calculating a subsequent estimated shake amount from the detected shake amount.
According to a ninth aspect of the present invention, in the photographing apparatus according to the sixth or seventh aspect, there is provided a shake amount prediction calculating means for calculating a subsequent shake amount from the detected shake amount. A second correction cycle set in advance is selected.
[0015]
According to a tenth aspect of the present invention, in the photographing apparatus according to the ninth aspect, when the magnitude relationship between the first correction cycle and the second correction cycle corresponding to the X and Y axes does not match, the two axes Or the correction operation is performed using the first correction period.
According to an eleventh aspect of the present invention, in the imaging device according to the ninth aspect, the imaging optical system has an information display function or a warning function by sound, light, or the like, and the first correction cycle and the second When the magnitude relations between the X and Y axes of the correction cycles do not match, the fact is displayed by the information display function or a warning is issued to the photographer by the warning function. .
According to a twelfth aspect of the present invention, in the photographing apparatus according to any one of the eighth to tenth aspects, a larger correction cycle of the predicted shake amount is shorter and a smaller correction cycle is longer. I do.
[0016]
According to a thirteenth aspect of the present invention, in the imaging device according to the third aspect, a shake monitoring unit that monitors a state of a shake of the imaging device is provided in the imaging device, and a change in the shake is predicted based on a detection result of the monitoring unit. A shake state prediction calculation means is provided, and a correction cycle is determined for each predetermined time based on a result of the prediction calculation.
According to a fourteenth aspect of the present invention, in the imaging device according to the thirteenth aspect, the correction cycle is determined in accordance with an absolute value of a change in shake or a ratio of a change in shake.
[0017]
According to a fifteenth aspect of the present invention, there is provided the image stabilization function according to the fourteenth aspect, further comprising setting means for allowing a photographer to determine a correspondence between the correction cycle and the absolute value of the change in the shake or the ratio of the change in the shake. An imaging device having
According to a sixteenth aspect of the present invention, in the photographing apparatus according to any one of the third to fifteenth aspects, an axis direction in which a shake is theoretically expected to be small, or an axis in which a detected shake amount or a predicted shake amount is smaller than a predetermined value. In the direction, the correction cycle is set to infinity, and the correction operation is substantially omitted.
According to a seventeenth aspect of the present invention, in the imaging apparatus according to the sixteenth aspect, the shake detection is continued at a predetermined cycle also in the axial direction where the correction operation is omitted, and the correction operation is performed when the detected shake amount becomes larger than the predetermined amount. Characterized by
[0018]
According to the eighteenth aspect of the present invention, in the photographing apparatus according to any one of the third to seventeenth aspects, a plurality of combinations of two-axis correction periods are prepared, and the photographer determines which combination is adopted. It is characterized by having such selection means.
According to a nineteenth aspect of the present invention, in the photographing apparatus according to any one of the first to eighteenth aspects, the number of correction operations is limited during exposure.
According to a twentieth aspect of the present invention, in the photographing device according to any one of the first to nineteenth aspects, a correction status is displayed.
[0019]
【Purpose】
Depending on the shooting situation, for example, the difference in the gripping state between the one-handed and the two-handed gripping state of the camera, or the difference in the attitude between the camera in the horizontal direction and the vertical direction, the X-axis A difference appears in the magnitude of the deflection around or around the Y axis.
[0020]
In an extreme case, if the shake around the Y axis is large and the shake around the X axis is negligibly small, the shake of the captured image is corrected by correcting only the shake around the Y axis.
[0021]
Alternatively, a normal correction operation can be performed around the axis with the larger shake, and the calculation formula of the correction operation can be simplified around the axis with the smaller shake to reduce the load on the CPU.
[0022]
Even in a less extreme case, more appropriate correction can be performed by making the correction frequency of the axis with the larger shake larger than the correction frequency of the other axis, that is, by shortening the correction cycle.
[0023]
In view of the above-mentioned conventional problems, the present invention reduces the load on the CPU performing the processing by changing the shake correction by the shake correction means according to the situation, particularly around the X axis and the Y axis individually. It is another object of the present invention to provide a photographing apparatus capable of improving processing performance and speed, suppressing waste of power consumption, and photographing a comfortable image with less blur due to shake.
[0024]
Alternatively, the object is to optimize the correction accuracy around both axes without increasing the load on the CPU.
Further, in the present invention, each correction cycle for shake correction around two axes is determined based on shake prediction information before exposure. Thus, the object of the present invention is to increase the effect of the shake correction as a whole without increasing the load on the system of the photographing apparatus.
[0025]
By the way, regarding the prediction of the magnitude of the shake, it is possible to predict the shake after the exposure based on the shake detection information detected by the shake detection means before the exposure.
Therefore, the magnitudes of the shakes around the two axes during the exposure are each predicted, the correction cycle around the axes predicted to have a large change in the shake is shortened to improve the correction accuracy, and the change in the shake is predicted to be small. The correction cycle around the axis is lengthened to reduce unnecessary correction accuracy.
As a result, it is possible to increase the accuracy of correction around the axis with large shake and increase the accuracy of correction around two axes without increasing the load on the system.
[0026]
【Example】
FIG. 2 shows an example of a functional configuration for realizing a basic shake correction method for the photographing apparatus 1. Reference numerals 4 and 5 indicate a shake detection sensor, and reference numeral 6 indicates an image sensor such as a CCD. The shake detection sensor 4 detects a shake around the X-axis of the imaging device, and the shake detection sensor 5 detects a shake around the Y-axis of the imaging device. These shake detection sensors 4, 5 and the shake detection sensor circuit 7 constitute a shake detection means 8. The output of the shake detecting means 8 is sent to the calculating means 9. The calculation means 9 is constituted by a microprocessor or the like, and calculates a shake correction amount according to the output from the shake detection means 8. A signal corresponding to the calculated shake correction amount is sent to the correction means drive circuit 10, whereby the shake correction means 11 and 12 are driven through the correction means drive circuit 10, and the image sensor 6 is moved in a direction to reduce the shake. Displace. The shake correcting means 11 is driven in a direction for correcting shake about the X axis, and the shake correcting means 12 is driven in a direction for correcting shake about the Y axis.
[0027]
Although hand shake caused by a human varies depending on how the photographing apparatus is held, as shown in FIG. 3, the shake in the vertical direction which is the shutter pressing direction is large, and the shake in the horizontal direction is small.
In this case, assuming that the repetition cycle of the shake correction is the same fixed cycle in both the X direction and the Y direction, in the vertical direction where the shake is large, the change amount ΔX per one cycle Δt as shown in FIG. _p Is large, the correction accuracy is low. On the other hand, in the horizontal direction where the shake is small, the change amount ΔX per unit time Δt as shown in FIG. _h Since the number of corrections is small, a phenomenon that the correction accuracy is unnecessarily high occurs.
[0028]
The shorter the repetition period of the shake correction, the higher the correction accuracy. However, the correction calculation by the calculation means 9 is not as simple as simply calculating the shake correction amount from the output data of the shake detection means 8, and corrects linear or non-linear characteristics caused by the mechanism of the shake correction means 11 and 12. There is also a demand for such calculations, and the correction calculation time tends to increase. Therefore, simply shortening the correction cycle has a large effect on the load on the imaging system, and has a limit.
Therefore, the assignment of the calculation time of the CPU is changed between the rotation around the axis that requires high accuracy and the rotation around the axis that does not need to achieve the desired correction accuracy without increasing the load on the CPU.
[0029]
In order to achieve this object, in the present invention, a simplified arithmetic expression is used for at least one of the X-axis direction and the Y-axis direction.
Normally, in the calculation of the shake correction amount, an expression for predicting the shake amount is used to compensate for the response delay of the correction means, and this expression is based on the data of the current shake amount and the data of the immediately preceding shake amount. The differential term representing the rate of change of the shake speed, that is, the term of acceleration, is added to the proportional expression for calculating the shake amount at the next point from the two points. Therefore, for the axis around which high precision is not considered necessary, the differential term of this equation is omitted, and simple calculations such as prediction using only a linear equation are performed, so that the load on the CPU is reduced and more appropriate. It is possible to perform a proper shake correction.
