JP4085729B2 - Signal recording apparatus and method, program, and recording medium - Google Patents

Description

【００３７】
ステップＳ４４において、閾値判定部５１は、積算値ｓｕｍが閾値Ｓ未満であるか否かを判別する。積算値ｓｕｍが閾値Ｓ未満である場合（Yes）にはステップＳ４６に進み、積算値ｓｕｍが閾値Ｓ以上である場合（No）にはステップＳ４５で閾値判定部５１がこの特徴ベクトルｆを不一致と判定して処理を終了する。このように、閾値判定部５１は、閾値Ｓを設け、積算演算の途中の階層で積算値ｓｕｍが閾値Ｓを上回った場合に階層的距離演算部５０での積算演算を打ち切ることで、処理の高速化を図っている。
【００３８】
ステップＳ４６では、成分番号ｉが特徴ベクトルｆ、ｇの次元数Ｎ以下であるか否かが判別される。成分番号ｉがＮ以下である場合（Yes）には、ステップＳ４７でｉをインクリメントしてステップＳ４３に戻る。一方、成分番号ｉがＮよりも大きい場合（No）には、特徴ベクトルｆ、ｇの最後の成分まで積算演算が終了しているため、ステップＳ４８で閾値判定部５１がこの特徴ベクトルｆを一致と判定して処理を終了する。
【００３９】
以上、図７のフローチャートでは、１つの特徴ベクトルｇに対する処理を示したが、１回の登録処理で複数の特徴ベクトルｇが登録されている場合には、それぞれについて同様の処理を行い、その何れかについて距離の積算値ｓｕｍが閾値Ｓを下回った全ての特徴ベクトルｆを検出する。この場合、例えば映像信号の特徴量と音響信号の特徴量との双方で最初に一致した時刻を最終的に一致が検出された時刻とすることもでき、また、映像信号の特徴量と音響信号の特徴量との何れか一方で最初に一致した時刻を最終的に一致が検出された時刻とすることもできる。
【００４０】
なお、以上説明した一致検出手法では、どのベクトル成分から検索を行うかはベクトルの並び順に依存するため、この並び順によって検索速度に差が生じる。そこで、特徴ベクトルｆ、ｇに対して正規直交変換行列を乗算して直交変換を行い、この直交変換後の特徴ベクトルｆ'、ｇ'を用いて有意性の高い順に検索を行うようにしても構わない。
【００４１】
直交変換としては、例えば順序行列を用いた変換、離散コサイン変換（Discrete Cosine Transform:DCT）、離散フーリエ変換（Discrete Fourier Transform:DFT）、ウォルシュ−アダマール（Walsh-Hadamard）変換、或いはカルーネン−レーベ（Karhunen-Loeve）変換等を用いることができ、このような変換を予め入力ベクトルに施して、有意性の高いベクトル成分、すなわち、上述の変換で分散や固有値の大きい成分、或いは低周波成分、低交番数（sequency）成分から順に積算演算を行うようにすることで、ベクトル成分の分布を考慮して、効率的且つ高速に特徴ベクトルｇと一致する特徴ベクトルｆを検出することができる。
【００４２】
以上説明した信号記録装置１の処理の具体例を図８に模式的に示す。図８（Ａ）に示すように、先ず、既に記録されているシリーズ番組について予め例えばタイトル部分の映像信号／音響信号の特徴量（特徴ベクトルｇ）を抽出して登録しておく。タイマー録画録音中には録画録音開始時刻ｔ_ｓ（放送開始予定時刻）から録画録音終了時刻ｔ_ｅ（放送終了予定時刻）までの録画録音が行われるが、図８（Ｂ）に示すように、映像信号／音響信号から離散時間ｔにおける特徴量（特徴ベクトルｆ）を順次抽出し、抽出された特徴量が登録された特徴量（特徴ベクトルｇ）と一致するか否かを検出する。なお、この図８（Ｂ）は、目的とするシリーズ番組が遅れて放送されている様子を表している。
【００４３】
一致位置が検出されると、図８（Ｃ）に示すように、一致した時刻ｔ_ｍと録画録音開始時刻ｔ_ｓとを比較する。そして、遅れ時間ｄ（＝ｔ_ｍ−ｔ_ｓ）が許容時間を越えている場合には、図８（Ｄ）に示すように、録画録音終了時刻ｔ_ｅを遅れ時間ｄだけ延長し、時刻ｔ'_ｅまで録画録音を行うようにする。
【００４４】
このように、第１の実施の形態における信号記録装置１は、タイマー録画録音中に、予め登録された特徴量と現在入力されている映像信号／音響信号の特徴量とを比較し、一致時刻が放送開始予定時刻から許容時間以上遅れている場合には、その遅れ分だけ録画録音終了時刻を延長する。これにより、何らかの理由により録画録音対象の番組が遅れた場合であっても、番組終了まで確実に録画録音することができる。
【００４５】
（２）第２の実施の形態
上述した第１の実施の形態では、図９（Ａ）に示すように、誤動作を防止するために許容時間を設定し、録画録音開始時刻から許容時間内に登録されたタイトル映像やタイトル音声が放送されれば放送の遅れがないと見なし、録画録音終了時刻を延長することなく録画録音することができた。
【００４６】
しかしながら、図９（Ｂ）に示すように、番組によっては、先ず番組のいわゆるイントロダクション部分を放送し、番組開始からやや長い時間が経過してからタイトル映像やタイトル音声が放送される場合がある。上述した手法でこのような場合にも対応するためには許容時間を十分に長く取る必要があるが、逆に長く取り過ぎると放送開始時刻の遅れがあった場合にも許容時間内として扱われ、自動延長が動作しないような場合が発生し得る。
【００４７】
そこで、以下に説明する第２の実施の形態では、（ａ）番組中にコマーシャルメッセージ（以下、単にコマーシャル又はＣＭという。）が放送される民間放送局の場合には、番組間に必ずコマーシャルが挿入されること、及び（ｂ）長いイントロ部分がある場合でも、ほぼ必ず最初の番組ブロックでタイトル映像やタイトル音声が放送されること、を用いることで、番組開始からやや長い時間が経過してからタイトル映像やタイトル音声が放送される場合にも対応可能とする。ここで番組ブロックとは、コマーシャルとコマーシャルとの間の番組部分である。
【００４８】
第２の実施の形態における信号記録装置の概略構成を図１０に示す。図１０に示すように、信号記録装置２の基本構成は図１に示した信号記録装置１と同様であるが、ＣＭ検出部２０を有し、放送された映像信号／音響信号から各コマーシャル開始時刻及びその長さを検出できる点に特徴を有する。したがって、先に図１に示した信号記録装置１と同様の構成については同一符号を付して詳細な説明を省略する。
【００４９】
なお、このコマーシャル検出の手法としては、例えば本件発明者が先に提案している特開２００２−１６８７３号公報に記載された技術を用いることができる。この技術は、以下のように要約される。
【００５０】
すなわち、日本国内においては、特殊な例を除くほぼ全てのコマーシャルが１５秒、３０秒、６０秒の長さで製作されており、また、各コマーシャルの前後では必然的に音声レベルが下がり、映像信号が切り替わるため、これらを検出の「必須条件」とする。また、規格に基づいて製作されるという制約、短い時間で宣伝効果を上げるためのものであるという制約、及び番組構成上の都合などによる制約の元で製作された結果として現れる一定の傾向を示す特徴を検出の「付加条件」とする。さらに、上記付加条件を満たす区間がオーバーラップして存在する場合には、少なくともその何れかの区間は正しいコマーシャル区間では有り得ないという条件を検出の「論理条件」とする。そして、「必須条件」に基づき決定論的にコマーシャル候補を抽出し、「付加条件」に基づくコマーシャルらしさの統計論的な評価により候補を選択し、「論理条件」により候補のオーバーラップ関係を解消することにより、高精度にコマーシャルを検出することができる。
【００５１】
この信号記録装置２の動作についても、先に図２に示したような３つのステップに大別される。すなわち、通常の録画録音処理が行われた後に、録画録音された映像信号／音響信号の一部区間を登録する登録処理が行われ、さらに自動延長付きのタイマー録画録音処理が行われる。
【００５２】
先ず、信号記録装置２の通常録画録音処理について図１１のフローチャートを用いて説明する。ステップＳ５０において、録画録音制御部１０（図１０）は、図示しない同復調装置により復調された映像信号／音響信号を入力し、録画録音制御情報に従って記録部１１に記録を開始する。ここで、録画録音制御情報は、上述したように録画録音開始時刻や録画録音終了時刻などの情報であり、タイマー等によって予め設定しておいてもよく、映像／音声を視聴中のユーザの操作によるものであってもよい。
【００５３】
録画録音中には、ステップＳ５１において、録画録音制御部１０が記録部１１に映像信号／音響信号を記録すると同時に、ステップＳ５２において、特徴抽出部１５が映像信号／音響信号の特徴量を抽出し、その特徴量を映像信号／音響信号と併せて記録部１１に記録する。さらにステップＳ５３において、ＣＭ検出部２０が映像信号／音響信号中の各コマーシャル開始時刻及びその長さを検出し、映像信号／音響信号と併せて記録部１１に記録する。
【００５４】
そしてステップＳ５４において、録画録音制御部１０は、録画録音制御情報に従って記録部１１への映像信号／音響信号の記録を終了し、同時に特徴抽出部１５も記録部１１への特徴量の記録を終了する。
【００５５】
次に、登録処理が行われるが、この登録処理は第１の実施の形態と同様であるため、説明を省略する。
【００５６】
続いて、信号記録装置２の自動延長付きタイマー録画録音処理について図１２のフローチャートを用いて説明する。ステップＳ６０において、録画録音制御部１０は、予め録画録音開始時刻、録画録音終了時刻、チャンネルなどの録画録音制御情報を設定し、さらにステップＳ６１において許容時間を設定する。ここで許容時間とは、登録されている映像信号／音響信号の直前のコマーシャルの終了時刻が録画録音開始時刻からこの時間範囲内であろうと予測される時間である。なお、毎回の録画録音動作で新たに設定する必要はなく、予め既定値を用意しておくこともできる。
【００５７】
次にステップＳ６２において、録画録音制御部１０は、録画録音開始時刻になったか否かを判別し、録画録音開始時刻でない場合（No）には判別処理を続け、録画録音開始時刻である場合（Yes）にはステップＳ６３に進む。
【００５８】
ステップＳ６３では、前述した通常の録画録音処理と同様の処理が行われる。すなわち、録画録音制御部１０は、記録部１１に映像信号／音響信号を記録し、同時に特徴抽出部１５は、映像信号／音響信号の特徴量を抽出し、その特徴量を映像信号／音響信号と併せて記録部１１に記録する。さらにＣＭ検出部２０は、映像信号／音響信号中の各コマーシャル開始時刻及びその長さを検出し、映像信号／音響信号と併せて記録部１１に記録する。
【００５９】
続いてステップＳ６４において、録画録音制御部１０は、録画録音終了時刻になったか否かを判別し、録画録音終了時刻である場合（Yes）にはステップＳ６５で録画録音処理を停止し、一連の処理を終了する。一方、録画録音終了時刻でない場合（No）にはステップＳ６６に進む。
【００６０】
ステップＳ６６において、一致検出部１３は、登録部１４に登録された特徴量と特徴抽出部１５で抽出された現在録画録音中の映像信号／音響信号の特徴量とを比較して一致しているか否かを判別し、一致していない場合（No）にはステップＳ６３に戻って録画録音処理を続ける。一方、一致している場合（Yes）にはステップＳ６７に進む。
【００６１】
ステップＳ６７において、録画録音制御部１０は、現在時刻の直前のコマーシャルの終了時刻を得る。
【００６２】
ステップＳ６８において、録画録音制御部１０は、直前のコマーシャルの終了時刻が録画録音開始時刻から許容時間内にあるか否かを判別し、許容時間内である場合（Yes）にはステップＳ６３に戻って録画録音処理を続ける。一方、許容時間を越えている場合（No）には何らかの理由により放送開始時刻が遅れていると判断し、ステップＳ６９で録画録音制御部１０が録画録音終了時刻を延長修正した後、ステップＳ６３に戻って録画録音処理を続ける。ここで、延長時間幅は、例えば直前のコマーシャルの終了時刻から録画録音開始時刻を減算した時間幅とすることで、番組終了まで確実に録画録音することができる。
【００６３】
以上説明した信号記録装置２の処理の具体例を図１３に模式的に示す。第１の実施の形態と同様にして一致位置が検出されると、図１３（Ａ）に示すように、一致した時刻ｔ_ｍの直前のコマーシャル終了時刻ｔ_ＣＭと録画録音開始時刻ｔ_ｓとを比較する。そして、遅れ時間ｄ（＝ｔ_ＣＭ−ｔ_ｓ）が許容時間を越えている場合には、図１３（Ｂ）に示すように、録画録音終了時刻ｔ_ｅを遅れ時間ｄだけ延長し、時刻ｔ'_ｅまで録画録音を行うようにする。
【００６４】
このように、第２の実施の形態における信号記録装置２は、タイマー録画録音中に、予め登録された特徴量と現在入力されている映像信号／音響信号の特徴量とを比較し、一致位置の直前のコマーシャル終了時刻が録画録音開始時刻から許容時間以上遅れている場合には、その遅れ分だけ録画録音終了時刻を延長する。これにより、何らかの理由により録画録音対象の番組が遅れた場合であっても、番組終了まで確実に録画録音することができる。
【００６５】
また、番組のイントロダクション部分が長い場合であっても許容時間を長くとる必要がないため、放送開始時刻の遅れがあった場合にも許容時間内として扱われて自動延長が動作しなくなるようなことが起こらない。
【００６６】
さらに、第１の実施の形態における信号記録装置１と比較して、延長修正時間が通常短くなるため、記録領域を節約することができる。
【００６７】
（３）第３の実施の形態
上述した第１、第２の実施の形態では、自動録画録音時間延長動作を行うためには、予め対象とする番組の一部区間、例えばシリーズ番組のタイトル映像やタイトル音声を登録する必要があった。これに対して、以下に説明する第３の実施の形態では、そのような登録動作を行うことなく、自動で録画録音時間延長動作を行う信号記録装置について説明する。
