JP4186543B2 - Encoding apparatus, encoding method, program, and recording medium - Google Patents

Description

【０１０１】
また、式（1０）において、式（１１）が成り立つ。
【数２】

【０１０２】
画像の符号化難易度が易しいものから難しいものへのシーンチェンジ、および、難しいものから易しいものへのシーンチェンジの、双方に対して仮想バッファ調整を行ってしまった場合、難しいものから易しいものへのシーンチェンジでは、エンコードに余裕があるはずの易画像においてわざわざ画質を悪くしてしまう結果となる場合がある。また、難しいものから易しいものへのシーンチェンジであっても、その変化の大きさ、あるいは、シーンチェンジ後の画像の難易度によっては、仮想バッファの調整を行うほうがよい場合がある。しかしながら、ＭＥ残差情報のみでは、シーンチェンジの有無を判定することはできるが、シーンチェンジの内容が、易しいものから難しいものへのシーンチェンジであるか、あるいは、難しいものから易しいものへのシーンチェンジであるかを判定することができない。
【０１０３】
そこで、イントラＡＣ算出部９１が、イントラＡＣを算出し、現在のイントラＡＣの値を示すmad_infoと、一つ前のイントラＡＣの値を示すprev_mad_infoとを、発生符号量制御部９２に出力することにより、発生符号量制御部９２は、シーンチェンジの状態を判定して、仮想バッファ調整を行うか否かを判断することができる。
【０１０４】
発生符号量制御部９２は、バッファ８２に格納される可変長符号化データの蓄積状態を常時監視しており、蓄積状態を表す占有量情報を基に量子化ステップサイズを決定するようになされている。
【０１０５】
また、発生符号量制御部９２は、イントラスライス部分の目標発生符号量データＴpiよりも実際に発生したマクロブロックの発生符号量データＢ（ｊ）が多い場合、発生符号量を減らすために量子化ステップサイズを大きくし、また、目標発生符号量データＴpiよりも実際の発生符号量データＢ（ｊ）が少ない場合、発生符号量を増やすために量子化ステップサイズを小さくするようになされている。
【０１０６】
更に、発生符号量制御部９２は、インタースライス部分の場合も同様に、目標発生符号量データＴppよりも実際に発生したマクロブロックの発生符号量データＢ（ｊ）が多い場合、発生符号量を減らすために量子化ステップサイズを大きくし、また、目標発生符号量データＴppよりも実際の発生符号量データＢ（ｊ）が少ない場合、発生符号量を増やすために量子化ステップサイズを小さくするようになされている。
【０１０７】
すなわち、発生符号量制御部９２は、デコーダ側に設けられたＶＢＶバッファに格納された可変長符号化データの蓄積状態の推移を想定することにより、図８に示されるように、ｊ番目のマクロブロックにおける仮想バッファのバッファ占有量ｄ（ｊ）を次の式（１２）によって表し、また、ｊ＋１番目のマクロブロックにおける仮想バッファのバッファ占有量ｄ（ｊ＋1）を次の式（１３）によって表し、（１２）式から（１３）式を減算することにより、ｊ＋１番目のマクロブロックにおける仮想バッファのバッファ占有量ｄ（ｊ＋1）を次の式（１４）として変形することができる。
【０１０８】
ｄ（ｊ）＝ｄ（０）＋Ｂ（ｊ−１）−{Ｔ×（ｊ−１）／MBcnt}・・・（１２）
【０１０９】
ここで、ｄ（０）は初期バッファ容量、Ｂ（ｊ）は、ｊ番目のマクロブロックにおける符号化発生ビット数、MBcntは、ピクチャ内のマクロブロック数、そして、Ｔは、ピクチャ単位の目標発生符号量である。
【０１１０】
ｄ（ｊ＋１）＝ｄ（０）＋Ｂ（ｉ）−（Ｔ×ｊ）／MBcnt・・・（１３）
【０１１１】
ｄ（ｊ＋１）＝ｄ（ｊ）＋{Ｂ（ｊ）−Ｂ（ｊ−１）}−Ｔ／MBcnt・・・（１４）
【０１１２】
続いて、発生符号量制御部９２は、ピクチャ内のマクロブロックがイントラスライス部分とインタースライス部分とに分かれているため、図９に示されるように、イントラスライス部分のマクロブロックとインタースライス部分の各マクロブロックに割り当てる目標発生符号量ＴpiおよびＴppをそれぞれ個別に設定する。
【０１１３】
グラフにおいて、マクロブロックのカウント数が０乃至ｓ、および、ｔ乃至endの間にあるとき、次の式（１５）に、インタースライスの目標発生符号量Ｔppを代入することにより、インタースライス部分におけるバッファ占有量ｄ（ｊ＋1）を得ることができる。
【０１１４】

【０１１５】
また、マクロブロックのカウント数がｓ乃至ｔの間にあるときに、次の式（１６）に、イントラスライスの目標発生符号量Ｔpiを代入することにより、イントラスライス部分におけるバッファ占有量ｄ（ｊ＋1）を得ることができる。
【０１１６】
ｄ（ｊ＋１）＝ｄ（ｊ）＋{Ｂ（ｊ）−Ｂ（ｊ−１）}−Ｔpi／（ｔ−ｓ）・・・（１６）
【０１１７】
従って、発生符号量制御部９２は、イントラスライス部分およびインタースライス部分におけるバッファ占有量ｄ（ｊ＋１）、および、式（１７）に示される定数ｒを、式（１８）に代入することにより、マクロブロック（ｊ＋１）の量子化インデックスデータＱ（ｊ＋１）を算出し、これを量子化部７５に供給する。
【０１１８】
ｒ＝（２×ｂｒ）／ｐｒ ・・・（１７）
Ｑ（ｊ＋１）＝ｄ（ｊ＋１）×（３１／ｒ） ・・・（１８）
ここで、ｂｒは、ビットレートであり、ｐｒは、ピクチャレートである。
【０１１９】
更に、発生符号量制御部９２は、シーンチェンジがあった場合、図１１乃至図１４のフローチャートを用いて後述する、Ｑマトリクス設定処理１乃至Ｑマトリクス設定処理４のうちのいずれかの処理により、必要に応じて、量子化部７５が量子化に用いるＱマトリクスの値を制御する。
【０１２０】
量子化部７５は、量子化インデックスデータＱ（ｊ＋１）、および、発生符号量制御部９２の処理により制御される値であるＱマトリクスに基づいて、次のマクロブロックにおけるイントラスライスまたはインタースライスに応じた量子化ステップサイズを決定し、量子化ステップサイズによってＤＣＴ係数データを量子化する。
【０１２１】
これにより、量子化部７５は、１つ前のピクチャのイントラスライス部分およびインタースライス部分における実際の発生符号量データＢ（ｊ）に基づいて算出された、次のピクチャのイントラスライス部分およびインタースライス部分における目標発生符号量ＴppおよびＴpiにとって最適な量子化ステップサイズによって、ＤＣＴ係数データを量子化することができる。
【０１２２】
かくして、量子化部７５では、バッファ８２のデータ占有量に応じて、バッファ８２がオーバーフローまたはアンダーフローしないように量子化し得るとともに、デコーダ側のＶＢＶバッファがオーバーフロー、またはアンダーフローしないように量子化した量子化ＤＣＴ係数データを生成することができる。
【０１２３】
例えば、従来の技術として上述した、特願平１１−２０５８０３では、通常のフィードバック型の量子化制御を行いながら、次に符号化処理するピクチャの絵柄が大きく変化する場合には、フィードバック型の量子化制御を止め、ＭＥ残差算出部９５から供給されるＭＥ残差情報に基づいて、仮想バッファの初期バッファ容量ｄ（０）を初期化し、新たな初期バッファ容量ｄ（０）を基に、イントラスライスおよびインタースライスごとに量子化インデックスデータＱ（ｊ＋１）を新たに算出するようになされている。
【０１２４】
しかしながら、従来における場合のように、ＭＥ残差のみで仮想バッファ調整を行うか否かを判定してしまうと、画像難易度が易しいものから難しいものに変わった場合、および難しいものから簡単なものに変わった場合の双方に対して、仮想バッファ調整を行ってしまう。すなわち、画像難易度が難しいものから簡単なものに変わった場合では、エンコードに余裕があるはずの簡単な画像において、わざわざ画質を悪くしてしまう結果となる。
【０１２５】
そこで、図７のビデオエンコーダ６１においては、例えば、イントラＡＣ算出部９１によって算出されるイントラＡＣなどの情報を用いて、画像難易度が易しいものから難しいものに変わるシーンチェンジの時にのみ、仮想バッファ調整を行うようにすることにより、簡単な画像での画質の劣化を防ぐようにすることができる。
【０１２６】
すなわち、発生符号量制御部９２は、通常のフィードバック型の量子化制御を行いながら、次に符号化処理するピクチャの絵柄が大きく変化する場合には、フィードバック型の量子化制御を止め、ＭＥ残差算出部９５から供給されるＭＥ残差情報ME_info、並びに、イントラＡＣ算出部９１から供給される、prev_mad_infoおよびmad_infoを基に、仮想バッファの初期バッファ容量ｄ（０）を初期化するか否かを判断し、仮想バッファの初期バッファ容量ｄ（０）を初期化する場合は、ＭＥ残差算出部９５から供給されるＭＥ残差情報ME_infoに基づいて、仮想バッファの初期バッファ容量ｄ（０）を初期化する。仮想バッファの初期バッファ容量ｄ（０）の初期化については、式（２）乃至式（５）を用いて説明した従来における場合と同様である。
【０１２７】
そして、発生符号量制御部９２は、新たな初期バッファ容量ｄ（０）を基に、イントラスライスおよびインタースライスごとに、式（１２）乃至式（１８）を用いて、量子化インデックスデータＱ（ｊ＋１）を新たに算出し、量子化部７５に供給する。
【０１２８】
更に、発生符号量制御部９２は、次に符号化処理するピクチャの絵柄が大きく変化する場合には、ビットレート、あるいは、画像難易度、もしくはビットレートおよび画像難易度を基に、量子化部７５を制御して、Ｑマトリクスの値を設定する。具体的には、発生符号量制御部９２は、符号化が困難な、低ビットレートや難易度の高い画像においては、粗く量子化を行うように、Ｑマトリクスの値を設定し、符号化がある程度容易な、高ビットレートや難易度の低い画像においては、レート制御に余裕があるにもかかわらず、画質を劣化させてしまうようなことがないように、Ｑマトリクスの値を設定する。
【０１２９】
図１０のフローチャートを参照して、イントラＡＣなどの画像難易度情報を用いて、シーンチェンジは、簡単な画像から難しい画像への変化であるか否かの判定を導入して仮想バッファの調整を行う、仮想バッファ更新処理について説明する。
【０１３０】
ステップＳ２１において、発生符号量制御部９２は、ＭＥ残差算出部９５から、ＭＥ残差情報ME_info を取得する。
【０１３１】
ステップＳ２２において、発生符号量制御部９２は、取得されたＭＥ残差情報から、ＭＥ残差情報の平均値avgを減算し、ME_info−avg > Ｄであるか否か、すなわち、算出された値が、所定の閾値Ｄよりも大きいか否かが判断される。ＭＥ残差情報の平均値avgは、後述するステップＳ２６において更新される値であり、上述した式（１）で示される。なお、所定の閾値Ｄは、画質を検討しながらチューニングされる性質の値である。
【０１３２】
ステップＳ２２において、算出された値は、所定の閾値Ｄより小さいと判断された場合、現在のピクチャにおける絵柄と、１つ前のピクチャにおける絵柄との差があまり無い、すなわちシーンチェンジがなかったと判断されるので、処理はステップＳ２６に進む。
【０１３３】
ステップＳ２２において、算出された値は、所定の閾値Ｄより大きいと判断された場合、現在のピクチャにおける絵柄と、１つ前のピクチャにおける絵柄との差が大きい、すなわち、シーンチェンジがあったと判断されるので、ステップＳ２３において、発生符号量制御部９２は、図１１乃至図１４を用いて説明するＱマトリクス設定処理１乃至４のうちのいずれかの処理を実行する。ここで、Ｑマトリクスの設定は、ビットレートやシーンチェンジ後の画像の難易度に基づいて設定される。
【０１３４】
ステップＳ２４において、発生符号量制御部９２は、イントラＡＣ算出部９１から取得される、このシーンチェンジの後のイントラＡＣの値であるmad_infoと、このシーンチェンジの前のイントラＡＣの値であるprev_mad_infoとを比較し、mad_info > prev_mad_infoであるか否かを判断する。
【０１３５】
ステップＳ２４において、mad_info > prev_mad_infoではないと判断された場合、このシーンチェンジは、難しい画像から、簡単な画像へのシーンチェンジであるので、処理は、ステップＳ２６に進む。
【０１３６】
ステップＳ２４において、mad_info > prev_mad_infoであると判断された場合、このシーンチェンジは、簡単な画像から、難しい画像へのシーンチェンジであるので、ステップＳ２５において、発生符号量制御部９２は、図１を用いて説明した従来における場合と同様の処理により、仮想バッファの初期バッファ容量ｄ（０）の更新を行う。
【０１３７】
すなわち、発生符号量制御部９２は、上述した式（２）、式（３）、式（４）および式（５）に基づいて、仮想バッファの初期バッファ容量ｄ（０）を算出し、仮想バッファを更新する。
【０１３８】
ステップＳ２２において、算出された値は、所定の閾値Ｄより小さいと判断された場合、ステップＳ２４において、mad_info > prev_mad_infoではないと判断された場合、もしくは、ステップＳ２５の処理の終了後、ステップＳ２６において、発生符号量制御部９２は、次に供給されるピクチャに備えて、ＭＥ残差情報の平均値avgを、上述した式（１）により更新し、処理は、ステップＳ２１に戻り、それ以降の処理が繰り返される。
【０１３９】
図１０のフローチャートを用いて説明した処理により、イントラＡＣを用いて、画像難易度が易しいものから難しいものに変更されるシーンチェンジの時にのみ仮想バッファ調整を行うようにしたので、エンコードに余裕があるはずの簡単な画像において、更に画質を悪くしてしまうことを防ぐことができる。
【０１４０】
また、シーンチェンジが発生した場合、図１１乃至図１４のフローチャートを用いて後述するＱマトリクス設定処理１乃至Ｑマトリクス設定処理４のうちのいずれかの処理が実行されて、画像のビットレートや難易度に適したＱマトリクスが設定される。
【０１４１】
次に、図１１のフローチャートを参照して、図１０のステップＳ２３において実行される、Ｑマトリクス設定処理１について説明する。Ｑマトリクス設定処理１においては、ビットレートを基にＱマトリクスを変更するか否かが判断される。
【０１４２】
ステップＳ４１において、発生符号量制御部９２は、目標符号量算出部９４から供給されるビットレートの値は、所定の閾値Ａより小さな値であるか否かを判断する。
【０１４３】
ステップＳ４１において、ビットレートの値は、所定の閾値Ａより小さな値であると判断された場合、ステップＳ４２において、発生符号量制御部９２は、Ｑマトリクスの値を、高周波成分を落とすことができる値に変更させるための制御信号を生成し、量子化部７５に出力する。量子化部７５は、発生符号量制御部９２から供給された制御信号に基づいて、量子化処理を実行する場合のＱマトリクスの値を、高周波成分を落とすことができる値に変更し、処理は、図１０のステップＳ２４に戻る。
【０１４４】
ステップＳ４１において、ビットレートの値は、所定の閾値Ａより小さな値ではないと判断された場合、ステップＳ４３において、発生符号量制御部９２は、Ｑマトリクスの値を変更させないので、制御信号を量子化部７５に出力しない。従って、量子化部７５は、量子化処理を実行するときに、通常のＱマトリクスの値を用いることとなる。ステップＳ４３の処理の終了後、処理は、図１０のステップＳ２４に戻る。
【０１４５】