[0030]
Further, in order to achieve the above object, in the present invention, the correction cycle in the vertical direction and the horizontal direction is not set to a fixed cycle, but the correction cycle around the axis where the shake is large is shortened to increase the correction accuracy, and the shake is further reduced. The balance of the change amount ΔY per cycle around the X-axis and the change amount ΔX per cycle around the Y-axis (ΔXＹΔY) is reduced by lengthening the correction cycle around the small axis to suppress unnecessary correction accuracy. Take.
[0031]
Further, in order to achieve the above object, the present invention determines a correction cycle in each of the shake correction directions around two axes based on shake prediction information before exposure. Thereby, the effect of the camera shake correction as a whole is increased without increasing the load on the system of the photographing apparatus.
[0032]
Furthermore, in order to achieve the above object, according to the present invention, the correction period around the X axis and the Y axis is not set to a fixed period over the entire period during the exposure, but the exposure period is set to a predetermined unit time. Dividing into sections, shortening the correction cycle around the axis where the change in shake is large for each section to increase the correction accuracy, further increasing the correction cycle around the axis where the change in shake is small, suppressing unnecessary correction accuracy, The balance between the amount of change per cycle around the axis with large vibration and the amount of change per cycle around the axis with small vibration is balanced.
[0033]
This series of shake correction operations is repeatedly performed during exposure of the CCD 6. Generally, the repetition period is the same for both axes. According to the present invention, more CPU processing time is given to one of the two axes that requires a higher degree of correction.
As one of the methods, in the first embodiment of the present invention, the correction cycle is the same for both axes as before, but the calculation time itself is different. The above-described ordinary arithmetic expression is applied to the axis around which higher-precision correction is required, and the above-described simplified arithmetic expression is used to the other axis. As an extreme case, the correction operation can be omitted for the axis around which the correction is determined to be unnecessary.
When the first embodiment is adopted, the correction cycle determining means 14 shown in FIG. 2 can be omitted.
[0034]
As a second embodiment of the present invention, even when a normal operation formula is used for the calculations around both axes, the calculation assignment time of the CPU is weighted by making the correction periods different from each other. Instead of making the vertical and horizontal correction cycles constant as in the example shown in FIG. 4, the vertical correction cycle Δt with large shake _p To improve the correction accuracy, and furthermore, the horizontal correction period Δt in which the shake is small. _h To reduce useless correction accuracy, and one period Δt in the vertical direction. _p Change amount per hit ΔX _p And one horizontal period Δt _p Change amount per hit ΔX _p (ΔX _p ≒ ΔX _h Take).
[0035]
Which direction has a larger rate of change depends to some extent on the holding state of the imaging device. In other words, the direction of the ease of shake differs depending on whether the camera is held vertically or horizontally. In the case where the photographing apparatus is set in the horizontal direction or the vertical direction as shown in FIG. 5, a mode selecting means for switching the photographing apparatus between the horizontal direction and the vertical direction may be provided. Further, a holding state detecting unit 13 for determining in which direction the camera is held, that is, whether the camera is held vertically or horizontally may be provided.
[0036]
FIG. 6 is a diagram for explaining one embodiment in a case where the imaging device with a shake correction function of the present invention is implemented. The photographing device 15 is held at an arbitrary inclination angle θ from a state of being held horizontally.
In the present embodiment, the holding state detecting means not only determines whether the imaging device holding direction is the vertical direction or the horizontal direction, but also detects the inclination angle of the imaging device 15.
Even in such a case, the correction cycle determination means 14 uses the tilt angle of the imaging device obtained from the holding state detection means 13 to calculate the X-axis rotation and Y-axis from the correction cycle table corresponding to each tilt angle inherent to the device. Determine the correction cycle around the axis.
[0037]
The manner in which the correction periods around both axes are made different will be described below using a modeled example.
The concept of the system load and the correction cycle will be described with reference to FIG. For example, the first example of FIG. 7A in which the ratio of the correction periods of the two correction units around the X axis and the Y axis is the same as 9: 9, and the ratio of the correction period of the two correction units is 6: 9. Consider the second example of FIG. 7B in which the correction cycle around the X-axis where the change in shake is large is 18 and the correction cycle around the Y-axis where the shake is small is long.
In FIG. 7, the horizontal axis is the time axis, and the cells X, Y, and E are each used for the shake correction processing time around the X axis, the shake correction processing time around the Y axis, and the system side during exposure with respect to the horizontal axis. Time.
[0038]
The shake correction processing time around the X-axis and the shake correction processing time around the Y-axis are the same as the shake correction processing contents except for the direction when the simple calculation formula is not used. The length in the direction is the same, and for convenience of explanation, the cell E has the same length as the cells X and Y, and this length is expressed as a unit time. The number described in the cell is the number of times each shake correction process has been repeated. Note that when the simplified arithmetic expression is used, the widths of the cell X and the cell Y are not equal.
[0039]
Here, within a certain fixed time T during exposure is considered. In the case of the first example in which the ratio of the correction periods of the two correction means for the X-axis and Y-axis shakes is 9: 9, the number of shake correction processes within the time T is two for the X axis and Y for the shake. The correction around the axis is performed twice, and the time that the system can use is 14 unit time. Further, in the case of the second example in which the ratio of the correction periods of the two correction means for the X-axis rotation and the Y-axis rotation is 6:18, the number of the vibration correction processing within the time T is three for the X-axis correction. , The correction about the Y axis is performed once, and the vibration correction precision about the X axis is 1.5 times and the vibration correction precision about the Y axis is 0.5 times as compared with the first example. It can be seen that the time available for the system is not different from 14 unit time.
[0040]
By using the simplified calculation formula for the calculation with the longer correction cycle, the difference in weighting around both axes can be further emphasized. Conversely, if the correction cycle is very short, the amount of shake generated per correction operation is considered to be considerably small, so that each calculation does not require much high accuracy. Therefore, it is possible to reduce the load on the CPU by applying the simplified arithmetic expression to the shorter correction cycle.
As described above, by appropriately determining the correction period of the two correction units in accordance with the magnitude of the shake around the X axis and the Y axis, the necessary correction accuracy can be obtained without increasing the load on the system side. be able to.
[0041]
FIG. 8 shows an example of a still camera capable of detecting a grip state. As shown in the figure, a grip portion of the camera body 101 has a grip state detection sensor 102 for determining that the photographer is gripping the camera. The other side of the main body also has a grip state detection sensor 103. These grip state detection sensors may be sensors that react by being shielded from light, such as a touch sensor or a photo interrupter. It is determined whether the camera is gripped with both hands or one hand based on whether a finger touches the grip state detection sensor.
[0042]
When the camera is gripped by both hands, the camera is held so as to sandwich the camera from the left and right with the same force, so that the shake around the Y-axis is small and the shake around the X-axis is likely to be large. In the case of holding with one hand, since the direction in which the force is applied is one direction, the deflection direction occurs independently of X and Y. Therefore, when it is determined that the hand is held by both hands, the CPU assigns a high weight to the calculation allocation time around the X axis. In the case of one hand, weighting is performed in consideration of other conditions. Any of the above-mentioned means may be used as the weighting means.
[0043]
FIG. 9 shows a horizontal thin camera having grip state detection sensors 104 and 105. In the case of this type of camera, the shake around the X-axis tends to be larger than that of the camera of the type shown in FIG. 8, and it is considered that the shake generated around the X-axis when held with both hands is larger. Also, even when holding the camera with both hands, in the case of a single-lens reflex camera, there are several possible ways to hold the camera. Considering the action direction, it is considered that the swing around the Y axis increases. Thus, in the case of the single-lens reflex type camera as described above, as shown in FIG. 10, the holding state detection sensors 106 and 107 are provided on the bottom of the camera body and also on the lower side of the lens barrel to hold both hands. The variation is determined, and the shake correction direction is determined.