【００６８】
第３の実施の形態における信号記録装置の概略構成を図１４に示す。図１４に示すように、信号記録装置３の基本構成は図１０に示した信号記録装置２と同様であるが、登録部がない点に特徴を有する。したがって、先に図１０に示した信号記録装置２と同様の構成については同一符号を付して詳細な説明を省略する。
【００６９】
この信号記録装置３の動作は、先に図２に示したような３つのステップのうち、録画録音された映像信号／音響信号の一部区間を登録する登録処理を省略したものである。すなわち、通常の録画録音処理が行われた後に、自動延長付きのタイマー録画録音処理が行われる。
【００７０】
先ず、信号記録装置３の通常録画録音処理が行われるが、この通常録画録音処理は第２の実施の形態と同様であるため、説明を省略する。
【００７１】
続いて、信号記録装置３の自動延長付きタイマー録画録音処理について図１５のフローチャートを用いて説明する。ステップＳ７０において、録画録音制御部１０は、予め録画録音開始時刻、録画録音終了時刻、チャンネルなどの録画録音制御情報を設定し、さらにステップＳ７１において許容時間を設定する。ここで許容時間とは、第２の実施の形態と同様に、登録されている映像信号／音響信号の直前のコマーシャルの終了時刻が録画録音開始時刻からこの時間範囲内であろうと予測される時間である。なお、毎回の録画録音動作で新たに設定する必要はなく、予め既定値を用意しておくこともできる。
【００７２】
次にステップＳ７２において、録画録音制御部１０は、録画録音開始時刻になったか否かを判別し、録画録音開始時刻でない場合（No）には判別処理を続け、録画録音開始時刻である場合（Yes）にはステップＳ７３に進む。
【００７３】
ステップＳ７３では、録画録音制御部１０は、記録部１１に映像信号／音響信号を記録し、同時に特徴抽出部１５は、映像信号／音響信号の特徴量を抽出する。さらにＣＭ検出部２０は、映像信号／音響信号中の各コマーシャル開始時刻及びその長さを検出する。そして、特徴抽出部１５は、コマーシャル部分を除いた番組部分についての特徴量を映像信号／音響信号と併せて記録部１１に記録し、ＣＭ検出部２０は、検出した各コマーシャル開始時刻及びその長さを記録部１１に記録する。
【００７４】
続いてステップＳ７４において、録画録音制御部１０は、録画録音終了時刻になったか否かを判別し、録画録音終了時刻である場合（Yes）にはステップＳ７５で録画録音処理を停止し、一連の処理を終了する。一方、録画録音終了時刻でない場合（No）にはステップＳ７６に進む。
【００７５】
ステップＳ７６において、一致検出部１３は、既に記録された対象番組の全ての特徴量と特徴抽出部１５で抽出された現在録画録音中の映像信号／音響信号の特徴量とを比較して、その何れかと一致しているか否かを判別する。そして、何れとも一致していない場合（No）にはステップＳ７３に戻って録画録音処理を続け、何れかと一致している場合（Yes）にはステップＳ７７に進む。
【００７６】
ステップＳ７７において、録画録音制御部１０は、現在時刻の直前のコマーシャルの終了時刻を得る。
【００７７】
ステップＳ７８において、録画録音制御部１０は、直前のコマーシャルの終了時刻が録画録音開始時刻から許容時間内にあるか否かを判別し、許容時間内である場合（Yes）にはステップＳ７３に戻って録画録音処理を続ける。一方、許容時間を越えている場合（No）には何らかの理由により放送開始時刻が遅れていると判断し、ステップＳ７９で録画録音制御部１０が録画録音終了時刻を延長修正した後、ステップＳ７３に戻って録画録音処理を続ける。ここで、延長時間幅は、例えば直前のコマーシャルの終了時刻から録画録音開始時刻を減算した時間幅とすることで、番組終了まで確実に録画録音することができる。
【００７８】
以上説明した信号記録装置３の処理の具体例を図１６に模式的に示す。図１６（Ａ）に示すように、先ず、シリーズ番組について予め例えばコマーシャル部分を除いた番組部分について、短時間フレーム毎に映像信号／音響信号の特徴量（特徴ベクトルｇ）を抽出して記録しておく。タイマー録画録音中には録画録音開始時刻ｔ_ｓ（放送開始予定時刻）から録画録音終了時刻ｔ_ｅ（放送終了予定時刻）までの録画録音が行われるが、図１６（Ｂ）に示すように、映像信号／音響信号から離散時間ｔにおける特徴量（特徴ベクトルｆ）を順次抽出し、抽出されたコマーシャル部分以外の特徴量（特徴ベクトルｆ）が記録された特徴量（特徴ベクトルｇ）の何れかと一致するか否かを検出する。なお、この図１６（Ｂ）は、目的とするシリーズ番組が遅れて放送されている様子を表している。
【００７９】
一致位置が検出されると、図１６（Ｃ）に示すように、一致した時刻ｔ_ｍとの直前のコマーシャル終了時刻ｔ_ＣＭと録画録音開始時刻ｔ_ｓとを比較する。そして、遅れ時間ｄ（＝ｔ_ＣＭ−ｔ_ｓ）が許容時間を越えている場合には、図１６（Ｄ）に示すように、録画録音終了時刻ｔ_ｅを遅れ時間ｄだけ延長し、時刻ｔ'_ｅまで録画録音を行うようにする。
【００８０】
このように、第３の実施の形態における信号記録装置３は、タイマー録画録音中に、予め記録された対象番組の特徴量と現在入力されている映像信号／音響信号の特徴量とを比較し、一致位置の直前のコマーシャル終了時刻が放送開始予定時刻から許容時間以上遅れている場合には、その遅れ分だけ録画録音時間を延長する。これにより、何らかの理由により録画録音対象の番組が遅れた場合であっても、番組終了まで確実に録画録音することができる。
【００８１】
また、抽出された特徴量と既に記録された番組全体の特徴量との一致検出を行うため、第１、第２の実施の形態よりも比較回数が増加するものの、登録処理を必要としないため、ユーザ負担が少なくなる。
【００８２】
なお、以上の説明では、第２の実施の形態と同様にコマーシャル終了時刻を用いるものとしたが、これに限定されるものではなく、第１の実施の形態のようにコマーシャル終了時刻を用いずに、一致が検出された時刻が許容時間内か否かを判別するようにしても構わない。
【００８３】
（４）特徴ベクトルの抽出
以下では、特徴抽出部１５において、音響信号、映像信号からそれぞれ音響特徴ベクトルａ、映像特徴ベクトルｖを抽出する手法について説明する。
【００８４】
（４−１）音響特徴ベクトルの抽出
（４−１−１）
音響信号に関する特徴量としてパワースペクトル係数を用いる場合の例について、図１７のフローチャートと図１８を用いて説明する。先ずステップＳ８０において、図１８に示すように、対象時区間内の音響信号から時間区間Ｔの音響信号を取得する。
【００８５】
次にステップＳ８１では、取得した音響信号に対して例えば高速フーリエ変換等のスペクトル演算を施し、短時間区間毎にパワースペクトル係数Ｓ_ｑ（ｑ＝０，１，…，Ｑ−１）を求める。ここで、ｑは離散周波数を表すインデックスであり、Ｑは最大離散周波数である。
【００８６】
続いてステップＳ８２では、対象時区間内の計算を終えたか否かが判別され、終えている場合（Yes）にはステップＳ８３に進み、終えていない場合（No）にはステップＳ８０に戻る。
【００８７】
ステップＳ８３では、求めたパワースペクトル係数Ｓ_ｑの平均スペクトルＳ'_ｑを計算し、ステップＳ８４においてこの平均スペクトルＳ'_ｑをベクトル化し、音響特徴ベクトルａを生成する。この音響特徴ベクトルａは、例えば以下の式（２）のように表される。
【００８８】
【数２】

【００８９】
なお、上述の例では、対象時区間内の音響信号を時間区間Ｔ毎に区切るものとして説明したが、対象時区間が短い場合には、時間区間Ｔ毎に区切らずに、スペクトル演算を施すようにしても構わない。
【００９０】
また、上述の例では、パワースペクトル係数を用いた例について説明したが、これに限定されるものではなく、例えば等価な情報を持つケプストラム係数等を用いることもできる。さらに、フーリエ変換を用いず、ＡＲ（Auto-Regressive）モデルを用いる線形予測係数を用いても同様の効果を得ることができる。
【００９１】
（４−１−２）
音響信号は膨大であるため、圧縮符号化されて記録、或いは伝送されることが多い。符号化された音響信号を復号してベースバンドに戻した後、上述の手法を用いて音響特徴ベクトルａを抽出することも可能であるが、部分的な復号のみで音響特徴ベクトルａを抽出できれば、抽出処理を効率化、高速化することができる。
【００９２】
ここで、一般的に用いられる符号化法である変換符号化では、図１９に示すように、原音となる音響信号が時間区間Ｔ毎にフレームに区切られる。そして、そのフレーム毎の音響信号に対して変更離散コサイン変換（Modified Discrete Cosine Transform:MDCT）等の直交変換が施され、その係数が量子化されて符号化される。この際、周波数帯域毎に、大きさの正規化係数であるスケールファクタが抽出され、別途符号化される。そこで、このスケールファクタのみを復号することにより、音響特徴ベクトルａとして用いることができる。
【００９３】
このように、音響信号に関する特徴量としてスケールファクタを用いる場合の例について、図２０のフローチャートと図２１を用いて説明する。先ずステップＳ９０において、対象時区間における時間区間Ｔ内の符号化音響信号が取得され、ステップＳ９１において、フレーム毎のスケールファクタが部分的に復号される。
【００９４】
続いてステップＳ９２では、対象時区間内の復号を終えたか否かが判別され、終えている場合（Yes）にはステップＳ９３に進み、終えていない場合（No）にはステップＳ９０に戻る。
【００９５】
ステップＳ９３では、対象時区間内のスケールファクタの中から各帯域毎に最大のスケールファクタを検出し、ステップＳ９４においてそれらをベクトル化し、音響特徴ベクトルａを生成する。
【００９６】
このようにして、符号化音響信号を完全に復号することなく、上述と等価な音響特徴ベクトルａを高速に抽出することができる。
【００９７】
（４−２）映像特徴ベクトルの抽出
（４−２−１）
映像信号に関する特徴量として輝度情報及び色情報を用いる場合の例について、図２２のフローチャート及び図２３を用いて説明する。先ずステップＳ１００において、図２３に示すように、対象時区間Ｔ内の映像信号から映像フレームを取得する。
【００９８】
次にステップＳ１０１では、取得した全ての映像フレームに基づいて、時間平均画像を作成する。
【００９９】
続いてステップＳ１０２では、作成された時間平均画像を横縦Ｘ×Ｙ個の小ブロックに分割し、各ブロック内の画素値を平均したブロック平均画像を作成する。
【０１００】
そしてステップＳ１０３では、これらを例えば左上から右下へ向かってＲ，Ｇ，Ｂの順に並べて、１次元の映像特徴ベクトルｖを生成する。この映像特徴ベクトルｖは、例えば以下の式（３）のように表される。
【０１０１】
【数３】

【０１０２】
なお、上述の例では、時間平均画像を分割したブロック平均画像の画素値を並べ替えて１次元の映像特徴ベクトルｖを生成する例について説明したが、これに限定されるものではなく、ブロック平均画像を作成せずに、時間平均画像の画素値を並べ替えて１次元の映像特徴ベクトルｖを生成するようにしても構わない。
【０１０３】
また、通常、映像信号の時間変化はあまり激しくないため、時間平均画像を作成せずに、対象時区間内の１フレームを代表画像として選択して代用しても、ほぼ同様の効果を得ることができる。
【０１０４】
（４−２−２）
全く同一な映像信号でなくても、例えばニュース映像の同じアングルから撮影されたスタジオ映像など、全画像に対する色の分布が類似する映像には何らかの関連があることが多く、これらを同一視して検索する要求もある。このような場合には、画像の空間依存性を排除し、色分布のヒストグラムを作成して比較すると効果的である。
【０１０５】
そこで、このように色分布のヒストグラムを特徴量として用いる場合の例について、図２４のフローチャート及び図２５を用いて説明する。先ずステップＳ１１０において、図２５に示すように、対象時区間Ｔ内の映像信号から映像フレームを取得する。
【０１０６】
次にステップＳ１１１では、各映像フレームの信号値から、各色、例えばＲ，Ｇ，Ｂの信号値に対するヒストグラムを作成する。
【０１０７】
そしてステップＳ１１２では、これらを例えばＲ，Ｇ，Ｂの順に並べて、１次元の映像特徴ベクトルｖを生成する。この映像特徴ベクトルｖは、例えば以下の式（４）のように表される。
【０１０８】
【数４】

【０１０９】
なお、上述の例では、Ｒ，Ｇ，Ｂの信号値に対するヒストグラムを作成するものとして説明したが、輝度（Ｙ）、色差（Ｃｂ、Ｃｒ）の信号値に対するヒストグラムを作成するようにしても、同様の効果を得ることができる。
【０１１０】
（４−２−３）
映像信号は膨大であるため、圧縮符号化されて記録、或いは伝送されることが多い。符号化された映像信号を復号してベースバンドに戻した後、上述の手法を用いて映像特徴ベクトルｖを抽出することも可能であるが、部分的な復号のみで映像特徴ベクトルｖを抽出できれば、抽出処理を効率化、高速化することができる。
【０１１１】
ＭＰＥＧ１（Moving Picture Experts Group 1）又はＭＰＥＧ２で圧縮符号化された映像信号から映像特徴ベクトルｖを抽出する場合の例について、図２６のフローチャートと図２７を用いて説明する。先ずステップＳ１２０において、ベクトル化する対象時区間Ｔに対して、その直近の符号化グループ（Group of Pictures:GOP）の符号化映像信号を取得し、そのＧＯＰ内のフレーム内符号化ピクチャ（Ｉピクチャ）を取得する。
【０１１２】
ここで、フレーム画像は、マクロブロック（１６×１６画素、又は８×８画素）を単位として符号化されており、また、離散コサイン変換（ＤＣＴ）が用いられている。このＤＣＴ変換されたＤＣ係数は、マクロブロック内画像の画素値の平均値にあたる。