すなわち、低いビットレートでは、符号化に余裕がなくなるため、Ｑマトリクスの値を、高周波成分を落とすことができる値に変更することにより、イントラスライスエンコードのシーンチェンジ時の画質を向上させることができるが、高いビットレートで同様の処理を行ってしまうと、符号化に余裕があるのにもかかわらず、粗く量子化することになってしまうため、画質を劣化してしまうことになってしまう。従って、ビットレートが所定の閾値よりも低い場合のみ、Ｑマトリクスを変更させるようにする。
【０１４６】
図１１のフローチャートを用いて説明した処理により、シーンチェンジが発生した場合、ビットレートに基づいて、Ｑマトリクスを変更するか否かを決定するようにしたので、イントラスライスエンコードのシーンチェンジ時の画質を向上させることができる。
【０１４７】
ここでは、ただ１つの閾値を用いて、ビットレートの高低を判断するようにしているが、複数の閾値を用いて、ビットレートの高低を判断するようにしても良いことは言うまでもない。この場合、ビットレートが低いと判断されるほど、粗い量子化が実行されるようなＱマトリクスが設定される。
【０１４８】
Ｑマトリクス設定処理１においては、ビットレートを基にＱマトリクスを変更するか否かが判断されたが、シーンチェンジ後の画像難易度を基に、Ｑマトリクスの設定値を変更するようにしてもよい。図１２のフローチャートを参照して、シーンチェンジ後の画像難易度を基に、Ｑマトリクスの設定値を変更するＱマトリクス設定処理２について説明する。
【０１４９】
ステップＳ５１において、発生符号量制御部９２は、イントラＡＣ算出部９１から供給されたmad_infoの値と、所定の閾値Ｂを比較して、mad_info＞閾値Ｂが成り立っているか否かを判断する。
【０１５０】
ステップＳ５１において、mad_info＞閾値Ｂが成り立っていると判断された場合、シーンチェンジ後の画像は、画像難易度が高いため符号化が困難な画像であるので、ステップＳ５２において、発生符号量制御部９２は、Ｑマトリクスの値を、高周波成分を落とすことができ、粗い量子化を実行する第１の値に変更させるための制御信号を生成し、量子化部７５に出力する。量子化部７５は、発生符号量制御部９２から供給された制御信号に基づいて、量子化処理を実行する場合のＱマトリクスの値を、高周波成分を落とすことができる第１の値に変更し、処理は、図１０のステップＳ２４に戻る。
【０１５１】
ステップＳ５１において、mad_info＞閾値Ｂが成り立っていないと判断された場合、シーンチェンジ後の画像は、画像難易度が低く、ある程度符号化が易しい画像であるので、ステップＳ５３において、発生符号量制御部９２は、Ｑマトリクスの値を、通常処理よりは高周波成分を落とすことができるが、第１の値と比較して細かく量子化することができる第２の値に変更させるための制御信号を生成し、量子化部７５に出力する。量子化部７５は、発生符号量制御部９２から供給された制御信号に基づいて、Ｑマトリクスの値を、高周波成分を落とすことができる第２の値に変更し、処理は、図１０のステップＳ２４に戻る。
【０１５２】
図１２のフローチャートを用いて説明した処理により、シーンチェンジが発生した場合、シーンチェンジ後の画像難易度に基づいて、Ｑマトリクスの値を設定するようにしたので、イントラスライスエンコードのシーンチェンジ時の画質を向上させることができる。
【０１５３】
なお、図１２においては、mad_info＞閾値Ｂが成り立っているか否かを基に、Ｑマトリクスの値を、第１の値にするか第２の値にするかを決定するようにしたが、例えば、mad_info＞閾値Ｂが成り立っているか否かを基に、Ｑマトリクスの値を変更するか否かを決定するようにしても良い。
【０１５４】
ここでは、ただ１つの閾値を用いて、シーンチェンジ後の画像難易度の高低を判断するようにしているが、複数の閾値を用いて、画像難易度の高低を判断するようにしても良いことは言うまでもない。この場合、画像難易度が高いと判断されるほど、粗い量子化が実行されるようなＱマトリクスが設定される。
【０１５５】
Ｑマトリクス設定処理１においては、ビットレートを基にＱマトリクスを変更するか否かが判断され、Ｑマトリクス設定処理２においては、シーンチェンジ後の画像難易度を基に、Ｑマトリクスの設定値を変更するようになされていたが、ビットレートおよびシーンチェンジ後の画像難易度の両方の値に基づいて、Ｑマトリクスの設定値を設定するようにしてもよい。図１３のフローチャートを参照して、シーンチェンジ後のビットレートおよび画像難易度を基に、Ｑマトリクスの設定値を変更するＱマトリクス設定処理３について説明する。
【０１５６】
ステップＳ６１において、発生符号量制御部９２は、目標符号量算出部９４から供給されるビットレートの値は、所定の閾値Ａより小さな値であるか（すなわち、符号化に余裕のない低いビットレートであるか）否かを判断する。
【０１５７】
ステップＳ６１において、ビットレートの値は、所定の閾値Ａより小さな値であると判断された場合、ステップＳ６２において、発生符号量制御部９２は、イントラＡＣ算出部９１から供給されたmad_infoの値と、所定の閾値Ｂを比較して、mad_info＞閾値Ｂが成り立っているか否かを判断する。
【０１５８】
ステップＳ６２において、mad_info＞閾値Ｂが成り立っていると判断された場合、シーンチェンジ後の画像は、画像難易度が高いため符号化が困難な画像であるので、ステップＳ６３において、発生符号量制御部９２は、Ｑマトリクスの値を、高周波成分を落とすことができ、粗い量子化を実行する第１の値に変更させるための制御信号を生成し、量子化部７５に出力する。量子化部７５は、発生符号量制御部９２から供給された制御信号に基づいて、量子化処理を実行するときのＱマトリクスの値を、高周波成分を落とすことができる第１の値に変更し、処理は、図１０のステップＳ２４に戻る。
【０１５９】
ステップＳ６２において、mad_info＞閾値Ｂが成り立っていないと判断された場合、シーンチェンジ後の画像は、画像難易度が低く、ある程度符号化が易しい画像であるので、ステップＳ６４において、発生符号量制御部９２は、Ｑマトリクスの値を、通常処理よりは高周波成分を落とすことができるが、第１の値と比較して細かく量子化することができる第２の値に変更させるための制御信号を生成し、量子化部７５に出力する。量子化部７５は、発生符号量制御部９２から供給された制御信号に基づいて、量子化処理を実行するときのＱマトリクスの値を、高周波成分を落とすことができる第２の値に変更し、処理は、図１０のステップＳ２４に戻る。
【０１６０】
ステップＳ６１において、ビットレートの値は、所定の閾値Ａより小さな値ではないと判断された場合、ステップＳ６５において、発生符号量制御部９２は、符号化に余裕があるのにもかかわらず、粗く量子化して、画像を劣化してしまうことを避けるため、Ｑマトリクスの値を変更させないので、制御信号を量子化部７５に出力しない。従って、量子化部７５は、通常のＱマトリクスの値を用いて量子化処理を実行するようになされる。ステップＳ６５の処理の終了後、処理は、図１０のステップＳ２４に戻る。
【０１６１】
図１３のフローチャートを用いて説明した処理により、ビットレートおよびシーンチェンジ後の画像難易度の両方の値に基づいて、Ｑマトリクスの設定値が設定されるので、必要以上に画像を劣化させてしまうことなく、シーンチェンジ後の画像の下部において、符号化に割り当てる符号量が足りなくなってしまう現象の発生を抑制することが可能となる。
【０１６２】
また、ビットレートおよび画像難易度の閾値を、それぞれ複数設けることにより、更に詳細に場合分けを行い、必要以上に画像を劣化させてしまうことなく、シーンチェンジ後の画像の下部において、符号化に割り当てる符号量が足りなくなってしまう現象の発生を効果的に抑制するようにすることができる。次に、図１４のフローチャートを参照して、ビットレートおよび画像難易度の閾値を複数設けるようにした、Ｑマトリクス設定処理４について説明する。
【０１６３】
ステップＳ７１において、発生符号量制御部９２は、目標符号量算出部９４から供給されるビットレートの値は、所定の閾値Ｅより小さな値であるか否かを判断する。ここで、閾値Ｅは、符号化が困難である小さなビットレートにおいて、更に詳細に符号化の余裕度を判断するために、図１０、または図１３のフローチャートを用いて説明した処理における閾値Ａより小さい値とするほうが好適である。
【０１６４】
ステップＳ７１において、ビットレートの値は、所定の閾値Ｅより小さな値であると判断された場合、ステップＳ７２において、発生符号量制御部９２は、イントラＡＣ算出部９１から供給されたmad_infoの値と、所定の閾値Ｇを比較して、mad_info＞閾値Ｇが成り立っているか否かを判断する。ここで、閾値Ｇは、符号化が困難である高ビットレートの画像において、比較的広い範囲で画像難易度の高い画像を選択することができる値に設定することが好ましく、基本的に、図１１、または図１３のフローチャートを用いて説明した処理における閾値Ｂと略同等、または、閾値Ｂよりも小さな値とすると好適である。
【０１６５】
ステップＳ７２において、mad_info＞閾値Ｇが成り立っていると判断された場合、シーンチェンジ後の画像は、画像難易度が非常に高いため符号化が困難な画像であるので、ステップＳ７３において、発生符号量制御部９２は、Ｑマトリクスの値を、高周波成分を落とすことができ、Ｑマトリクス設定処理４のうち、最も粗い量子化を実行する第３の値に変更させるための制御信号を生成し、量子化部７５に出力する。量子化部７５は、発生符号量制御部９２から供給された制御信号に基づいて、量子化処理を実行するときのＱマトリクスの値を、第３の値に変更し、処理は、図１０のステップＳ２４に戻る。
【０１６６】
ステップＳ７２において、mad_info＞閾値Ｇが成り立っていないと判断された場合、シーンチェンジ後の画像は、画像難易度が低く、ある程度符号化が易しい画像であるので、ステップＳ７４において、発生符号量制御部９２は、Ｑマトリクスの値を、高周波成分を落とすことができ、Ｑマトリクス設定処理４のうち、２番目に粗い量子化を実行する第４の値に変更させるための制御信号を生成し、量子化部７５に出力する。量子化部７５は、発生符号量制御部９２から供給された制御信号に基づいて、量子化処理を実行するときのＱマトリクスの値を、第４の値に変更し、処理は、図１０のステップＳ２４に戻る。
【０１６７】
ステップＳ７１において、ビットレートの値は、所定の閾値Ｅより小さな値ではないと判断された場合、ステップＳ７５において、発生符号量制御部９２は、目標符号量算出部９４から供給されるビットレートの値は、所定の閾値Ｆより小さな値であるか否かを判断する。ここで、符号化の余裕度を判断するための閾値である閾値Ｆは、閾値Ｅよりも大きな値であり、基本的に、図１０、または図１３のフローチャートを用いて説明した処理における閾値Ａと略同等の値とすると好適である。
【０１６８】
ステップＳ７５において、ビットレートの値は、所定の閾値Ｆより小さな値であると判断された場合、ステップＳ７６において、発生符号量制御部９２は、イントラＡＣ算出部９１から供給されたmad_infoの値と、所定の閾値Ｈを比較して、mad_info＞閾値Ｈが成り立っているか否かを判断する。
【０１６９】
ここで、閾値Ｈは、閾値Ｇより大きな値である。すなわち、ビットレートが閾値Ｅよりも小さな、低ビットレートの画像においては、画面下部まで正しく符号化させるために、難易度の高い画像を充分広い範囲で選択して、選択された画像におけるＱマトリクスの値を、粗い量子化を実行させる第３の値に設定させる必要がある。これに対して、ビットレートが、閾値Ｅよりは大きく、閾値Ｆより小さな画像においては、ビットレートが高い分だけ、レート制御に余裕が生じるので、ビットレートが閾値Ｅよりも小さな場合よりも高い難易度の画像のみ、画面下部で符号量が不足する。従って、閾値Ｈを閾値Ｇよりも大きな値にしなければ、レート制御に余裕がある画像において、必要以上に量子化を粗くして、画質を低下させてしまう。
【０１７０】
ステップＳ７６において、mad_info＞閾値Ｈが成り立っていると判断された場合、シーンチェンジ後の画像は、画像難易度が非常に高いため符号化が困難な画像であるので、ステップＳ７７において、発生符号量制御部９２は、Ｑマトリクスの値を、Ｑマトリクス設定処理４のうち、３番目に粗い量子化を実行する第５の値に変更させるための制御信号を生成し、量子化部７５に出力する。量子化部７５は、発生符号量制御部９２から供給された制御信号に基づいて、量子化処理を実行するときのＱマトリクスの値を、第５の値に変更し、処理は、図１０のステップＳ２４に戻る。
【０１７１】
ステップＳ７６において、mad_info＞閾値Ｈが成り立っていないと判断された場合、シーンチェンジ後の画像は、画像難易度が低く、ある程度符号化が易しい画像であるので、ステップＳ７８において、発生符号量制御部９２は、Ｑマトリクスの値を、変更なしの場合の量子化よりは粗い量子化を実行するが、Ｑマトリクス設定処理４によって変更されて設定されるうちでは、もっとも細かい量子化を実行する第６の値に変更させるための制御信号を生成し、量子化部７５に出力する。量子化部７５は、発生符号量制御部９２から供給された制御信号に基づいて、量子化処理を実行するときのＱマトリクスの値を、第６の値に変更し、処理は、図１０のステップＳ２４に戻る。
【０１７２】
ステップＳ７５において、ビットレートの値は、所定の閾値Ｆより小さな値ではないと判断された場合、ステップＳ７９において、発生符号量制御部９２は、符号化に余裕があるのにもかかわらず、粗く量子化して、画像を劣化してしまうことを避けるため、Ｑマトリクスの値を変更させないので、制御信号を量子化部７５に出力しない。従って、量子化部７５は、量子化処理を実行するとき、通常のＱマトリクスの値を用いる。ステップＳ７９の終了後、処理は、図１０のステップＳ２４に戻る。
【０１７３】
図１４のフローチャートを用いて説明した処理により、ビットレートおよび画像難易度の閾値を複数設けて、更に詳細に場合分けを行うようにすることで、必要以上に画像を劣化させてしまうことなく、シーンチェンジ後の画像の下部において、符号化に割り当てる符号量が足りなくなってしまう現象の発生を効果的に抑制するようにすることができる。
【０１７４】
なお、図１４のフローチャートを用いて説明した処理においては、ビットレートおよび画像難易度の閾値をそれぞれ２つ設けた場合について説明したが、ビットレートおよび画像難易度の閾値を、それぞれ、２つ以上設けるようにしても良いことは言うまでもない。ビットレートおよび画像難易度の閾値を、画像符号化が困難な部分においてより細かく設定する（多くの閾値を設定する）ことにより、画像を劣化させることなく、シーンチェンジ後の画像の下部において、符号化に割り当てる符号量が足りなくなってしまう現象の発生を、更に効果的に抑制するようにすることができる。
【０１７５】
また、上述の実施の形態においては、ローディレイコーディングとしてナンバ０乃至１１の各フレーム画像を全てＰピクチャとし、例えば、横４５マクロブロック、縦２４マクロブロックの画枠サイズの中でフレーム画像の上段から縦２マクロブロックおよび横４５マクロブロック分の領域を１つのイントラスライス部分、他を全てインタースライス部分として設定するようにした場合について述べたが、本発明はこれに限らず、例えば、イントラスライス部分を縦１マクロブロック、横４５マクロブロック分の領域とするなど、他の種々の大きさの領域で形成するようにしても良い。
【０１７６】