[0044]
When the camera is in the shake correction mode, the CPU detects, for example, shake around the X axis, calculates the amount of correction, and performs correction, and then detects shake around the Y axis, and calculates and corrects the amount of correction. Are alternately performed, and during that time, operations such as zoom and focus are sequentially processed. Therefore, while the shake correction is being performed, the number of processes performed by the CPU increases, which affects the processing speed of the CPU and, consequently, the responsiveness and followability of the shake correction. Therefore, in the present invention, the correction of the shake correction direction that is determined to be unnecessary is omitted by using the gripping method of two hands, one hand, and the output result of the grip state detection sensor indicating whether the grip is normal grip or bottom grip. Thus, the load on the CPU can be reduced, the power consumption can be reduced, and appropriate shake correction can be performed.
[0045]
In the shake correcting means according to the present invention, the grip state detection sensor attached to the imaging apparatus body determines whether the hand is held by both hands or one hand, and determines a correction cycle for each shake correction direction that is predetermined according to the held state. Make a decision.
[0046]
For example, when both the grip state detection sensors 102 and 103 are on in FIG. 8, the CPU executes only the correction operation around the X axis at a predetermined cycle, whereby only the shake correction means 11 is driven, and A series of operations up to data output to the shake correcting means 12 after the output of the shaft rotation detection sensor 5 is captured is omitted. Thereby, the load on the CPU 8 is reduced. The example of the sensor described here is only an example, and other combinations are conceivable.
[0047]
Even if the shake correction direction is determined around the X-axis as in the above example and the CPU 8 performs the correction operation for the shake correction means at a predetermined cycle, the CPU 8 continuously checks around the Y-axis. Good. That is, the process from the detection of the shake to the calculation of the shake amount may be performed in a predetermined cycle, for example, in the same cycle as the correction operation around the X axis. By doing so, when the shake in the direction determined to omit the processing by the previous determination of the correction direction exceeds a predetermined value, the shake correction in that direction can be started.
In this case, as for the shake in which the processing is omitted, a simplified arithmetic expression may be used for the shake correction calculation in the subsequent check.
[0048]
As shown in FIG. 2, a holding state detecting means 13 for detecting a holding state of the photographing apparatus, and a correction cycle in the X direction and the Y direction are determined based on the holding state of the photographing apparatus detected by the holding state detecting means 13. By having the correction cycle determining means 14, the correction cycle around the vertical axis can be shortened and the correction cycle around the horizontal axis can be lengthened, regardless of whether the imaging apparatus is set to be landscape or portrait. , X-axis and Y-axis. It should be noted that there is no problem if the correction cycle determining means 14 is included in the calculating means 9.
[0049]
The procedure for determining the correction cycle around the X axis and Y axis in the case of the present embodiment will be described with reference to the flowchart shown in FIG. The photographing device 15 always checks whether or not a photographing instruction signal has been generated by the photographer operating a start instruction such as a release button (S1). When the photographing instruction signal is detected (Yes in S1), the holding state of the photographing device is captured (S2). It is determined whether the photographing device is held horizontally or vertically based on the held information of the photographing device taken in S2 (S3). When it is determined that the imaging device 15 is held in the horizontal direction (Yes in S3), the direction around the X axis is the vertical direction, and A is set for the correction period around the X axis and B is set for the correction period around the Y axis. (S31). If it is determined that the photographing device 15 is held in the portrait direction (No in S3), the direction around the Y axis is the vertical direction, B is set for the correction period around the X axis, and A is set for the correction period around the Y axis. (S32). Here, the correction period A is smaller than the correction period B, and A <B. The shake correction operation is started simultaneously with the exposure according to the correction cycle around the X axis and the Y axis determined in this way (S4).
[0050]
In the capturing of the held information of the photographing device in S2, for example, as shown in FIG. 1, when the shutter pressing direction is one direction and the photographing device is always held horizontally, the held information which the photographing device has by default is read. In the case where the photographing apparatus is set in the horizontal direction or the vertical direction as shown in FIG. 5, information set by the photographer is read by the mode selecting means for switching between the horizontal direction and the vertical direction.
Further, in the case of having the holding state detecting means 13 for determining in which direction the camera is held, that is, whether the camera is held vertically or horizontally, the holding state detecting means Based on the output information, it is determined whether the photographing device is held in the horizontal direction or the vertical direction.
[0051]
A camera posture determination sensor may be provided inside the camera, and the direction of the shake correction may be determined from two factors, the posture of the camera and the gripping state of the user. FIG. 12 shows a flowchart corresponding to the first embodiment. Specifically, when the camera is held with one hand and the camera is held in a horizontally long state, the degree of touch varies depending on the user, but it is necessary to perform shake correction in two directions in the vertical and horizontal directions in consideration of the degree of force. When holding the camera in the portrait orientation with one hand, the camera supports the downward movement regardless of whether the release button is held up or down. Easy to swing around Y axis). At this time, shake correction around the Y axis is determined. In a case where the camera is held with both hands and the camera is held in the vertically long direction, the swing in the horizontal direction (in this case, around the X axis) becomes large due to the direction of the force. At this time, shake correction around the X axis is performed.
Note that, in this flowchart, the example of the first embodiment in which the simplified calculation formula is used as the weight for the CPU calculation time allocation is shown, but the simple calculation may be replaced with the omission of the correction. In addition, the second embodiment in which the correction cycle is changed may be adopted based on the same concept.
[0052]
Next, a flow of a shake prediction about two axes, determination of a correction cycle, and a correction operation according to the third embodiment of the present invention will be described with reference to FIG. The shake prediction calculation means 20 calculates shake prediction information about the two axes after exposure based on the amount of shake detection about the two axes detected by the shake detection means 8 immediately before the exposure and during a predetermined period before the exposure. On the basis of the shake prediction information about the two axes calculated by the shake prediction calculation means 20, the correction cycle determination means 14 corrects the shake displacement about the two axes orthogonal to the optical axis of the photographing apparatus. Twelve correction cycles are determined. As a result, the correcting means driving circuit 10 is controlled through the CPU of the calculating means 9, and the shake correcting means 11 and 12 repeat the correcting operation at an appropriate correcting cycle. Note that there is no problem even if the shake prediction calculation means 20 and the correction cycle determination means 14 are included in the calculation means 9.
[0053]
As shown in FIG. 14, as a shake prediction information calculation method of the shake prediction calculation means 20, there is a method based on a shake detection amount around two axes detected by the shake detection means 8 immediately before exposure. As described above, by using the latest shake detection amount immediately before the exposure, the shake after the exposure can be easily and accurately predicted.
Further, as shown in FIG. 15, there is a method based on a shake detection amount around two axes detected by the shake detection means 8 during a predetermined period before exposure. As described above, by accurately grasping the nature of the shake from the shake trajectory based on the data of a plurality of times during the predetermined period immediately before the exposure, the shake after the exposure can be predicted very accurately.
[0054]
A procedure for determining a correction cycle around the X axis and the Y axis in the case of the present embodiment will be described with reference to the flowchart shown in FIG. 17 and FIG. First, the shake detecting means 8 detects the shake state of the photographing device 15 at any time (S0), and writes (updates and stores) in the memory by the storage means (S1). Therefore, the memory holds the shake detection information for a predetermined time, that is, a predetermined number of times.
[0055]
On the other hand, the photographing device 15 always checks whether or not a photographing instruction signal is generated by a photographer's operation of starting instruction such as a release button (S2). When the photographing instruction signal is detected (Yes in S2), the amount of memory data storing the shake detection information detected by the shake detecting means 8 is confirmed (S3). In this check, if it is confirmed that a predetermined amount of data necessary for prediction is stored (Yes in S3), the shake information stored in the memory is fetched from the memory into the calculating means 9 (S4). As shown in FIG. 16A, the slope of the shake data for a predetermined period is calculated by the least square method or regression line calculation (S5), and the slope θ _x , Θ _y Or its ratio θ _x / Θ _y The correction cycle determination means 14 determines the correction cycles of the two shake correction means for correcting the shake displacement about two axes orthogonal to the optical axis of the photographing apparatus based on the above.