【０１１３】
そこで、ステップＳ１２１では、このＤＣ係数を取得し、続くステップＳ１２２では、これらを例えばＹ，Ｃｂ，Ｃｒの順に並べて、１次元の映像特徴ベクトルｖを生成する。この映像特徴ベクトルｖは、例えば以下の式（５）のように表される。
【０１１４】
【数５】

【０１１５】
このようにして、符号化映像信号を完全に復号することなく、映像特徴ベクトルｖを高速に抽出することができる。
【０１１６】
なお、上述の例では、ＭＰＥＧ１又はＭＰＥＧ２で圧縮符号化された映像信号を用いるものとして説明したが、他の圧縮符号化方式にも適用可能であることは勿論である。
【０１１７】
（５）その他
本発明は上述した実施の形態のみに限定されるものではなく、本発明の要旨を逸脱しない範囲において種々の変更が可能であることは勿論である。
【０１１８】
例えば、上述の実施の形態では、映像信号と音響信号との両方を用いるものとして説明したが、これに限定されるものではなく、例えばラジオ放送等に対しては、音響信号のみを用いるなど、何れか一方を用いても同様の効果が得られる。
【０１１９】
また、上述の実施の形態では、ハードウェアの構成として説明したが、これに限定されるものではなく、任意の処理を、ＣＰＵ（Central Processing Unit）にコンピュータプログラムを実行させることにより実現することも可能である。この場合、コンピュータプログラムは、記録媒体に記録して提供することも可能であり、また、インターネットその他の伝送媒体を介して伝送することにより提供することも可能である。
【０１２０】
【発明の効果】
以上詳細に説明したように本発明に係る信号記録装置は、信号を記録する記録手段と、上記信号から特徴量を抽出する特徴抽出手段と、所望の放送番組の信号を上記記録手段に記録する際の記録開始時刻及び記録終了時刻を設定する時刻設定手段と、予め設定された比較対象信号の特徴量と記録中の上記所望の放送番組の信号から上記特徴抽出手段によって抽出される特徴量との一致を検出する一致検出手段と、上記一致検出手段の検出結果に応じて、上記記録終了時刻を修正する時刻修正手段とを備え、上記時刻修正手段は、上記一致検出手段にて最初に特徴量が検出された番組ブロックの先頭の時刻と記録開始時刻との差分を時間差とし、さらに上記時刻修正手段は、所定の許容時間を設定し、上記時間差が上記許容時間以内であれば修正を行わず、上記許容時間を越えていれば上記時間差に応じて上記記録終了時刻を修正する。
【０１２１】
ここで、信号記録装置は、上記比較対象信号の一部区間における特徴量を登録する登録手段を備えるようにしてもよく、この場合、上記一致検出手段は、上記登録手段によって登録された上記一部区間における特徴量と記録中の上記所望の信号から抽出される特徴量との一致を検出する。
【０１２２】
また、上記時刻修正手段は、上記一致検出手段によって一致が検出された時刻と上記記録開始時刻との時間差を求め、当該時間差に応じて上記記録終了時刻を修正するようにしてもよく、上記所望の信号内におけるコマーシャルメッセージを検出するコマーシャルメッセージ検出手段を備えている場合には、上記一致検出手段によって一致が検出された時刻の直前にあたるコマーシャルメッセージの終了時刻と上記記録開始時刻との時間差を求め、当該時間差に応じて上記記録終了時刻を修正するようにしてもよい。
【０１２３】
さらに、上記時刻修正手段は、所定の許容時間を設定し、上記時間差が上記許容時間以内であれば修正を行わないようにすることもできる。
【０１２４】
また、本発明に係る信号記録方法は、所望の放送番組の信号を記録手段に記録する際の記録開始時刻及び記録終了時刻を設定する時刻設定工程と、上記所望の放送番組の信号から特徴量を抽出する特徴抽出工程と、予め設定された比較対象信号の特徴量と記録中の上記所望の放送番組の信号から上記特徴抽出工程にて抽出される特徴量との一致を検出する一致検出工程と、上記一致検出工程における検出結果に応じて、上記記録終了時刻を修正する時刻修正工程とを有し、上記時刻修正工程では、上記一致検出工程にて最初に特徴量が検出された番組ブロックの先頭の時刻と記録開始時刻との差分を時間差とし、さらに上記時刻修正工程では、所定の許容時間を設定し、上記時間差が上記許容時間以内であれば修正を行わず、上記許容時間を越えていれば上記時間差に応じて上記記録終了時刻を修正する。
【０１２５】
ここで、信号記録方法は、上記比較対象信号の一部区間における特徴量を登録する登録工程を有するようにしてもよく、この場合、上記一致検出工程では、上記登録工程にて登録された上記一部区間における特徴量と記録中の上記所望の信号から抽出される特徴量との一致が検出される。
【０１２６】
また、上記時刻修正工程は、上記一致検出工程にて一致が検出された時刻と上記記録開始時刻との時間差を求め、当該時間差に応じて上記記録終了時刻を修正するものであってもよく、上記所望の信号内におけるコマーシャルメッセージを検出するコマーシャルメッセージ検出工程を有している場合には、上記一致検出工程にて一致が検出された時刻の直前にあたるコマーシャルメッセージの終了時刻と上記記録開始時刻との時間差を求め、当該時間差に応じて上記記録終了時刻を修正するものであってもよい。
【０１２７】
このような信号記録装置及びその方法によれば、例えば放送された映像信号及び／又は音響信号を記録手段に記録する場合に、予め設定された映像信号及び／又は音響信号の特徴量と記録中の映像信号及び／又は音響信号の特徴量との一致を検出し、その検出結果に応じて、記録終了時刻を修正することができる。これにより、何らかの理由により記録対象の映像信号及び／又は音響信号の開始が遅れた場合であっても、その終了まで確実に記録することが可能となる。
【０１２８】
また、本発明に係るプログラムは、上述した信号記録処理をコンピュータに実行させるものであり、本発明に係る記録媒体は、そのようなプログラムが記録されたコンピュータ読み取り可能なものである。
【０１２９】
このようなプログラム及び記録媒体によれば、上述した信号記録処理をソフトウェアにより実現することができる。
【図面の簡単な説明】
【図１】第１の実施の形態における信号記録装置の概略構成を説明する図である。
【図２】同信号記録装置の全体的な動作を説明するフローチャートである。
【図３】同信号記録装置における通常録画録音処理を説明するフローチャートである。
【図４】同信号記録装置における登録処理を説明するフローチャートである。
【図５】同信号記録装置における自動延長付きタイマー録画録音処理を説明するフローチャートである。
【図６】同一致検出部の概略構成を説明する図である。
【図７】同一致検出部における一致検出処理を説明するフローチャートである。
【図８】同信号記録装置の処理の具体例を模式的に示す図であり、同図（Ａ）は、特徴量を登録する処理を示し、同図（Ｂ）は、タイマー録画録音中の処理を示し、同図（Ｃ）は、遅れ時間を判断する処理を示し、同図（Ｄ）は、録画録音時間を延長する処理を示す。
【図９】番組の放送形態の例を示す図であり、同図（Ａ）は、番組開始から許容時間以内にタイトルが放送される例を示し、同図（Ｂ）は、タイトル放送前に番組の一部が放送される例を示す。
【図１０】第２の実施の形態における信号記録装置の概略構成を説明する図である。
【図１１】同信号記録装置における通常録画録音処理を説明するフローチャートである。
【図１２】同信号記録装置における自動延長付きタイマー録画録音処理を説明するフローチャートである。
【図１３】同信号記録装置の処理の具体例を模式的に示す図であり、同図（Ａ）は、遅れ時間を判断する処理を示し、同図（Ｂ）は、録画録音時間を延長する処理を示す。
【図１４】第３の実施の形態における信号記録装置の概略構成を説明する図である。
【図１５】同信号記録装置における自動延長付きタイマー録画録音処理を説明するフローチャートである。
【図１６】同信号記録装置の処理の具体例を模式的に示す図であり、同図（Ａ）は、対象番組の特徴量を記録する処理を示し、同図（Ｂ）は、タイマー録画録音中の処理を示し、同図（Ｃ）は、遅れ時間を判断する処理を示し、同図（Ｄ）は、録画録音時間を延長する処理を示す。
【図１７】音響信号から音響特徴ベクトルを抽出する処理の一例を説明するフローチャートである。
【図１８】音響信号から音響特徴ベクトルを抽出する処理の一例を説明する図である。
【図１９】音響信号における変換符号化を説明する図である。
【図２０】符号化音響信号から音響特徴ベクトルを抽出する処理の一例を説明するフローチャートである。
【図２１】符号化音響信号から音響特徴ベクトルを抽出する処理の一例を説明する図である。
【図２２】映像信号から映像特徴ベクトルを抽出する処理の一例を説明するフローチャートである。
【図２３】映像信号から映像特徴ベクトルを抽出する処理の一例を説明する図である。
【図２４】映像信号から映像特徴ベクトルを抽出する処理の他の例を説明するフローチャートである。
【図２５】映像信号から映像特徴ベクトルを抽出する処理の他の例を説明する図である。
【図２６】符号化映像信号から映像特徴ベクトルを抽出する処理の他の例を説明するフローチャートである。
【図２７】符号化映像信号から映像特徴ベクトルを抽出する処理の他の例を説明する図である。
【図２８】従来の信号記録処理を説明する図であり、同図（Ａ）は、所望の番組についての録画録音予約区間を定めた様子を示し、同図（Ｂ）は、タイマー録画録音時に、その番組の放送開始時刻が遅れた様子を示す。
【符号の説明】
１，２，３ 信号記録装置、１０ 録画録音制御部、１１ 記録部、１２ 再生部、１３ 一致検出部、１４ 登録部、１５ 特徴抽出部、２０ ＣＭ検出部、５０ 階層的距離演算部、５１ 閾値判定部[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a signal recording apparatus and method for recording a signal such as a broadcast video signal and / or an audio signal, a program, and a recording medium.
[0002]
[Prior art]
Conventionally, video and audio such as broadcast programs are recorded and recorded and viewed at a time convenient for the user (viewer). In such a case, recording / recording is generally performed by presetting the program start time, end time, and broadcast channel as the recording / recording information.
[0003]
[Problems to be solved by the invention]
However, as shown in FIG. 28 (A), the recording / recording start time t is synchronized with the original scheduled broadcast time. _s Recording recording end time t _e Even if is set, the program is often broadcast after the scheduled broadcast start time, as shown in Fig. 28 (B), such as when the previous program is extended or specially organized. There was a problem of causing failure in recording and recording the program.
[0004]
The present invention has been proposed in view of such a conventional situation, for example, a signal recording apparatus and method for reliably recording and recording even when the broadcast start time is delayed, It is another object of the present invention to provide a program for causing a computer to execute signal recording processing and a recording medium on which such a program is recorded.
[0005]
[Means for Solving the Problems]
