JP2004078297A - Information processor, information processing method and environment control device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an information processor and an information processing method that can precisely extract an area of a human portion from a thermal image, and an environment control device that can control an appropriate heating environment suitable to actual conditions in a space to be controlled. <P>SOLUTION: Two temporally different thermal images ID1 and ID2 collected by an infrared camera are subjected to pixel-by-pixel subtraction to produce a differential image SD. An area in the differential image SD constituted of pixels whose value is outside a given range including zero is extracted as an unmatched area of a human between the two thermal images. From the temperature indicated by the thermal image corresponding to the unmatched area, thresholds (Bt ± TH2) for extraction of an area showing the same temperature as the unmatched area are derived. Either of the two thermal images ID1 and ID2 is subjected to pixel-by-pixel comparison about the thresholds, and a human area in the thermal image is extracted. <P>COPYRIGHT: (C)2004,JPO

Description

【００７８】
また、（３）式におけるｈｃ及びｆｃｌは次の式により算出される。
【００７９】
【数２】

【００８０】
但し、Ｍ：代謝率（Ｗ／ｍ^２）
Ｗ：外部仕事（Ｗ／ｍ^２）・・・通常は０
Ｉｃｌ：衣服の熱抵抗（ｍ^２Ｋ／Ｗ）
ｆｃｌ：人間の裸の場合に対する着衣時の表面積の比
ｔａ：室温（℃）
ｔｒ：平均輻射温度（℃）
ｖａｒ：風速（ｍ／ｓ）
Ｐａ：水蒸気分圧（Ｐａ）
ｈｃ：対流熱伝導率（Ｗ／ｍ^２Ｋ）
ｔｃｌ：着衣の表面温度（℃）
である。
【００８１】
ここで、代謝率Ｍ、室温ｔａ、平均輻射温度ｔｒ、風速ｖａｒ、及び着衣の表面温度ｔｃｌは、以上の処理によって得られている。
【００８２】
従って、ここでは、衣服の熱抵抗Ｉｃｌ及び水蒸気分圧Ｐａを算出する必要があるが、熱抵抗Ｉｃｌは上記ステップ１０２において得られた情報によって示される着衣量（ｃｌｏ）を用いた周知の演算式により導出することができ、水蒸気分圧Ｐａは室温ｔａ及び湿度センサ４０によって得られた湿度を用いた周知の演算式により導出することができる。
【００８３】
次のステップ１２０では、以上の処理によって得られた各種パラメータを（３）式に代入することによりＰＭＶ値を算出する。
【００８４】
ところで、通常、着衣の表面温度ｔｃｌは次の（４）式で表わされる。
【００８５】
【数３】

【００８６】
しかしながら、（４）式を（３）式に代入すると、（３）式は未知数ｔｃｌ、ｈｃに関する非線形方程式となり、収束演算（繰り返し演算）を行う必要がある。
【００８７】
これに対し、本実施の形態に係る環境制御システム１０では、温熱画像情報に基づく着衣の表面温度ｔｃｌ及び平均輻射温度ｔｒの測定値を直接（３）式に代入しているので、上記収束演算を省略することができる。
【００８８】
次のステップ１２２では、上記ステップ１２０において算出したＰＭＶ値に基づいて空調装置４４の作動を制御する空調制御処理を実行する。
【００８９】
なお、本実施の形態に係る空調制御処理は、図８に示すように、ＰＭＶ値が＋０．５以上である（１）の領域にあるときは、空調装置４４から吹き出される空気の温度を所定温度だけ低下させると共に、当該空気の湿度を所定湿度だけ低下させ、かつ当該空気の風量を所定量だけ増加させる。これに対し、ＰＭＶ値が−０．５以下である（２）の領域にあるときは、空調装置４４から吹き出される空気の温度を所定温度だけ上昇させると共に、当該空気の湿度を所定湿度だけ上昇させ、かつ当該空気の風量を所定量だけ増加させる。
【００９０】
次のステップ１２４では、本環境制御処理を終了させるタイミングか否かを判定し、否定判定の場合は上記ステップ１０６に戻って、再び上記ステップ１０６〜ステップ１２２の処理を実行し、肯定判定となった時点で本環境制御処理を終了する。なお、上記の本環境制御処理を終了させるタイミングか否かの判定は、ＰＣ２０Ａの操作者により本環境制御処理を終了させる旨の指示入力があった否かを判定することや、予め定められた運転時間を経過したか否かを判定すること等によって行うことができる。また、上記ステップ１０６〜ステップ１２４の処理を繰り返し実行する際にステップ１１０では、その直前にステップ１０８において収集された温熱画像情報と、その前回にステップ１０８において収集された温熱画像情報とを処理対象とする温熱画像情報とするようにする。
【００９１】
上記ステップ１０６〜ステップ１２４の繰り返し処理により、的確なタイミングで空調動作を行うことにより、必要以上の空調動作を行うことなくＰＭＶ値を一例として−０．５から＋０．５までの範囲内である快適域（図８参照）に入るようにすることができ、この結果として、快適化及び省エネルギー化を推進することができる。
【００９２】
以上詳細に説明したように、本実施の形態に係る環境制御システムでは、処理対象とする空間内を２次元平面として捉え、当該２次元平面の予め定められた分割領域（本実施の形態では、画素）毎の温度を示す温熱画像情報を収集する赤外線カメラ１２により収集された時間的に異なる二つの温熱画像情報における上記分割領域毎の減算によって当該分割領域毎の差分画像情報を導出し、当該差分画像情報における零を含む所定範囲の外の値となる分割領域により構成される領域を人間の上記二つの温熱画像情報間の非重複領域として抽出し、当該非重複領域に対応する上記温熱画像情報により示される温度に基づいて、当該非重複領域と同一の温度を示す領域を抽出するための閾値を導出し、上記二つの温熱画像情報の何れかと上記閾値との上記分割領域毎の比較により当該温熱画像情報における上記人間に対応する領域を抽出しているので、温熱画像から人間部分の領域を高精度に抽出することができる。
【００９３】
また、本実施の形態に係る環境制御システムでは、以上のように抽出された人間に対応する領域と、当該領域を抽出する際に用いた温熱画像情報と、に基づいて得られるパラメータを用いて処理対象とする空間内の快適性を示す快適性指標を導出し、当該快適性指標に基づいて当該空間内の環境を制御しているので、制御対象とする空間内の実態に即した適切な温熱環境の制御を行うことができる。
【００９４】
なお、本実施の形態では、人間領域を検出する際に用いる閾値を、後で収集された温熱画像情報（図７に示す例では温熱画像ＩＤ２）の非重複領域に対応する温度値に基づいて導出した場合について説明したが、本発明はこれに限定されるものではなく、先に収集された温熱画像情報（図７に示す例では温熱画像ＩＤ１）の非重複領域（図７に示す例では非重複領域Ａ’）に対応する温度値に基づいて導出する形態とすることもでき、双方の温熱画像情報の非重複領域（図７に示す例では、非重複領域Ａ’及びＢ’）に対応する温度値に基づいて導出する形態とすることもできる。これらの場合も、本実施の形態と同様の効果を奏することができる。
【００９５】
また、本実施の形態では、後で収集された温熱画像情報から人間領域を抽出した場合について説明したが、本発明はこれに限定されるものではなく、先に収集された温熱画像情報から人間領域を抽出する形態とすることもできる。この場合も、本実施の形態と同様の効果を奏することができる。
【００９６】
〔第２実施形態〕
上記第１実施形態では、収集された温熱画像情報から直接人間領域を抽出する場合の形態について説明したが、本第２実施形態では、収集時間の異なる二つの温熱画像情報に基づいて生成される合成画像情報を用いて、人間の当該二つの温熱画像情報間の重複領域を抽出し、当該重複領域と前述の非重複領域とを合成することにより人間領域を抽出する場合の形態、すなわち、人間領域の抽出手法に請求項２及び請求項６に記載の発明を適用した場合の実施の形態について説明する。なお、本第２実施形態に係る環境制御システム１０の構成は、上記第１実施形態に係る環境制御システム１０と同様であるので、ここでの説明は省略する。
【００９７】
以下、本第２実施形態に係る環境制御システム１０の環境制御処理実行時の作用を説明する。なお、本第２実施形態に係る環境制御処理は、上記第１実施形態に係る環境制御処理（図４参照）に比較して、ステップ１１０において実行される人間領域検出処理のみが異なるので、ここでは、本第２実施形態に係る人間領域検出処理のみについて、図９及び図１０を参照しつつ説明する。ここで、図９は、本第２実施形態に係る人間領域検出処理を行う際に、ＰＣ２０ＡのＣＰＵ２２によって実行される人間領域検出処理プログラムの処理の流れを示すフローチャートであり、当該プログラムもＲＯＭ２６の所定領域に予め記憶されている。また、図９における図６と同一の処理を行うステップについては図６と同一のステップ番号を付して、その説明を省略する。
【００９８】
図９のステップ２０２’では、上記ステップ２００で得られた差分画像情報において負の値を示す画素により構成される領域と、正の値を示す画素により構成される領域とを、非重複領域として抽出する。これにより、図１０に示す例では、非重複領域Ａ’及び非重複領域Ｂ’のみが抽出されることになる。なお、同図における温熱画像ＩＤ１、温熱画像ＩＤ２及び差分画像ＳＤは、図７と同様のものである。
【００９９】
次のステップ２０４’では、上記負の値を示す画素により構成される領域として抽出した非重複領域の画素に対応する画素値（温度値）と、上記正の値を示す画素により構成される領域として抽出した非重複領域の画素に対応する画素値とを、各々、先に収集された温熱画像情報（図１０に示す例では温熱画像ＩＤ１）と、後で収集された温熱画像情報（図１０に示す例では温熱画像ＩＤ２）とから抽出することにより、各温熱画像情報における非重複領域の温度値を抽出する。これにより、図１０に示す例では、非重複領域Ａ’及び非重複領域Ｂ’の各領域の温度値が領域別に抽出されることになる。
【０１００】
そして、ステップ２０４’では、抽出した各非重複領域毎の温度値に基づいて、次の（５）式及び（６）式により、二つの異なるタイミングで収集された温熱画像情報の各々に含まれる人間の当該二つの温熱画像情報間の重複領域を抽出するための二つの閾値（下限閾値ＴＨＬ及び上限閾値ＴＨＨ）を導出する。
【０１０１】
ＴＨＬ＝Ａｔ＋Ｂｔ−ＴＨ２　　　・・・（５）
ＴＨＨ＝Ａｔ＋Ｂｔ＋ＴＨ２　　　・・・（６）
ここで、Ａｔは本ステップ２０４’で負の値を示す画素に対応するものとして抽出した温度値の平均値を、Ｂｔは本ステップ２０４’で正の値を示す画素に対応するものとして抽出した温度値の平均値を、ＴＨ２は各閾値ＴＨＬ、ＴＨＨの許容誤差（例えば、０．１℃）を、各々示す。
【０１０２】
次のステップ２０８では、上記ステップ１０４においてハードディスク２８に記憶した温熱画像情報（先に収集された温熱画像情報）と、上記ステップ１０８においてハードディスク２８に記憶した温熱画像情報（後で収集された温熱画像情報）と、の画素毎の加算を行うことにより、合成画像情報を導出する。
【０１０３】
例えば、図１０に示す例では、本ステップ２０８で得られる合成画像情報により示される画像は合成画像ＧＤのようになる。同図に示すように、本ステップ２０８で得られる合成画像情報により示される合成画像ＧＤは、発熱体以外の静止物の領域Ｈについては、当該領域Ｈに対応する二つの温熱画像ＩＤ１、ＩＤ２の各々の温度値が加算された値（図１０では、‘２Ｈ’と表現。）を示す画像となり、静止物である発熱体の領域Ｗについては、当該領域Ｗに対応する二つの温熱画像ＩＤ１、ＩＤ２の温度値が加算された値（図１０では、‘２Ｗ’と表現。）を示す画像となる。また、人間の上記二つの温熱画像ＩＤ１、ＩＤ２間の非重複領域Ａ’、Ｂ’については、当該領域Ａ’、Ｂ’に対応する人間の領域の温度値と静止物の領域の温度値とが加算された値（図１０では、‘Ａ＋Ｈ’及び‘Ｂ＋Ｈ’と表現。）を示す画像となり、人間の上記二つの温熱画像ＩＤ１、ＩＤ２間の重複領域については、当該領域に対応する二つの温熱画像情報ＩＤ１、ＩＤ２の各々の温度値が加算された値（図１０では、‘Ａ＋Ｂ’と表現。）を示す画像となる。
【０１０４】
次のステップ２１０では、上記ステップ２０８において導出された合成画像情報から、上記ステップ２０４’で導出された下限閾値ＴＨＬから上限閾値ＴＨＨまでの範囲内の値を示す画素を抽出することにより、当該合成画像情報における人間の上記重複領域を抽出する。これにより、図１０に示す例では、重複領域のみ（‘Ａ＋Ｂ’の領域）が抽出されることになる。
【０１０５】
次のステップ２１２では、上記ステップ２０２’において画素値が正の値であるものとして抽出した非重複領域と、上記ステップ２１０において抽出した重複領域とを合成することにより、人間領域を抽出する。これによって、図１０に示す例では、人間領域Ｂのみが得られることになる。
【０１０６】
このように、本実施の形態に係る人間領域検出処理でも、上記第１実施形態と同様に、人間は、例えば椅子に座っているような状態であっても、顔、腕、肩等の一部は動くものである、という点に着目し、差分画像情報を導出することによって人間の動いている領域（非重複領域）を抽出し、抽出した領域に対応する温熱画像情報に基づいて人間領域を抽出するための閾値（下限閾値ＴＨＬ及び上限閾値ＴＨＨ）を導出しており、この閾値は、処理対象空間内の気温や季節等の環境状態に応じた実際の人間の温度に基づくものであるので、処理対象空間内に人間の温度と略等しい温度の発熱体が存在する場合であっても、当該発熱体の温度が僅かでも人間の温度と異なる場合には、当該発熱体を抽出対象から除外することができ、人間部分の領域のみを高精度に抽出することができる。
【０１０７】
上記ステップ２１２の処理が終了すると本人間領域検出処理を終了し、環境制御処理プログラム（図４参照）のステップ１１２に移行する。
【０１０８】
本実施の形態に係る人間領域検出処理プログラムにおけるステップ２００の処理が請求項２記載の発明の差分画像導出手段に、ステップ２０２’の処理が請求項２記載の発明の非重複領域抽出手段に、ステップ２０４’の処理が請求項２記載の発明の閾値導出手段に、ステップ２０８の処理が請求項２記載の発明の合成画像導出手段に、ステップ２１０の処理が請求項２記載の発明の重複領域抽出手段に、ステップ２１２の処理が請求項２記載の発明の人間領域抽出手段に、各々相当する。
【０１０９】
図１１及び図１２には、本第２実施形態に係る環境制御システム１０による環境制御処理の実行途中で実際に得られた各種画像情報により示される画像の一例が示されている。なお、図１１では、便宜上、温熱画像における温度の差異を線の密度で５段階程度で表現しているが、実際には、例えば、５０℃の領域を赤とし、５０℃から１０℃にかけて温度が低くなるに従って予め定められた温度範囲毎に多段階で寒色系の色に徐々に変化する、というように、各温熱画像は温度を示す色が付された状態とされている。
【０１１０】
図１１（Ａ）は、環境制御処理（図４）のステップ１０４の処理で得られた温熱画像情報により示される温熱画像を、図１１（Ｂ）は、環境制御処理のステップ１０８の処理で得られた温熱画像情報により示される温熱画像を、各々示している。なお、同図に示すように、ここでは、処理対象空間内に多数の人間が着席した状態で存在し、発熱体としてＰＣ及びプリンタが固定的に設置されている場合について示されている。
【０１１１】
この場合、ＰＣ及びプリンタは静止しているため、図１１（Ａ）及び図１１（Ｂ）とも略同様の状態となっているが、人間については若干の動きがあるため、図１１（Ａ）と図１１（Ｂ）とで若干異なる温度分布となっている。
【０１１２】
これに対し、図１１（Ｃ）は、人間領域検出処理（図９）のステップ２００の処理によって得られた差分画像情報により示される差分画像を示しており、図１２（Ａ）は、人間領域検出処理のステップ２０２’の処理によって抽出された非重複領域の画像を示している。なお、図１２（Ａ）では、図１１（Ｃ）に示す画像を２値化することにより非重複領域を抽出した場合の２値化画像が示されている。