[0056]
Alternatively, as shown in FIG. 16B, a difference Δu between the maximum displacement and the minimum displacement of the shake data for a predetermined period. _x , Δu _y Or its ratio Δu _x / Δu _y The correction cycle determination means 14 determines the correction cycles of the two shake correction means for correcting the shake displacement about both axes based on the above. Alternatively, as shown in FIG. 16C, the shorter the period of the shake during the predetermined period is, the longer the change of the shake is the steep slope, etc. _x , Δt _y Or its ratio Δt _x / Δt _y The correction cycle determination means 14 determines the correction cycles of the two shake correction means for correcting the shake displacement about both axes based on the above. When determining the correction periods of the two shake correction means in this way, the shake correction period around the axis where the change in shake is large is short (period A in S71 and S72), and the axis around the axis where the change in shake is small is small. The shake correction cycle is set to be long (cycle B in S71 and S72). Thereafter, at the same time as the start of the exposure, the two shake correcting means start the shake correcting operation at the set cycle, respectively (S10).
The flowchart shown in FIG. 18 is a case of shake prediction (S41) based on one shake detection information about two axes detected by the shake detection means 8 immediately before exposure, and S7, S71, and S72 are the same as those in FIG. This is the same as the case of the shake prediction based on the shake detection information for a predetermined plurality of times described above.
[0057]
According to the concept shown in the above example, the shake correction cycle around the axis where the change in shake is large is shortened, and the shake correction cycle around the axis where the change in shake is small is set long. The shake correction cycle cannot be determined if the absolute value of the change in the shake is the same or the ratio alone is used, and the absolute values of the change in the shake around the two axes are almost the same. As described above, in order to actually determine the shake correction cycle, in addition to the magnitude of the change in the shake around the two axes and the ratio thereof, it is necessary to consider several factors such as balance so as not to exceed the limit of the system load. The effective determination means are combined to appropriately determine the correction cycle in each of the two-axis shake correction directions.
[0058]
In the flowchart shown in FIG. 19, when the difference in the magnitude of the shake change based on the shake prediction information about the two axes calculated from the shake prediction calculation means 20 is equal to or smaller than the shake change difference reference value (S6), the system load is reduced. Considering the limit, the shake correction cycle around the X axis and the Y axis is set to the same cycle C (cycle A <cycle C <cycle B) (S61).
[0059]
As described above, the camera shake caused by a human varies depending on how the photographing apparatus is held, but when the shutter 25 is pressed while the photographing apparatus is held in the landscape orientation as shown in FIG. Is generally large, and the change in shake around the Y axis is small. Therefore, the result predicted by the shake prediction calculation means 20 should be judged to have a large change in the shake around the X axis and a small change in the shake around the Y axis. Therefore, if the tendency of the magnitude of the change in the shake predicted by the attitude of the photographing apparatus is different from the tendency of the magnitude of the change in the shake determined by the shake prediction calculating means 20, that is, the magnitude relationship does not match. In the case, there is a problem in the holding form or the photographing posture of the photographing device of the photographer, and the prediction of the change in the shake around the X-axis and around the Y-axis due to the shake before the exposure by the shake prediction calculating means 20 is greatly deviated. There is. In this case, when determining the shake correction cycle around the X axis and the Y axis, the shake prediction information by the shake prediction calculation unit 20 is not taken into consideration, or the photographer is warned about the holding form and the shooting posture of the shooting device. By doing so, it is possible to prevent photographing with reduced shake correction accuracy from being performed.
[0060]
Many digital photographing apparatuses have display means for displaying image information or character information for displaying photographing state and other information. Also in the photographing apparatus of the present invention, if the photographing apparatus main body is provided with an image display means, the amount of shake correction, the direction of shake correction, and the like can be displayed, and the above-mentioned cautions and warnings can be displayed, which is convenient. In addition, it is also possible to configure so that a warning can be issued by light emission or sound generation.
[0061]
The photographing device 15 shown in FIG. 5 has a release button 16 when the camera is held in the horizontal direction and a release button 17 when the camera is held in the vertical direction. When the photographing device 15 is held in the horizontal direction as shown in FIG. 5A, as described above, it is general that the change in the shake around the X axis is large and the change in the shake around the Y axis is small. is there. In addition, when the photographing device 15 is held in the vertically long direction as shown in FIG. 5B, the change in the shake around the X axis is generally small, and the change in the shake around the Y axis is generally large. As described above, the tendency of the magnitude of the fluctuation of the imaging device 15 about the X axis and the Y axis varies depending on how the imaging device is held.
[0062]
Therefore, the photographing device has a mode setting unit that allows the photographer to set the photographing device holding state in order to determine whether the photographing device holding direction is the portrait direction or the landscape direction, or as shown in FIG. A holding state detecting unit for detecting a holding state of the photographing apparatus, and by constantly monitoring the holding state of the photographing apparatus, a tendency of a magnitude of a change in shake predicted according to how the photographing apparatus is held; It is possible to determine whether or not the tendency of the magnitude of the change in the shake determined in step 20 matches or does not match.
[0063]
In the flowchart shown in FIG. 20, the correction direction-based cycle determination determined based on the magnitude of the change in the shake determined by the shake prediction calculation unit 20 and the correction direction-based cycle determination determined from the imaging apparatus holding information are different, When the magnitude relationship is reversed (S9), the shake correction cycle around the X axis and the Y axis is set to the same cycle C (cycle A <cycle C <cycle B) in consideration of the system load limit (S91). ). However, the processing of S91 is not limited to this, and for example, a cycle determined from the photographing apparatus holding information may be adopted.
[0064]
Further, in the flowchart shown in FIG. 21, the case where the cycle determination for each correction direction determined based on the magnitude of the change in the shake determined by the shake prediction calculation unit 20 is different from the cycle determination for each correction direction determined from the image capturing apparatus holding information ( S9), the photographer is warned about the holding state of the photographing device or the photographing posture (S92).
The capture of the held information of the photographing apparatus in S8 is the same as in the case of FIG. 8 described above.
[0065]
An example of a functional configuration according to the fourth embodiment of the present invention will be described with reference to FIG. The shake monitoring sensor 22 detects a shake of the imaging device around the X axis, and the shake monitoring sensor 23 detects a shake of the imaging device around the Y axis. The shake monitoring sensors 24 and 23 and the shake monitoring sensor circuit 21 constitute a shake monitoring unit 24.
The shake prediction calculating means 20 calculates shake prediction information based on the amount of shake detection around two axes detected by the shake monitoring means 24 during a predetermined period during exposure. On the basis of the shake prediction information about the two axes calculated by the shake prediction calculation means 20, the correction cycle determination means 14 corrects the shake displacement about the two axes orthogonal to the optical axis of the photographing apparatus. Twelve correction cycles are determined. As a result, the correcting means driving circuit 10 is controlled through the CPU of the calculating means 9, and the shake correcting means 11 and 12 repeat the correcting operation at an appropriate correcting cycle.
[0066]
In the present embodiment, the exposure period is divided into several sections by a predetermined unit time, and the correction cycle is changed for each section. Note that there is no problem even if the shake monitoring means 24 is also used as the shake detection means 8. Further, there is no problem even if the prediction calculation means 20 and the correction cycle determination means are included in the calculation means 9.
[0067]
Instead of making the correction cycle around the X axis and the Y axis a fixed cycle over the entire period during exposure, as shown in FIG. 23, the exposure period is divided into several sections by a predetermined unit time, The correction period Δt around the axis where the fluctuation of the shake is large for each section ₁ To improve the correction accuracy, and furthermore, the correction period Δt around the axis where the change of the shake is small. ₂ To reduce unnecessary correction accuracy, and one cycle Δt around the axis with large shake. ₁ Change Δu per hit ₁ And one cycle Δt around the axis with small runout ₂ Change Δu per hit ₂ (Δu ₁ ≒ Δu ₂ Take).