In order to achieve the above-described object, a signal recording apparatus according to the present invention includes a recording unit that records a signal, a feature extraction unit that extracts a feature amount from the signal, and a signal of a desired broadcast program in the recording unit. Time setting means for setting a recording start time and a recording end time at the time of recording, and a feature extracted by the feature extraction means from a preset feature amount of the comparison target signal and the signal of the desired broadcast program being recorded A coincidence detecting means for detecting a coincidence with the amount, and a time correcting means for correcting the recording end time according to a detection result of the coincidence detecting means. The time correction means sets the difference between the start time of the program block whose feature value is first detected by the coincidence detection means and the recording start time as a time difference, and the time correction means sets a predetermined allowable time. If the time difference is within the allowable time, the correction is not performed. If the time difference is exceeded, the recording end time is corrected according to the time difference. .
[0006]
Here, the signal recording apparatus may include registration means for registering a feature amount in a partial section of the comparison target signal. In this case, the coincidence detection means is the one registered by the registration means. A coincidence between the feature quantity in the partial section and the feature quantity extracted from the desired signal being recorded is detected.
[0007]
The time correction means is , In the case where commercial message detection means for detecting a commercial message in the desired signal is provided, the time difference between the end time of the commercial message immediately before the time when the match is detected by the match detection means and the recording start time The recording end time may be corrected according to the time difference.
[0008]
Further, the time correction means may set a predetermined allowable time, and may not perform correction if the time difference is within the allowable time.
[0009]
In order to achieve the above-described object, a signal recording method according to the present invention includes a time setting step for setting a recording start time and a recording end time when recording a signal of a desired broadcast program on a recording unit, A feature extraction step of extracting a feature amount from a desired broadcast program signal, a feature amount of a preset comparison target signal, and a feature amount extracted from the desired broadcast program signal being recorded in the feature extraction step A coincidence detection step for detecting a coincidence with the time, and a time correction step for correcting the recording end time according to the detection result in the coincidence detection step. In the time correction step, the difference between the start time of the program block in which the feature amount is first detected in the coincidence detection step and the recording start time is set as a time difference. In the time correction step, a predetermined allowable time is set. If the time difference is within the permissible time, no correction is made, and if the time exceeds the permissible time, the recording end time is corrected according to the time difference. .
[0010]
Here, the signal recording method may include a registration step of registering a feature amount in a partial section of the comparison target signal. In this case, in the coincidence detection step, the registration registered in the registration step is performed. The coincidence between the feature value in the partial section and the feature value extracted from the desired signal being recorded is detected.
[0011]
In addition, the time correction process is , In the case of having a commercial message detection step for detecting a commercial message in the desired signal, the end time of the commercial message immediately before the time when the match is detected in the match detection step, and the recording start time May be obtained, and the recording end time may be corrected according to the time difference.
[0012]
According to such a signal recording apparatus and its method, for example, when a broadcast video signal and / or audio signal is recorded in the recording means, a preset feature amount of the video signal and / or audio signal and recording are being performed. The coincidence with the feature amount of the video signal and / or the audio signal is detected, and the recording end time can be corrected according to the detection result.
[0013]
A program according to the present invention causes a computer to execute the signal recording process described above, and a recording medium according to the present invention is a computer-readable medium on which such a program is recorded.
[0014]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, specific embodiments to which the present invention is applied will be described in detail with reference to the drawings. In this embodiment, the present invention is applied to a signal recording apparatus and method for reliably recording and recording even when the broadcast start time is delayed.
[0015]
(1) First embodiment
In the first embodiment described below, (a) in many series programs, title videos and title sounds are inserted immediately after the start of the program or within a short time in order to inform the viewer of the program name at the time of broadcasting. In order to intuitively impress that it is a series program, (b) the same title video / audio is often used for each broadcast. That is, when the broadcast start time is delayed, the title video / audio is broadcast with a delay from the scheduled broadcast start time. Therefore, the recording / recording end time is extended by detecting the delay.
[0016]
FIG. 1 shows a schematic configuration of the signal recording apparatus according to the first embodiment. As shown in FIG. 1, the signal recording apparatus 1 includes a recording / recording control unit 10, a recording unit 11, a reproduction unit 12, a coincidence detection unit 13, a registration unit 14, and a feature extraction unit 15. .