【０１１３】
図１１（Ｃ）及び図１２（Ａ）に示す画像では、静止しているＰＣ及びプリンタの領域と、人間の重複領域については殆どの画像が消去され、人間の非重複領域の殆どの画像が残っている。従って、図１１（Ｃ）に示すような差分画像となる差分画像情報に基づいて得られた閾値によって人間領域を抽出することにより、人間領域を高精度に抽出できることがわかる。
【０１１４】
一方、図１２（Ｂ）は、人間領域検出処理のステップ２１０の処理によって抽出された重複領域の画像を示しており、図１２（Ｃ）は、人間領域検出処理のステップ２１２の処理によって抽出された人間領域の画像を示している。なお、図１２（Ｂ）及び図１２（Ｃ）では、抽出された重複領域ないし人間領域のみを黒レベルとした場合の２値化画像が示されている。
【０１１５】
図１２（Ｂ）に示す画像では、人間の重複領域のみが的確に抽出できている。この結果、図１２（Ｃ）に示されるように、人間領域のみを高精度に抽出することができる。
【０１１６】
以上詳細に説明したように、本実施の形態に係る環境制御システムでは、非重複領域に対応する温熱画像情報により示される温度に基づいて、人間の上記二つの温熱画像情報間の重複領域を抽出するための閾値を導出すると共に、上記二つの温熱画像情報における分割領域（本実施の形態では、画素）毎の加算によって当該分割領域毎の合成画像情報を導出し、当該合成画像情報と上記閾値との上記分割領域毎の比較により人間の上記二つの温熱画像情報間の重複領域を抽出し、上記非重複領域と上記重複領域とを合成することにより人間に対応する領域を抽出しているので、単に温熱画像情報から抽出する場合に比較して、より高精度に人間に対応する領域を抽出することができる。
【０１１７】
また、本実施の形態に係る環境制御システムでは、以上のように抽出された人間に対応する領域と、当該領域を抽出する際に用いた温熱画像情報と、に基づいて得られるパラメータを用いて処理対象とする空間内の快適性を示す快適性指標を導出し、当該快適性指標に基づいて当該空間内の環境を制御しているので、制御対象とする空間内の実態に即した適切な温熱環境の制御を行うことができる。
【０１１８】
なお、本実施の形態では、下限閾値ＴＨＬ及び上限閾値ＴＨＨを（５）式及び（６）式を用いて算出した場合について説明したが、本発明はこれに限定されるものではなく、例えば、次の（７）式及び（８）式を用いて算出する形態とすることもできる。また、（７）式及び（８）式のＢｔをＡｔに置き換えて算出することもできる。
【０１１９】
ＴＨＬ＝２×Ｂｔ−ＴＨ２　　　・・・（７）
ＴＨＨ＝２×Ｂｔ＋ＴＨ２　　　・・・（８）
これらの場合には、当該閾値を算出するために、一例として図１０に示される領域Ａ’及び領域Ｂ’の何れか一方のみを抽出すればよいので、本実施の形態に比較して、演算負荷を低減することができる。
【０１２０】
また、本実施の形態では、人間領域検出処理のステップ２０８における合成画像情報を二つの温熱画像情報の画素毎の加算によって導出した場合について説明したが、本発明はこれに限定されるものではなく、例えば、二つの温熱画像情報の画素毎の乗算によって導出する形態とすることもできる。この場合は、本実施の形態に比較して、生成された合成画像情報における人間領域の重複領域の情報と、発熱体の重複領域の情報との差を大きくすることができるので、本実施の形態に比較して、より高精度に人間領域の重複領域を抽出することができる。
【０１２１】
〔第３実施形態〕
上記第１、第２実施形態では、差分画像情報に基づいて導出した閾値を用いることにより、人間領域を抽出する場合の形態について説明したが、本第３実施形態では、温熱画像情報から人間の領域の候補となる領域を抽出し、当該領域から差分画像情報に基づいて抽出された非重複領域に近接する領域を人間領域であるものとして抽出する場合の形態、すなわち、人間領域の抽出手法として請求項３及び請求項７に記載の発明を適用した場合の形態について説明する。なお、本第３実施形態に係る環境制御システム１０の構成は、上記第１、第２実施形態に係る環境制御システム１０と同様であるので、ここでの説明は省略する。
【０１２２】
以下、本第３実施形態に係る環境制御システム１０の環境制御処理実行時の作用を説明する。なお、本第３実施形態に係る環境制御処理も、上記第１実施形態に係る環境制御処理（図４参照）に比較して、ステップ１１０において実行される人間領域検出処理のみが異なるので、ここでも、本第３実施形態に係る人間領域検出処理のみについて、図１３及び図１４を参照しつつ説明する。なお、図１３は、本第３実施形態に係る人間領域検出処理を行う際に、ＰＣ２０ＡのＣＰＵ２２によって実行される人間領域検出処理プログラムの処理の流れを示すフローチャートであり、当該プログラムもＲＯＭ２６の所定領域に予め記憶されている。また、図１３における図６と同一の処理を行うステップについては図６と同一のステップ番号を付して、その説明を省略する。
【０１２３】
図１３のステップ２１４では、後で収集された温熱画像情報（図１４に示す例では温熱画像ＩＤ２）から、人間の温度に対応する領域を抽出するために予め定めた下限閾値ＴＨＬ（ここでは、３０℃）から上限閾値ＴＨＨ（ここでは、３７℃）までの範囲内の値を示す画素により構成される領域を人間候補領域として抽出する。これにより、図１４に示す例では、人間領域Ｂ及び人間の温度域と略同様の温度となる発熱体の領域Ｗが人間候補領域として抽出されることになる。なお、図１４における温熱画像ＩＤ１、温熱画像ＩＤ２、差分画像ＳＤ及び非重複領域Ｂ’は、図７と同様のものである。
【０１２４】
次のステップ２１６では、上記ステップ２１４で抽出された人間候補領域のうち、上記ステップ２０２の処理によって抽出された非重複領域から所定距離以内に位置する領域のみを人間領域として抽出する。なお、上記非重複領域から所定距離以内に位置する領域には、非重複領域内に位置する領域（非重複領域と重複する領域）、及び非重複領域に接する領域が含まれる。また、ここでは、上記所定距離に対応する距離として、非重複領域と人間候補領域との間の最も近接する画素間の距離を適用するものとする。
【０１２５】
これにより、図１４に示す例では、非重複領域Ｂ’から所定距離以内に位置する領域として、非重複領域Ｂ’と重複する領域及び非重複領域Ｂ’に接する領域から構成される人間領域Ｂのみが的確に抽出される。
【０１２６】
このように、本実施の形態に係る人間領域検出処理でも、人間は、例えば椅子に座っているような状態であっても、顔、腕、肩等の一部は動くものである、という点に着目し、差分画像情報を導出することによって確実に人間の領域（非重複領域）を抽出し、抽出した領域に近接する人間候補領域を人間領域として抽出しており、処理対象空間内に人間の温度と略等しい温度の発熱体が存在する場合であっても、当該発熱体を抽出対象から除外することができ、人間部分の領域のみを高精度に抽出することができるようにしている。
【０１２７】
上記ステップ２１６の処理が終了すると本人間領域検出処理を終了し、環境制御処理プログラム（図４参照）のステップ１１２に移行する。
【０１２８】
本実施の形態に係る人間領域検出処理プログラムにおけるステップ２００の処理が請求項３記載の発明の差分画像導出手段に、ステップ２０２の処理が請求項３記載の発明の非重複領域抽出手段に、ステップ２１４の処理が請求項３記載の発明の候補領域抽出手段に、ステップ２１６の処理が請求項３記載の発明の人間領域抽出手段に、各々相当する。
【０１２９】
以上詳細に説明したように、本実施の形態に係る環境制御システムでは、処理対象とする空間内を２次元平面として捉え、当該２次元平面の予め定められた分割領域（本実施の形態では、画素）毎の温度を示す温熱画像情報を収集する赤外線カメラ１２により収集された時間的に異なる二つの温熱画像情報における上記分割領域毎の減算によって当該分割領域毎の差分画像情報を導出し、当該差分画像情報における零を含む所定範囲の外の値となる分割領域により構成される領域を人間の上記二つの温熱画像情報間の非重複領域として抽出し、当該二つの温熱画像情報の何れかと、人間の温度に対応する領域を抽出するために予め定めた閾値との上記分割領域毎の比較により当該温熱画像情報における人間の領域の候補となる領域を抽出し、抽出した領域のうち、上記非重複領域から所定距離以内に位置する領域のみを人間に対応する領域として抽出しているので、温熱画像から人間部分の領域を高精度に抽出することができる。
【０１３０】
また、本実施の形態に係る環境制御システムでは、以上のように抽出された人間に対応する領域と、当該領域を抽出する際に用いた温熱画像情報と、に基づいて得られるパラメータを用いて処理対象とする空間内の快適性を示す快適性指標を導出し、当該快適性指標に基づいて当該空間内の環境を制御しているので、制御対象とする空間内の実態に即した適切な温熱環境の制御を行うことができる。
【０１３１】
なお、本実施の形態では、後で収集された温熱画像情報（図１４に示す例では温熱画像ＩＤ２）から非重複領域を抽出した場合について説明したが、本発明はこれに限定されるものではなく、先に収集された温熱画像情報（図１４に示す例では温熱画像ＩＤ１）から非重複領域（図１４に示す例では非重複領域Ａ’）を抽出する形態とすることもでき、双方の温熱画像情報から非重複領域（図１４に示す例では、非重複領域Ａ’及びＢ’）を抽出する形態とすることもできる。これらの場合も、本実施の形態と同様の効果を奏することができる。
【０１３２】
また、本実施の形態では、後で収集された温熱画像情報から人間候補領域を抽出した場合について説明したが、本発明はこれに限定されるものではなく、先に収集された温熱画像情報から人間候補領域を抽出する形態とすることもできる。この場合も、本実施の形態と同様の効果を奏することができる。
【０１３３】
更に、本実施の形態では、人間候補領域を抽出するために用いる閾値として予め定められた閾値を適用した場合について説明したが、本発明はこれに限定されるものではなく、例えば、上記第１実施形態に係る人間領域検出処理（図６）のステップ２０４において導出した閾値や、上記第２実施形態に係る人間領域検出処理（図９）のステップ２０４’において導出した閾値を適用する形態とすることもできる。この場合、これらの閾値は、処理対象空間内の気温や季節等の環境状態に応じた実際の人間の温度に基づくものであるので、処理対象空間内に人間の温度と略等しい温度の発熱体が存在する場合であっても、当該発熱体の温度が僅かでも人間の温度と異なる場合には、当該発熱体を抽出対象から除外することができ、より的確に人間候補領域を抽出することができる。
【０１３４】
〔第４実施形態〕
上記第３実施形態では、収集された温熱画像情報から直接人間候補領域を抽出する場合の形態について説明したが、本第４実施形態では、収集時間の異なる二つの温熱画像情報に基づいて生成される合成画像情報を用いて、人間の当該二つの温熱画像情報間の重複領域を人間候補領域として抽出する場合の形態、すなわち、人間領域の抽出手法に請求項４及び請求項８に記載の発明を適用した場合の実施の形態について説明する。なお、本第４実施形態に係る環境制御システム１０の構成も、上記第１実施形態に係る環境制御システム１０と同様であるので、ここでの説明は省略する。
【０１３５】
以下、本第４実施形態に係る環境制御システム１０の環境制御処理実行時の作用を説明する。なお、本第４実施形態に係る環境制御処理も、上記第１実施形態に係る環境制御処理（図４参照）に比較して、ステップ１１０において実行される人間領域検出処理のみが異なるので、ここでは、本第４実施形態に係る人間領域検出処理のみについて、図１５及び図１６を参照しつつ説明する。ここで、図１５は、本第４実施形態に係る人間領域検出処理を行う際に、ＰＣ２０ＡのＣＰＵ２２によって実行される人間領域検出処理プログラムの処理の流れを示すフローチャートであり、当該プログラムもＲＯＭ２６の所定領域に予め記憶されている。また、図１５における図６と同一の処理を行うステップについては図６と同一のステップ番号を付して、その説明を省略する。
【０１３６】
図１５のステップ２１８では、第２実施形態に係る人間領域検出処理（図９）のステップ２０８と同様にして合成画像情報を導出する。これにより、図１６に示す例では、合成画像ＧＤを示す合成画像情報を得ることができる。なお、同図における温熱画像ＩＤ１、温熱画像ＩＤ２、差分画像ＳＤ及び合成画像ＧＤは、図１０と同様のものである。
【０１３７】
次のステップ２２０では、上記ステップ２１８において導出された合成画像情報と、人間の異なるタイミングで収集された二つの温熱画像情報間の重複領域を抽出するためのものとして予め定めた下限閾値ＴＨＬ（ここでは、６０℃）から上限閾値ＴＨＨ（ここでは、７４℃）までの範囲内の値を示す画素により構成される領域を人間候補領域として抽出する。これにより、図１６に示す例では、人間の重複領域、及び人間の温度域と略同様の温度となる発熱体の領域Ｗが人間候補領域として抽出されることになる。
【０１３８】
次のステップ２２２では、上記ステップ２２０で抽出された人間候補領域のうち、上記ステップ２０２の処理によって抽出された非重複領域から所定距離以内に位置する領域のみを人間領域として抽出する。なお、上記非重複領域から所定距離以内に位置する領域には、上記第３実施形態と同様に、非重複領域内に位置する領域（非重複領域と重複する領域）、及び非重複領域に接する領域が含まれる。また、ここでも、上記所定距離に対応する距離として、非重複領域と人間候補領域との間の最も近接する画素間の距離を適用するものとする。これにより、図１６に示す例では、非重複領域Ｂ’から所定距離以内に位置する領域として、人間の重複領域のみが抽出される。
【０１３９】
そして、本ステップ２２２では、抽出された領域（人間の重複領域）と上記ステップ２０２の処理によって抽出された非重複領域とを合成することによって人間領域を抽出する。これにより、図１６に示す例では、人間領域Ｂのみが高精度に抽出される。
【０１４０】
本実施の形態に係る人間領域検出処理プログラムにおけるステップ２００の処理が請求項４記載の発明の差分画像導出手段に、ステップ２０２の処理が請求項４記載の発明の非重複領域抽出手段に、ステップ２１８の処理が請求項４記載の発明の合成画像導出手段に、ステップ２２０の処理が請求項４記載の発明の候補領域抽出手段に、ステップ２２２の処理が請求項４記載の発明の人間領域抽出手段に、各々相当する。
【０１４１】
以上詳細に説明したように、本実施の形態に係る環境制御システムでは、二つの温熱画像情報における分割領域（本実施の形態では、画素）毎の加算によって当該分割領域毎の合成画像情報を導出し、当該合成画像情報と、人間の上記二つの温熱画像情報間の重複領域を抽出するために予め定めた閾値との上記分割領域毎の比較により人間の領域の候補となる領域を抽出し、抽出した領域のうちの非重複領域から所定距離以内に位置する領域と、当該非重複領域とを合成することによって人間に対応する領域を抽出しているので、単に温熱画像情報から人間の領域の候補となる領域を抽出する場合に比較して、より高精度に人間に対応する領域を抽出することができる。
【０１４２】
また、本実施の形態に係る環境制御システムでは、以上のように抽出された人間に対応する領域と、当該領域を抽出する際に用いた温熱画像情報と、に基づいて得られるパラメータを用いて処理対象とする空間内の快適性を示す快適性指標を導出し、当該快適性指標に基づいて当該空間内の環境を制御しているので、制御対象とする空間内の実態に即した適切な温熱環境の制御を行うことができる。
【０１４３】
なお、本実施の形態では、後で収集された温熱画像情報（図１６に示す例では温熱画像ＩＤ２）から非重複領域を抽出した場合について説明したが、本発明はこれに限定されるものではなく、先に収集された温熱画像情報（図１６に示す例では温熱画像ＩＤ１）から非重複領域（図１６に示す例では非重複領域Ａ’）を抽出する形態とすることもでき、双方の温熱画像情報から非重複領域（図１６に示す例では、非重複領域Ａ’及びＢ’）を抽出する形態とすることもできる。これらの場合も、本実施の形態と同様の効果を奏することができる。
【０１４４】
更に、本実施の形態では、人間候補領域を抽出するために用いる閾値として予め定められた閾値を適用した場合について説明したが、本発明はこれに限定されるものではなく、例えば、上記第１実施形態に係る人間領域検出処理（図６）のステップ２０４において導出した閾値や、上記第２実施形態に係る人間領域検出処理（図９）のステップ２０４’において導出した閾値を適用する形態とすることもできる。この場合、これらの閾値は、処理対象空間内の気温や季節等の環境状態に応じた実際の人間の温度に基づくものであるので、処理対象空間内に人間の温度と略等しい温度の発熱体が存在する場合であっても、当該発熱体の温度が僅かでも人間の温度と異なる場合には、当該発熱体を抽出対象から除外することができ、より的確に人間候補領域を抽出することができる。
【０１４５】
また、上記各実施の形態では、快適域をＰＭＶ値が＋０．５から−０．５までの範囲であるものとした場合について説明したが、本発明はこれに限定されるものではなく、例えば、＋０．６から−０．６までの範囲や、＋０．４から−０．４までの範囲等、当該快適域を示す値は、環境制御の対象とする空間の条件等に応じて適宜変更することができることは言うまでもない。