また、ここでは、ローディレイエンコードを行う場合を例として説明したが、本発明は、例えば、１５フレームを、フレーム内符号化画像（以下、Ｉピクチャと称する）、フレーム間順方向予測符号化画像（以下、Ｐピクチャと称する）、もしくは、双方向予測符号化画像（以下、Ｂピクチャと称する）の３つの画像タイプのうちのいずれの画像タイプとして処理するかを指定し、指定されたフレーム画像の画像タイプ（Ｉピクチャ、Ｐピクチャ、あるいは、Ｂピクチャ）に応じて、フレーム画像を符号化するような場合にも適用可能である。
【０１７７】
更に、上述の実施の形態においては、本発明をＭＰＥＧ方式によって圧縮符号化する符号化装置としてのビデオエンコーダ６１に適用するようにした場合について述べたが、本発明はこれに限らず、他の種々の画像圧縮方式による符号化装置に適用するようにしても良い。
【０１７８】
上述した一連の処理は、ハードウエアにより実行させることもできるが、ソフトウエアにより実行させることもできる。この場合、例えば、ビデオエンコーダ６１は、図１５に示されるようなパーソナルコンピュータ１０１により構成される。
【０１７９】
図１５において、ＣＰＵ１１１は、ＲＯＭ１１２に記憶されているプログラム、または記憶部１１８からＲＡＭ１１３にロードされたプログラムに従って、各種の処理を実行する。ＲＡＭ１１３にはまた、ＣＰＵ１１１が各種の処理を実行する上において必要なデータなども適宜記憶される。
【０１８０】
ＣＰＵ１１１、ＲＯＭ１１２、およびＲＡＭ１１３は、バス１１４を介して相互に接続されている。このバス１１４にはまた、入出力インタフェース１１５も接続されている。
【０１８１】
入出力インタフェース１１５には、キーボード、マウスなどよりなる入力部１１６、ディスプレイやスピーカなどよりなる出力部１１７、ハードディスクなどより構成される記憶部１１８、モデム、ターミナルアダプタなどより構成される通信部１１９が接続されている。通信部１１９は、インターネットを含むネットワークを介しての通信処理を行う。
【０１８２】
入出力インタフェース１１５にはまた、必要に応じてドライブ１２０が接続され、磁気ディスク１３１、光ディスク１３２、光磁気ディスク１３３、あるいは、半導体メモリ１３４などが適宜装着され、それらから読み出されたコンピュータプログラムが、必要に応じて記憶部１１８にインストールされる。
【０１８３】
一連の処理をソフトウエアにより実行させる場合には、そのソフトウエアを構成するプログラムが、専用のハードウエアに組み込まれているコンピュータ、または、各種のプログラムをインストールすることで、各種の機能を実行することが可能な、例えば汎用のパーソナルコンピュータなどに、ネットワークや記録媒体からインストールされる。
【０１８４】
この記録媒体は、図１５に示されるように、装置本体とは別に、ユーザにプログラムを供給するために配布される、プログラムが記憶されている磁気ディスク１３１（フロッピディスクを含む）、光ディスク１３２（ＣＤ-ＲＯＭ（Compact Disk-Read Only Memory），ＤＶＤ（Digital Versatile Disk）を含む）、光磁気ディスク１３３（ＭＤ（Mini-Disk）（商標）を含む）、もしくは半導体メモリ１３４などよりなるパッケージメディアにより構成されるだけでなく、装置本体に予め組み込まれた状態でユーザに供給される、プログラムが記憶されているＲＯＭ１１２や、記憶部１１８に含まれるハードディスクなどで構成される。
【０１８５】
なお、本明細書において、記録媒体に記憶されるプログラムを記述するステップは、含む順序に沿って時系列的に行われる処理はもちろん、必ずしも時系列的に処理されなくとも、並列的あるいは個別に実行される処理をも含むものである。
【０１８６】
【発明の効果】
本発明によれば、画像データをエンコードすることができる。
また、本発明によれば、量子化に用いられる係数（Ｑマトリクス）の設定を、シーンチェンジが発生したと判断された場合、ビットレートが高いときよりも低いときのほうが、高周波成分を落とす値となるように変更するようにしたので、シーンチェンジ時に必要以上に画像を劣化させることなく、シーンチェンジ後の画像の下部において、符号化に割り当てる符号量が足りなくなってしまう現象の発生を、効果的に抑制するようにすることができる。
また、他の本発明によれば、画像データをエンコードすることができる他に、量子化に用いられる係数（Ｑマトリクス）の設定を、シーンチェンジが発生したと判断された場合、画像難易度が低いときよりも高いときのほうが、高周波成分を落とす値となるように変更するようにしたので、シーンチェンジ時に必要以上に画像を劣化させることなく、シーンチェンジ後の画像の下部において、符号化に割り当てる符号量が足りなくなってしまう現象の発生を、効果的に抑制するようにすることができる。
【図面の簡単な説明】
【図１】ＭＰＥＧ２方式によって映像データを圧縮符号化する場合、および圧縮符号化された画像データを復号する場合の処理について説明する図である。
【図２】ＶＢＶバッファについて説明する図である。
【図３】ローディレイコーディングについて説明する図である。
【図４】ＶＢＶバッファについて説明する図である。
【図５】シーンチェンジについて説明する図である。
【図６】従来の仮想バッファ更新処理について説明するフローチャートである。
【図７】本発明を適用したビデオエンコーダの構成を示すブロック図である。
【図８】仮想バッファのバッファ占有量について説明する図である。
【図９】イントラスライスおよびインタースライス毎の、仮想バッファのバッファ占有量について説明する図である。
【図１０】本発明を適用した仮想バッファ更新処理について説明するフローチャートである。
【図１１】Ｑマトリクス設定処理１について説明するフローチャートである。
【図１２】Ｑマトリクス設定処理２について説明するフローチャートである。
【図１３】Ｑマトリクス設定処理３について説明するフローチャートである。
【図１４】Ｑマトリクス設定処理４について説明するフローチャートである。
【図１５】パーソナルコンピュータの構成を示すブロック図である。
【符号の説明】
６１ ビデオエンコーダ， ７１ 前処理部， ７２ 演算部， ７３ 動きベクトル検出部， ７４ ＤＣＴ部， ７５ 量子化部， ７７ ＶＬＣ部， ７８ 逆量子化部， ７９ 逆ＤＣＴ部， ８０ 演算部， ８１ 動き補償部， ８２ バッファ， ８３ 量子化制御部， ８４ フレームメモリ， ９１イントラＡＣ算出部， ９２ 発生符号量制御部， ９３ ＧＣ算出部， ９４ 目標符号量算出部， ９５ ＭＥ残差算出部[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an encoding device, an encoding method, a program, and a recording medium, and more particularly, to an encoding device, an encoding method, a program, and a recording medium that are suitable for performing low delay coding.
[0002]
[Prior art]
In recent years, various compression encoding methods have been proposed as methods for reducing the amount of information by compressing video data and audio data, and a representative one is MPEG2 (Moving Picture Experts Group Phase 2).
[0003]
With reference to FIG. 1, a description will be given of processing when video data is compression-encoded according to the MPEG2 system and when image data that has been compression-encoded is decoded.
[0004]
The encoder 1 on the transmission side converts the frame images 11 of numbers 0 to 11 into intra-frame encoded images (hereinafter referred to as I pictures), inter-frame forward prediction encoded images (hereinafter referred to as P pictures), or Specifies which of the three image types of the bi-predictive encoded image (hereinafter referred to as B picture) is to be processed, and the image type of the specified frame image (I picture, P picture, or , B picture), the reordering is performed so that the frame images are rearranged in the encoding order, the encoding processing is performed on each frame image in that order, and the encoded frame 12 is generated. Transmit to.
[0005]
The decoder 2 on the receiving side decodes the frame image encoded by the encoder 1 and then reorders again, restores the image frame to the original order, restores the frame image 13, and displays the reproduced image. .
[0006]
In the encoder 1, since the encoding process is performed after the reordering, the frame image of the number 2 must be encoded before the encoding of the frame image of the number 0, and a delay (( Hereinafter, this will be referred to as reordering delay).
[0007]
Since the decoder 2 also performs reordering after decoding, the frame image of the number 2 must be decoded before the frame image of the number 0 is decoded and displayed, and the reordering delay is correspondingly increased. It will occur.
[0008]
As described above, since both the encoder 1 and the decoder 2 perform reordering, a reordering delay of 3 frames occurs between the time when the image data is encoded and the time when the reproduced image is displayed.
[0009]
In addition, when encoded data compressed and encoded by the MPEG2 system is transmitted, the encoded data transmitted from the compression encoding device on the transmission side is converted into a video STD (System Target Decoder) buffer (so-called VBV) on the reception side. (Video Buffer Verifier) is stored for each picture.
[0010]
As shown in FIG. 2, the buffer size (capacity) of the VBV buffer is determined, and the encoded data is sequentially stored for each picture in the VBV buffer. In this case, the encoded data of I picture, P picture, and B picture are respectively stored in the VBV buffer at a constant transmission rate, and are extracted by the decoder at the decoding timing at the end of storage (one frame period). . Since the I picture has a larger amount of encoded data than the B picture, it takes more time than the B picture to be stored in the VBV buffer.
[0011]
At this time, when the encoded data is stored in the VBV buffer of the decoder 2 and when the encoded data is extracted from the VBV buffer, the encoder 1 on the data transmission side causes overflow and underflow in the VBV buffer. In order to prevent this from occurring, it is necessary to control (rate control) the generated code amount of the encoded data generated based on the buffer occupation ratio of the VBV buffer. However, since the amount of generated code of the I picture necessary for the screen update is large, a long transmission time of the image data of the I picture is required, and this time is delayed.