[0017]
Here, the operation of the signal recording apparatus 1 is roughly divided into three steps shown in FIG. That is, a normal recording / recording process is performed in step S1 of FIG. 2, and a registration process for registering a part of the recorded video / audio signal is performed in step S2. In step S3, a timer recording / recording process with automatic extension is performed. Therefore, hereinafter, the operation of each unit in these three steps will be described in order.
[0018]
First, the normal recording / recording process (step S1 in FIG. 2) of the signal recording apparatus 1 will be described with reference to the flowchart of FIG. In step S10, the recording / recording control unit 10 (FIG. 1) inputs the video signal / audio signal demodulated by the demodulator (not shown) and starts recording in the recording unit 11 according to the recording / recording control information. Here, the recording / recording control information is information such as a recording / recording start time and a recording / recording end time, which may be set in advance by a timer or the like, and is based on an operation of a user who is viewing the video / audio. May be.
[0019]
During recording and recording, in step S11, the recording and recording control unit 10 records the video signal / audio signal in the recording unit 11, and at the same time, in step S12, the feature extraction unit 15 extracts the feature amount of the video signal / audio signal. The feature amount is recorded in the recording unit 11 together with the video signal / audio signal. The details of the video signal / audio signal feature amount extraction processing in the feature extraction unit 15 will be described later, but simply, feature amounts are extracted for each short-time frame and a feature vector obtained by vectorizing them is configured. It is.
[0020]
In step S13, the recording / recording control unit 10 ends the recording of the video signal / audio signal to the recording unit 11 in accordance with the recording / recording control information, and at the same time, the feature amount extraction unit 12 also records the feature amount to the recording unit 11. finish.
[0021]
The recorded video signal / audio signal can be viewed through the playback unit 12.
[0022]
Next, the registration process (step S2 in FIG. 2) of the signal recording apparatus 1 will be described using the flowchart in FIG. In step S20, the reproducing unit 12 reproduces the video signal / audio signal recorded in the recording unit 11.
[0023]
In step S21, the registration unit 14 copies the video signal / audio signal at the position instructed by the user (viewer) through a remote control button or the like (not shown) and the corresponding feature amount from the recording unit 11 and holds them. . Here, the part to be registered is, for example, a video signal / acoustic signal / feature amount in a section before and after a position in response to a user instruction.
[0024]
Note that the registration unit 14 may hold only a pointer to the video signal / audio signal / feature amount recorded in the recording unit 11 instead of actually holding a copy of the data. Here, the pointer is information indicating at which position in the recording unit 11 the recorded video signal / audio signal / feature amount is recorded.
[0025]
Next, timer recording / recording processing with automatic extension (step S3 in FIG. 2) of the signal recording apparatus 1 will be described with reference to the flowchart in FIG. In step S30, the recording / recording control unit 10 sets recording / recording control information such as a recording / recording start time, a recording / recording end time, and a channel in advance, and further sets an allowable time in step S31. Here, the allowable time is a time when the registered video signal / audio signal is predicted to be broadcast within this time range from the recording / recording start time. In addition, it is not necessary to newly set every recording and recording operation, and a default value can be prepared in advance.
[0026]
Next, in step S32, the recording / recording control unit 10 determines whether or not the recording / recording start time has come, and if it is not the recording / recording start time (No), the determination processing is continued, and if it is the recording / recording start time ( If yes, go to step S33.
[0027]
In step S33, processing similar to the normal recording and recording processing described above is performed. That is, the recording / recording control unit 10 records the video signal / audio signal in the recording unit 11, and at the same time, the feature extraction unit 15 extracts the feature quantity of the video signal / audio signal, and the feature quantity is converted into the video signal / audio signal. Together with the recording in the recording unit 11.
[0028]
Subsequently, in step S34, the recording / recording control unit 10 determines whether or not the recording / recording end time has come, and if it is the recording / recording end time (Yes), the recording / recording process is stopped in step S35, and a series of steps are performed. The process ends. On the other hand, if it is not the recording / recording end time (No), the process proceeds to step S36.
[0029]
In step S <b> 36, the coincidence detection unit 13 compares the feature amount registered in the registration unit 14 with the feature amount of the video signal / audio signal currently being recorded and recorded extracted by the feature extraction unit 15, to determine whether or not they match. If it does not match (No), the process returns to step S33 to continue the recording and recording process. On the other hand, if they match (Yes), the process proceeds to step S37. Details of processing in the coincidence detection unit 13 will be described later.
[0030]
In step S37, the recording / recording control unit 10 determines whether or not the current time is within the allowable time from the recording / recording start time. If the current time is within the allowable time (Yes), the process returns to step S33 to perform the recording / recording process. Continue. On the other hand, if the allowable time is exceeded (No), it is determined that the broadcast start time is delayed for some reason, and the recording / recording processing unit 10 extends and corrects the recording / recording end time in step S38. Go back and continue the recording process. Here, by setting the extended time width to a time width obtained by subtracting the recording / recording start time from the current time, for example, it is possible to reliably record and record until the program ends.
[0031]
Here, an example of the coincidence detection process in the coincidence detection unit 13 will be described in detail. Of course, the coincidence detection process is not limited to this example.
[0032]
In the following, feature amounts (feature vectors) extracted from the currently input video signal / audio signal and registered feature amounts (feature vectors) are assumed to be feature vectors f and g, respectively. The i-th component is expressed as f [i] and g [i], respectively. Also, let N be the number of dimensions of the feature vectors f and g.
[0033]
As shown in FIG. 6, the coincidence detection unit 13 receives feature vectors f and g and detects a coincidence between the vectors, and includes a hierarchical distance calculation unit 50 and a threshold determination unit 51. The
[0034]
The coincidence detection process in the coincidence detection unit 13 will be described with reference to the flowchart of FIG. First, in step S40, the threshold determination unit 51 (FIG. 6) sets a distance threshold S, and in the subsequent step S41, the hierarchical distance calculation unit 50 inputs the feature vector f and the registration unit 14 (FIG. 1). ) Is acquired.
[0035]
Subsequently, in step S42, the hierarchical distance calculation unit 50 sets the component number i, which is an internal variable, to 1 and the integrated value sum of distances to 0. In step S43, the i-th component f [ An integration calculation as shown in the following equation (1) is performed between i] and the i-th component g [i] of the feature vector g.
[0036]
[Expression 1]

[0037]
In step S44, the threshold determination unit 51 determines whether or not the integrated value sum is less than the threshold S. If the integrated value sum is less than the threshold value S (Yes), the process proceeds to step S46. If the integrated value sum is equal to or greater than the threshold value S (No), the threshold value determination unit 51 determines that the feature vector f does not match in step S45. Determine and end the process. As described above, the threshold value determination unit 51 provides the threshold value S, and when the integrated value sum exceeds the threshold value S in a layer in the middle of the integration calculation, the threshold value determination unit 51 terminates the integration calculation in the hierarchical distance calculation unit 50. We are trying to speed up.
[0038]
In step S46, it is determined whether or not the component number i is equal to or less than the dimension number N of the feature vectors f and g. If the component number i is N or less (Yes), i is incremented in step S47 and the process returns to step S43. On the other hand, if the component number i is greater than N (No), since the integration operation has been completed up to the last component of the feature vectors f and g, the threshold determination unit 51 matches the feature vector f in step S48. To end the process.
[0039]
As described above, the process for one feature vector g is shown in the flowchart of FIG. 7, but when a plurality of feature vectors g are registered in one registration process, the same process is performed for each of them. All feature vectors f for which the integrated value sum of the distances falls below the threshold S are detected. In this case, for example, the time when the feature amount of the video signal and the feature amount of the audio signal first match may be the time when the match is finally detected. It is also possible to set the time at which the first coincidence with any one of the feature quantities is finally the time when the coincidence is detected.
[0040]
In the coincidence detection method described above, the vector component from which the search is performed depends on the vector arrangement order, and therefore the search speed varies depending on the arrangement order. Therefore, the orthogonal transformation is performed by multiplying the feature vectors f and g by the orthonormal transformation matrix, and the feature vectors f ′ and g ′ after the orthogonal transformation are used to perform the search in descending order of significance. I do not care.
[0041]
As the orthogonal transform, for example, a transform using an order matrix, a discrete cosine transform (DCT), a discrete Fourier transform (DFT), a Walsh-Hadamard transform, or a carunen-label ( Karhunen-Loeve) conversion can be used, and such conversion is applied to the input vector in advance to obtain a highly significant vector component, that is, a component having a large variance or eigenvalue, or a low-frequency component, low By performing the integration operation sequentially from the alternating component, the feature vector f matching the feature vector g can be detected efficiently and at high speed in consideration of the distribution of vector components.
[0042]
A specific example of the processing of the signal recording apparatus 1 described above is schematically shown in FIG. As shown in FIG. 8A, first, for example, a feature amount (feature vector g) of a video signal / audio signal of a title portion is extracted and registered in advance for a series program already recorded. Recording start time t during timer recording _s Recorded recording end time t from (broadcast scheduled start time) _e Recording and recording are performed until (scheduled broadcast end time). As shown in FIG. 8B, feature quantities (feature vectors f) at discrete times t are sequentially extracted from the video signal / audio signal and extracted. It is detected whether or not the feature quantity matches the registered feature quantity (feature vector g). FIG. 8B shows a state in which the target series program is broadcast with a delay.
[0043]
When the coincidence position is detected, as shown in FIG. _m And recording start time t _s And compare. And the delay time d (= t _m -T _s ) Exceeds the allowable time, as shown in FIG. _e Is extended by the delay time d, and the time t ′ _e To record until.
[0044]
As described above, the signal recording apparatus 1 according to the first embodiment compares the feature quantity registered in advance with the feature quantity of the video signal / audio signal currently input during the timer recording and recording, and the coincidence time. Is delayed by more than the allowable time from the scheduled broadcast start time, the recording / recording end time is extended by that delay. As a result, even if the program to be recorded is delayed for some reason, it is possible to reliably record and record until the end of the program.
[0045]
(2) Second embodiment
In the first embodiment described above, as shown in FIG. 9A, an allowable time is set to prevent malfunction, and the title video and title audio registered within the allowable time from the recording start time are recorded. If it was broadcast, it was considered that there was no delay in broadcasting, and it was possible to record and record without extending the recording and recording end time.
[0046]
However, as shown in FIG. 9B, depending on the program, a so-called introduction part of the program is first broadcast, and a title video and title audio may be broadcast after a little longer time has elapsed since the start of the program. In order to cope with such a case with the above-described method, it is necessary to take a sufficiently long allowable time, but conversely, if it is excessively long, even if there is a delay in the broadcast start time, it is treated as within the allowable time. In some cases, automatic extension does not work.
[0047]
Therefore, in the second embodiment described below, (a) in the case of a commercial broadcasting station that broadcasts a commercial message (hereinafter simply referred to as a commercial or CM) during a program, there is always a commercial between the programs. By using the fact that the title video and the title audio are almost always broadcast in the first program block, even if there is a long intro part, a little longer time has passed since the program started. It is possible to cope with the case where title video and title audio are broadcast from. Here, the program block is a program portion between commercials.