【０１４６】
また、上記各実施の形態では、空調制御処理（図４のステップ１２２参照）において、空調装置４４から吹き出される空気の温度、当該空気の湿度、及び当該空気の風量の三つのパラメータを全て調整する場合について説明したが、本発明はこれに限定されるものではなく、これら三つのパラメータの一つ又は二つを調整する形態とすることもできる。この場合には、上記各実施の形態に比較して調整対象の数が減るので、当該調整によって快適性が変化するまでに要する時間が長時間化するものの、調整のための制御を簡易化することができる。
【０１４７】
また、上記各実施の形態では、録画装置１６を介して温熱画像情報を収集することができるＰＣ２０Ａによって環境制御処理を実行する場合について説明したが、本発明はこれに限定されるものではなく、録画装置１６を介することなく温熱画像情報を収集することのできるＰＣ２０Ｂによって環境制御処理を実行する形態とすることもできることはいうまでもない。この場合も、上記各実施の形態と同様の効果を奏することができる。
【０１４８】
また、上記各実施の形態では、本発明の情報処理装置及び情報処理方法を温熱環境を制御する環境制御装置に適用した場合について説明したが、本発明はこれに限定されるものではなく、一例として図１７に示すように、防災・施設維持管理装置５０を用いた防災・施設維持管理システムに適用する形態とすることもできる。
【０１４９】
この場合、上記各実施の形態に係るＰＣ２０Ａは、人間領域検出処理によって人間領域が検出された場合に、処理対象空間内に人間が存在することを示す信号及び人間の存在位置を示す信号を防災・施設維持管理装置５０に出力する。これに応じて防災・施設維持管理装置５０では、この時点が人間の存在しない時間帯（深夜や休日等）であるときには、不審者が処理対象空間内に侵入している可能性があるものと判断して、当該空間内を確認する、警報を発する、警備会社に通報する、といった対応を行う。
【０１５０】
なお、このようなシステムでは、通常、処理対象空間内の生画像をビデオカメラを用いて収集する、といったことを行うが、この場合、当該空間内に存在する個人が特定されるので、プライバシーの問題があった。しかしながら、本発明に係る情報処理装置を適用したシステムでは、生画像を用いることなくシステムを構築することができるので、このようなプライバシーの問題が生じることはない。
【０１５１】
更に、上記各実施の形態では、本発明の快適性指標としてＰＭＶ値を適用した場合について説明したが、本発明はこれに限定されるものではなく、例えば、Ｇａｇｇｅらによって開発され、ＡＳＨＲＡＥ（米国空調学会）により採用されたものであり、発汗を含めた熱平衡式に基づく温熱指標であるＥＴ＊（ＥＴスター：新有効温度）を適用する形態とすることもできる。この場合にも、上記各実施の形態と同様の効果を奏することができる。
【０１５２】
【発明の効果】
請求項１に記載の情報処理装置及び請求項５に記載の情報処理方法によれば、処理対象とする空間内を２次元平面として捉え、当該２次元平面の予め定められた分割領域毎の温度を示す温熱画像情報を収集する収集手段により収集された時間的に異なる二つの温熱画像情報における上記分割領域毎の減算によって当該分割領域毎の差分画像情報を導出し、当該差分画像情報における零を含む所定範囲の外の値となる分割領域により構成される領域を人間の上記二つの温熱画像情報間の非重複領域として抽出し、当該非重複領域に対応する上記温熱画像情報により示される温度に基づいて、当該非重複領域と同一の温度を示す領域を抽出するための閾値を導出し、上記二つの温熱画像情報の何れかと上記閾値との上記分割領域毎の比較により当該温熱画像情報における上記人間に対応する領域を抽出しているので、温熱画像から人間部分の領域を高精度に抽出することができる、という効果が得られる。
【０１５３】
また、請求項２に記載の情報処理装置及び請求項６に記載の情報処理方法によれば、非重複領域に対応する温熱画像情報により示される温度に基づいて、人間の上記二つの温熱画像情報間の重複領域を抽出するための重複領域抽出閾値を導出すると共に、上記二つの温熱画像情報における上記分割領域毎の加算又は乗算によって当該分割領域毎の合成画像情報を導出し、当該合成画像情報と上記重複領域抽出閾値との上記分割領域毎の比較により人間の上記二つの温熱画像情報間の重複領域を抽出し、上記非重複領域と上記重複領域とを合成することにより人間に対応する領域を抽出しているので、単に温熱画像情報から抽出する場合に比較して、より高精度に人間に対応する領域を抽出することができる、という効果が得られる。
【０１５４】
また、請求項３に記載の情報処理装置及び請求項７に記載の情報処理方法によれば、処理対象とする空間内を２次元平面として捉え、当該２次元平面の予め定められた分割領域毎の温度を示す温熱画像情報を収集する収集手段により収集された時間的に異なる二つの温熱画像情報における上記分割領域毎の減算によって当該分割領域毎の差分画像情報を導出し、当該差分画像情報における零を含む所定範囲の外の値となる分割領域により構成される領域を人間の上記二つの温熱画像情報間の非重複領域として抽出し、当該二つの温熱画像情報の何れかと、人間の温度に対応する領域を抽出するために予め定めた閾値との上記分割領域毎の比較により当該温熱画像情報における人間の領域の候補となる領域を抽出し、抽出した領域のうち、上記非重複領域から所定距離以内に位置する領域のみを人間に対応する領域として抽出しているので、温熱画像から人間部分の領域を高精度に抽出することができる、という効果が得られる。
【０１５５】
また、請求項４に記載の情報処理装置及び請求項８に記載の情報処理方法によれば、二つの温熱画像情報における上記分割領域毎の加算又は乗算によって当該分割領域毎の合成画像情報を導出し、当該合成画像情報と、人間の上記二つの温熱画像情報間の重複領域を抽出するために予め定めた閾値との上記分割領域毎の比較により人間の領域の候補となる領域を抽出し、抽出した領域のうちの非重複領域から所定距離以内に位置する領域と、当該非重複領域とを合成することによって人間に対応する領域を抽出しているので、単に温熱画像情報から人間の領域の候補となる領域を抽出する場合に比較して、より高精度に人間に対応する領域を抽出することができる、という効果が得られる。
【０１５６】
更に、請求項９に記載の環境制御装置によれば、本発明に係る情報処理装置の人間領域抽出手段によって抽出された人間に対応する領域と、当該領域を抽出する際に用いた温熱画像情報と、に基づいて得られるパラメータを用いて処理対象とする空間内の快適性を示す快適性指標を導出し、当該快適性指標に基づいて当該空間内の環境を制御しているので、制御対象とする空間内の実態に即した適切な温熱環境の制御を行うことができる、という効果が得られる。
【図面の簡単な説明】
【図１】実施の形態に係る人間部分の非重複領域の抽出手法の説明に供する概略図である。
【図２】実施の形態に係る環境制御システム１０の全体構成を示す概略ブロック図である。
【図３】実施の形態に係るＰＣ２０Ａの電気系の構成を示すブロック図である。
【図４】実施の形態に係る環境制御処理プログラムの処理の流れを示すフローチャートである。
【図５】固定パラメータ設定画面の一例を示す概略図である。
【図６】第１実施形態に係る人間領域検出処理プログラムの処理の流れを示すフローチャートである。
【図７】第１実施形態に係る人間領域検出処理の処理手順の説明に供する説明図である。
【図８】実施の形態に係る環境制御処理の説明に供する、ＰＭＶ値と不満足者率との関係、及び快適域を示すグラフである。
【図９】第２実施形態に係る人間領域検出処理プログラムの処理の流れを示すフローチャートである。
【図１０】第２実施形態に係る人間領域検出処理の処理手順の説明に供する説明図である。
【図１１】第２実施形態に係る環境制御システム１０による環境制御処理の実行途中で実際に得られた温熱画像情報により示される温熱画像の一例を示す図である。
【図１２】第２実施形態に係る環境制御システム１０による環境制御処理の実行途中で実際に得られた画像情報により示される２値化画像の一例を示す図である。
【図１３】第３実施形態に係る人間領域検出処理プログラムの処理の流れを示すフローチャートである。
【図１４】第３実施形態に係る人間領域検出処理の処理手順の説明に供する説明図である。
【図１５】第４実施形態に係る人間領域検出処理プログラムの処理の流れを示すフローチャートである。
【図１６】第４実施形態に係る人間領域検出処理の処理手順の説明に供する説明図である。
【図１７】本発明の情報処理装置及び情報処理方法の他の適用例を示すブロック図である。
【図１８】従来技術の説明に供する模式図である。
【符号の説明】
１０　　環境制御システム
１２　　赤外線カメラ（収集手段）
１８Ａ、１８Ｂ　　キャプチャボード
２０Ａ、２０Ｂ　　パーソナル・コンピュータ
２２　　ＣＰＵ[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an information processing apparatus, an information processing method, and an environment control apparatus, and more specifically, views a space to be processed as a two-dimensional plane, and calculates a temperature of each predetermined divided area of the two-dimensional plane. The present invention relates to an information processing apparatus and an information processing method for performing processing on the thermal image information shown, and an environment control apparatus that controls the interior of the space to be a comfortable environment based on information obtained by the information processing apparatus.
[0002]
[Prior art]
2. Description of the Related Art In recent years, from the viewpoint of protection of useful natural resources such as fossil fuels and protection of the global environment, demands for energy saving of buildings have been increasing. On the other hand, the demand for comfortable living environments for buildings has been increasing year by year.
[0003]
In view of these points, conventionally, environmental factors such as temperature, humidity, wind speed, and radiation temperature in a space to be controlled, and the amount of clothes, metabolism (activity), etc. of a person existing in the space. And a comfort index (for example, a PMV (Predicted Mean Vote) value) indicating the comfort in the space based on the two elements of the space and the space, so that the comfort index is included in the comfort zone. There is an environmental control device that controls an air conditioner, a heating device, and the like provided in the inside. According to this device, since the control of the air conditioner, the heating device, and the like is appropriately performed based on the two elements of the environmental element and the element related to the person, when the control is performed based only on the environmental element, In comparison, useless operation of these devices can be suppressed, and as a result, energy can be saved and the living environment can be comfortable.
[0004]
The PMV value is developed by Fanger and adopted as ISO-7730, and is obtained by a comfort equation that predicts a thermal neutral temperature based on the amount of heat exchange between a human and the environment.
[0005]
By the way, when controlling the thermal environment in the space using the above-mentioned comfort index, conventionally, as for the surface temperature of the clothes of a person, a predicted value obtained by calculation or a fixed value given in advance has been applied, There is a problem that the actual condition of the person existing in the space to be controlled is not reflected, and the control of the thermal environment cannot always be performed appropriately.
[0006]
In view of this problem, in the technology described in JP-A-6-180139, a sensor is used to measure the temperature of the occupants in the room and the environment to obtain a thermal image, and the human image is obtained from the thermal image. Has been detected, the clothing surface temperature of a person has been acquired based on the temperature indicated by the thermal image corresponding to the region, and a comfort index has been derived using the clothing surface temperature.
[0007]