[0068]
A correction cycle of the two shake correction means 11 and 12 for correcting shake displacement around two axes orthogonal to the optical axis of the photographing device determined by the correction cycle determination means 14 and a rotation about two axes detected by the shake monitoring means 24 Will be described with reference to FIG.
Note that the detection of the shake by the shake detecting means 8 is performed for each shake correction operation, so that the shake detection cycle changes for each change section of the correction cycle. It is desirable that the monitoring of the shake due to is performed at a constant cycle. Of course, the detection of the shake by the shake detecting means 8 and the monitoring of the shake by the shake monitoring means 24 detect the same shake, and the shake transition of the shake detection result is the same as a whole. 8 and the shake monitoring means 24 are the same, and there is no problem if the shake monitoring is performed using the shake detection data for each shake correction operation. Rather, considering a slight increase in system load due to vibration monitoring, it is desirable to use the vibration detection means 8 and the vibration monitoring means 24 as the same and use the vibration detection data for each vibration correction operation as the vibration monitoring data.
[0069]
FIG. 24 shows the transition of the shake around the X-axis and around the Y-axis in any four continuous sections a, b, c and d during exposure. Section a is a section in which the change in the shake around the X-axis and around the Y-axis are both small, and the correction cycle is set long both around the X-axis and around the Y-axis. The section b is a section in which the change in the shake around the X axis is small and the change in the shake around the Y axis is large, and the correction cycle around the X axis is set long and the correction cycle around the Y axis is set short. In the section c, the change in the shake around the X axis is large, and the change in the shake around the Y axis is small. The correction cycle around the X axis is set short, and the correction cycle around the Y axis is set long. The section d is a section in which the change in shake around the X axis and around the Y axis is large, and the correction cycle is set short both around the X axis and around the Y axis.
[0070]
The relationship between the determination of the correction cycle around the X axis and the Y axis in these sections a to d and the system load will be described with reference to FIG. The horizontal axis and the definition of each symbol are the same as those in FIG. Further, in this example, it is assumed that the system load needs to be available for at least 14 unit times within a certain time T as the limit of the system load.
[0071]
Consider a certain time T in the section a. The ratio of the correction periods of the two correction means for the X-axis and Y-axis shakes is 18:18, and the number of shake correction processes within the time T is one for the X-axis and one for the Y-axis. Is once, and the time that the system can use is 16 unit time.
Consider a certain time T in the section b. The ratio of the correction periods of the two correction units for the X-axis and Y-axis shakes is 18: 6, and the number of shake correction processes within the time T is one for the X-axis and one for the Y-axis. Is three times, and the time that the system can use is 14 unit time.
Consider a certain time T in the section c. The ratio of the correction periods of the two correction units for the X-axis and Y-axis shakes is 6:18, and the number of shake correction processes within the time T is three for the X-axis and three for the Y-axis. Is one time, and the time that the system can use is 14 unit time.
Consider a period d within a fixed time T. The ratio of the correction periods of the two correction units for the X-axis rotation and the Y-axis rotation is 9: 9, and the number of vibration correction processes within the time T is two for the X-axis and two for the Y-axis. Is twice, and the time that the system can use is 14 unit time.
[0072]
In a section d in which the change in the shake around the X-axis and the shake around the Y-axis are both large, the number of times of the shake correction processing within the fixed time T is twice for the shake correction about the X-axis and twice for the Y-axis. The time that can be used by the system is set at 14 unit time, which is the system load limit.
In the section b, the number of times of shake correction processing within the fixed time T is one for the X axis shake correction and three for the Y axis correction, and the shake correction accuracy around the X axis is 0.5 compared to the section d. And the shake correction accuracy around the Y axis is 1.5 times. Further, in the section c, the number of times of the shake correction processing within the fixed time T is three times for the shake correction about the X axis and one for the Y axis, and the accuracy of the shake correction about the X axis is 1 compared to the section d. 0.5 times, and the accuracy of shake correction around the Y axis is 0.5 times. As described above, in the sections b and c, the shake correction accuracy around the X axis and the Y axis is different from that of the section d. However, even in this case, the time that can be used by the system side is 14 unit times. The same system load limit is set.
[0073]
Further, in order to realize the control described with reference to FIGS. 24 and 25, by limiting the number of times of executing the shake correction around the two axes, the correction cycle is determined in consideration of the system load. In this way, by limiting the number of times of performing the shake correction about the two axes in accordance with the magnitude of the change of the shake about the X axis and the Y axis in each section, the system load is reduced for each section. The correction cycle of the two correction means for the shake about the X axis and the shake about the Y axis is appropriately determined so as to be less than the limit. This makes it possible to obtain necessary correction accuracy around the X axis and around the Y axis without increasing the load on the system side during the exposure period.
[0074]
By the way, the shake detection amount data detected by the shake monitoring means 24 for a predetermined period necessary for the prediction is stored in, for example, a memory. Therefore, the prediction calculation unit 20 performs the prediction calculation using the shake detection amount data of the predetermined period immediately before the corresponding section, which is stored in the memory or the like. As this prediction calculation method, for example, primary and secondary differentiation of a series of shake detection amount data for a predetermined period is performed to calculate a shake speed and a shake acceleration immediately before the relevant section, and the magnitude of the change in the shake in the relevant section is calculated. For example, there is a method of calculating the degree.
[0075]
Based on the magnitude of the change in the shake around the two axes in the corresponding section predicted based on the concept shown in the above example, the shake displacement around the two axes is calculated based on the absolute value of the change in the shake around the two axes. The correction cycle of the two shake correction means to be corrected is determined, and the correction cycle of the two shake correction means for correcting the shake displacement about the two axes is determined based on the ratio of the magnitude of the change of the shake about the two axes. be able to. In addition to this, by combining some effective judgment means such as considering a balance that does not exceed the limit of the system load, the correction cycle in each of the shake correction directions around the two axes can be appropriately set for each section. decide.
The combination of the correction cycle corresponding to the absolute value of the magnitude of the change in the shake or the ratio of the magnitude of the change in the shake may be fixed. Selection means may be provided so that the user can make a selection.
[0076]
The procedure for determining the correction cycle around the X-axis and Y-axis in the case of the present embodiment will be described based on the flowcharts shown in FIGS.
First, FIG. 26 illustrates the pre-exposure operation. The shake monitoring means 24 detects the shake state of the photographing device 15 at any time (S0), and writes (updates and stores) in the memory by the storage means (update). S1). Therefore, the memory holds the shake monitoring information for a predetermined time, that is, for a plurality of predetermined times.
[0077]
On the other hand, the photographing device 15 always checks whether or not a photographing instruction signal is generated by a photographer's operation of starting instruction such as a release button (S2). When the photographing instruction signal is detected (Yes in S2), it is confirmed that the shake monitoring data detected by the shake monitoring unit 24 is stored (Yes in S3), and the shake information stored in the memory is obtained. From the memory to the calculating means 9 (S4), for example, the first and second derivatives of a series of shake detection amount data for a predetermined period are calculated to calculate a shake speed and a shake acceleration. The size is predicted (S5). On the basis of the predicted shake information calculated by the processing in step 5, the correction cycle determining means 14 corrects the shake displacement around two axes orthogonal to the optical axis of the photographing apparatus in the section immediately after the exposure. The correction cycle is determined, and at the same time as the exposure starts, the shake correction is started at the correction cycles determined around the X axis and the Y axis, respectively (S10).
[0078]
FIG. 27 explains the shake correction operation during exposure. The output from the shake detection means 8 is taken in (S11), and the shake correction amount is calculated according to this output (S12). Then, a signal corresponding to the calculated shake correction amount is sent to the correction unit drive circuit 10, and the shake correction units 11 and 12 are driven through the correction unit drive circuit 10 (S13). This series of shake correction operations is repeatedly performed during exposure at different periods around the X axis and around the Y axis (S19).
[0079]