[0048]
FIG. 10 shows a schematic configuration of a signal recording apparatus according to the second embodiment. As shown in FIG. 10, the basic configuration of the signal recording device 2 is the same as that of the signal recording device 1 shown in FIG. 1, but has a CM detection unit 20 and starts each commercial from a broadcast video signal / audio signal. It is characterized in that the time and its length can be detected. Therefore, the same components as those of the signal recording apparatus 1 previously shown in FIG.
[0049]
As a commercial detection method, for example, the technique described in Japanese Patent Laid-Open No. 2002-16873 previously proposed by the present inventor can be used. This technique is summarized as follows.
[0050]
In other words, in Japan, almost all commercials except for special cases are produced with lengths of 15 seconds, 30 seconds, and 60 seconds, and the audio level is inevitably lowered before and after each commercial. Since the signals are switched, these are set as “indispensable conditions” for detection. In addition, it shows a certain tendency that appears as a result of being produced under the restriction that it is produced based on the standard, the restriction that it is to increase the advertising effect in a short time, and the restriction due to the circumstances of the program structure etc. The feature is set as an “additional condition” for detection. Further, when there are overlapping sections that satisfy the additional condition, a condition that at least one of the sections cannot be a correct commercial section is set as a “logical condition” for detection. Then, commercial candidates are extracted deterministically based on “required conditions”, candidates are selected based on statistical evaluation of commercial quality based on “additional conditions”, and candidate overlap relationships are resolved based on “logical conditions” By doing so, commercials can be detected with high accuracy.
[0051]
The operation of the signal recording apparatus 2 is also roughly divided into three steps as shown in FIG. That is, after the normal recording / recording process is performed, a registration process for registering a part of the recorded video / audio signal is performed, and a timer recording / recording process with automatic extension is performed.
[0052]
First, the normal recording / recording process of the signal recording apparatus 2 will be described with reference to the flowchart of FIG. In step S50, the recording / recording control unit 10 (FIG. 10) inputs the video signal / audio signal demodulated by the demodulator (not shown) and starts recording in the recording unit 11 according to the recording / recording control information. Here, the recording / recording control information is information such as the recording / recording start time and the recording / recording end time as described above, and may be set in advance by a timer or the like. It may be due to.
[0053]
During the recording, the recording / recording control unit 10 records the video signal / audio signal in the recording unit 11 in step S51, and at the same time, the feature extraction unit 15 extracts the feature amount of the video signal / audio signal in step S52. The feature amount is recorded in the recording unit 11 together with the video signal / audio signal. Further, in step S53, the CM detection unit 20 detects each commercial start time and its length in the video signal / audio signal, and records them in the recording unit 11 together with the video signal / audio signal.
[0054]
In step S54, the recording / recording control unit 10 ends the recording of the video signal / audio signal to the recording unit 11 according to the recording / recording control information, and at the same time, the feature extraction unit 15 also ends the recording of the feature amount to the recording unit 11. To do.
[0055]
Next, registration processing is performed. Since this registration processing is the same as that of the first embodiment, description thereof is omitted.
[0056]
Next, the timer recording / recording process with automatic extension of the signal recording apparatus 2 will be described with reference to the flowchart of FIG. In step S60, the recording / recording control unit 10 sets recording / recording control information such as a recording / recording start time, a recording / recording end time, and a channel in advance, and further sets an allowable time in step S61. Here, the allowable time is a time at which the commercial end time immediately before the registered video signal / audio signal is predicted to be within this time range from the recording / recording start time. Note that it is not necessary to newly set each recording and recording operation, and a default value can be prepared in advance.
[0057]
Next, in step S62, the recording / recording control unit 10 determines whether or not the recording / recording start time has come, and if it is not the recording / recording start time (No), the determination processing is continued, and if it is the recording / recording start time ( If yes, go to step S63.
[0058]
In step S63, processing similar to the normal recording and recording processing described above is performed. That is, the recording / recording control unit 10 records the video signal / audio signal in the recording unit 11, and at the same time, the feature extraction unit 15 extracts the feature quantity of the video signal / audio signal, and the feature quantity is converted into the video signal / audio signal. Together with the recording in the recording unit 11. Further, the CM detection unit 20 detects each commercial start time and its length in the video signal / audio signal, and records them in the recording unit 11 together with the video signal / audio signal.
[0059]
Subsequently, in step S64, the recording / recording control unit 10 determines whether or not the recording / recording end time has come, and if it is the recording / recording end time (Yes), the recording / recording process is stopped in step S65, and a series of steps are performed. The process ends. On the other hand, if it is not the recording / recording end time (No), the process proceeds to step S66.
[0060]
In step S66, the coincidence detection unit 13 compares the feature amount registered in the registration unit 14 with the feature amount of the video signal / audio signal currently recorded and recorded extracted by the feature extraction unit 15 to determine whether or not they match. If it does not match (No), the process returns to step S63 to continue the recording and recording process. On the other hand, if they match (Yes), the process proceeds to step S67.
[0061]
In step S67, the recording / recording control unit 10 obtains the commercial end time immediately before the current time.
[0062]
In step S68, the recording / recording control unit 10 determines whether or not the end time of the immediately preceding commercial is within the allowable time from the recording / recording start time. If it is within the allowable time (Yes), the process returns to step S63. To continue the recording process. On the other hand, if the allowable time is exceeded (No), it is determined that the broadcast start time is delayed for some reason, and the recording / recording control unit 10 extends and corrects the recording / recording end time in step S69, and then the process proceeds to step S63. Go back and continue the recording process. Here, for example, by setting the extended time width to a time width obtained by subtracting the recording / recording start time from the end time of the immediately preceding commercial, it is possible to reliably record and record until the end of the program.
[0063]
A specific example of the processing of the signal recording apparatus 2 described above is schematically shown in FIG. When a coincidence position is detected in the same manner as in the first embodiment, as shown in FIG. _m Commercial end time t just before _CM And recording start time t _s And compare. And the delay time d (= t _CM -T _s ) Exceeds the allowable time, as shown in FIG. _e Is extended by the delay time d, and the time t ′ _e To record until.
[0064]
As described above, the signal recording device 2 according to the second embodiment compares the feature quantity registered in advance with the feature quantity of the video signal / audio signal currently input during the timer recording and recording, and matches the matching position. When the commercial end time immediately before is delayed more than the allowable time from the recording / recording start time, the recording / recording end time is extended by the delay. As a result, even if the program to be recorded is delayed for some reason, it is possible to reliably record and record until the end of the program.
[0065]
Also, even if the introduction part of the program is long, there is no need to increase the allowable time, so even if there is a delay in the broadcast start time, it will be treated as within the allowable time and automatic extension will not work Does not happen.
[0066]
Furthermore, since the extended correction time is usually shorter than that of the signal recording apparatus 1 in the first embodiment, the recording area can be saved.
[0067]
(3) Third embodiment
In the first and second embodiments described above, in order to perform the automatic recording and recording time extension operation, it is necessary to register in advance a partial section of the target program, for example, the title video and title audio of the series program. It was. On the other hand, in a third embodiment described below, a signal recording apparatus that automatically performs a recording / recording time extension operation without performing such a registration operation will be described.
[0068]
FIG. 14 shows a schematic configuration of a signal recording apparatus according to the third embodiment. As shown in FIG. 14, the basic configuration of the signal recording device 3 is the same as that of the signal recording device 2 shown in FIG. 10, but is characterized in that there is no registration unit. Therefore, the same components as those of the signal recording apparatus 2 previously shown in FIG.
[0069]
The operation of the signal recording device 3 is obtained by omitting the registration process for registering a part of the recorded video / audio signal among the three steps as shown in FIG. That is, after the normal recording / recording process is performed, the timer recording / recording process with automatic extension is performed.
[0070]
First, the normal recording / recording process of the signal recording device 3 is performed. Since the normal recording / recording process is the same as that of the second embodiment, the description thereof is omitted.
[0071]
Next, timer recording / recording processing with automatic extension of the signal recording apparatus 3 will be described with reference to the flowchart of FIG. In step S70, the recording / recording control unit 10 sets recording / recording control information such as a recording / recording start time, a recording / recording end time, and a channel in advance, and further sets an allowable time in step S71. Here, as in the second embodiment, the allowable time is a time during which the commercial end time immediately before the registered video signal / audio signal is predicted to be within this time range from the recording start time. It is. In addition, it is not necessary to newly set every recording and recording operation, and a default value can be prepared in advance.
[0072]
Next, in step S72, the recording / recording control unit 10 determines whether or not the recording / recording start time has been reached. If yes, go to step S73.
[0073]
In step S73, the recording / recording control unit 10 records the video signal / audio signal in the recording unit 11, and at the same time, the feature extraction unit 15 extracts the feature amount of the video signal / audio signal. Further, the CM detection unit 20 detects each commercial start time and its length in the video signal / audio signal. Then, the feature extraction unit 15 records the feature amount of the program part excluding the commercial part in the recording unit 11 together with the video signal / audio signal, and the CM detection unit 20 detects each commercial start time and its length. Is recorded in the recording unit 11.
[0074]
Subsequently, in step S74, the recording / recording control unit 10 determines whether or not the recording / recording end time has come. If it is the recording / recording end time (Yes), the recording / recording process is stopped in step S75, and The process ends. On the other hand, if it is not the recording / recording end time (No), the process proceeds to step S76.
[0075]
In step S76, the coincidence detection unit 13 compares all feature quantities of the target program that have already been recorded with the feature quantities of the video signal / audio signal that is currently being recorded and extracted, which is extracted by the feature extraction unit 15. It is determined whether or not it matches any one. If it does not match (No), the process returns to step S73 to continue the recording / recording process. If it matches any (Yes), the process proceeds to step S77.
[0076]
In step S77, the recording / recording control unit 10 obtains the commercial end time immediately before the current time.
[0077]
In step S78, the recording / recording control unit 10 determines whether or not the end time of the previous commercial is within the allowable time from the recording / recording start time, and if within the allowable time (Yes), the process returns to step S73. To continue the recording process. On the other hand, if the allowable time is exceeded (No), it is determined that the broadcast start time is delayed for some reason, and the recording / recording control unit 10 extends and corrects the recording / recording end time in step S79, and then the process proceeds to step S73. Go back and continue the recording process. Here, for example, by setting the extended time width to a time width obtained by subtracting the recording / recording start time from the end time of the immediately preceding commercial, it is possible to reliably record and record until the end of the program.