[0012]
When performing real-time transmission that requires real-time performance, such as video data for videophones and video conferences, as described above, if a delay due to transmission time or a reordering delay occurs, the transmission side There is a time difference between receiving the encoded data sent on the receiving side and displaying the reproduced image. On the other hand, in order to reduce such a delay, in the MPEG2 system, a method called low delay coding (Low Delay Coding) that shortens the delay time to 150 [ms] or less is prepared by the standard.
[0013]
In low delay coding, a B picture that causes reordering delay and an I picture with a large amount of generated code are not used, but only a P picture is used, and this P picture is converted into an intra slice consisting of several slices. By dividing into an inter slice consisting of all remaining slices, encoding can be performed without reordering.
[0014]
The intra slice is an image portion in which the image data of the slice portion is encoded in the frame, and the inter slice is the difference data between the image data of the slice portion and the reference image data of the same region in the previous frame image. This is the image part.
[0015]
In the low delay coding, for example, as shown in FIG. 3, the encoder 1 uses all the frame images 11 of the numbers 0 to 11 as P pictures, for example, within an image frame size of 45 macro blocks horizontally and 24 macro blocks vertically. The area for the vertical 2 macroblocks and the horizontal 45 macroblocks from the top of the frame image of number 0 is set as the intra slice I0, and all other areas are set as the inter slice P0.
[0016]
Then, in the frame image of the next number 1, the encoder 1 sets the intra slice I1 in the area of the same area at the position continuing in the downward direction of the intra slice I0 of the frame image of the number 0, and all others are inter slices. Set to P1. Similarly, an intra slice and an inter slice are set for each frame image, and an intra slice I11 and an inter slice P11 are also set for the frame image of the last number 11.
[0017]
The encoder 1 encodes the intra slices I0 to I11 of each frame image as transmission data as they are, and encodes the other inter slices P0 to P11 based on the difference data from the reference image in the same area of the previous frame image. (However, since there is no frame image before the reference image of the inter slice P0 at the start of encoding, this is not the case only at the start of encoding). Then, the same encoding process is repeatedly performed for the frame image of the number 0 to the frame image of the number 11, whereby the encoder 1 encodes the image data of the entire screen in one P picture and encodes the encoded frame 21. Can be generated.
[0018]
In this case, the image data sizes of the intra slices I0 to I11 in each frame image are all uniform. Of course, the image data size of the inter slices P0 to P11 is also uniform. A constant fixed rate.
[0019]
As a result, as shown in FIG. 4, all frame images of the P picture become encoded data having the same generated code amount, and the encoded data in the VBV buffer when stored in the VBV buffer and when extracted are extracted. All transitions are the same. As a result, the encoder 1 on the transmission side can easily control the generated code amount of the encoded data without causing an underflow and overflow in the VBV buffer of the decoder 2, and the I-picture with a large generated code amount can be controlled. Problems caused by such delays and reordering delays can be solved, and a reproduced image can be displayed without delay.
[0020]
By the way, in the compression coding apparatus having the configuration described above, the intra slices I0 to I11 are encoded as transmission data as they are, and the inter slices P0 to P11 are difference data from the reference image in the same region in the previous frame image. Therefore, the actual generated code amount when the image data portions of the intra slices I0 to I11 are compression-encoded is large, and the actual data amount when the image data portions of the inter-slices P0 to P11 are compression-encoded is encoded. The amount of generated code is reduced.
[0021]
However, although the amount of generated code for the entire picture is specified, the amount of generated code assigned to each of the intra slices I0 to I11 and the inter slices P0 to P11 is not specified. In other words, an image in which the amount of generated code when encoding is performed as in inter slices P0 through P11 is not so large even for an image portion in which the amount of generated code when encoding is performed as in intra slices I0 through I11. The generated code amount is equally allocated to the data portion.
[0022]
Therefore, the generated code amount allocated to the intra slices I0 to I11 having a large amount of data is small, and the generated code amount allocated to the inter slices P0 to P11 having a small data amount may be increased. In addition, there is a problem that the image as a whole picture is distorted.
[0023]
Specifically, as shown in FIG. 5, when there is an image 32 with a high image encoding difficulty following an image 31 with a low image encoding difficulty, the image 31 with a low encoding difficulty is present. Since the image is easy to encode, the Q scale becomes small. However, in the conventional method, encoding of the subsequent image 32 having a high degree of difficulty in encoding the image starts with a small Q scale. For this reason, a given amount of bits is consumed by the middle of the screen, and the previous picture remains at the lower end of the screen. This phenomenon has an effect until the intra slice next appears at the problem location at the bottom of the screen.
[0024]
In order to solve this problem, the applicant of the present invention disclosed in Japanese Patent Laid-Open No. 11-205803, an encoding device capable of generating encoded data capable of reproducing a high-quality image on the decoder side even in the low delay mode, and An encoding method is proposed.
[0025]
That is, when performing normal feedback type quantization control to determine the optimal quantization step size for each intra slice and inter slice and performing quantization control, the next picture is the same as the previous picture and the picture. When a significantly different scene change occurs, instead of using the quantization index data Q (j + 1) calculated based on the previous picture, based on the ME residual information of the picture to be encoded, By updating the initial buffer capacity d (0) of the virtual buffer, the quantization index data Q (j + 1) is newly calculated again. Thereby, even when a scene change occurs, an optimal quantization step size is determined for each intra slice and inter slice, and quantization control is performed.
[0026]
The ME residual is calculated on a picture-by-picture basis, and is a total value of luminance difference values between the previous picture and the next picture. Therefore, when the ME residual information shows a large value, it means that the picture of the previous picture is greatly different from the picture of the next picture to be encoded (so-called scene change).
[0027]
This encoding method will be described with reference to the flowchart of FIG.
[0028]
In step S1, for example, ME residual information obtained when detecting a motion vector is acquired. The ME residual information acquired here is referred to as ME_info.
[0029]
In step S2, the average value avg of the ME residual information is subtracted from the acquired ME residual information, and it is determined whether or not the calculated value is larger than a predetermined threshold value D. The average value avg of the ME residual information is a value updated in step S4 to be described later, and is represented by the following formula (1).
[0030]
avg = 1/2 (avg + ME_info) (1)
[0031]
If it is determined in step S2 that the calculated value is smaller than the predetermined threshold D, it is determined that there is not much difference between the pattern in the current picture and the pattern in the previous picture, that is, there has been no scene change. Thus, the process proceeds to step S4.
[0032]
If it is determined in step S2 that the calculated value is greater than the predetermined threshold value D, it is determined that the difference between the pattern in the current picture and the pattern in the previous picture is large, that is, there has been a scene change. Therefore, in step S3, the initial buffer capacity d (0) of the virtual buffer is calculated based on the equations (2), (3), (4), and (5), and the virtual buffer is updated. Is done.
[0033]
X representing the difficulty GC (Global Complexity) of the picture unit picture is represented by the following equation (2).
X = T × Q (2)
Here, T is a generated code amount in units of pictures, and Q is an average value of quantization step sizes in units of pictures.
[0034]
When the difficulty X of the picture unit picture is equal to the ME residual information ME_info, that is, when the following equation (3) is satisfied, the quantization index data Q of the entire picture is represented by the equation ( 4).
[0035]
X = ME_info (3)
Q = {d (0) × 31} / {2 × (br / pr)} (4)
However, br is a bit rate and pr is a picture rate.
[0036]
The initial buffer capacity d (0) of the virtual buffer in equation (4) is expressed by the following equation (5).
d (0) = 2 × {(ME_info × br / pr) / 31 × T} (5)
[0037]
By substituting the initial buffer capacity d (0) of the virtual buffer into the equation (4) again, the quantization index data Q of the entire picture is calculated.
[0038]
If it is determined in step S2 that the calculated value is smaller than the predetermined threshold value D, or after the processing in step S3 is completed, in step S4, the ME residual information is prepared for the next picture to be supplied. The average value avg is calculated and updated by the above-described equation (1), the process returns to step S1, and the subsequent processes are repeated.
[0039]
When the scene described in the flowchart of FIG. 6 causes a scene change in which the next picture is significantly different from the previous picture, based on the ME residual information ME_info of the picture to be encoded. Thus, the initial buffer capacity d (0) of the virtual buffer is updated, and the quantized index data Q (j + 1) is newly calculated based on this value, so that the intra slice and inter slice are corresponding to the scene change. An optimal quantization step size is determined for each.
[0040]
[Problems to be solved by the invention]
However, when the method described in Japanese Patent Application Laid-Open No. 11-205803 is used, the same encoding is performed even when the scene changes from an image with a high (difficult) encoding difficulty to an image with a low (easy) encoding difficulty. Since this process is performed, the image quality is adversely affected.
[0041]
Specifically, since the virtual buffer adjustment is performed both when the scene changes from an easy image to a difficult image and when the scene changes from a difficult image to an easy image, the scene changes from a difficult image to an easy image. In the case of changing, there is a case where the image quality is bothered in an image with a low encoding difficulty level that should have a margin for encoding.
[0042]
Also, if many intra macroblocks are generated due to the scene change, the encoding process by updating the conventional virtual buffer will have the maximum quantization scale when encoding images with very high image difficulty. Since the generated code amount cannot be suppressed at the upper part of the screen, the code amount assigned to the encoding at the lower part of the screen becomes insufficient as in the case of the image 32 described with reference to FIG. The phenomenon occurs.
[0043]
This is because the quantization index that controls quantization is stuck at a substantially maximum value.
[0044]
The present invention has been made in view of such circumstances, and is intended to improve image quality at the time of a scene change in intra slice encoding in response to various scene changes.
[0045]
[Means for Solving the Problems]
The first encoding device of the present invention is , Detecting a change in the pattern between the first picture that is the previous picture and the second picture that is the next picture to be encoded First detection means When, First detection means Quantization index data using determination means for determining whether or not a scene change has occurred between the first picture and the second picture based on the detection result of the first picture and the initial buffer capacity value of the virtual buffer A determination means for determining When it is determined by the determination means that a scene change has occurred, the value when the bit rate of the frame image is lower than when the bit rate is high is a value that drops the high frequency component. Based on the setting means for setting the Q matrix, the quantization index data determined by the determination means, and the Q matrix set by the setting means, the quantization means for performing quantization, and the quantization means Encoding means for encoding the quantized coefficient data Prepare It is characterized by that.
[0047]
Setting means , If the bit rate is less than the predetermined threshold, To be a value that drops high-frequency components, A Q matrix can be set.
[0048]
The setting means compares a plurality of threshold values with the bit rate, and based on the comparison result, the lower the bit rate, To be a value that drops high-frequency components, A Q matrix can be set.
[0049]
First detection means In this case, an index indicating a difference between the pattern of the first picture and the pattern of the second picture can be calculated, and a change in the pattern can be detected based on the index.
[0050]
Of the second picture image Detect difficulty Second detection means In the setting means, when it is determined by the determining means that a scene change has occurred, The Q matrix can be set so that the higher the bit rate is, the lower the high frequency component is. When the bit rate is lower than a predetermined threshold, the second detection means Of the second picture detected by image difficulty When the value is higher than when the value is low , Q matrix can be set.
[0051]
Second detection means Is The difficulty of the pattern of the frame image and the complexity of the frame image, which is a statistic that correlates with the amount of data after compression, is used as an indicator of the difficulty of the image. And calculate the second picture based on the index image The degree of difficulty can be detected.
[0053]
Setting means ,image If the difficulty level is higher than a predetermined threshold, It is a value that drops high frequency components than when the image difficulty is lower than a predetermined threshold In this way, the Q matrix can be set.
[0054]
The setting means includes a bit rate smaller than the first threshold value. If the image difficulty level is larger than the second threshold value, the Q matrix can be set so as to be the first Q matrix that reduces the highest frequency component, and the bit rate is smaller than the first threshold value. When the image difficulty level is smaller than the second threshold value, the Q matrix can be set to be a second Q matrix that performs quantization more finely than the first Q matrix, Bit rate is the first threshold When, A third threshold value greater than the first threshold value If the image difficulty level is larger than a fourth threshold value that is larger than the second threshold value, the third Q matrix is quantized more finely than the second Q matrix. , The Q matrix can be set, the bit rate is a value between the first threshold value and the third threshold value, and the image difficulty level is smaller than the fourth threshold value, the third Q matrix To become the fourth Q matrix that performs quantization more finely , Q matrix can be set.
[0055]
All the frame images can be inter-frame forward prediction encoded images.
[0056]
The first encoding method of the present invention is , Detecting a change in the pattern between the first picture that is the previous picture and the second picture that is the next picture to be encoded Detection step When, Detection step Based on the detection result of the above process, a quantization step is performed using a determination step for determining whether or not a scene change has occurred between the first picture and the second picture, and the value of the initial buffer capacity of the virtual buffer. A determination step for determining index data; When it is determined that a scene change has occurred as a result of the processing in the determination step, when the bit rate of the frame image is lower than when the bit rate is high, the high frequency component is reduced. A setting step for setting a Q matrix, a quantization step for performing quantization based on the quantization index data determined by the processing of the determination step, and the Q matrix set by the processing of the setting step, and a quantization step An encoding step for encoding the quantized coefficient data quantized by the processing of Include It is characterized by that.
[0057]
The program recorded on the first recording medium of the present invention is , Detecting a change in the pattern between the first picture that is the previous picture and the second picture that is the next picture to be encoded Detection step When, Detection step Based on the detection result of the above process, a quantization step is performed using a determination step for determining whether or not a scene change has occurred between the first picture and the second picture, and the value of the initial buffer capacity of the virtual buffer. A determination step for determining index data; When it is determined that a scene change has occurred as a result of the processing in the determination step, when the bit rate of the frame image is lower than when the bit rate is high, the high frequency component is reduced. A setting step for setting a Q matrix, a quantization step for performing quantization based on the quantization index data determined by the processing of the determination step, and the Q matrix set by the processing of the setting step, and a quantization step An encoding step for encoding the quantized coefficient data quantized by the processing of Include It is characterized by Let the computer execute the process .