Here, in the technique described in JP-A-6-180139, in a technique for extracting a region that is a human temperature range from a thermal image as a region of a human part, heat is generated at substantially the same temperature as a human. In order to avoid the problem of erroneously extracting the region of the device as the region of the human part, attention is paid to the fact that humans move, and as shown in FIG. A difference image is obtained by taking the difference between two temporally different thermal images A and B obtained from the space. In the difference image, only the moving person remains, and a stationary object such as a personal computer (in FIG. 18, represented by “PC”) is canceled and erased regardless of the heat generation temperature. Therefore, the image area remaining in the difference image is detected as a human area.
[0008]
[Problems to be solved by the invention]
However, in the technique described in Japanese Patent Application Laid-Open No. 6-180139, the human portion is detected based on a difference image obtained by calculating a difference between two temporally different thermal images. If a human overlaps in the same area of the two temporally different thermal images, the overlapping area cannot be detected, and the human area cannot be properly detected. , There was a problem.
[0009]
That is, usually, a person in a room rarely moves at all times, and in most cases, only a part of a body such as an arm, a foot, and a neck moves. In this case, the human torso portion is present in the same region of the two temporally different thermal images in an overlapping manner, and the region of the torso portion is erased in the difference image. Therefore, in this case, it is not possible to detect a body part which is a main part of a human. The problem is that when acquiring the surface temperature of a person's clothing based on the temperature indicated by the thermal image corresponding to the area of the aforementioned human part, the clothing has a point that most of the clothing is located on the human torso. Considering this, it is a serious problem.
[0010]
The present invention has been made to solve the above problems, and has as its first object to provide an information processing apparatus and an information processing method capable of extracting a human part region from a thermal image with high accuracy. It is a second object of the present invention to provide an environment control device capable of appropriately controlling a thermal environment in accordance with the actual state of a space to be controlled.
[0011]
[Means for Solving the Problems]
In order to achieve the first object, the information processing apparatus according to claim 1 captures a space to be processed as a two-dimensional plane and indicates a temperature of each predetermined divided area of the two-dimensional plane. Collecting means for collecting thermal image information, and difference image deriving means for deriving difference image information for each of the divided areas by subtraction for each of the divided areas in two temporally different thermal image information collected by the collecting means; A non-overlapping region extracting means for extracting a region formed by a divided region having a value outside a predetermined range including zero in the difference image information as a non-overlapping region between the two thermal image information items of a human; Threshold deriving means for deriving a threshold for extracting an area having the same temperature as the non-overlapping area based on the temperature indicated by the thermal image information corresponding to the overlapping area; Comprises of the human region extraction means for extracting a region corresponding to the human in the thermal image information by comparing the divided every region of either said threshold thermal image information.
[0012]
According to the information processing apparatus of the first aspect, the collection unit that captures the inside of the space to be processed as a two-dimensional plane and collects thermal image information indicating the temperature of each predetermined divided region of the two-dimensional plane is provided. Difference image information for each of the divided regions is derived by difference image deriving means by subtraction for each of the divided regions in the collected two thermal image information items that differ in time. The collecting means includes all devices that can obtain thermal image information, such as an infrared camera and a radiation thermometer.
[0013]
Here, in the present invention, the non-overlapping area extracting means uses the difference image information to define an area formed by divided areas having a value outside a predetermined range including zero in the difference image information. A threshold for extracting an area showing the same temperature as the non-overlapping area is derived by the threshold deriving means, based on the temperature indicated by the thermal image information corresponding to the non-overlapping area, by the threshold deriving means. The human region extracting means extracts a region corresponding to the human in the thermal image information by comparing one of the two thermal image information with the threshold value for each of the divided regions.
[0014]
That is, the present invention focuses on the fact that a part of the face, arms, shoulders, and the like is moving even when the person is sitting in a chair, for example, A region where a person is moving (non-overlapping region) is extracted based on the image information, and a threshold for extracting a human region is derived based on thermal image information corresponding to the extracted region. This threshold value is highly accurate based on the actual human temperature according to the environmental conditions such as the temperature and the season in the space to be processed, which is collected by the collection unit. By applying the threshold as a threshold for extracting a region, even if a heating element having a temperature substantially equal to a human temperature exists in the space to be processed, even if the temperature of the heating element is small, If it is different from the above, the heating element can be excluded from the extraction target, and only the human part region can be extracted with high accuracy.
[0015]
Thus, according to the information processing apparatus of the first aspect, the inside of the space to be processed is regarded as a two-dimensional plane, and the thermal image information indicating the temperature of each predetermined divided area of the two-dimensional plane is obtained. The difference image information for each of the divided regions is derived by subtraction for each of the divided regions in the two temporally different thermal image information collected by the collecting means to be collected, and the difference image information outside the predetermined range including zero in the difference image information is obtained. A region formed by the divided regions having a value is extracted as a non-overlapping region between the two thermal image information items of the human, and the non-overlapping region is determined based on the temperature indicated by the thermal image information corresponding to the non-overlapping region. A threshold value for extracting a region having the same temperature as the region is derived, and one of the two thermal image information is compared with the threshold value for each of the divided regions, and the thermal image information is compared with the threshold value. Since the extracted region corresponding to the human, it is possible to extract a region of a human part with high accuracy from the thermal image.