Incidentally, the shake detection amount data (S0) detected by the shake monitoring means 24 is stored (updated) in a storage means such as a memory at any time as in the pre-exposure operation (S1). Further, the microprocessor of the calculating means 9 judges the transition between the sections corresponding to the section a to the section d illustrated in FIG. 24 (S15), and if it is the transition of the section, immediately before the next section stored in the memory. (S4), and a prediction calculation is performed by the prediction calculation means 18 using the data (S5). Based on the predicted shake information calculated in the process of S5, the correction cycle determination unit 14 corrects the shake displacement around the two axes orthogonal to the optical axis of the imaging device in the section immediately after the exposure by the two shake correction units. The cycle is determined, and the shake correction is started at the correction cycle determined around the X axis and around the Y axis simultaneously with the start of the next section (Yes in S19). In determining the correction cycle around the X axis and the Y axis for each section, the correction cycle is determined in consideration of not only the shake prediction information but also the system load. It is desirable that the start of the shake correction operation around the X axis and the start of the shake correction operation around the Y axis is the same timing as the Q timing.
[0080]
As described above, in the present embodiment, the case of the shake correction method in which the shake correction unit drives the image sensor in the direction opposite to the shake has been described, but the case where the shake correction unit is a correction lens included in the shooting optical system, The same can be said for a variable vertex angle prism provided on the optical axis of the system.
[0081]
【The invention's effect】
According to a first aspect of the present invention, in the photographing apparatus having the shake correcting function, the CPU assigns a weight to the calculation allocation time required for the correction processing of the shake correcting means around two axes orthogonal to the optical axis of the photographing apparatus. Since the determination is made, it is possible to perform shake correction with appropriate correction accuracy in accordance with shake characteristics around each axis without increasing the load on the CPU. Therefore, the shake correction effect of the captured image can be improved as a whole.
[0082]
According to a second aspect of the present invention, in the photographing apparatus according to the first aspect, a normal arithmetic expression and a simplified arithmetic expression are prepared as arithmetic expressions for calculating the correction amounts given to the two shake correcting means by the CPU. By using a normal calculation formula for the correction means and using a simplified calculation formula for the other correction means, weighting of the CPU calculation allocation time is performed. The shake correction can be performed with an appropriate correction accuracy in accordance with the shake characteristic around the axis. Therefore, the shake correction effect of the captured image can be improved as a whole.
[0083]
According to a third aspect of the present invention, in the photographing apparatus according to the first aspect, the correction periods of the two shake correcting units for correcting shake displacement around two axes orthogonal to the optical axis of the image capturing apparatus are individually set. As a result, the CPU assigns a weight to the calculation assignment time, so that the shake can be corrected with appropriate correction accuracy in accordance with the shake characteristics around each axis without increasing the load on the CPU. Therefore, the shake correction effect of the captured image can be improved as a whole.
[0084]
According to a fourth aspect of the present invention, in the photographing apparatus according to the third aspect, the correction period of the correction means in the direction in which the shake largely swings among the shake directions around two axes orthogonal to the optical axis of the image pickup apparatus is set. The correction cycle is shortened and the correction cycle of the correction means in the direction in which the shake is small is lengthened, thereby suppressing unnecessary correction accuracy. This means, for example, that the surplus time obtained by suppressing the correction accuracy in the horizontal direction is used for the correction time in the vertical direction. The correction effect can be improved.
[0085]
According to a fifth aspect of the present invention, in the photographing apparatus according to the third or fourth aspect, in the correction amount calculation around the two axes having different correction periods, the correction amount given to the two shake correction means by the CPU is used. A normal operation expression and a simplified operation expression are prepared as the operation expression for the operation, and the simplified operation expression is used for the operation on one axis, so that the load on each axis is not increased without increasing the load on the CPU. Shake correction can be performed with appropriate correction accuracy in accordance with shake characteristics. Therefore, the shake correction effect of the captured image can be improved as a whole.
[0086]
According to a sixth aspect of the present invention, in the imaging device according to the third, fourth, or fifth aspect, the imaging device having a shake correction function includes a grip state detection sensor of the imaging device, and based on a detection result, By having the correction cycle determining means for determining the first correction cycle of each of the two shake correction means for correcting shake displacements about two axes orthogonal to the optical axis of the photographing apparatus, it is possible to adjust the shake characteristics around each axis. In addition, shake correction can be performed with appropriate correction accuracy. Therefore, the shake correction effect of the captured image can be improved as a whole.
[0087]
According to a seventh aspect of the present invention, in the photographing apparatus according to the third, fourth or fifth aspect, the posture of the photographing apparatus, that is, whether the holding state is a vertically long direction or a horizontally long direction is determined, and the photographing apparatus is orthogonal to the optical axis of the photographing apparatus. By having the correction cycle determining means for determining the first correction cycle of each of the two shake correction means for correcting the shake displacement around the two axes to be performed, the correction accuracy can be adjusted appropriately in accordance with the shake characteristics around each axis. Can be corrected. Therefore, the shake correction effect of the captured image can be improved as a whole.
[0088]
According to an eighth aspect of the present invention, in the photographing apparatus according to the second, third, fourth, or fifth aspect, based on the amount of shake detection around the two axes detected by the shake detection means before the exposure, the rotation around the two axes after the exposure is performed. A shake prediction calculating means for calculating a predicted shake amount, and two shakes for correcting a shake displacement about two axes orthogonal to the optical axis of the photographing apparatus based on the predicted shake amount about the two axes calculated by the shake prediction calculation means. A correction cycle determining means for determining a correction cycle of the correcting means, whereby a magnitude of a change in a shake in a shake direction around two axes after exposure is determined based on a predicted shake amount detected by a shake detection means before exposure. The correction cycle around the axis where the change in vibration is large is shortened to increase the correction accuracy, the correction cycle around the axis where the change in the vibration is small is long, the unnecessary correction accuracy is suppressed, and the load on the shooting system is reduced. Without increasing It is possible to improve the shake correction effect of the captured image as a body.
[0089]
According to a ninth aspect of the present invention, in the photographing apparatus according to the sixth or seventh aspect, a predicted shake amount around the two axes after the exposure is based on a shake detection amount around the two axes detected by the shake detection unit before the exposure. And two shake correction means for correcting a shake displacement about two axes orthogonal to the optical axis of the photographing apparatus based on the predicted shake amount about the two axes calculated by the shake prediction calculation means. A correction cycle determining means for determining each of the second correction cycles, so that the magnitude of the change in the shake in the shake directions around the two axes after the exposure is determined based on the predicted shake amount detected by the shake detection means before the exposure. The correction cycle around the axis where the change in vibration is large is shortened to increase the correction accuracy, the correction cycle around the axis where the change in the vibration is small is long, the unnecessary correction accuracy is suppressed, and the load on the shooting system is reduced. Without increasing It is possible to improve the shake correction effect of the captured image as a body.
[0090]
According to a tenth aspect of the present invention, in the photographing apparatus of the ninth aspect, when the magnitude relations of the first correction cycle and the second correction cycle corresponding to the X and Y axes do not match, two-axis correction is performed. Equalize the cycles or use the first correction cycle. Further, such a situation is displayed on the photographing device or a warning is given to the photographer by sound or light, so that photographing in a state where the shake correction accuracy is reduced can be prevented.
[0091]