[0058]
The first program of the present invention is , Detecting a change in the pattern between the first picture that is the previous picture and the second picture that is the next picture to be encoded Detection step When, Detection step Based on the detection result of the above process, a quantization step is performed using a determination step for determining whether or not a scene change has occurred between the first picture and the second picture, and the value of the initial buffer capacity of the virtual buffer. A determination step for determining index data; When it is determined that a scene change has occurred as a result of the processing in the determination step, when the bit rate of the frame image is lower than when the bit rate is high, the high frequency component is reduced. A setting step for setting a Q matrix, a quantization step for performing quantization based on the quantization index data determined by the processing of the determination step, and the Q matrix set by the processing of the setting step, and a quantization step An encoding step for encoding the quantized coefficient data quantized by the processing of Include It is characterized by Let the computer execute the process .
[0059]
The second encoding apparatus according to the present invention is configured to output a first picture that is a picture to be encoded next. image First detection means for detecting the difficulty level, second detection means for detecting a change in the pattern of the second picture, which is the previous picture, and the first picture, and second detection means Quantization index data using a determination means for determining whether or not a scene change has occurred between the second picture and the first picture based on the detection result of the first picture, and the initial buffer capacity value of the virtual buffer A determination means for determining When it is determined by the determining means that a scene change has occurred, the higher the image difficulty level detected by the first detecting means, the lower the high frequency component. Based on the setting means for setting the Q matrix, the quantization index data determined by the determination means, and the Q matrix set by the setting means, the quantization means for performing quantization, and the quantization means Encoding means for encoding the quantized coefficient data Prepare It is characterized by that.
[0061]
Setting means ,image If the difficulty level is higher than a predetermined threshold, A value that drops high-frequency components. In this way, the Q matrix can be set.
[0062]
The setting means includes a plurality of predetermined threshold values and image Compare the difficulty and based on the comparison result, image The higher the difficulty level A value that drops high-frequency components. In this way, the Q matrix can be set.
[0063]
The second detection means can calculate an index indicating a difference between the pattern of the first picture and the pattern of the second picture, and can detect a change in the pattern based on the index.
[0064]
The first detection means includes The difficulty of the pattern of the frame image and the complexity of the frame image, which is a statistic that correlates with the amount of data after compression, is used as an indicator of the difficulty of the image. And calculate the first picture based on the index image The degree of difficulty can be detected.
[0065]
All the frame images can be inter-frame forward prediction encoded images.
[0066]
In the second encoding method of the present invention, the first picture which is a picture to be encoded next is processed. image A first detection step for detecting a difficulty level, a second detection step for detecting a change in the pattern of the second picture, which is the previous picture, and the first picture, and a second detection step Based on the detection result of the above process, a determination step for determining whether or not a scene change has occurred between the second picture and the first picture, and a quantization using the initial buffer capacity value of the virtual buffer A determination step for determining index data; When it is determined that a scene change has occurred due to the processing of the determination step, the value at which the high frequency component is dropped is higher when the image difficulty level detected by the processing of the first detection step is higher than when the image difficulty level is low. A setting step for setting a Q matrix, a quantization step for performing quantization based on the quantization index data determined by the processing of the determination step, and the Q matrix set by the processing of the setting step, and a quantization step An encoding step for encoding the quantized coefficient data quantized by the processing of Include It is characterized by that.
[0067]
The program recorded on the second recording medium of the present invention is the first picture that is the picture to be encoded next. image A first detection step for detecting a difficulty level, a second detection step for detecting a change in the pattern of the second picture, which is the previous picture, and the first picture, and a second detection step Based on the detection result of the above process, a determination step for determining whether or not a scene change has occurred between the second picture and the first picture, and a quantization using the initial buffer capacity value of the virtual buffer A determination step for determining index data; When it is determined that a scene change has occurred due to the processing of the determination step, the value at which the high frequency component is dropped is higher when the image difficulty level detected by the processing of the first detection step is higher than when the image difficulty level is low. A setting step for setting a Q matrix, a quantization step for performing quantization based on the quantization index data determined by the processing of the determination step, and the Q matrix set by the processing of the setting step, and a quantization step An encoding step for encoding the quantized coefficient data quantized by the processing of Include It is characterized by Let the computer execute the process .
[0068]
The second program of the present invention stores the first picture that is the picture to be encoded next. image A first detection step for detecting a difficulty level, a second detection step for detecting a change in the pattern of the second picture, which is the previous picture, and the first picture, and a second detection step Based on the detection result of the above process, a determination step for determining whether or not a scene change has occurred between the second picture and the first picture, and a quantization using the initial buffer capacity value of the virtual buffer A determination step for determining index data; When it is determined that a scene change has occurred due to the processing in the determination step, the value at which the high-frequency component is dropped is higher when the image difficulty level detected by the processing in the first detection step is higher than when the image difficulty level is low. A setting step for setting a Q matrix, a quantization step for performing quantization based on the quantization index data determined by the processing of the determination step, and the Q matrix set by the processing of the setting step, and a quantization step An encoding step for encoding the quantized coefficient data quantized by the processing of Include It is characterized by Let the computer execute the process .
[0069]
In the first encoding device, encoding method, and program of the present invention, , A change in pattern between the first picture that is the previous picture and the second picture that is the next picture to be encoded is detected, and a scene change occurs between the first picture and the second picture The quantization index data is determined using the value of the initial buffer capacity of the virtual buffer, When it is determined that a scene change has occurred, when the bit rate is lower than when the bit rate is high, the high frequency component is reduced. Q matrix is set, quantization is performed based on quantization index data and Q matrix, and quantized quantized coefficient data is encoded Be done .
[0070]
In the second encoding device, encoding method, and program of the present invention, the difficulty level of the first picture that is the next picture to be encoded is detected, and the second picture that is the previous picture is detected. And a change in the pattern with the first picture is detected, it is determined whether or not a scene change has occurred between the second picture and the first picture, and the value of the initial buffer capacity of the virtual buffer is used. The quantization index data is determined, When it is determined that a scene change has occurred, when the image difficulty level is higher than when the image difficulty level is low, the high frequency component is reduced. Q matrix is set, quantization is performed based on quantization index data and Q matrix, and quantized quantized coefficient data is encoded Be done .
[0071]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0072]
FIG. 7 is a block diagram showing the configuration of the video encoder 61.
[0073]
The video encoder 61 is configured to encode image data by a low delay coding method using all P pictures. The preprocessing unit 71 of the video encoder 61 receives input of image data supplied from the outside.
[0074]
The pre-processing unit 71 divides each frame image (in this case, all P pictures) of sequentially input image data into macroblocks composed of a luminance signal of 16 pixels × 16 lines and a color difference signal corresponding to the luminance signal. Then, this is supplied as macro block data to the calculation unit 72, the motion vector detection unit 73, and the intra AC calculation unit 91 of the quantization control unit 83.
[0075]
The motion vector detection unit 73 receives the input of the macroblock data, calculates the motion vector of each macroblock based on the macroblock data and the reference image data stored in the frame memory 84, and serves as motion vector data. And sent to the motion compensation unit 81.
[0076]
The calculation unit 72 uses the intra mode for the intra slices I0 to I11 and the inter slices P0 to P11 for the macro slice data supplied from the preprocessing unit 71 based on the image type of each macro block. Motion compensation is performed in the forward prediction mode.
[0077]
Here, the intra mode is a method in which a frame image to be encoded is used as transmission data as it is, and the forward prediction mode is a prediction residual between a frame image to be encoded and a past reference image as transmission data. It is a method. In the video encoder 61, only the P picture is used, and the intra slices I0 to I11 and the inter slices P0 to P11 are encoded separately.
[0078]
First, when the macroblock data is one of the intra slices I0 to I11, the macroblock data is processed in the intra mode. In other words, the calculation unit 72 sends the macroblock of the input macroblock data to the DCT (Discrete Cosine Transform) unit 74 as calculation data as it is. The DCT unit 74 performs DCT conversion processing on the input operation data to generate DCT coefficients, and sends this to the quantization unit 75 as DCT coefficient data.
[0079]
The quantization unit 75 performs quantization processing on the input DCT coefficient data by multiplying the quantization index data Q (j + 1) supplied from the generated code amount control unit 92 by the coefficient of the Q matrix. Then, it is sent to the VLC (Variable Length Code) unit 77 and the inverse quantization unit 78 as quantized DCT coefficient data. Here, the quantization unit 75 adjusts the quantization step size in the quantization process according to the quantization index data Q (j + 1) supplied from the generated code amount control unit 92, thereby reducing the generated code amount. It is made to control.
[0080]
Further, the value of the Q matrix to be multiplied with the quantization index data Q (j + 1) will be described later with reference to the flowcharts of FIGS. 11 to 14 executed by the generated code amount control unit 92 when there is a scene change. The value is controlled as necessary by any one of the Q matrix setting process 1 to the Q matrix setting process 4.
[0081]
The quantized DCT coefficient data sent to the inverse quantization unit 78 undergoes an inverse quantization process with the same quantization step size as that of the quantization unit 75 and is sent to the inverse DCT unit 79 as DCT coefficient data. The inverse DCT unit 79 performs inverse DCT processing on the supplied DCT coefficient data, and the generated computation data is sent to the computation unit 80 and stored in the frame memory 84 as reference image data.
[0082]
When the macroblock data is one of the inter slices P0 to P11, the calculation unit 72 performs motion compensation processing in the forward prediction mode on the macroblock data.
[0083]
The motion compensation unit 81 performs motion compensation on the reference image data stored in the frame memory 84 according to the motion vector data, and calculates forward prediction image data. The calculation unit 72 performs subtraction processing on the macroblock data using the forward prediction image data supplied from the motion compensation unit 81.
[0084]
That is, in the forward prediction mode, the motion compensation unit 81 reads the reference image data by shifting the read address of the frame memory 84 according to the motion vector data, and uses the reference image data as the forward prediction image data. It supplies to the calculating part 80. The computing unit 72 subtracts the forward prediction image data from the supplied macroblock data to obtain difference data as a prediction residual. Then, the calculation unit 72 sends the difference data to the DCT unit 74.
[0085]
Further, forward prediction image data is supplied from the motion compensation unit 81 to the calculation unit 80, and the calculation unit 80 adds the forward prediction image data to the calculation data supplied from the inverse DCT unit. The reference image data is locally reproduced and output to the frame memory 84 for storage.
[0086]
Thus, the image data input to the video encoder 61 undergoes motion compensation prediction processing, DCT processing, and quantization processing, and is supplied to the VLC unit 77 as quantized DCT coefficient data. The VLC unit 77 performs variable length encoding processing based on a predetermined conversion table for the quantized DCT coefficient data, sends the resulting variable length encoded data to the buffer 82, and encodes each macroblock. Generated code amount data B (j) representing the number of generated bits is sent to the generated code amount control unit 92 and the GC (Global Complexity) calculation unit 93 of the quantization control unit 83, respectively.
[0087]
The GC calculation unit 93 sequentially accumulates the generated code amount data B (j) for each macroblock, and when all the generated code amount data B (j) for one picture is accumulated, the generated code amount data B (j) is generated for all macroblocks. The generated code amount for one picture is calculated by cumulatively adding the code amount data B (j).
[0088]
Then, the GC calculation unit 93 calculates the product of the generated code amount of the intra slice portion and the average value of the quantization step size in the intra slice portion of one picture using the following equation (6). Thus, the GC data Xi representing the difficulty of the image of the intra slice portion (hereinafter referred to as GC) is obtained and supplied to the target code amount calculation unit 94.
[0089]
Xi = (Ti / Ni) × Qi (6)
Here, Ti is the generated code amount of the intra slice, Ni is the number of intra slices, and Qi is the average value of the quantization step size of the intra slice.
[0090]
At the same time, the GC calculation unit 93 uses the following equation (7) to calculate the generated code amount of the inter slice portion of one picture and the average value of the quantization step size in the inter slice portion. By calculating the product, the GC data Xp in the inter-slice part is obtained and supplied to the target code amount calculation unit 94.
[0091]
Xp = (Tp / Np) × Qp (7)
Here, Tp is the generated code amount of the inter slice, Np is the number of inter slices, and Qp is the average value of the quantization step size of the inter slice.
[0092]
Based on the GC data Xi supplied from the GC calculator 93, the target code amount calculator 94 calculates the target generated code amount data Tpi of the intra slice portion in the next picture using the following equation (8). At the same time, based on the GC data Xp supplied from the GC calculation unit 93, the target generated code amount data Tpp of the inter slice portion in the next picture is calculated based on the following equation (9), and the calculated target generated code The quantity data Tpi and Tpp are sent to the generated code quantity control unit 92, respectively.
[0093]
Tpi = {(Ni × Xi) / (Np × Xp) + (Np × Xi)} × Xp (8)
[0094]
Tpp = {(Np × Xp) / (Np × Xp) + (Ni × Xi)} × Xp (9)
[0095]
The target code amount calculation unit 94 receives an input of the bit rate value input by the user using the operation input unit 85 and supplies the bit rate value to the code amount generation unit 92.
[0096]
The ME residual calculation unit 95 calculates ME residual information ME_info based on the input macroblock data and outputs the ME residual information to the generated code amount control unit 92. Here, the ME residual information ME_info is calculated in units of pictures and is a total value of luminance difference values between the previous picture and the next picture. Therefore, when the ME residual information ME_info shows a large value, it means that the picture of the previous picture is greatly different from the picture of the picture to be encoded next (so-called scene change).