[0016]
Further, in the information processing apparatus according to claim 2, in the invention according to claim 1, the threshold deriving means is configured to determine the threshold value of the human based on the temperature indicated by the thermal image information corresponding to the non-overlapping area. The overlap region extraction threshold for extracting the overlap region between two thermal image information is derived instead of the threshold, and the composite image information for each of the divided regions is obtained by adding or multiplying the two thermal image information for each of the divided regions. Synthetic image deriving means for deriving, and an overlapping area extracting means for extracting an overlapping area between the two thermal image information pieces of the human by comparing the synthesized image information and the overlapping area extraction threshold for each of the divided areas, The human region extracting unit further includes: combining the non-overlapping region extracted by the non-overlapping region extracting unit with the overlapping region extracted by the overlapping region extracting unit. And it extracts a region corresponding to the human by.
[0017]
According to the information processing apparatus of the second aspect, the threshold deriving means extracts the overlapping area between the two thermal image information of the human based on the temperature indicated by the thermal image information corresponding to the non-overlapping area. For the overlapping region extraction threshold for is derived, by the combined image deriving means, by the addition or multiplication of each of the divided regions in the two thermal image information, the combined image information for each of the divided regions is derived, by the overlapping region extraction means, By comparing each of the divided regions with the composite image information and the overlapping region extraction threshold, the overlapping region between the two thermal image information of the human is extracted, and further, by the human region extracting unit, by the non-overlapping region extracting unit By combining the extracted non-overlapping area and the overlapping area extracted by the overlapping area extracting means, the area corresponding to the person is extracted.
[0018]
That is, in the present invention, the combined image information for each of the divided regions is derived by adding or multiplying the two thermal image information for each of the divided regions. Therefore, this combined image information is obtained when the heating element having a temperature substantially equal to the human temperature exists in the space to be processed and the temperature of the heating element is slightly different from the human temperature. The difference between the values of the overlapping area of the human part between the image information and the overlapping area of the heating element is increased. Therefore, by extracting the overlapping region of the human part from the composite image information using the overlapping region extraction threshold derived by the threshold deriving means, the accuracy is higher than when simply extracting from the thermal image information. The overlapping area can be extracted, and by combining the overlapping area and the non-overlapping area, an area corresponding to a human can be extracted with higher accuracy.
[0019]
As described above, according to the information processing apparatus of the second aspect, based on the temperature indicated by the thermal image information corresponding to the non-overlapping region, the overlapping region between the two thermal image information of the human is extracted. And deriving composite image information for each of the divided regions by adding or multiplying each of the divided regions in the two thermal image information, and deriving the composite image information and the overlap region extraction threshold. By extracting the overlapping region between the two thermal image information of the human by the comparison for each divided region, and extracting the region corresponding to the human by combining the non-overlapping region and the overlapping region, simply An area corresponding to a human can be extracted with higher accuracy than when extracting from thermal image information.
[0020]
On the other hand, in order to achieve the first object, the information processing apparatus according to claim 3 captures a space to be processed as a two-dimensional plane, and sets a temperature for each predetermined divided area of the two-dimensional plane. And a difference image deriving unit that derives difference image information for each of the divided regions by subtraction for each of the divided regions in the two temporally different thermal image information collected by the collecting unit. Means, a non-overlapping area extracting means for extracting a region constituted by divided regions having values outside a predetermined range including zero in the difference image information as a non-overlapping area between the two thermal image information pieces of a human, A human region in the thermal image information is obtained by comparing each of the two thermal image information with a predetermined threshold for extracting a region corresponding to the human temperature for each of the divided regions. Candidate area extracting means for extracting a candidate area, and, of the areas extracted by the candidate area extracting means, only areas located within a predetermined distance from the non-overlapping area extracted by the non-overlapping area extracting means. And a human area extracting means for extracting the area corresponding to the human.
[0021]
According to the information processing apparatus of the third aspect, the collection unit that captures the inside of the space to be processed as a two-dimensional plane and collects thermal image information indicating the temperature of each predetermined divided region of the two-dimensional plane is provided. Difference image information for each of the divided regions is derived by difference image deriving means by subtraction for each of the divided regions in the collected two thermal image information items that differ in time. The collecting means includes all devices that can obtain thermal image information, such as an infrared camera and a radiation thermometer.
[0022]
Here, in the present invention, the non-overlapping area extracting means uses the difference image information to define an area formed by divided areas having a value outside a predetermined range including zero in the difference image information. The thermal image is extracted by the candidate region extracting means by comparing each of the two thermal image information with a predetermined threshold for extracting a region corresponding to the human temperature for each of the divided regions. A region which is a candidate for a human region in the information is extracted, and furthermore, by a human region extracting unit, of the regions extracted by the candidate region extracting unit, within a predetermined distance from the non-overlapping region extracted by the non-overlapping region extracting unit. Is extracted as a region corresponding to a human.
[0023]
The “region located within a predetermined distance from the non-overlapping region” includes a region located in the non-overlapping region (a region overlapping with the non-overlapping region) and a region in contact with the non-overlapping region. Further, as the distance corresponding to the predetermined distance, the distance between the closest divided region between the non-overlapping region and the candidate region of the human region, and the region that is a candidate of the non-overlapping region and the human region And the distance between the respective centers of gravity can be applied.
[0024]
That is, even in the present invention, the human focuses on the fact that a part of the face, arms, shoulders, and the like is moving even in a state of sitting on a chair, for example. A region where a human is moving (non-overlapping region) is reliably extracted based on image information, and a candidate region of a human region close to the extracted region is extracted as a region corresponding to a human. Even if there is a heating element having a temperature substantially equal to the temperature of a person in the space, if the heating element is relatively far from the human, the heating element should be excluded from the extraction target. This makes it possible to extract only the human part region with high accuracy.
[0025]
As described above, according to the information processing apparatus of the third aspect, the inside of the space to be processed is regarded as a two-dimensional plane, and the thermal image information indicating the temperature of each predetermined divided area of the two-dimensional plane is obtained. The difference image information for each of the divided regions is derived by subtraction for each of the divided regions in the two temporally different thermal image information collected by the collecting means to be collected, and the difference image information outside the predetermined range including zero in the difference image information is obtained. In order to extract a region constituted by the divided regions having a value as a non-overlapping region between the two thermal image information of a human, and to extract any of the two thermal image information and a region corresponding to the human temperature, A region that is a candidate for a human region in the thermal image information is extracted by comparing each of the divided regions with a predetermined threshold, and within the extracted regions, a predetermined distance from the non-overlapping region is extracted. Since by extracting only a region located as a region corresponding to a human, it is possible to extract a region of a human part with high accuracy from the thermal image.
[0026]
According to a fourth aspect of the present invention, in the information processing apparatus according to the third aspect, a combined image deriving method is configured to derive combined image information for each of the divided regions by adding or multiplying the two thermal image information for each of the divided regions. The candidate area extracting means further comprises means for comparing the combined image information and a threshold value for extracting an overlapping area between the two thermal image information pieces of the human for each of the divided areas. A candidate region of a human region is extracted, and the human region extracting unit is within a predetermined distance from the non-overlapping region extracted by the non-overlapping region extracting unit among the regions extracted by the candidate region extracting unit. And the area corresponding to the human being is extracted by combining the area located at.
[0027]
According to the information processing apparatus of claim 4, the composite image deriving means derives the composite image information for each of the divided areas by adding or multiplying the two thermal image information for each of the divided areas, and the candidate area extracting means By the above, the candidate image of the human region is extracted by comparing the composite image information and the threshold for each of the divided regions with a predetermined threshold for extracting the overlapping region between the two thermal image information of the human, Further, by the human region extracting unit, a region located within a predetermined distance from the non-overlapping region extracted by the non-overlapping region extracting unit out of the regions extracted by the candidate region extracting unit, and the non-overlapping region An area corresponding to a human is extracted by being combined.
[0028]
That is, in the present invention, the combined image information for each of the divided regions is derived by adding or multiplying the two thermal image information for each of the divided regions. Therefore, this combined image information is obtained when the heating element having a temperature substantially equal to the human temperature exists in the space to be processed and the temperature of the heating element is slightly different from the human temperature. The difference between the values of the overlapping area of the human part between the image information and the overlapping area of the heating element is increased.
[0029]
Therefore, by extracting a candidate area of a human area based on a predetermined threshold for extracting the overlapping area from the composite image information, as compared with a case where the area is simply extracted from thermal image information, The candidate area can be extracted with higher accuracy, and as a result, an area corresponding to a human can be extracted with higher accuracy.
[0030]
Thus, according to the information processing apparatus according to claim 4, composite image information for each of the divided regions is derived by addition or multiplication for each of the divided regions in the two thermal image information, and the combined image information and A candidate area of a human is extracted by comparing each of the divided areas with a predetermined threshold value to extract an overlapping area between the two thermal image information of the human, and a non-overlapping area of the extracted area is extracted. Since a region corresponding to a human is extracted by combining a region located within a predetermined distance from the region and the non-overlapping region, a region to be a candidate for a human region is simply extracted from thermal image information. In this case, a region corresponding to a human can be extracted with higher accuracy than in the case of.
[0031]
On the other hand, in order to achieve the first object, the information processing method according to claim 5, wherein a space to be processed is regarded as a two-dimensional plane, and a temperature of each predetermined divided region of the two-dimensional plane is determined. Is collected at predetermined time intervals, and the difference image information for each of the divided regions is derived by subtraction for each of the divided regions in the collected two different thermal image information in time, and the difference image information in the difference image information is obtained. A region composed of divided regions having values outside a predetermined range including zero is extracted as a non-overlapping region between the two thermal image information of a human, and is indicated by the thermal image information corresponding to the non-overlapping region. Based on the temperature, derive a threshold for extracting an area indicating the same temperature as the non-overlapping area, and compare any of the two thermal image information with the threshold for each of the divided areas. And it extracts a region corresponding to the human in the thermal image information.
[0032]
Therefore, according to the information processing method of the fifth aspect, the operation is performed in the same manner as the first aspect of the present invention, and thus, similarly to the first aspect of the present invention, the region of the human part is extracted with high accuracy from the thermal image. can do.
[0033]
According to a sixth aspect of the present invention, in the information processing method according to the fifth aspect, based on the temperature indicated by the thermal image information corresponding to the non-overlapping area, the information between the two thermal image information of the human is provided. Deriving an overlapping region extraction threshold for extracting the overlapping region in place of the threshold, and deriving composite image information for each of the divided regions by addition or multiplication for each of the divided regions in the two thermal image information, By comparing the composite image information and the overlapping region extraction threshold for each of the divided regions, extracting the overlapping region between the two thermal image information of the human, and combining the non-overlapping region and the overlapping region, An area corresponding to a human is extracted.
[0034]
Therefore, according to the information processing method of the sixth aspect, the operation is the same as that of the second aspect of the invention. Thus, a region corresponding to a human can be extracted with higher accuracy.
[0035]
On the other hand, in order to achieve the first object, an information processing method according to claim 7, wherein a space to be processed is regarded as a two-dimensional plane, and a temperature of each predetermined divided region of the two-dimensional plane is determined. Is collected at predetermined time intervals, and the difference image information for each of the divided regions is derived by subtraction for each of the divided regions in the collected two different thermal image information in time, and the difference image information in the difference image information is obtained. A region constituted by divided regions having values outside a predetermined range including zero is extracted as a non-overlapping region between the two thermal image information of a human, and any one of the two thermal image information, By comparing each of the divided regions with a predetermined threshold value for extracting a corresponding region, a candidate region of a human region in the thermal image information is extracted, and among the extracted regions, the non-overlapping region is extracted. Only the area located within a predetermined distance from and extracts a region corresponding to the human.
[0036]
Therefore, according to the information processing method of the seventh aspect, the operation is performed in the same manner as the third aspect of the present invention, and thus, similarly to the third aspect of the present invention, the region of the human part is accurately extracted from the thermal image. can do.
[0037]
Further, in the information processing method according to claim 8, in the invention according to claim 7, the composite image information for each of the divided regions is derived by adding or multiplying the two thermal image information for each of the divided regions, and Image information and a region that is a candidate for the human region is extracted and extracted by comparing each of the divided regions with a predetermined threshold to extract an overlapping region between the two thermal image information of the human. An area corresponding to the person is extracted by combining an area located within a predetermined distance from the non-overlapping area with the non-overlapping area.