[0078]
A specific example of the processing of the signal recording apparatus 3 described above is schematically shown in FIG. As shown in FIG. 16A, first, for a program part excluding, for example, a commercial part, a video signal / audio signal feature quantity (feature vector g) is extracted and recorded for each short-time frame. Keep it. Recording start time t during timer recording _s Recorded recording end time t from (broadcast scheduled start time) _e Recording and recording are performed up to (scheduled broadcast end time). As shown in FIG. 16B, feature quantities (feature vectors f) at discrete times t are sequentially extracted from the video signal / audio signal and extracted. It is detected whether or not the feature quantity (feature vector f) other than the commercial part matches any of the recorded feature quantities (feature vector g). Note that FIG. 16B shows a state in which a target series program is broadcast with a delay.
[0079]
When the coincidence position is detected, as shown in FIG. _m Commercial end time t just before _CM And recording start time t _s And compare. And the delay time d (= t _CM -T _s ) Exceeds the allowable time, as shown in FIG. _e Is extended by the delay time d, and the time t ′ _e To record until.
[0080]
As described above, the signal recording device 3 according to the third embodiment compares the feature amount of the target program recorded in advance with the feature amount of the currently input video signal / audio signal during timer recording. When the commercial end time immediately before the coincidence position is delayed by an allowable time or more from the scheduled broadcast start time, the recording / recording time is extended by the delay. As a result, even if the program to be recorded is delayed for some reason, it is possible to reliably record and record until the end of the program.
[0081]
In addition, since the coincidence detection between the extracted feature quantity and the feature quantity of the entire recorded program is performed, the number of comparisons is increased as compared with the first and second embodiments, but registration processing is not required. , User burden is reduced.
[0082]
In the above description, the commercial end time is used as in the second embodiment. However, the present invention is not limited to this, and the commercial end time is not used as in the first embodiment. In addition, it may be determined whether the time when the match is detected is within the allowable time.
[0083]
(4) Feature vector extraction
Hereinafter, a method of extracting the acoustic feature vector a and the video feature vector v from the acoustic signal and the video signal in the feature extraction unit 15 will be described.
[0084]
(4-1) Extraction of acoustic feature vector
(4-1-1)
An example in which a power spectrum coefficient is used as a feature amount related to an acoustic signal will be described with reference to the flowchart of FIG. 17 and FIG. First, in step S80, as shown in FIG. 18, the acoustic signal in the time interval T is acquired from the acoustic signal in the target time interval.
[0085]
Next, in step S81, the acquired acoustic signal is subjected to spectrum calculation such as fast Fourier transform, and the power spectrum coefficient S is calculated for each short time interval. _q (Q = 0, 1,..., Q-1) is obtained. Here, q is an index representing a discrete frequency, and Q is a maximum discrete frequency.
[0086]
Subsequently, in step S82, it is determined whether or not the calculation within the target time interval has been completed. If completed (Yes), the process proceeds to step S83, and if not completed (No), the process returns to step S80.
[0087]
In step S83, the obtained power spectrum coefficient S _q Average spectrum S ' _q And this average spectrum S ′ in step S84. _q To generate an acoustic feature vector a. This acoustic feature vector a is represented, for example, by the following equation (2).
[0088]
[Expression 2]

[0089]
In the above-described example, the acoustic signal in the target time interval is described as being divided for each time interval T. However, when the target time interval is short, the spectrum calculation is performed without being divided for each time interval T. It doesn't matter.
[0090]
In the above-described example, the example using the power spectrum coefficient has been described. However, the present invention is not limited to this. For example, a cepstrum coefficient having equivalent information can be used. Furthermore, the same effect can be obtained by using a linear prediction coefficient using an AR (Auto-Regressive) model without using Fourier transform.
[0091]
(4-1-2)
Since acoustic signals are enormous, they are often compressed and recorded or transmitted. After decoding the encoded acoustic signal and returning it to the baseband, it is possible to extract the acoustic feature vector a using the above-described method. However, if the acoustic feature vector a can be extracted only by partial decoding, The extraction process can be made more efficient and faster.
[0092]
Here, in transform coding, which is a commonly used coding method, as shown in FIG. 19, an acoustic signal as an original sound is divided into frames for each time interval T. Then, an orthogonal transform such as a modified discrete cosine transform (MDCT) is performed on the acoustic signal for each frame, and the coefficients are quantized and encoded. At this time, a scale factor, which is a normalization coefficient of the size, is extracted for each frequency band and separately encoded. Therefore, by decoding only this scale factor, it can be used as the acoustic feature vector a.
[0093]
As described above, an example in which the scale factor is used as the feature amount related to the acoustic signal will be described with reference to the flowchart of FIG. 20 and FIG. First, in step S90, an encoded acoustic signal in the time interval T in the target time interval is acquired, and in step S91, the scale factor for each frame is partially decoded.
[0094]
Subsequently, in step S92, it is determined whether or not the decoding within the target time interval has been completed. If completed (Yes), the process proceeds to step S93. If not completed (No), the process returns to step S90.
[0095]
In step S93, the maximum scale factor is detected for each band from the scale factors in the target time interval, and in step S94 they are vectorized to generate an acoustic feature vector a.
[0096]
In this way, the acoustic feature vector a equivalent to the above can be extracted at high speed without completely decoding the encoded acoustic signal.
[0097]
(4-2) Video feature vector extraction
(4-2-1)
An example in the case where luminance information and color information are used as feature amounts relating to a video signal will be described with reference to the flowchart in FIG. 22 and FIG. First, in step S100, as shown in FIG. 23, a video frame is acquired from the video signal in the target time interval T.
[0098]
Next, in step S101, a time average image is created based on all the acquired video frames.
[0099]
Subsequently, in step S102, the created time average image is divided into X × Y small blocks in the horizontal and vertical directions, and a block average image is created by averaging the pixel values in each block.
[0100]
In step S103, these are arranged in the order of R, G, B from the upper left to the lower right, for example, to generate a one-dimensional video feature vector v. This video feature vector v is expressed, for example, by the following equation (3).
[0101]
[Equation 3]

[0102]
In the above example, the example in which the pixel values of the block average image obtained by dividing the time average image are rearranged to generate the one-dimensional video feature vector v has been described. However, the present invention is not limited to this. The one-dimensional video feature vector v may be generated by rearranging the pixel values of the time average image without creating an image.
[0103]
Also, since the time change of the video signal is usually not so severe, even if one frame in the target time interval is selected and used as a representative image without creating a time average image, almost the same effect can be obtained. Can do.
[0104]
(4-2-2)
Even if the video signals are not exactly the same, there are often some relationships between videos with similar color distributions for all images, such as studio videos shot from the same angle of news videos. There is also a request to search. In such a case, it is effective to eliminate the spatial dependence of the image and create a histogram of the color distribution for comparison.
[0105]
Thus, an example of using a color distribution histogram as a feature amount will be described with reference to the flowchart of FIG. 24 and FIG. First, in step S110, as shown in FIG. 25, a video frame is acquired from the video signal in the target time interval T.
[0106]
In step S111, a histogram for each color, for example, R, G, B signal values, is created from the signal values of each video frame.
[0107]
In step S112, these are arranged in the order of R, G, B, for example, to generate a one-dimensional video feature vector v. This video feature vector v is expressed, for example, by the following equation (4).
[0108]
[Expression 4]

[0109]
In the example described above, the histogram for the R, G, and B signal values is created. However, the histogram for the luminance (Y) and chrominance (Cb, Cr) signal values may be created. Similar effects can be obtained.
[0110]
(4-2-3)
Since video signals are enormous, they are often compressed and recorded or transmitted. After decoding the encoded video signal and returning it to the baseband, it is possible to extract the video feature vector v using the above-described method. However, if the video feature vector v can be extracted only by partial decoding, The extraction process can be made more efficient and faster.
[0111]
An example of extracting a video feature vector v from a video signal compressed and encoded by MPEG1 (Moving Picture Experts Group 1) or MPEG2 will be described with reference to the flowchart of FIG. 26 and FIG. First, in step S120, an encoded video signal of the latest encoding group (Group of Pictures: GOP) is acquired for the target time interval T to be vectorized, and an intra-frame encoded picture (I picture) in the GOP is acquired. ) To get.
[0112]
Here, the frame image is encoded in units of macroblocks (16 × 16 pixels or 8 × 8 pixels), and discrete cosine transform (DCT) is used. The DC coefficient subjected to the DCT conversion is an average value of pixel values of the image in the macroblock.
[0113]
Therefore, in step S121, this DC coefficient is acquired, and in the subsequent step S122, these are arranged in the order of, for example, Y, Cb, and Cr to generate a one-dimensional video feature vector v. This video feature vector v is expressed, for example, by the following equation (5).
[0114]
[Equation 5]

[0115]
In this way, the video feature vector v can be extracted at high speed without completely decoding the encoded video signal.
[0116]
In the above-described example, the video signal compressed and encoded by MPEG1 or MPEG2 has been described. However, it is needless to say that the present invention can be applied to other compression encoding methods.
[0117]
(5) Other
The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the scope of the present invention.
[0118]
For example, in the above-described embodiment, it has been described that both a video signal and an audio signal are used. However, the present invention is not limited to this. For example, only radio signals are used for radio broadcasting, etc. The same effect can be obtained by using either one.
[0119]
In the above-described embodiment, the hardware configuration has been described. However, the present invention is not limited to this, and arbitrary processing may be realized by causing a CPU (Central Processing Unit) to execute a computer program. Is possible. In this case, the computer program can be provided by being recorded on a recording medium, or can be provided by being transmitted via the Internet or another transmission medium.
[0120]
【The invention's effect】
As described above in detail, the signal recording apparatus according to the present invention records a signal, a feature extraction unit that extracts a feature amount from the signal, and records a signal of a desired broadcast program in the recording unit. A time setting means for setting a recording start time and a recording end time at the time, a feature quantity of a preset comparison target signal, and a feature quantity extracted by the feature extraction means from a signal of the desired broadcast program being recorded A coincidence detecting means for detecting the coincidence of the recording medium and a time correcting means for correcting the recording end time according to the detection result of the coincidence detecting means The time correction means sets the difference between the start time of the program block whose feature value is first detected by the coincidence detection means and the recording start time as a time difference, and the time correction means sets a predetermined allowable time. If the time difference is within the allowable time, the correction is not performed. If the time difference is exceeded, the recording end time is corrected according to the time difference. .
[0121]
Here, the signal recording apparatus may include registration means for registering a feature amount in a partial section of the comparison target signal. In this case, the coincidence detection means is the one registered by the registration means. A coincidence between the feature quantity in the partial section and the feature quantity extracted from the desired signal being recorded is detected.
[0122]
The time correcting means may obtain a time difference between the time when the coincidence is detected by the coincidence detecting means and the recording start time, and may correct the recording end time according to the time difference. If a commercial message detection means for detecting a commercial message in the signal is provided, the time difference between the end time of the commercial message immediately before the time when the match is detected by the match detection means and the recording start time is obtained. The recording end time may be corrected according to the time difference.
[0123]
Further, the time correction means may set a predetermined allowable time, and may not perform correction if the time difference is within the allowable time.