[0097]
When the picture of the previous picture is different from the picture of the picture to be encoded next, the quantum generated based on the target generated code amount data Tpi and Tpp calculated using the image data of the previous picture It is not appropriate to determine the quantization step size of the quantization unit 75 based on the quantization index data Q (j + 1). Therefore, when a scene change occurs, the target generated code amount data Tpi and Tpp may be newly calculated again.
[0098]
The intra AC calculation unit 91 calculates an intra AC, and outputs mad_info indicating the current intra AC value and prev_mad_info indicating the previous intra AC value to the generated code amount control unit 92 To do.
[0099]
Intra AC is a parameter defined as the sum of variance values of video data for each DCT block in the DCT processing unit in the MPEG system, and indicates the complexity of the video. Correlate with data volume. That is, the intra AC is the sum in the screen of the absolute value sum obtained by subtracting the average value of the pixel values for each block from the pixel value of each pixel for each DCT block. IntraAC is represented by the following equation (10).
[0100]
[Expression 1]

[0101]
In addition, in Expression (10), Expression (11) is established.
[Expression 2]

[0102]
If the virtual buffer adjustment is performed for both scene changes from easy to difficult encoding of images and scene changes from difficult to easy, from difficult to easy In the case of the scene change, there is a case where the image quality of the easy image that should have a margin for encoding is bothered. Even if the scene changes from a difficult one to an easy one, it may be better to adjust the virtual buffer depending on the magnitude of the change or the difficulty of the image after the scene change. However, it is possible to determine the presence or absence of a scene change using only the ME residual information. Cannot determine whether it is a change.
[0103]
Therefore, the intra AC calculation unit 91 calculates the intra AC, and outputs mad_info indicating the current intra AC value and prev_mad_info indicating the previous intra AC value to the generated code amount control unit 92. Thus, the generated code amount control unit 92 can determine the state of the scene change and determine whether or not to perform virtual buffer adjustment.
[0104]
The generated code amount control unit 92 constantly monitors the accumulation state of the variable-length encoded data stored in the buffer 82, and determines the quantization step size based on the occupation amount information indicating the accumulation state. Yes.
[0105]
Also, the generated code amount control unit 92 performs quantization to reduce the generated code amount when the generated code amount data B (j) of the macro block actually generated is larger than the target generated code amount data Tpi of the intra slice portion. When the step size is increased and the actual generated code amount data B (j) is smaller than the target generated code amount data Tpi, the quantization step size is decreased to increase the generated code amount.
[0106]
Furthermore, similarly in the case of the inter slice part, the generated code amount control unit 92 determines the generated code amount when the generated code amount data B (j) of the macro block actually generated is larger than the target generated code amount data Tpp. In order to reduce the quantization step size, if the actual generated code amount data B (j) is smaller than the target generated code amount data Tpp, the quantization step size is decreased to increase the generated code amount. Has been made.
[0107]
That is, the generated code amount control unit 92 assumes the transition of the accumulation state of the variable-length encoded data stored in the VBV buffer provided on the decoder side, and as shown in FIG. The buffer occupancy d (j) of the virtual buffer in the block is represented by the following equation (12), and the buffer occupancy d (j + 1) of the virtual buffer in the j + 1-th macroblock is represented by the following equation (13): By subtracting the equation (13) from the equation (12), the buffer occupation amount d (j + 1) of the virtual buffer in the j + 1-th macroblock can be modified as the following equation (14).
[0108]
d (j) = d (0) + B (j−1) − {T × (j−1) / MBcnt} (12)
[0109]
Here, d (0) is the initial buffer capacity, B (j) is the number of encoding generation bits in the j-th macroblock, MBcnt is the number of macroblocks in the picture, and T is the target generation in units of pictures. Code amount.
[0110]
d (j + 1) = d (0) + B (i) − (T × j) / MBcnt (13)
[0111]
d (j + 1) = d (j) + {B (j) -B (j-1)}-T / MBcnt (14)
[0112]
Subsequently, since the generated code amount control unit 92 divides the macroblock in the picture into an intra slice portion and an inter slice portion, as shown in FIG. 9, the macro code in the intra slice portion and the inter slice portion Target generated code amounts Tpi and Tpp assigned to each macroblock are individually set.
[0113]
In the graph, when the number of macroblock counts is between 0 and s and between t and end, by substituting the target generated code amount Tpp of the inter slice into the following equation (15), A buffer occupancy d (j + 1) can be obtained.
[0114]

[0115]
Further, when the macroblock count is between s and t, the buffer occupancy d (j + 1) in the intra slice portion is substituted by substituting the target generated code amount Tpi of the intra slice into the following equation (16). ) Can be obtained.
[0116]
d (j + 1) = d (j) + {B (j) -B (j-1)}-Tpi / (ts) (16)
[0117]
Therefore, the generated code amount control unit 92 substitutes the buffer occupancy d (j + 1) in the intra slice portion and the inter slice portion and the constant r shown in the equation (17) into the equation (18), so that the macro The quantization index data Q (j + 1) of the block (j + 1) is calculated and supplied to the quantization unit 75.
[0118]
r = (2 × br) / pr (17)
Q (j + 1) = d (j + 1) × (31 / r) (18)
Here, br is a bit rate and pr is a picture rate.
[0119]
Further, when there is a scene change, the generated code amount control unit 92 performs any one of the Q matrix setting process 1 to the Q matrix setting process 4 described later using the flowcharts of FIGS. As needed, the quantization unit 75 controls the value of the Q matrix used for quantization.
[0120]
Based on the quantization index data Q (j + 1) and the Q matrix that is a value controlled by the processing of the generated code amount control unit 92, the quantization unit 75 responds to the intra slice or inter slice in the next macroblock. The quantization step size is determined, and the DCT coefficient data is quantized according to the quantization step size.
[0121]
Thereby, the quantization unit 75 calculates the intra slice portion and the inter slice of the next picture calculated based on the actual generated code amount data B (j) in the intra slice portion and the inter slice portion of the previous picture. The DCT coefficient data can be quantized with the optimum quantization step size for the target generated code amounts Tpp and Tpi in the portion.
[0122]
Thus, the quantization unit 75 can quantize the buffer 82 so that the buffer 82 does not overflow or underflow according to the data occupancy of the buffer 82, and quantize the decoder side VBV buffer so that it does not overflow or underflow. Quantized DCT coefficient data can be generated.
[0123]
For example, in Japanese Patent Application No. 11-205803 described above as the prior art, when the picture of a picture to be encoded next changes greatly while performing normal feedback quantization control, the feedback quantum Control is initialized, the initial buffer capacity d (0) of the virtual buffer is initialized based on the ME residual information supplied from the ME residual calculation unit 95, and based on the new initial buffer capacity d (0), The quantization index data Q (j + 1) is newly calculated for each intra slice and inter slice.
[0124]
However, as in the conventional case, if it is determined whether or not the virtual buffer adjustment is performed only with the ME residual, the image difficulty level changes from easy to difficult, and difficult to simple In both cases, the virtual buffer adjustment is performed. That is, when the image difficulty level is changed from a difficult one to a simple one, the result is that the image quality is bothered in a simple image that should have a margin for encoding.
[0125]
Therefore, in the video encoder 61 of FIG. 7, for example, using information such as the intra AC calculated by the intra AC calculation unit 91, the virtual buffer is used only at the time of a scene change that changes from easy to difficult image. By performing the adjustment, it is possible to prevent deterioration in image quality in a simple image.
[0126]
That is, the generated code amount control unit 92 performs the normal feedback type quantization control, and stops the feedback type quantization control when the picture of the picture to be encoded next changes greatly, and the ME remaining Whether to initialize the initial buffer capacity d (0) of the virtual buffer based on the ME residual information ME_info supplied from the difference calculation unit 95 and the prev_mad_info and mad_info supplied from the intra AC calculation unit 91 Is determined, and the initial buffer capacity d (0) of the virtual buffer is initialized based on the ME residual information ME_info supplied from the ME residual calculation unit 95. Is initialized. The initialization of the initial buffer capacity d (0) of the virtual buffer is the same as in the conventional case described using the equations (2) to (5).
[0127]
Then, the generated code amount control unit 92 uses the expression (12) to the expression (18) for each intra slice and inter slice based on the new initial buffer capacity d (0), and uses the quantization index data Q ( j + 1) is newly calculated and supplied to the quantization unit 75.
[0128]
Further, the generated code amount control unit 92, when the picture of the picture to be encoded next changes greatly, the quantization unit based on the bit rate, the image difficulty level, or the bit rate and the image difficulty level 75 is set to set the value of the Q matrix. Specifically, the generated code amount control unit 92 sets the value of the Q matrix so as to coarsely quantize an image that is difficult to encode and has a low bit rate or a high difficulty level. For an image that is easy to some extent and has a high bit rate and low difficulty level, the value of the Q matrix is set so that the image quality is not deteriorated even though there is a margin in rate control.
[0129]
Referring to the flowchart of FIG. 10, using image difficulty level information such as Intra AC, a scene change is introduced to determine whether or not a change from a simple image to a difficult image is performed, and the virtual buffer is adjusted. The virtual buffer update process to be performed will be described.
[0130]
In step S <b> 21, the generated code amount control unit 92 acquires ME residual information ME_info from the ME residual calculation unit 95.
[0131]
In step S22, the generated code amount control unit 92 subtracts the average value avg of the ME residual information from the acquired ME residual information, and whether ME_info−avg> D, that is, the calculated value. Is greater than a predetermined threshold value D. The average value avg of the ME residual information is a value updated in step S <b> 26 described later, and is represented by the above-described formula (1). Note that the predetermined threshold D is a value that is tuned while considering image quality.
[0132]
If it is determined in step S22 that the calculated value is smaller than the predetermined threshold value D, it is determined that there is not much difference between the pattern in the current picture and the pattern in the previous picture, that is, there has been no scene change. Thus, the process proceeds to step S26.
[0133]
If it is determined in step S22 that the calculated value is greater than the predetermined threshold D, it is determined that there is a large difference between the pattern in the current picture and the pattern in the previous picture, that is, there has been a scene change. Therefore, in step S23, the generated code amount control unit 92 executes any one of the Q matrix setting processes 1 to 4 described with reference to FIGS. Here, the Q matrix is set based on the bit rate and the difficulty level of the image after the scene change.
[0134]
In step S24, the generated code amount control unit 92 acquires from the intra AC calculation unit 91 the mad_info value of the intra AC after the scene change and the prev_mad_info value of the intra AC before the scene change. And mad_info> prev_mad_info is determined.
[0135]
If it is determined in step S24 that mad_info> prev_mad_info is not satisfied, this scene change is a scene change from a difficult image to a simple image, so the process proceeds to step S26.
[0136]
If it is determined in step S24 that mad_info> prev_mad_info, this scene change is a scene change from a simple image to a difficult image. Therefore, in step S25, the generated code amount control unit 92 performs FIG. The initial buffer capacity d (0) of the virtual buffer is updated by the same processing as in the conventional case described above.
[0137]
That is, the generated code amount control unit 92 calculates the initial buffer capacity d (0) of the virtual buffer based on the above-described Expression (2), Expression (3), Expression (4), and Expression (5). Update the buffer.
[0138]
If it is determined in step S22 that the calculated value is smaller than the predetermined threshold D, if it is determined in step S24 that mad_info> prev_mad_info is not satisfied, or after the processing of step S25 is completed, in step S26 The generated code amount control unit 92 updates the average value avg of the ME residual information according to the above-described equation (1) in preparation for the next picture to be supplied, and the process returns to step S21, and thereafter The process is repeated.
[0139]
With the processing described with reference to the flowchart of FIG. 10, the virtual buffer adjustment is performed only at the time of a scene change in which the image difficulty level is changed from the easy one to the difficult one using the intra AC. In a simple image that should be, it is possible to prevent the image quality from further deteriorating.
[0140]
When a scene change occurs, any one of Q matrix setting processing 1 to Q matrix setting processing 4, which will be described later with reference to the flowcharts of FIGS. A Q matrix suitable for the degree is set.
[0141]
Next, the Q matrix setting process 1 executed in step S23 of FIG. 10 will be described with reference to the flowchart of FIG. In the Q matrix setting process 1, it is determined whether or not to change the Q matrix based on the bit rate.
[0142]
In step S41, the generated code amount control unit 92 determines whether or not the bit rate value supplied from the target code amount calculation unit 94 is smaller than a predetermined threshold A.
[0143]
If it is determined in step S41 that the bit rate value is smaller than the predetermined threshold A, in step S42, the generated code amount control unit 92 can reduce the high frequency component of the Q matrix value. A control signal for changing to a value is generated and output to the quantization unit 75. Based on the control signal supplied from the generated code amount control unit 92, the quantization unit 75 changes the value of the Q matrix when the quantization process is performed to a value that can reduce the high-frequency component. Return to step S24 in FIG.
[0144]
If it is determined in step S41 that the bit rate value is not smaller than the predetermined threshold A, in step S43, the generated code amount control unit 92 does not change the value of the Q matrix, so that the control signal is quantized. The data is not output to the conversion unit 75. Therefore, the quantization unit 75 uses a normal Q matrix value when performing the quantization process. After the process of step S43 ends, the process returns to step S24 of FIG.
[0145]
That is, since there is no room for encoding at a low bit rate, the image quality at the time of intra-slice encoding scene change can be improved by changing the value of the Q matrix to a value that can reduce high-frequency components. However, if the same processing is performed at a high bit rate, the image quality is deteriorated because the quantization is coarsely performed although there is a margin for encoding. Therefore, the Q matrix is changed only when the bit rate is lower than the predetermined threshold.
[0146]
When a scene change occurs by the process described with reference to the flowchart of FIG. 11, it is determined whether to change the Q matrix based on the bit rate. Therefore, the image quality at the time of the intra-slice encoding scene change is determined. Can be improved.