[0038]
Therefore, according to the information processing method of the eighth aspect, since it operates in the same manner as the invention of the fourth aspect, similarly to the fourth aspect of the present invention, it becomes a candidate for a human area simply from the thermal image information. An area corresponding to a human can be extracted with higher accuracy than when an area is extracted.
[0039]
On the other hand, in order to achieve the second object, an environment control device according to claim 9 includes an information processing device according to any one of claims 1 to 4 and the human area of the information processing device. Comfort indicating the comfort in the space to be processed using parameters obtained based on the region corresponding to the person extracted by the extraction means and the thermal image information used in extracting the region. A comfort index deriving unit that derives an index, and an environment control unit that controls an environment in the space based on the comfort index are provided.
[0040]
According to the ninth aspect of the present invention, the area corresponding to the human extracted by the human area extracting means of the information processing apparatus according to any one of the first to fourth aspects by the comfort index deriving means. And a parameter obtained based on the thermal image information used in extracting the region, a comfort index indicating the comfort in the space to be processed is derived, and the environment control unit determines the comfort index. The environment in the space to be processed is controlled based on the sex index.
[0041]
In addition, the above parameters include a human clothing surface temperature obtained as an average value of the thermal image information of the region corresponding to the human, a temperature indicated by the thermal image information of the region corresponding to the human, and a normal human body temperature. And the average radiation temperature obtained as the average value of the thermal image information of the area other than the area corresponding to the person. The comfort index includes a PMV value, ET *, and the like. Further, the control target by the environment control means includes all controls relating to comfort, such as temperature control, humidity control, and wind speed control in a space to be processed.
[0042]
That is, in the present invention, a region corresponding to a human extracted with high accuracy by the information processing apparatus of the present invention, and a thermal image collected from the space to be actually used for extracting the region. Using the information and parameters obtained based on the information, a comfort index indicating the comfort in the space to be processed is derived to control the thermal environment, and the predicted value obtained by calculation for the parameter As compared with the case where a comfort index is derived by applying a fixed value given in advance, more accurate control of the thermal environment can be performed according to the actual situation in the space to be controlled. ing.
[0043]
As a result, for example, when the environmental control device according to the present invention is applied as a device for controlling a thermal environment in a building, high quality, human-friendly and soft (cold / hot shock) in consideration of unsteadiness Less) building environment.
[0044]
As described above, according to the environment control device of the ninth aspect, the region corresponding to the person extracted by the human region extraction means of the information processing device according to the present invention and the heat used when extracting the region Since a comfort index indicating the comfort in the space to be processed is derived using the parameters obtained based on the image information and the comfort index, and the environment in the space is controlled based on the comfort index, Appropriate thermal environment control can be performed in accordance with the actual conditions in the space to be controlled.
[0045]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In each of the embodiments described below, a method of extracting a partial region of a human part (a non-overlapping region described later) is common to all embodiments.
[0046]
That is, in each embodiment, as shown in FIG. 1, a difference image C is derived by taking the difference between two temporally different thermal images A and B obtained from the real space. Is extracted as a non-overlapping area of the human part (the shaded area of the difference image C).
[0047]
In FIG. 1, the thermal image A and the thermal image B are represented by line images for convenience, but actually, for example, a region of 50 ° C. is set to red, and as the temperature decreases from 50 ° C. to 10 ° C. Each thermal image is represented by a color indicating the temperature, for example, gradually changing to a cool color for each predetermined temperature range.
[0048]
Here, a case will be described in which the information processing device, the information processing method, and the environment control device according to the present invention are applied to an environment control system that controls a thermal environment by an air conditioner.
[0049]
[First Embodiment]
First, the overall configuration of the environment control system 10 according to the present embodiment will be described with reference to FIG. As shown in the figure, the environment control system 10 according to the present embodiment is capable of collecting a temperature distribution in a substantially rectangular parallelepiped real space S to be controlled in a thermal environment over a wide area as much as possible (for example, a real position). It is installed near any wall surface of the space S and at a position substantially at the center of the wall surface.) The real space S is regarded as a two-dimensional plane, and a predetermined divided area (here, , Pixels) as thermal image information as moving image information.
[0050]
The environment control system 10 includes a display 14 that displays a thermal image indicated by the thermal image information collected by the infrared camera 12 as a moving image, and a recording device 16 that sequentially records the thermal image information collected by the infrared camera 12. And a capture board 18A having a function of capturing arbitrary information from thermal image information recorded in the recording device 16, and controlling capture by the capture board 18A, and thermal image information captured by the capture board 18A. And a personal computer (hereinafter, referred to as “PC”) 20 </ b> A that is capable of acquiring a personal computer.
[0051]
The environment control system 10 further includes a PC 20B to which a capture board 18B having a function of capturing arbitrary information from thermal image information collected by the infrared camera 12 is mounted.
[0052]
Although not shown in FIG. 2, the environment control system 10 includes an air conditioner (hereinafter, referred to as an “air conditioner”) provided in or near the real space S in addition to the above components. ), A temperature sensor, a humidity sensor, and a wind speed sensor for detecting the temperature, humidity, and wind speed in the real space S.
[0053]
Next, the electrical configuration of the PC 20A according to the present embodiment will be described with reference to FIG. As shown in FIG. 1, the PC 20A includes a CPU (Central Processing Unit) 22 that controls the entire operation of the PC 20A, a RAM (Random Access Memory) 24 used as a work area when the CPU 22 executes various processing programs, A ROM (Read Only Memory) 26 storing a processing program and various parameters, a hard disk 28 used for storing various information, a display 30 used for displaying various information, a keyboard and a mouse. The input device 32 is connected to the capture board 18A, the temperature sensor 38, the humidity sensor 40, and the wind speed sensor 42, and an interface (hereinafter, referred to as an “I / F”) for controlling transmission and reception of various information to and from these units. 34) and the aforementioned An I / F 36 for controlling the exchange of various information with the air conditioner 44 is connected to each other by a system bus BUS.
[0054]
Therefore, the CPU 22 accesses the RAM 24, the ROM 26, and the hard disk 28, controls the display of various information on the display 30, obtains various information via the input device 32, controls the capture board 18A via the I / F 34, The acquisition of information indicating the detection results of the temperature sensor 38, the humidity sensor 40, and the wind speed sensor 42 via the F34, and the control of the air conditioner 44 via the I / F 36 can be performed.
[0055]
Note that the configuration of the PC 20B is the same as that of the PC 20A except that the capture board 18A becomes the capture board 18B, and a description thereof will be omitted.
[0056]
Next, the operation of the environment control system 10 according to the present embodiment will be described with reference to FIG. FIG. 4 is a flowchart showing a process flow of an environment control processing program executed by the CPU 22 of the PC 20A when controlling the thermal environment in the environment control system 10. The program is stored in a predetermined area of the ROM 26 in advance. Have been. Here, a case will be described where thermal image information in the real space S is obtained as moving image information by the infrared camera 12, and the thermal image information is sequentially input to the capture board 18A via the recording device 16. I do.
[0057]
In step 100 of FIG. 4, the display 30 is controlled so as to display a predetermined fixed parameter setting screen. In the next step 102, input of predetermined information via the input device 32 is waited.
[0058]
FIG. 5 shows a fixed parameter setting screen displayed on the display 30 by the processing of step 100 described above. As shown in the figure, on the fixed parameter setting screen according to the present embodiment, a message prompting input of the metabolic rate and the amount of clothing is displayed, and a rectangular frame for inputting these parameters is displayed. You. When the fixed parameter setting screen as shown in the figure is displayed on the display 30, the operator of the PC 20A operates the keyboard and the mouse constituting the input device 32 to change the metabolic rate and the real space S in the real space S. After inputting each piece of information on the existing human clothing amount in the corresponding rectangular frame, the mouse is used to point to the “input end” button displayed at the bottom of the screen. As a result, information indicating the metabolic rate and the amount of clothing input by the operator and information indicating that the “input end” button has been specified are input to the CPU 22, and the above-described step 102 is affirmatively determined. Move to step 104.
[0059]
The metabolic rate indicates the amount of human activity, and is usually represented by a value called a Met value. For example, the Met value is 0.8 when lying down and relaxing in a lying position, 1.0 when sitting quietly in a chair, and 1.4 when relaxing in a standing position. The value is 1.6 when light work is performed at a speed, 2.0 when walking at 3 km / h, and 3.0 when running at 5 km / h or moving violently. In the present embodiment, the metabolic rate is [W / m ² ] As a unit, 1Met = 58 [W / m] ² ] Is applied.
[0060]
In step 104, by controlling the capture board 18A to capture the thermal image information input to the capture board 18A, the thermal image information collected by the infrared camera 12 at this time is collected and In a predetermined area (hereinafter, referred to as a “first area”), and in the next step 106, the elapse of a predetermined time (one minute in the present embodiment) is waited, and in the next step 108, Similarly to step 104, the thermal image information collected by the infrared camera 12 at this time is collected and stored in an area of the hard disk 28 different from the first area (hereinafter, referred to as "second area"). .
[0061]
In the next step 110, a human area detection process is performed using the thermal image information stored in the hard disk 28 in step 104 and the thermal image information stored in the hard disk 28 in step 108. Hereinafter, the human region detection processing according to the present embodiment will be described in detail with reference to FIG. FIG. 6 is a flowchart showing a processing flow of a human area detection processing program executed by the CPU 22 of the PC 20A when performing the human area detection processing, and the program is also stored in a predetermined area of the ROM 26 in advance. .
[0062]
In step 200 of FIG. 6, difference image information is derived by performing a subtraction for each pixel between the thermal image information stored in the hard disk 28 in step 104 and the thermal image information stored in the hard disk 28 in step 108. I do. In addition, this subtraction may be performed by subtracting the other thermal image information from either thermal image information, but here, a case is described in which the previously acquired thermal image information is subtracted from the later acquired thermal image information. explain.
[0063]
In deriving the difference image information, in the human area detection process according to the present embodiment, the absolute value of the value (value indicating the temperature difference) for each pixel obtained by the above subtraction is determined by a predetermined threshold value. If it is smaller than TH1 (here, 0.1 ° C.), the pixel value is set to 0 (zero).
[0064]
For example, as shown in FIG. 7, when the image indicated by the previously acquired thermal image information is the thermal image ID1 and the image represented by the thermal image information acquired later is the thermal image ID2, The image indicated by the difference image information obtained at 200 becomes like a difference image SD. In FIG. 7, A and B denote a human region, W denotes a region of a heating element (for example, a PC, a printer, or the like) which is a stationary object, and H denotes a stationary object other than the heating element (for example, an inner wall of a processing target space). , Desks, chairs, etc.).
[0065]
As shown in FIG. 7, the difference image SD indicated by the difference image information obtained in step 200 is such that only the non-overlapping region between the two thermal images ID1 and ID2 of the human corresponds to the human region corresponding to the region. Is an image showing the temperature of the difference between the temperature of the area of the stationary object and the temperature of the area of the stationary object, and the pixel values of all the other areas are 0 (zero). When (temperature of area A> temperature of area H) and (temperature of area B> temperature of area H), the environment control system 10 according to the present embodiment uses the human area A included in the thermal image ID1. Is a negative value, and the pixel value of the non-overlapping region corresponding to the human region B included in the thermal image ID2 is a positive value.
[0066]
Note that the temperature of the stationary object may slightly change between the time when one thermal image information is collected and the time when the other thermal image information is collected. If the absolute value of each value is smaller than a predetermined threshold value TH1, the pixel value is set to 0 (zero), so that the pixel value of the area of the stationary object can be set to 0. Therefore, the threshold value TH1 is set to a larger value as the change amount is larger, in accordance with the change amount of the temperature of the stationary object existing in the processing target space.
[0067]
In the next step 202, an area composed of pixels having a positive value in the difference image information obtained in step 200 is extracted as a non-overlapping area. As a result, in the example shown in FIG. 7, only the non-overlapping area B ′ is extracted.
[0068]
In the next step 204, a pixel value (temperature value) corresponding to the extracted pixel of the non-overlapping region is extracted from the thermal image information collected later (the thermal image ID2 in the example shown in FIG. 7), and the extracted temperature is extracted. Based on the values, two thresholds (lower threshold THL and upper threshold THH) for extracting an area having the same temperature as the non-overlap area are derived from the following equations (1) and (2).
[0069]
THL = Bt-TH2 (1)
THH = Bt + TH2 (2)
Here, Bt indicates an average value of the temperature values extracted in step 204, and TH2 indicates an allowable error (for example, 0.1 ° C.) of the threshold values THL and THH.