[0124]
The signal recording method according to the present invention includes a time setting step for setting a recording start time and a recording end time when recording a signal of a desired broadcast program on a recording means, and a feature amount from the signal of the desired broadcast program. And a feature detection step for detecting a match between a feature amount of a preset comparison target signal and a feature amount extracted in the feature extraction step from the signal of the desired broadcast program being recorded And a time correction step for correcting the recording end time according to the detection result in the coincidence detection step. In the time correction step, the difference between the start time of the program block in which the feature amount is first detected in the coincidence detection step and the recording start time is set as a time difference. In the time correction step, a predetermined allowable time is set. If the time difference is within the permissible time, no correction is made, and if the time exceeds the permissible time, the recording end time is corrected according to the time difference. .
[0125]
Here, the signal recording method may include a registration step of registering a feature amount in a partial section of the comparison target signal. In this case, in the coincidence detection step, the registration registered in the registration step is performed. The coincidence between the feature value in the partial section and the feature value extracted from the desired signal being recorded is detected.
[0126]
Further, the time correction step may be to obtain a time difference between the time when the match is detected in the match detection step and the recording start time, and to correct the recording end time according to the time difference, In the case of having a commercial message detection step for detecting a commercial message in the desired signal, the end time of the commercial message immediately before the time when the match is detected in the match detection step, and the recording start time May be obtained, and the recording end time may be corrected according to the time difference.
[0127]
According to such a signal recording apparatus and its method, for example, when a broadcast video signal and / or audio signal is recorded in the recording means, a preset feature amount of the video signal and / or audio signal and recording are being performed. The coincidence with the feature amount of the video signal and / or the audio signal is detected, and the recording end time can be corrected according to the detection result. Thereby, even if the start of the video signal and / or audio signal to be recorded is delayed for some reason, it is possible to record reliably until the end.
[0128]
A program according to the present invention causes a computer to execute the signal recording process described above, and a recording medium according to the present invention is a computer-readable medium on which such a program is recorded.
[0129]
According to such a program and recording medium, the signal recording process described above can be realized by software.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating a schematic configuration of a signal recording apparatus according to a first embodiment.
FIG. 2 is a flowchart illustrating an overall operation of the signal recording apparatus.
FIG. 3 is a flowchart for explaining normal recording / recording processing in the signal recording apparatus;
FIG. 4 is a flowchart illustrating registration processing in the signal recording apparatus.
FIG. 5 is a flowchart for explaining timer recording / recording processing with automatic extension in the signal recording apparatus;
FIG. 6 is a diagram illustrating a schematic configuration of the coincidence detection unit.
FIG. 7 is a flowchart illustrating a match detection process in the match detection unit.
FIG. 8 is a diagram schematically showing a specific example of processing of the signal recording apparatus. FIG. 8A shows processing for registering feature amounts, and FIG. 8B shows during recording by timer recording. FIG. 4C shows processing for determining the delay time, and FIG. 4D shows processing for extending the recording / recording time.
FIG. 9 is a diagram showing an example of a broadcast form of a program. FIG. 9A shows an example in which a title is broadcast within an allowable time from the start of the program, and FIG. An example in which a part of a program is broadcast is shown.
FIG. 10 is a diagram illustrating a schematic configuration of a signal recording apparatus according to a second embodiment.
FIG. 11 is a flowchart for explaining normal recording / recording processing in the signal recording apparatus;
FIG. 12 is a flowchart for explaining timer recording / recording processing with automatic extension in the signal recording apparatus;
FIG. 13 is a diagram schematically showing a specific example of processing of the signal recording apparatus. FIG. 13A shows processing for determining a delay time, and FIG. 13B extends recording recording time. The processing to be performed is shown.
FIG. 14 is a diagram illustrating a schematic configuration of a signal recording apparatus according to a third embodiment.
FIG. 15 is a flowchart for explaining timer recording / recording processing with automatic extension in the signal recording apparatus;
FIG. 16 is a diagram schematically showing a specific example of processing of the signal recording apparatus, where FIG. 16A shows processing for recording the feature quantity of the target program, and FIG. 16B shows timer recording; The process during recording is shown. FIG. 10C shows the process for determining the delay time, and FIG. 10D shows the process for extending the recording time.
FIG. 17 is a flowchart illustrating an example of processing for extracting an acoustic feature vector from an acoustic signal.
FIG. 18 is a diagram illustrating an example of processing for extracting an acoustic feature vector from an acoustic signal.
FIG. 19 is a diagram for explaining transform coding in an acoustic signal.
FIG. 20 is a flowchart illustrating an example of processing for extracting an acoustic feature vector from an encoded acoustic signal.
FIG. 21 is a diagram illustrating an example of processing for extracting an acoustic feature vector from an encoded acoustic signal.
FIG. 22 is a flowchart illustrating an example of processing for extracting a video feature vector from a video signal.
FIG. 23 is a diagram illustrating an example of processing for extracting a video feature vector from a video signal.
FIG. 24 is a flowchart illustrating another example of processing for extracting a video feature vector from a video signal.
FIG. 25 is a diagram illustrating another example of processing for extracting a video feature vector from a video signal.
FIG. 26 is a flowchart illustrating another example of processing for extracting a video feature vector from an encoded video signal.
FIG. 27 is a diagram illustrating another example of processing for extracting a video feature vector from an encoded video signal.
FIG. 28 is a diagram for explaining a conventional signal recording process, in which FIG. (A) shows a state in which a recording / recording reserved section for a desired program is defined, and FIG. This shows how the broadcast start time of the program is delayed.
[Explanation of symbols]
1, 2, 3 signal recording device, 10 recording / recording control unit, 11 recording unit, 12 playback unit, 13 coincidence detection unit, 14 registration unit, 15 feature extraction unit, 20 CM detection unit, 50 hierarchical distance calculation unit, 51 Threshold judgment unit

Claims (12)

Recording means for recording signals;
Feature extraction means for extracting a feature quantity from the signal;
Time setting means for setting a recording start time and a recording end time when recording a signal of a desired broadcast program in the recording means;
A coincidence detecting means for detecting a coincidence between a preset feature quantity of the comparison target signal and a feature quantity extracted by the feature extracting means from the signal of the desired broadcast program being recorded;
Time correction means for correcting the recording end time according to the detection result of the coincidence detection means ,
The time correction means uses the difference between the start time of the program block in which the feature amount is first detected by the coincidence detection means and the recording start time as a time difference,
Further, the time correction means sets a predetermined allowable time, and does not correct if the time difference is within the allowable time, and corrects the recording end time according to the time difference if the time exceeds the allowable time. A signal recording apparatus characterized by the above. Registration means for registering a feature quantity in a partial section of the comparison target signal;
The coincidence detection unit detects a coincidence between a feature amount in the partial section registered by the registration unit and a feature amount extracted from a signal of the desired broadcast program being recorded. 1. The signal recording apparatus according to 1. A registration unit for registering information indicating a recording position in which a feature amount of the comparison target signal is recorded in advance in the recording unit, and a feature amount in a partial section of the comparison target signal is recorded;
The coincidence detecting means detects a coincidence between a feature amount in the partial section indicated by the information indicating the recording position and a feature amount extracted from a signal of the desired broadcast program being recorded. The signal recording apparatus according to claim 1. Commercial message detection means for detecting a commercial message in the signal of the desired broadcast program,
The signal recording apparatus according to claim 1, wherein the time correction unit corrects the recording end time based on a detection result of the coincidence detection unit and a detection result of the commercial message detection unit. The time correction means obtains a time difference between the end time of the commercial message immediately before the time when the match is detected by the match detection means and the recording start time, and corrects the recording end time according to the time difference. 5. The signal recording apparatus according to claim 4, wherein The time correction means sets a predetermined allowable time, does not correct if the time difference is within the allowable time, and corrects the recording end time according to the time difference if the time exceeds the allowable time. The signal recording apparatus according to claim 5 . A time setting step for setting a recording start time and a recording end time when recording a signal of a desired broadcast program in the recording means;
A feature extraction step of extracting a feature amount from the signal of the desired broadcast program;
A coincidence detecting step for detecting a coincidence between a preset feature amount of the comparison target signal and the feature amount extracted in the feature extracting step from the signal of the desired broadcast program being recorded;
Depending on the detection result of the coincidence detection step, it possesses a time correction step of correcting the recording end time,
In the time correction step, the difference between the start time of the program block in which the feature amount is first detected in the coincidence detection step and the recording start time is defined as a time difference.
Further, in the time adjustment step, a predetermined allowable time is set, and if the time difference is within the allowable time, correction is not performed, and if the allowable time is exceeded, the recording end time is corrected according to the time difference. A signal recording method characterized by the above. A registration step of registering a feature amount in a partial section of the comparison target signal;
In the coincidence detection step, a coincidence between the feature amount in the partial section registered in the registration step and the feature amount extracted from the signal of the desired broadcast program being recorded is detected. The signal recording method according to claim 7 . A commercial message detecting step of detecting a commercial message in the signal of the desired broadcast program,
8. The signal recording method according to claim 7 , wherein, in the time correction step, the recording end time is corrected based on a detection result in the coincidence detection step and a detection result in the commercial message detection step. In the time correction step, a time difference between the end time of the commercial message immediately before the time when the match is detected in the match detection step and the recording start time is obtained, and the recording end time is corrected according to the time difference. The signal recording method according to claim 9 . In a program for causing a computer to execute a predetermined process,
A time setting step for setting a recording start time and a recording end time when recording a signal of a desired broadcast program in the recording means;
A feature extraction step of extracting a feature amount from the signal of the desired broadcast program;
A coincidence detecting step for detecting a coincidence between a preset feature amount of the comparison target signal and the feature amount extracted in the feature extracting step from the signal of the desired broadcast program being recorded;
Depending on the detection result of the coincidence detection step, it possesses a time correction step of correcting the recording end time,
In the time correction step, the difference between the start time of the program block in which the feature amount is first detected in the coincidence detection step and the recording start time is defined as a time difference.
Further, in the time adjustment step, a predetermined allowable time is set, and if the time difference is within the allowable time, correction is not performed, and if the allowable time is exceeded, the recording end time is corrected according to the time difference. A program characterized by this. In a computer-readable recording medium on which a program for causing a computer to execute a predetermined process is recorded,
A time setting step for setting a recording start time and a recording end time when recording a signal of a desired broadcast program in the recording means;
A feature extraction step of extracting a feature amount from the signal of the desired broadcast program;
A coincidence detecting step for detecting a coincidence between a preset feature amount of the comparison target signal and the feature amount extracted in the feature extracting step from the signal of the desired broadcast program being recorded;
Depending on the detection result of the coincidence detection step, it possesses a time correction step of correcting the recording end time,
In the time correction step, the difference between the start time of the program block in which the feature amount is first detected in the coincidence detection step and the recording start time is defined as a time difference.
Further, in the time adjustment step, a predetermined allowable time is set, and if the time difference is within the allowable time, correction is not performed, and if the allowable time is exceeded, the recording end time is corrected according to the time difference. A recording medium on which a program is recorded.

Images