[0147]
Here, only one threshold value is used to determine the bit rate level, but it goes without saying that the bit rate level may be determined using a plurality of threshold values. In this case, a Q matrix is set such that coarser quantization is performed as the bit rate is determined to be lower.
[0148]
In the Q matrix setting process 1, it is determined whether or not to change the Q matrix based on the bit rate. However, the setting value of the Q matrix may be changed based on the image difficulty after the scene change. Good. With reference to the flowchart of FIG. 12, the Q matrix setting process 2 for changing the set value of the Q matrix based on the image difficulty after the scene change will be described.
[0149]
In step S51, the generated code amount control unit 92 compares the value of mad_info supplied from the intra AC calculation unit 91 with a predetermined threshold B, and determines whether or not mad_info> threshold B is satisfied.
[0150]
When it is determined in step S51 that mad_info> threshold value B is satisfied, the image after the scene change is an image that is difficult to encode because of high image difficulty. In step S52, the generated code amount control unit 92 generates a control signal for changing the value of the Q matrix to a first value that can drop a high-frequency component and performs coarse quantization, and outputs the control signal to the quantization unit 75. Based on the control signal supplied from the generated code amount control unit 92, the quantization unit 75 changes the value of the Q matrix when performing the quantization process to a first value that can reduce high-frequency components. The processing returns to step S24 in FIG.
[0151]
If it is determined in step S51 that mad_info> threshold value B does not hold, the image after the scene change is an image that has a low image difficulty level and is easily encoded to some extent. In step S53, the generated code amount control unit 92 generates a control signal for changing the value of the Q matrix to a second value that can reduce high-frequency components compared to the normal processing, but can be quantized more finely than the first value. And output to the quantization unit 75. Based on the control signal supplied from the generated code amount control unit 92, the quantization unit 75 changes the value of the Q matrix to a second value that can reduce the high-frequency component, and the processing is performed in the steps of FIG. Return to S24.
[0152]
When a scene change occurs by the processing described with reference to the flowchart of FIG. 12, the Q matrix value is set based on the image difficulty level after the scene change. Image quality can be improved.
[0153]
In FIG. 12, whether the value of the Q matrix is set to the first value or the second value is determined based on whether or not mad_info> threshold value B is satisfied. Based on whether or not mad_info> threshold B holds, it may be determined whether or not to change the value of the Q matrix.
[0154]
Here, only one threshold value is used to determine the level of image difficulty after a scene change. However, a plurality of threshold values may be used to determine the level of image difficulty level. Needless to say. In this case, a Q matrix is set such that coarser quantization is performed as the image difficulty level is determined to be higher.
[0155]
In the Q matrix setting process 1, it is determined whether or not to change the Q matrix based on the bit rate. In the Q matrix setting process 2, the set value of the Q matrix is determined based on the image difficulty after the scene change. However, the setting value of the Q matrix may be set based on both the bit rate and the image difficulty level after the scene change. With reference to the flowchart of FIG. 13, the Q matrix setting process 3 for changing the setting value of the Q matrix based on the bit rate after the scene change and the image difficulty level will be described.
[0156]
In step S61, the generated code amount control unit 92 determines whether the bit rate value supplied from the target code amount calculation unit 94 is smaller than a predetermined threshold A (that is, a low bit rate with no margin for encoding). It is determined whether or not.
[0157]
When it is determined in step S61 that the bit rate value is smaller than the predetermined threshold A, in step S62, the generated code amount control unit 92 determines the value of mad_info supplied from the intra AC calculation unit 91. The predetermined threshold B is compared to determine whether or not mad_info> threshold B is satisfied.
[0158]
If it is determined in step S62 that mad_info> threshold value B is satisfied, the image after the scene change is an image that is difficult to encode because the image difficulty level is high. In step S63, the generated code amount control unit 92 generates a control signal for changing the value of the Q matrix to a first value that can drop a high-frequency component and performs coarse quantization, and outputs the control signal to the quantization unit 75. Based on the control signal supplied from the generated code amount control unit 92, the quantization unit 75 changes the value of the Q matrix when executing the quantization process to a first value that can reduce high-frequency components. The processing returns to step S24 in FIG.
[0159]
If it is determined in step S62 that mad_info> threshold value B does not hold, the image after the scene change is an image that has a low image difficulty level and is easily encoded to some extent. In step S64, the generated code amount control unit 92 generates a control signal for changing the value of the Q matrix to a second value that can reduce high-frequency components compared to the normal processing, but can be quantized more finely than the first value. And output to the quantization unit 75. Based on the control signal supplied from the generated code amount control unit 92, the quantization unit 75 changes the value of the Q matrix when performing the quantization process to a second value that can reduce high-frequency components. The processing returns to step S24 in FIG.
[0160]
If it is determined in step S61 that the value of the bit rate is not smaller than the predetermined threshold A, the generated code amount control unit 92 is coarse in step S65 even though there is a margin for encoding. In order to avoid quantizing and degrading the image, the value of the Q matrix is not changed, so that the control signal is not output to the quantization unit 75. Therefore, the quantization unit 75 performs the quantization process using the normal Q matrix values. After the process of step S65 ends, the process returns to step S24 of FIG.
[0161]
The processing described with reference to the flowchart of FIG. 13 sets the Q matrix setting values based on both the bit rate and the image difficulty level after the scene change, and thus deteriorates the image more than necessary. Therefore, it is possible to suppress the occurrence of a phenomenon that the amount of code allocated to encoding is insufficient in the lower part of the image after the scene change.
[0162]
Also, by providing multiple thresholds for bit rate and image difficulty level, more detailed classification is performed, and encoding is performed at the bottom of the image after the scene change without degrading the image more than necessary. It is possible to effectively suppress the occurrence of a phenomenon that the code amount to be allocated becomes insufficient. Next, the Q matrix setting process 4 in which a plurality of thresholds for the bit rate and the image difficulty level are provided will be described with reference to the flowchart of FIG.
[0163]
In step S <b> 71, the generated code amount control unit 92 determines whether the bit rate value supplied from the target code amount calculation unit 94 is smaller than a predetermined threshold E. Here, the threshold value E is smaller than the threshold value A in the process described with reference to the flowchart of FIG. 10 or FIG. 13 in order to determine the margin of encoding in more detail at a small bit rate where encoding is difficult. A smaller value is preferable.
[0164]
When it is determined in step S71 that the bit rate value is smaller than the predetermined threshold value E, the generated code amount control unit 92 determines in step S72 the value of mad_info supplied from the intra AC calculation unit 91. The predetermined threshold G is compared to determine whether or not mad_info> threshold G holds. Here, the threshold value G is preferably set to a value that can select an image with a high image difficulty level in a relatively wide range in an image with a high bit rate that is difficult to encode. 11 or the threshold value B in the processing described with reference to the flowchart of FIG. 13 or a value smaller than the threshold value B is preferable.
[0165]
If it is determined in step S72 that mad_info> threshold G holds, the image after the scene change is an image that is difficult to encode because the image difficulty level is very high. The control unit 92 can reduce the high frequency component of the value of the Q matrix, and generates a control signal for changing the value to the third value for executing the coarsest quantization in the Q matrix setting process 4. To the conversion unit 75. Based on the control signal supplied from the generated code amount control unit 92, the quantization unit 75 changes the value of the Q matrix when executing the quantization process to the third value. The process returns to step S24.
[0166]
If it is determined in step S72 that mad_info> threshold G does not hold, the image after the scene change is an image having a low image difficulty level and is easy to encode to some extent. In step S74, the generated code amount control unit 92 can generate a control signal for changing the value of the Q matrix to a fourth value for executing the second coarsest quantization in the Q matrix setting process 4 in which a high frequency component can be dropped. To the conversion unit 75. Based on the control signal supplied from the generated code amount control unit 92, the quantization unit 75 changes the value of the Q matrix when executing the quantization process to the fourth value. The process returns to step S24.
[0167]
When it is determined in step S71 that the bit rate value is not smaller than the predetermined threshold value E, the generated code amount control unit 92 in step S75 determines the bit rate supplied from the target code amount calculation unit 94. It is determined whether or not the value is smaller than a predetermined threshold value F. Here, the threshold value F, which is a threshold value for judging the encoding margin, is larger than the threshold value E. Basically, the threshold value A in the processing described with reference to the flowchart of FIG. 10 or FIG. It is preferable that the value is substantially equivalent to.
[0168]
When it is determined in step S75 that the bit rate value is smaller than the predetermined threshold value F, in step S76, the generated code amount control unit 92 determines the value of mad_info supplied from the intra AC calculation unit 91. The predetermined threshold value H is compared to determine whether or not mad_info> threshold value H is satisfied.
[0169]
Here, the threshold value H is larger than the threshold value G. That is, in a low bit rate image whose bit rate is smaller than the threshold value E, an image having a high degree of difficulty is selected in a sufficiently wide range in order to correctly encode to the bottom of the screen, and the Q matrix in the selected image is selected. Must be set to a third value that causes coarse quantization to be performed. On the other hand, in an image whose bit rate is larger than the threshold value E and smaller than the threshold value F, there is a margin in rate control by the amount of the higher bit rate, so that the bit rate is higher than when the bit rate is smaller than the threshold value E. Only for difficulty images, the code amount is insufficient at the bottom of the screen. Accordingly, unless the threshold value H is set to a value larger than the threshold value G, the image quality is deteriorated by coarsening the quantization more than necessary in an image having a margin for rate control.
[0170]
If it is determined in step S76 that mad_info> threshold value H holds, the image after the scene change is an image that is difficult to encode because the image difficulty level is very high. In step S77, the generated code amount The control unit 92 generates a control signal for changing the value of the Q matrix to the fifth value for executing the third coarsest quantization in the Q matrix setting process 4, and outputs the control signal to the quantization unit 75. . Based on the control signal supplied from the generated code amount control unit 92, the quantization unit 75 changes the value of the Q matrix when executing the quantization process to the fifth value. The process returns to step S24.
[0171]
If it is determined in step S76 that mad_info> threshold value H does not hold, the image after the scene change is an image that has a low image difficulty level and is easily encoded to some extent. In step S78, the generated code amount control unit No. 92 executes a coarser quantization of the value of the Q matrix than the quantization in the case where there is no change, but the sixth quantization which executes the finest quantization among the values changed and set by the Q matrix setting process 4 A control signal for changing to the value is generated and output to the quantization unit 75. Based on the control signal supplied from the generated code amount control unit 92, the quantization unit 75 changes the value of the Q matrix when executing the quantization process to the sixth value. The process returns to step S24.
[0172]
When it is determined in step S75 that the bit rate value is not smaller than the predetermined threshold value F, in step S79, the generated code amount control unit 92 is coarse although there is a margin for encoding. In order to avoid quantizing and degrading the image, the value of the Q matrix is not changed, so that the control signal is not output to the quantization unit 75. Therefore, the quantization unit 75 uses a normal Q matrix value when performing the quantization process. After step S79 ends, the process returns to step S24 in FIG.
[0173]
By providing a plurality of thresholds for the bit rate and the image difficulty level by the processing described with reference to the flowchart of FIG. 14 and performing more detailed classification, the image is not deteriorated more than necessary. It is possible to effectively suppress the occurrence of a phenomenon that the amount of code allocated to encoding becomes insufficient at the lower part of the image after the scene change.
[0174]
In the process described with reference to the flowchart of FIG. 14, the case where two threshold values for the bit rate and the image difficulty level are provided has been described. However, two or more threshold values for the bit rate and the image difficulty level are provided. Needless to say, it may be provided. By setting the bit rate and image difficulty threshold values more finely in areas where image encoding is difficult (by setting many threshold values), the code is reduced in the lower part of the image after the scene change without degrading the image. It is possible to more effectively suppress the occurrence of a phenomenon that the amount of codes allocated to the conversion becomes insufficient.
[0175]
Further, in the above-described embodiment, all the frame images of numbers 0 to 11 are set as P pictures as low delay coding. For example, the upper part of the frame image in the frame size of 45 macro blocks horizontally and 24 macro blocks vertically. In the above description, the region corresponding to 2 macroblocks in the vertical direction and 45 macroblocks in the horizontal direction is set as one intra slice portion, and all the others are set as inter slice portions. The portion may be formed with regions of various other sizes, such as a region for one vertical macroblock and 45 horizontal macroblocks.
[0176]
Further, here, the case where low delay encoding is performed has been described as an example. However, in the present invention, for example, 15 frames are encoded into an intra-frame encoded image (hereinafter referred to as an I picture), an inter-frame forward prediction encoded image. (Hereinafter referred to as a P picture), or a bi-predictive encoded image (hereinafter referred to as a B picture). The present invention can also be applied to a case where a frame image is encoded according to the image type (I picture, P picture, or B picture).
[0177]
Furthermore, in the above-described embodiment, the case where the present invention is applied to the video encoder 61 as an encoding apparatus that performs compression encoding by the MPEG system has been described. You may make it apply to the encoding apparatus by various image compression systems.
[0178]
The series of processes described above can be executed by hardware, but can also be executed by software. In this case, for example, the video encoder 61 is configured by a personal computer 101 as shown in FIG.
[0179]
In FIG. 15, the CPU 111 executes various processes according to a program stored in the ROM 112 or a program loaded from the storage unit 118 to the RAM 113. The RAM 113 also appropriately stores data necessary for the CPU 111 to execute various processes.
[0180]
The CPU 111, ROM 112, and RAM 113 are connected to each other via a bus 114. An input / output interface 115 is also connected to the bus 114.
[0181]
The input / output interface 115 includes an input unit 116 including a keyboard and a mouse, an output unit 117 including a display and a speaker, a storage unit 118 including a hard disk, and a communication unit 119 including a modem and a terminal adapter. It is connected. The communication unit 119 performs communication processing via a network including the Internet.