[0070]
In the next step 206, from the thermal image information collected later, a region composed of pixels indicating values within the range from the lower threshold value THL to the upper threshold value THH derived in step 204 is extracted. Thus, in the example shown in FIG. 7, only the human area B is extracted.
[0071]
As described above, in the human area detection processing according to the present embodiment, a part of a face, an arm, a shoulder, and the like is moved even when a human is, for example, sitting in a chair. To extract a moving area (non-overlapping area) of a human by deriving difference image information as described above, and to extract a human area based on thermal image information corresponding to the extracted area. Thresholds (lower threshold THL and upper threshold THH) are derived, and since these thresholds are based on actual human temperatures according to environmental conditions such as temperature and seasons in the processing target space, the processing target space Even if there is a heating element having a temperature substantially equal to the temperature of a human in the case, if the temperature of the heating element is slightly different from the human temperature, the heating element can be excluded from the extraction target. , Only the human area It can be extracted to.
[0072]
When the process of step 206 is completed, the human region detection process is completed, and the process proceeds to step 112 of the environment control processing program (see FIG. 4).
[0073]
The processing of step 200 in the human area detection processing program according to the present embodiment is performed by the difference image deriving means of the invention described in claim 1, and the processing of step 202 is performed by the non-overlapping area extraction means of the invention described in claim 1. The processing in step 204 corresponds to the threshold deriving means according to the first aspect of the present invention, and the processing in step 206 corresponds to the human area extracting means in the first aspect.
[0074]
In step 112, a pixel value (temperature value) corresponding to the human area obtained by the human area detection processing is extracted from the thermal image information (thermal image ID2 in the example shown in FIG. 7) collected later, and the temperature is extracted. The average of the values is calculated as the surface temperature (° C.) of the clothing of the person, and in the next step 114, the temperature value of the area other than the human area obtained by the human area detection processing is obtained from the thermal image information collected later. The average value of the extracted temperature values is calculated as an average radiation temperature (° C.).
[0075]
In the next step 116, information indicating the room temperature from the temperature sensor 38, information indicating the humidity from the humidity sensor 40, and information indicating the wind speed from the wind speed sensor 42 are obtained. In the next step 118, the above processing is performed. The parameters required to calculate the comfort index other than the obtained parameters (metabolism rate, clothing surface temperature, average radiation temperature) are derived.
[0076]
In the environment control system 10 according to the present embodiment, as an example, a PMV value calculated by the following equation (3) is applied as the comfort index.
[0077]
(Equation 1)

[0078]
Hc and fcl in the equation (3) are calculated by the following equations.
[0079]
(Equation 2)

[0080]
Where M: metabolic rate (W / m ² )
W: External work (W / m ² ) ・ ・ ・ Normally 0
Icl: thermal resistance of clothes (m ² K / W)
fcl: the ratio of the surface area at the time of clothing to the naked human case
ta: room temperature (° C)
tr: average radiation temperature (° C)
var: wind speed (m / s)
Pa: partial pressure of water vapor (Pa)
hc: convective thermal conductivity (W / m ² K)
tcl: surface temperature of clothing (° C)
It is.
[0081]
Here, the metabolic rate M, the room temperature ta, the average radiant temperature tr, the wind speed var, and the clothing surface temperature tcl are obtained by the above processing.
[0082]
Therefore, here, it is necessary to calculate the thermal resistance Icl of the clothing and the water vapor partial pressure Pa, and the thermal resistance Icl is a well-known arithmetic expression using the clothing amount (clo) indicated by the information obtained in step 102. And the water vapor partial pressure Pa can be derived by a well-known arithmetic expression using the room temperature ta and the humidity obtained by the humidity sensor 40.
[0083]
In the next step 120, the PMV value is calculated by substituting the various parameters obtained by the above processing into equation (3).
[0084]
By the way, usually, the clothing surface temperature tcl is represented by the following equation (4).
[0085]
[Equation 3]

[0086]
However, if the equation (4) is substituted into the equation (3), the equation (3) becomes a nonlinear equation related to the unknowns tcl and hc, and it is necessary to perform a convergence operation (iterative operation).
[0087]
On the other hand, in the environment control system 10 according to the present embodiment, the measured values of the surface temperature tcl and the average radiation temperature tr of the clothing based on the thermal image information are directly substituted into the expression (3). Can be omitted.
[0088]
In the next step 122, an air conditioning control process for controlling the operation of the air conditioner 44 based on the PMV value calculated in the above step 120 is executed.
[0089]
As shown in FIG. 8, the air conditioning control process according to the present embodiment reduces the temperature of the air blown out from air conditioner 44 when the PMV value is in the region (1) where the PMV value is +0.5 or more. At the same time, the air temperature is lowered by a predetermined temperature, the humidity of the air is lowered by a predetermined humidity, and the air volume of the air is increased by a predetermined amount. On the other hand, when the PMV value is in the range of (2) where the PMV value is equal to or less than -0.5, the temperature of the air blown from the air conditioner 44 is raised by a predetermined temperature, and the humidity of the air is reduced by the predetermined humidity. And the air volume of the air is increased by a predetermined amount.
[0090]
In the next step 124, it is determined whether or not it is time to end the present environmental control processing. If a negative determination is made, the process returns to step 106, and the processing of steps 106 to 122 is executed again. At this point, the present environmental control process ends. The determination as to whether or not it is time to terminate the present environmental control process may be made by determining whether or not an instruction to terminate the present environmental control process has been input by the operator of the PC 20A, or may be determined in advance. This can be performed by determining whether or not the operation time has elapsed. Further, when the processing of the above-mentioned steps 106 to 124 is repeatedly executed, in step 110, the thermal image information collected in step 108 immediately before that and the thermal image information previously collected in step 108 are processed. Thermal image information.
[0091]
By performing the air-conditioning operation at an appropriate timing by repeating the above-described steps 106 to 124, the PMV value is within the range of -0.5 to +0.5 as an example without performing the air-conditioning operation more than necessary. It is possible to enter the comfortable area (see FIG. 8), and as a result, it is possible to promote comfort and energy saving.
[0092]
As described above in detail, in the environment control system according to the present embodiment, the space to be processed is regarded as a two-dimensional plane, and a predetermined divided area of the two-dimensional plane (in the present embodiment, The difference image information for each of the divided areas is derived by subtraction for each of the divided areas in the two different thermal image information temporally collected by the infrared camera 12 that collects the thermal image information indicating the temperature for each pixel). A region constituted by divided regions having values outside a predetermined range including zero in the difference image information is extracted as a non-overlapping region between the two thermal image information of a human, and the thermal image corresponding to the non-overlapping region is extracted. Based on the temperature indicated by the information, to derive a threshold for extracting a region indicating the same temperature as the non-overlapping region, and either one of the two thermal image information and the threshold Since the extracted region corresponding to the human in the thermal image information by comparing each of the divided regions, it is possible to extract a region of a human part with high accuracy from the thermal image.
[0093]
Further, in the environment control system according to the present embodiment, using the parameters obtained based on the region corresponding to the human extracted as described above and the thermal image information used when extracting the region, A comfort index indicating the comfort in the space to be processed is derived, and the environment in the space is controlled based on the comfort index. Control of thermal environment can be performed.
[0094]
In the present embodiment, the threshold used for detecting the human region is determined based on the temperature value corresponding to the non-overlapping region of the thermal image information (the thermal image ID2 in the example shown in FIG. 7) collected later. Although the case of derivation has been described, the present invention is not limited to this, and the non-overlapping area of the previously acquired thermal image information (the thermal image ID1 in the example shown in FIG. 7) (the example shown in FIG. 7). It is also possible to adopt a mode in which the thermal image information is derived based on the temperature value corresponding to the non-overlapping area A ′). The non-overlapping areas (non-overlapping areas A ′ and B ′ in the example shown in FIG. It is also possible to adopt a form that is derived based on the corresponding temperature value. In these cases, the same effects as in the present embodiment can be obtained.
[0095]
Further, in the present embodiment, a case has been described in which a human region is extracted from thermal image information collected later. However, the present invention is not limited to this, and a human region is extracted from previously collected thermal image information. It is also possible to adopt a mode of extracting a region. In this case, the same effect as in the present embodiment can be obtained.
[0096]
[Second embodiment]
In the first embodiment, the case where the human region is directly extracted from the collected thermal image information has been described. However, in the second embodiment, the human region is generated based on two thermal image information having different collection times. Using the composite image information, the overlapping region between the two thermal image information of the human is extracted, and the human region is extracted by combining the overlapping region and the non-overlapping region, that is, the human An embodiment in a case where the inventions according to claims 2 and 6 are applied to an area extraction method will be described. The configuration of the environment control system 10 according to the second embodiment is the same as that of the environment control system 10 according to the first embodiment, and a description thereof will not be repeated.
[0097]
Hereinafter, the operation of the environmental control system 10 according to the second embodiment when executing the environmental control processing will be described. The environment control process according to the second embodiment differs from the environment control process according to the first embodiment (see FIG. 4) only in the human area detection process executed in step 110. Now, only the human region detection processing according to the second embodiment will be described with reference to FIGS. Here, FIG. 9 is a flowchart showing a flow of a human area detection processing program executed by the CPU 22 of the PC 20A when performing the human area detection processing according to the second embodiment. It is stored in a predetermined area in advance. Steps in FIG. 9 that perform the same processing as in FIG. 6 are assigned the same step numbers as in FIG. 6, and descriptions thereof are omitted.
[0098]
In step 202 ′ of FIG. 9, a region formed by pixels having a negative value and a region formed by pixels having a positive value in the difference image information obtained in step 200 are defined as non-overlapping regions. Extract. Thus, in the example shown in FIG. 10, only the non-overlapping area A ′ and the non-overlapping area B ′ are extracted. Note that the thermal image ID1, the thermal image ID2, and the difference image SD in FIG. 7 are the same as those in FIG.
[0099]
In the next step 204 ', the pixel value (temperature value) corresponding to the pixel of the non-overlapping area extracted as the area composed of the pixel having the negative value and the area composed of the pixel having the positive value The pixel values corresponding to the pixels of the non-overlapping region extracted as the thermal image information (the thermal image ID1 in the example shown in FIG. 10) and the thermal image information (FIG. 10 In the example shown in (1), the temperature value of the non-overlapping area in each thermal image information is extracted by extracting from the thermal image ID2). Thus, in the example shown in FIG. 10, the temperature values of the non-overlapping area A ′ and the non-overlapping area B ′ are extracted for each area.
[0100]
Then, in step 204 ′, based on the extracted temperature value for each non-overlapping area, the thermal image information is collected in two different timings according to the following equations (5) and (6). Two thresholds (lower threshold THL and upper threshold THH) for extracting an overlapping region between the two thermal image information of the human are derived.
[0101]
THL = At + Bt-TH2 (5)
THH = At + Bt + TH2 (6)
Here, At is an average value of the temperature values extracted as corresponding to the pixel having a negative value in the present step 204 ′, and Bt is extracted as corresponding to the pixel having a positive value in the present step 204 ′. TH2 indicates an average value of the temperature values, and TH2 indicates an allowable error (eg, 0.1 ° C.) of each of the threshold values THL and THH.
[0102]
In the next step 208, the thermal image information stored in the hard disk 28 in the above step 104 (the previously collected thermal image information) and the thermal image information stored in the hard disk 28 in the above step 108 (the thermal image collected later) ) Is derived for each pixel to derive composite image information.
[0103]
For example, in the example shown in FIG. 10, the image indicated by the composite image information obtained in step 208 becomes a composite image GD. As shown in the drawing, the composite image GD indicated by the composite image information obtained in step 208 is a two-dimensional thermal image ID1 and ID2 corresponding to the region H of the stationary object other than the heating element. An image indicating a value (expressed as “2H” in FIG. 10) obtained by adding the respective temperature values is obtained. For the region W of the heating element that is a stationary object, two thermal images ID1 and ID2 corresponding to the region W are generated. The image shows a value (expressed as “2W” in FIG. 10) to which the temperature value of ID2 is added. Further, regarding the non-overlapping areas A ′ and B ′ between the two thermal images ID1 and ID2 of the human, the temperature values of the human area and the stationary object area corresponding to the areas A ′ and B ′ are Are added (represented as “A + H” and “B + H” in FIG. 10), and for the overlapping region between the two thermal images ID1 and ID2 of the human, two images corresponding to the region The image indicates a value obtained by adding the temperature values of the thermal image information ID1 and ID2 (expressed as “A + B” in FIG. 10).