According to a twelfth aspect of the present invention, in the photographing device according to any one of the eighth to tenth aspects, the correction cycle determining unit calculates a correction shake amount around two axes calculated by the shake prediction calculation unit. Of the correction cycles of the two shake correction means for correcting shake displacement around two axes orthogonal to the optical axis of the photographing apparatus, the correction cycle around the axis with a large change in shake is shortened, and the correction cycle around the axis with a small change in shake is reduced. The correction cycle around the axis predicted to have a large change in vibration is increased by changing the correction cycle long, and the correction cycle around the axis predicted to be small when the change in vibration is small is increased to reduce unnecessary correction precision. By suppressing this, the correction accuracy in the direction of large shake can be increased without increasing the load on the system, and the correction accuracy around the two axes can be made uniform, so that the effect of correcting the camera shake of the captured image as a whole can be improved. It can be.
[0092]
According to a thirteenth aspect of the present invention, in the photographing apparatus according to the third aspect, a shake monitoring means for monitoring a shake state around two axes orthogonal to an optical axis of the imaging apparatus during exposure, and the shake monitoring means detects the shake state. A shake prediction calculating means for calculating a predicted shake amount around the two axes after exposure based on the detected shake amount around the two axes, and a predicted shake amount around the two axes calculated by the shake prediction calculating means. Correction cycle determining means for changing a correction cycle of each of the two shake correction means for correcting shake displacements about two axes orthogonal to the optical axis of the photographing apparatus at predetermined intervals, so that the X axis and the Y axis can be changed. Obtaining the necessary correction accuracy around the X-axis and around the Y-axis by appropriately determining the correction period of the two correction means around the X-axis and around the Y-axis according to the magnitude of the fluctuation of the surrounding vibration An imaging device capable of performing the above operation can be provided.
[0093]
According to a fourteenth aspect of the present invention, in the photographing apparatus according to the thirteenth aspect, the light of the photographing apparatus is determined by the correction cycle determining means based on shake prediction information about two axes calculated by the shake prediction calculation means. The correction periods of the two shake correcting means for correcting the shake displacement about two axes perpendicular to the axis are changed based on the absolute value of the magnitude of the change of the shake about the two axes based on the shake prediction information about the two axes. This makes it easy and accurate to determine the magnitude of the change in the biaxial direction by making the change based on the ratio of the magnitude of the change in the biaxial direction based on the shake prediction information about the two axes. It is possible to increase the correction accuracy around the axis predicted to have a large change in runout, and to prolong the correction cycle around the axis predicted to have a small change in the runout, thereby suppressing unnecessary correction accuracy. ,system Without increasing the load, since the direction of larger correction accuracy of shake like, also may be uniform around two axes of correction accuracy, thereby improving the image stabilization effect of overall photographic image.
[0094]
According to a fifteenth aspect of the present invention, in the imaging device according to the fourteenth aspect, the imaging device is configured such that an absolute value of a magnitude of a change in shake due to a predicted shake amount around two axes calculated by the shake prediction calculation means. The relationship between the ratio of the magnitude of the change in shake due to the predicted shake amount about the two axes and the correction cycle of the two shake correction means for correcting the shake displacement about the two axes determined by the correction cycle determination means. Is set by the photographer, the size of the change in the shake in the biaxial directions can be determined more finely.
[0095]
According to a sixteenth aspect of the present invention, in the photographing apparatus according to any one of the third to fifteenth aspects, an axis direction in which a shake is theoretically expected to be small, or an axis in which a detected shake amount or a predicted shake amount is smaller than a predetermined value. Since the direction shake correction is omitted, the processing amount of the CPU is reduced, the processing speed is improved, and the power consumption is further reduced.
According to a seventeenth aspect of the present invention, in the imaging device according to the sixteenth aspect, the amount of axial shake in which the shake correction operation is omitted is continuously measured, and correction is performed as necessary. You can shoot clear and easy
[0096]
According to an eighteenth aspect of the present invention, in the photographing apparatus according to any one of the third to seventeenth aspects, a plurality of combinations of two-axis correction periods are prepared, and the photographer determines which combination is to be adopted. Is provided, the photographer can select an optimal combination according to the actual photographing situation, and more accurate correction can be performed.
[0097]
According to a nineteenth aspect of the present invention, in the photographing apparatus according to any one of the first to eighteenth aspects, two shake correcting means for correcting shake displacement around two axes orthogonal to the optical axis of the photographing apparatus during exposure. The number of times of executing the shake correction around the two axes by the above is limited, so that the correction period of the two correction means of the shake around the X axis and the shake around the Y axis is appropriately set so that the system load is equal to or less than the system load limit. It is possible to obtain the necessary correction accuracy around the X axis and around the Y axis without increasing the load on the system side during the exposure period, thereby improving the effect of correcting the camera shake of the captured image as a whole. be able to.
[0098]
According to a twentieth aspect of the present invention, in the photographing apparatus according to any one of the first to nineteenth aspects, the state of correction is displayed. You can know if it has been done and use it as a reference for subsequent shooting.
[Brief description of the drawings]
FIG. 1 is a schematic perspective view of a photographing apparatus for describing an embodiment of the present invention.
FIG. 2 is a block diagram illustrating a functional configuration example according to a second embodiment of the present invention.
FIG. 3 is a diagram showing the relationship between a conventional two-way camera shake displacement and a correction cycle.
FIG. 4 is a diagram illustrating a relationship between a camera shake displacement in two directions and a correction cycle according to the present invention.
FIG. 5 is a schematic perspective view showing a state in which the imaging device is held in the horizontal direction and the vertical direction.
FIG. 6 is a schematic perspective view of a photographing device for describing an embodiment of the present invention.
FIG. 7 is a diagram for explaining a relationship between a system load and a correction cycle according to the present invention.
FIG. 8 shows a still camera of a general shape having a grip state detection sensor.
FIG. 9 shows a thin horizontal camera having a grip state detection sensor.
FIG. 10 shows a camera provided with a grip state detection sensor on the bottom surface and also below the lens barrel.
FIG. 11 is a schematic flowchart showing a correction cycle determination procedure according to the second embodiment of the present invention.
FIG. 12 is a flowchart in which a posture and a gripping state of the photographing device are considered in the first embodiment of the present invention.
FIG. 13 is a block diagram illustrating a functional configuration example according to a third embodiment of the present invention.
FIG. 14 is a diagram for explaining a third embodiment of the present invention.
FIG. 15 is a diagram for explaining a third embodiment of the present invention.
FIG. 16 is a diagram for explaining a third embodiment of the present invention.
FIG. 17 is a schematic flowchart showing a correction cycle determination procedure according to the third embodiment of the present invention.
FIG. 18 is a schematic flowchart showing a correction cycle determination procedure according to the third embodiment of the present invention.
FIG. 19 is a schematic flowchart showing a correction cycle determination procedure according to the third embodiment of the present invention.
FIG. 20 is a schematic flowchart showing a correction cycle determination procedure according to the third embodiment of the present invention.
FIG. 21 is a schematic flowchart showing a correction cycle determination procedure according to the third embodiment of the present invention.
FIG. 22 is a block diagram illustrating a functional configuration example according to a fourth embodiment of the present invention.
FIG. 23 is a diagram illustrating a relationship between a temporal change of a shake amount and a correction cycle according to the present invention.
FIG. 24 is a diagram for explaining a fourth embodiment of the present invention.
FIG. 25 is a diagram for explaining a relationship between a system load and a correction cycle according to the present invention.
FIG. 26 is a schematic flowchart showing a correction cycle determination procedure of the embodiment of the table 4 of the present invention.
FIG. 27 is a schematic flowchart showing a correction cycle determination procedure of the embodiment of the table 4 of the present invention.
[Explanation of symbols]
1 Photographing device
2 Shooting lens
3 Shooting optical system
4 Runout detection sensor (around X axis)
5 Runout detection sensor (around Y axis)
6 Image sensor
7 Runout detection sensor circuit
8 Shake detection means
9 Calculation means
10 Correction means drive circuit
11 Shake correction means (around X axis)
12 Shake correction means (around Y axis)
13 Holding state detecting means
14 Correction cycle determination means
15 Imaging equipment
16 Release button (when held horizontally)
17 Release button (for vertical orientation)
20 Shake prediction calculation means
21 Run-out monitoring sensor circuit
22 Runout monitoring sensor (around X axis)
23 Runout monitoring sensor (Y axis)
24 Runout monitoring means
25 Release button
101 Camera
102, 103, 104, 105, 106, 107 gripping state detection sensor