[0182]
A drive 120 is connected to the input / output interface 115 as necessary, and a magnetic disk 131, an optical disk 132, a magneto-optical disk 133, a semiconductor memory 134, or the like is appropriately mounted, and a computer program read from them is loaded. If necessary, it is installed in the storage unit 118.
[0183]
When a series of processing is executed by software, a program constituting the software executes various functions by installing a computer incorporated in dedicated hardware or various programs. For example, a general-purpose personal computer is installed from a network or a recording medium.
[0184]
As shown in FIG. 15, this recording medium is distributed to supply a program to a user separately from the apparatus main body, and includes a magnetic disk 131 (including a floppy disk) and an optical disk 132 (including a floppy disk). Package media including CD-ROM (compact disk-read only memory), DVD (including digital versatile disk), magneto-optical disk 133 (including MD (mini-disk) (trademark)), or semiconductor memory 134 In addition to being configured, it is configured by a ROM 112 in which a program is stored and a hard disk included in the storage unit 118 supplied to the user in a state of being incorporated in the apparatus main body in advance.
[0185]
In the present specification, the step of describing the program stored in the recording medium is not limited to the processing performed in chronological order in the order in which it is included, but is not necessarily processed in chronological order, either in parallel or individually. The process to be executed is also included.
[0186]
【The invention's effect】
According to the present invention, image data can be encoded.
Moreover, according to the present invention, To quantize Setting of coefficient (Q matrix) to be used When it is determined that a scene change has occurred, the value at which the high frequency component is dropped is lower when the bit rate is lower than when the bit rate is high. Since the change is made, it is possible to effectively suppress the occurrence of a phenomenon in which the amount of code allocated to encoding is insufficient at the bottom of the image after the scene change without degrading the image more than necessary at the time of the scene change. Can be.
According to another aspect of the present invention, in addition to encoding image data, To quantize Set the coefficient (Q matrix) used When it is determined that a scene change has occurred, the high frequency component is reduced when the image difficulty level is higher than when the image difficulty level is low. Since the change is made, it is possible to effectively suppress the occurrence of a phenomenon in which the amount of code allocated to encoding is insufficient at the bottom of the image after the scene change without degrading the image more than necessary at the time of the scene change. Can be.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating processing when video data is compression-encoded according to the MPEG2 system and when image data that has been compression-encoded is decoded.
FIG. 2 is a diagram illustrating a VBV buffer.
FIG. 3 is a diagram illustrating low delay coding.
FIG. 4 is a diagram illustrating a VBV buffer.
FIG. 5 is a diagram illustrating a scene change.
FIG. 6 is a flowchart illustrating a conventional virtual buffer update process.
FIG. 7 is a block diagram showing a configuration of a video encoder to which the present invention is applied.
FIG. 8 is a diagram for explaining a buffer occupation amount of a virtual buffer;
FIG. 9 is a diagram illustrating the buffer occupancy of a virtual buffer for each intra slice and inter slice.
FIG. 10 is a flowchart illustrating a virtual buffer update process to which the present invention is applied.
FIG. 11 is a flowchart for describing Q matrix setting processing 1;
FIG. 12 is a flowchart illustrating Q matrix setting processing 2;
FIG. 13 is a flowchart illustrating Q matrix setting processing 3;
FIG. 14 is a flowchart illustrating Q matrix setting processing 4;
FIG. 15 is a block diagram illustrating a configuration of a personal computer.
[Explanation of symbols]
61 video encoder, 71 preprocessing unit, 72 calculation unit, 73 motion vector detection unit, 74 DCT unit, 75 quantization unit, 77 VLC unit, 78 inverse quantization unit, 79 inverse DCT unit, 80 calculation unit, 81 motion compensation Unit, 82 buffer, 83 quantization control unit, 84 frame memory, 91 intra AC calculation unit, 92 generated code amount control unit, 93 GC calculation unit, 94 target code amount calculation unit, 95 ME residual calculation unit

Claims (21)

In an encoding device for encoding a frame image
First detection means for detecting a change in pattern between a first picture that is the previous picture and a second picture that is the next picture to be encoded;
Determining means for determining whether or not a scene change has occurred between the first picture and the second picture based on a detection result by the first detecting means ;
Determining means for determining quantization index data using the value of the initial buffer capacity of the virtual buffer;
Setting means for setting a Q matrix so that when the determination means determines that the scene change has occurred, the frame image has a value for dropping high-frequency components when the bit rate of the frame image is lower than when the frame image is high ; ,
Quantization means for performing quantization based on the quantization index data determined by the determination means and the Q matrix set by the setting means;
Encoding means for encoding the quantized coefficient data quantized by the quantizing means;
Encoding device, characterized in that it comprises a. The setting means, if the bit rate is less than the predetermined threshold value, so that the value of dropping the high frequency component, the encoding device according to claim 1, characterized in that to set the Q matrix. The setting means compares a plurality of threshold values with the bit rate, and sets the Q matrix based on the comparison result so that the lower the bit rate, the lower the high frequency component. The encoding device according to claim 1 . The first detection means calculates an index indicating a difference between the pattern of the first picture and the pattern of the second picture, and detects a change in the pattern based on the index. The encoding device according to claim 1. A second detection means for detecting an image difficulty level of the second picture;
The setting unit sets the Q matrix so that the lower the bit rate is, the lower the high frequency component is when the determination unit determines that the scene change has occurred, and the bit rate is predetermined. When the value is lower than the threshold value, the Q is such that the high frequency component is dropped when the image difficulty level of the second picture detected by the second detection means is higher than when the image difficulty level is low. The encoding apparatus according to claim 1, wherein a matrix is set. The second detection means calculates the complexity of the frame image, which is a statistic having a correlation with the difficulty of the pattern of the frame image and the data amount after compression, as an index indicating the image difficulty level , 6. The encoding apparatus according to claim 5 , wherein an image difficulty level of the second picture is detected based on an index. The setting means, when the image difficulty is higher than the predetermined threshold, that the image difficulty to set the Q matrix such that the value of dropping the high frequency component than is lower than the predetermined threshold The encoding device according to claim 5 . The setting means includes
When the bit rate is smaller than the first threshold and the image difficulty level is larger than the second threshold, the Q matrix is set to be the first Q matrix that reduces the highest frequency component,
When the bit rate is smaller than the first threshold and the image difficulty level is smaller than the second threshold, the second Q matrix is quantized more finely than the first Q matrix. Set the Q matrix,
The bit rate is a value between the first threshold value and a third threshold value greater than the first threshold value, and the image difficulty level is greater than a fourth threshold value greater than the second threshold value. may have a size, such that the third Q matrix performing finer quantization than said second Q matrix, and set the Q matrix,
When the bit rate is a value between the first threshold and the third threshold, and the image difficulty level is smaller than the fourth threshold, the bit rate is quantized more finely than the third Q matrix. The encoding apparatus according to claim 5 , wherein the Q matrix is set so as to be a fourth Q matrix that performs . The encoding apparatus according to claim 1, wherein all the frame images are inter-frame forward prediction encoded images. In an encoding method of an encoding device for encoding a frame image,
A detection step of detecting a change in pattern between a first picture that is a previous picture and a second picture that is a picture to be encoded next;
Based on the detection result by the processing of said detecting step, a determining step of the scene change between the first and the second picture from the picture, it is determined whether or not occurred,
A determination step of determining quantization index data using a value of an initial buffer capacity of the virtual buffer;
Setting that sets the Q matrix so that when it is determined that the scene change has occurred by the processing of the determination step, the value of dropping the high-frequency component is lower when the bit rate of the frame image is lower than when the frame image is high Steps,
A quantization step for performing quantization based on the quantization index data determined by the determination step processing and the Q matrix set by the setting step processing;
An encoding step for encoding the quantized coefficient data quantized by the quantization step;
The encoding method characterized by including . A program for an encoding device for encoding a frame image,
A detection step of detecting a change in pattern between a first picture that is a previous picture and a second picture that is a picture to be encoded next;
Based on the detection result by the processing of said detecting step, a determining step of the scene change between the first and the second picture from the picture, it is determined whether or not occurred,
A determination step of determining quantization index data using a value of an initial buffer capacity of the virtual buffer;
Setting that sets the Q matrix so that when it is determined that the scene change has occurred by the processing of the determination step, the value of dropping the high-frequency component is lower when the bit rate of the frame image is lower than when the frame image is high Steps,
A quantization step for performing quantization based on the quantization index data determined by the determination step processing and the Q matrix set by the setting step processing;
An encoding step for encoding the quantized coefficient data quantized by the quantization step;
A recording medium on which is recorded a program that causes a computer to execute processing characterized by including: A computer-executable program for controlling an encoding device for encoding a frame image,
A detection step of detecting a change in pattern between a first picture that is a previous picture and a second picture that is a picture to be encoded next;
Based on the detection result by the processing of said detecting step, a determining step of the scene change between the first and the second picture from the picture, it is determined whether or not occurred,
A determination step of determining quantization index data using a value of an initial buffer capacity of the virtual buffer;
Setting to set the Q matrix so that when it is determined that the scene change has occurred by the processing of the determination step, the value of dropping the high-frequency component is lower when the bit rate of the frame image is lower than when the frame image is high Steps,
A quantization step for performing quantization based on the quantization index data determined by the determination step processing and the Q matrix set by the setting step processing;
An encoding step for encoding the quantized coefficient data quantized by the quantization step;
A program for causing a computer to execute a process characterized by including: In an encoding device for encoding a frame image,
A first detecting means for detecting an image difficulty level of a first picture that is a picture to be encoded next;
Second detection means for detecting a change in pattern between a second picture that is a previous picture and the first picture;
Determining means for determining whether or not a scene change has occurred between the second picture and the first picture based on a detection result by the second detecting means;
Determining means for determining quantization index data using the value of the initial buffer capacity of the virtual buffer;
When it is determined by the determining means that the scene change has occurred, the higher the image difficulty level detected by the first detecting means than when the image difficulty level is lower, the lower the high-frequency component value. Setting means for setting a Q matrix;
Quantization means for performing quantization based on the quantization index data determined by the determination means and the Q matrix set by the setting means;
Encoding means for encoding the quantized coefficient data quantized by the quantizing means;
Encoding device, characterized in that it comprises a. The setting means, when the image difficulty is higher than the predetermined threshold value, the encoding apparatus according to claim 13, characterized in that to set the Q matrix such that the value of dropping the high frequency components. The setting means compares a plurality of the predetermined threshold values with the image difficulty level, and sets the Q matrix so that the higher the image difficulty level, the lower the high frequency component based on the comparison result. The encoding apparatus according to claim 13 . The second detection means calculates an index indicating a difference between the pattern of the first picture and the pattern of the second picture, and detects a change in the pattern based on the index. The encoding device according to claim 13 . The first detection means calculates the complexity of the frame image, which is a statistical quantity having a correlation with the difficulty of the pattern of the frame image and the amount of data after compression, as an index indicating the degree of difficulty of the image , The encoding device according to claim 13 , wherein an image difficulty level of the first picture is detected based on an index. The encoding apparatus according to claim 13 , wherein all the frame images are inter-frame forward prediction encoded images. In an encoding method of an encoding device for encoding a frame image,
A first detection step of detecting an image difficulty level of a first picture that is a picture to be encoded next;
A second detection step of detecting a change in pattern between a second picture that is a previous picture and the first picture;
A determination step of determining whether a scene change has occurred between the second picture and the first picture based on a detection result obtained by the processing of the second detection step;
A determination step of determining quantization index data using a value of an initial buffer capacity of the virtual buffer;
When it is determined that the scene change has occurred by the process of the determination step, a value for dropping the high frequency component is higher when the image difficulty level detected by the process of the first detection step is higher than when the image difficulty level is low. so that, a setting step of setting a Q matrix,
A quantization step for performing quantization based on the quantization index data determined by the determination step processing and the Q matrix set by the setting step processing;
An encoding step for encoding the quantized coefficient data quantized by the quantization step;
The encoding method characterized by including . A program for an encoding device for encoding a frame image,
A first detection step of detecting an image difficulty level of a first picture that is a picture to be encoded next;
A second detection step of detecting a change in pattern between a second picture that is a previous picture and the first picture;
A determination step of determining whether a scene change has occurred between the second picture and the first picture based on a detection result obtained by the processing of the second detection step;
A determination step of determining quantization index data using a value of an initial buffer capacity of the virtual buffer;
When it is determined that the scene change has occurred by the process of the determination step, a value for dropping the high frequency component is higher when the image difficulty level detected by the process of the first detection step is higher than when the image difficulty level is low. so that, a setting step of setting a Q matrix,
A quantization step for performing quantization based on the quantization index data determined by the determination step processing and the Q matrix set by the setting step processing;
An encoding step for encoding the quantized coefficient data quantized by the quantization step;
A recording medium on which is recorded a program that causes a computer to execute processing characterized by including: A computer-executable program for controlling an encoding device for encoding a frame image,
A first detection step of detecting an image difficulty level of a first picture that is a picture to be encoded next;
A second detection step of detecting a change in pattern between a second picture that is a previous picture and the first picture;
A determination step of determining whether a scene change has occurred between the second picture and the first picture based on a detection result obtained by the processing of the second detection step;
A determination step of determining quantization index data using a value of an initial buffer capacity of the virtual buffer;
When it is determined that the scene change has occurred by the process of the determination step, a value for dropping the high frequency component is higher when the image difficulty level detected by the process of the first detection step is higher than when the image difficulty level is low. so that, a setting step of setting a Q matrix,
A quantization step for performing quantization based on the quantization index data determined by the determination step processing and the Q matrix set by the setting step processing;
An encoding step for encoding the quantized coefficient data quantized by the quantization step;
A program for causing a computer to execute a process characterized by including:

Images