Claims

An imaging optical system, imaging means for receiving light from a subject passing through the imaging optical system, shake detection means for detecting shake around two axes orthogonal to the optical axis of the imaging optical system, and shake detection A photographing apparatus comprising: a CPU for performing a predetermined process by receiving a detected shake amount detected by the means; and two shake correcting means for correcting a shake displacement around the two axes based on a processing result of the CPU. A photographing apparatus having a shake correction function, wherein a weight is assigned to a calculation time assigned by a CPU required for processing for correction by shake correction means around two axes.

2. The photographing apparatus according to claim 1, wherein a normal calculation formula and a simplified calculation formula are prepared as calculation formulas for calculating a correction amount given to the two shake correction units by the CPU, and one of the normal calculation formulas is provided for one of the correction units. A photographing apparatus having a shake correction function, wherein the calculation assignment time of the CPU is weighted by using a simple calculation equation for the other correction means using the above calculation equation.

2. The imaging apparatus according to claim 1, wherein a weighting of a calculation assignment time of a CPU is performed by making correction cycles of the two shake correction units different from each other. 3.

4. The photographing apparatus according to claim 3, wherein a correction cycle of the larger detected shake amount is shortened and a correction cycle of the smaller detected shake amount is lengthened. 5.

5. The photographing apparatus according to claim 3, wherein in the correction amount calculation around two axes having different correction periods, a calculation expression for calculating a correction amount given to the two shake correction units by the CPU. An imaging apparatus having a shake correction function, wherein a normal operation expression and a simplified operation expression are prepared as the first and second operations, and the simplified operation expression is used for the operation on one axis.

6. The photographing apparatus according to claim 3, wherein the photographing apparatus main body includes a grasping state detection sensor for detecting a grasping state of the photographing apparatus, and is set in advance according to the detection result. An imaging device having a shake correction function, wherein a first correction cycle is selected.

6. The photographing device according to claim 3, wherein the CPU determines a posture of the photographing device, and selects a first correction cycle set in advance corresponding to the posture. An imaging device having a shake correction function.

6. The photographing apparatus according to claim 2, further comprising a shake amount predicting operation means for calculating a predicted shake amount from the detected shake amount. apparatus.

8. The photographing apparatus according to claim 6, further comprising a shake amount prediction calculation unit configured to calculate a subsequent estimated shake amount from the detected shake amount, wherein a second predetermined shake amount is set in advance corresponding to the estimated shake amount. An imaging apparatus having a shake correction function, wherein a correction cycle is selected.

10. The photographing apparatus according to claim 9, wherein when the magnitude relations of the first correction cycle and the second correction cycle corresponding to the X and Y axes do not match, whether the correction cycles of the two axes are equal. Alternatively, a photographing apparatus having a shake correction function, wherein a correction operation is performed using the first correction cycle.

The photographing apparatus according to claim 9, wherein the photographing optical system has an information display function or a warning function by sound, light, or the like, and the first correction cycle and the second correction cycle are different from each other. When the magnitude relations corresponding to the X and Y axes do not match, an indication to that effect is displayed by the information display function, or a warning is issued to the photographer by the warning function, and a photographing apparatus having a shake correction function. .

11. The imaging apparatus according to claim 8, wherein a longer correction cycle and a shorter correction cycle of the predicted shake amount are lengthened. 12. apparatus.

4. The photographing apparatus according to claim 3, further comprising: a shake monitor for monitoring a shake state of the photographing apparatus in the photographing apparatus; An image pickup apparatus having a shake correction function, wherein a correction cycle is determined for each predetermined time based on a result of a prediction calculation.

14. The photographing apparatus according to claim 13, wherein the correction cycle is determined according to an absolute value of a change in the shake or a ratio of the change in the shake.

15. The photographing apparatus according to claim 14, further comprising a setting unit for allowing a photographer to determine a correspondence between the correction cycle and an absolute value of a change in shake or a ratio of a change in shake. Shooting equipment.

16. The photographing apparatus according to claim 3, wherein the correction cycle is infinite in an axial direction in which the shake is theoretically expected to be small or an axial direction in which the detected shake amount or the predicted shake amount is smaller than a predetermined value. , And substantially eliminates the correction operation.

17. The photographing apparatus according to claim 16, wherein the shake detection is continued at a predetermined cycle also in the axial direction where the correction operation is omitted, and the correction operation is performed when the detected shake amount becomes larger than the predetermined amount. An imaging device having a shake correction function.

18. The photographing apparatus according to claim 3, wherein a plurality of combinations of two-axis correction periods are prepared, and the photographer has a selecting means for determining which combination is to be adopted. A photographing apparatus having a shake correction function.

19. The photographing apparatus according to claim 1, wherein the number of correction operations is limited during exposure.

20. The photographing apparatus according to claim 1, wherein a state of the correction is displayed.