[0104]
In the next step 210, from the combined image information derived in step 208, a pixel indicating a value within the range from the lower threshold value THL to the upper threshold value THH derived in step 204 ′ is extracted, thereby performing the synthesis. The above-mentioned overlapping region of the human in the image information is extracted. As a result, in the example shown in FIG. 10, only the overlapping area (the area of 'A + B') is extracted.
[0105]
In the next step 212, a human region is extracted by combining the non-overlapping region extracted as a pixel value having a positive value in the above-mentioned step 202 'with the overlapping region extracted in the above-mentioned step 210. Thus, in the example shown in FIG. 10, only the human area B is obtained.
[0106]
As described above, in the human area detection process according to the present embodiment, similarly to the first embodiment, even when the human is in a state of sitting on a chair, for example, the face, arm, shoulder, etc. Paying attention to the fact that the part is moving, the area where the person is moving (non-overlapping area) is extracted by deriving the difference image information, and the human area is extracted based on the thermal image information corresponding to the extracted area. Are derived (lower threshold THL and upper threshold THH), and the threshold is based on the actual human temperature according to the environmental state such as the temperature and the season in the processing target space. Therefore, even if a heating element having a temperature substantially equal to a human temperature exists in the processing target space, if the temperature of the heating element is slightly different from the human temperature, the heating element is extracted from the extraction target. Can be excluded in the human part It can be extracted frequency only with high accuracy.
[0107]
When the process of step 212 is completed, the human region detection process is completed, and the process proceeds to step 112 of the environment control processing program (see FIG. 4).
[0108]
The processing of step 200 in the human area detection processing program according to the present embodiment is performed by the difference image deriving means of the invention described in claim 2, and the processing of step 202 'is performed by the non-overlapping area extraction means of the invention described in claim 2. The processing of step 204 'is performed by the threshold deriving means of the invention described in claim 2, the processing of step 208 is performed by the synthesized image deriving means of the invention described in claim 2, and the processing of step 210 is performed by the overlapping area of the invention described in claim 2. The processing of step 212 corresponds to the human area extracting means of the present invention.
[0109]
FIGS. 11 and 12 show an example of an image represented by various kinds of image information actually obtained during the execution of the environmental control process by the environmental control system 10 according to the second embodiment. In FIG. 11, for convenience, the temperature difference in the thermal image is represented by about five levels of line density, but in actuality, for example, the region of 50 ° C. is set to red, and the temperature is increased from 50 ° C. to 10 ° C. Each thermal image is given a color indicating the temperature, for example, the color gradually changes to a cool color in multiple stages for each predetermined temperature range as the temperature decreases.
[0110]
FIG. 11A shows a thermal image indicated by the thermal image information obtained in step 104 of the environmental control process (FIG. 4), and FIG. 11B shows a thermal image obtained in step 108 of the environmental control process. Each of the thermal images indicated by the obtained thermal image information is shown. As shown in the figure, here, a case is shown in which a large number of people are present in a space to be processed in a seated state, and a PC and a printer are fixedly installed as heating elements.
[0111]
In this case, since the PC and the printer are stationary, they are almost in the same state as in FIGS. 11A and 11B. However, since there is a slight movement for humans, FIG. And FIG. 11B has a slightly different temperature distribution.
[0112]
On the other hand, FIG. 11C shows a difference image indicated by the difference image information obtained by the processing of step 200 of the human area detection processing (FIG. 9), and FIG. It shows an image of a non-overlapping area extracted by the processing of step 202 ′ of the detection processing. Note that FIG. 12A shows a binarized image in which a non-overlapping region is extracted by binarizing the image shown in FIG. 11C.
[0113]
In the images shown in FIGS. 11 (C) and 12 (A), almost all images in the stationary PC and printer area and the overlapping area of the human are deleted, and almost all the images in the non-overlapping area of the human are removed. Remaining. Therefore, it can be seen that the human region can be extracted with high accuracy by extracting the human region with the threshold value obtained based on the difference image information that becomes the difference image as shown in FIG.
[0114]
On the other hand, FIG. 12B shows an image of the overlapping area extracted by the processing of step 210 of the human area detection processing, and FIG. 12C shows the image of the overlapping area extracted by the processing of step 212 of the human area detection processing. 2 shows an image of a human region. FIGS. 12B and 12C show binarized images in which only the extracted overlapping region or human region is set to the black level.
[0115]
In the image shown in FIG. 12B, only the overlapping region of the human can be accurately extracted. As a result, as shown in FIG. 12C, only the human region can be extracted with high accuracy.
[0116]
As described above in detail, in the environment control system according to the present embodiment, based on the temperature indicated by the thermal image information corresponding to the non-overlapping area, the overlapping area between the two thermal image information of the human is extracted. And a composite image information for each of the divided regions (pixels in the present embodiment) in the two thermal image information is derived, and the composite image information and the threshold value are calculated. Since the overlap region between the two thermal image information of the human is extracted by comparing the divided regions with the above, and the region corresponding to the human is extracted by combining the non-overlap region and the overlap region. It is possible to extract a region corresponding to a human with higher accuracy than when simply extracting from thermal image information.
[0117]
Further, in the environment control system according to the present embodiment, using the parameters obtained based on the region corresponding to the human extracted as described above and the thermal image information used when extracting the region, A comfort index indicating the comfort in the space to be processed is derived, and the environment in the space is controlled based on the comfort index. Control of thermal environment can be performed.
[0118]
In the present embodiment, the case where the lower threshold THL and the upper threshold THH are calculated using the equations (5) and (6) has been described. However, the present invention is not limited to this. It is also possible to adopt a form in which calculation is performed using the following equations (7) and (8). Further, it is also possible to calculate by replacing Bt in Expressions (7) and (8) with At.
[0119]
THL = 2 × Bt−TH2 (7)
THH = 2 × Bt + TH2 (8)
In these cases, in order to calculate the threshold value, only one of the area A ′ and the area B ′ shown in FIG. 10 may be extracted as an example. The load can be reduced.
[0120]
Further, in the present embodiment, the case has been described where the combined image information in step 208 of the human region detection process is derived by adding two thermal image information for each pixel, but the present invention is not limited to this. For example, it is also possible to adopt a mode in which the information is derived by multiplying two thermal image information for each pixel. In this case, the difference between the information on the overlapping region of the human region and the information on the overlapping region of the heating element in the generated composite image information can be increased as compared with the present embodiment. As compared with the form, the overlapping area of the human area can be extracted with higher accuracy.
[0121]
[Third embodiment]
In the first and second embodiments, the case where the human region is extracted by using the threshold derived based on the difference image information has been described. However, in the third embodiment, the human image is extracted from the thermal image information. A form in which a candidate area is extracted and an area adjacent to a non-overlapping area extracted from the area based on the difference image information is extracted as a human area, that is, as a human area extraction method. An embodiment in which the inventions of claims 3 and 7 are applied will be described. The configuration of the environment control system 10 according to the third embodiment is the same as that of the environment control systems 10 according to the first and second embodiments, and a description thereof will not be repeated.
[0122]
Hereinafter, the operation of the environmental control system 10 according to the third embodiment when executing the environmental control processing will be described. The environment control process according to the third embodiment differs from the environment control process according to the first embodiment (see FIG. 4) only in the human area detection process executed in step 110. However, only the human region detection process according to the third embodiment will be described with reference to FIGS. FIG. 13 is a flowchart showing the flow of a human area detection processing program executed by the CPU 22 of the PC 20A when performing the human area detection processing according to the third embodiment. It is stored in the area in advance. Steps in FIG. 13 that perform the same processing as in FIG. 6 are assigned the same step numbers as in FIG. 6, and descriptions thereof are omitted.
[0123]
In step 214 in FIG. 13, a predetermined lower limit threshold THL (here, in FIG. 14) for extracting a region corresponding to human temperature from thermal image information collected later (the thermal image ID2 in the example shown in FIG. 14). An area constituted by pixels indicating values within a range from 30 ° C.) to an upper threshold THH (here, 37 ° C.) is extracted as a human candidate area. Thereby, in the example shown in FIG. 14, the heating element region W having a temperature substantially similar to the human region B and the human temperature region is extracted as the human candidate region. Note that the thermal image ID1, the thermal image ID2, the difference image SD, and the non-overlapping area B 'in FIG. 14 are the same as those in FIG.
[0124]
In the next step 216, of the human candidate areas extracted in step 214, only those areas located within a predetermined distance from the non-overlapping area extracted by the processing in step 202 are extracted as human areas. The area located within a predetermined distance from the non-overlap area includes an area located in the non-overlap area (an area overlapping with the non-overlap area) and an area adjacent to the non-overlap area. Here, the distance between the closest pixels between the non-overlapping area and the human candidate area is applied as the distance corresponding to the predetermined distance.
[0125]
As a result, in the example shown in FIG. 14, as the area located within a predetermined distance from the non-overlapping area B ′, the human area B including the area overlapping with the non-overlapping area B ′ and the area adjacent to the non-overlapping area B ′ Only those are accurately extracted.
[0126]
As described above, even in the human area detection processing according to the present embodiment, a part of a face, an arm, a shoulder, and the like moves even when a human is sitting in a chair, for example. By extracting difference image information, a human area (non-overlapping area) is reliably extracted, and a human candidate area close to the extracted area is extracted as a human area. Even if there is a heating element having a temperature substantially equal to the temperature, the heating element can be excluded from the extraction target, so that only the human part region can be extracted with high accuracy.
[0127]
When the process of step 216 is completed, the human region detection process is completed, and the process proceeds to step 112 of the environment control processing program (see FIG. 4).
[0128]
The processing of step 200 in the human area detection processing program according to the present embodiment is performed by the difference image deriving means of the invention described in claim 3, and the processing of step 202 is performed by the non-overlapping area extraction means of the invention described in claim 3. The processing in step 214 corresponds to the candidate area extracting means of the present invention, and the processing in step 216 corresponds to the human area extracting means in the present invention.
[0129]
As described above in detail, in the environment control system according to the present embodiment, the space to be processed is regarded as a two-dimensional plane, and a predetermined divided area of the two-dimensional plane (in the present embodiment, The difference image information for each of the divided areas is derived by subtraction for each of the divided areas in the two different thermal image information temporally collected by the infrared camera 12 that collects the thermal image information indicating the temperature for each pixel). A region formed by a divided region having a value outside a predetermined range including zero in the difference image information is extracted as a non-overlapping region between the two thermal image information pieces of a human, and either of the two thermal image information pieces, Extract a region that is a candidate for a human region in the thermal image information by comparing each of the divided regions with a predetermined threshold to extract a region corresponding to human temperature, Of out area, since extracted as a region corresponding to human only area located within a predetermined distance from the non-overlapping regions, it is possible to extract a region of a human part with high accuracy from the thermal image.
[0130]
Further, in the environment control system according to the present embodiment, using the parameters obtained based on the region corresponding to the human extracted as described above and the thermal image information used when extracting the region, A comfort index indicating the comfort in the space to be processed is derived, and the environment in the space is controlled based on the comfort index. Control of thermal environment can be performed.
[0131]
In the present embodiment, a case has been described in which a non-overlapping region is extracted from thermal image information collected later (thermal image ID2 in the example shown in FIG. 14), but the present invention is not limited to this. Alternatively, a non-overlapping area (non-overlapping area A ′ in the example shown in FIG. 14) may be extracted from the previously collected thermal image information (the thermal image ID1 in the example shown in FIG. 14). The non-overlapping area (non-overlapping areas A ′ and B ′ in the example shown in FIG. 14) may be extracted from the thermal image information. In these cases, the same effects as in the present embodiment can be obtained.
[0132]
Further, in the present embodiment, the case where the human candidate area is extracted from the thermal image information collected later has been described. However, the present invention is not limited to this, and the thermal image information collected earlier may be used. A form in which a human candidate region is extracted can also be used. In this case, the same effect as in the present embodiment can be obtained.
[0133]
Furthermore, in the present embodiment, a case has been described where a predetermined threshold is applied as a threshold used for extracting a human candidate region. However, the present invention is not limited to this. The threshold value derived in step 204 of the human region detection process (FIG. 6) according to the embodiment or the threshold value derived in step 204 ′ of the human region detection process (FIG. 9) according to the second embodiment is applied. You can also. In this case, since these thresholds are based on the actual human temperature according to the environmental conditions such as the temperature and the season in the processing target space, the heating element having a temperature substantially equal to the human temperature in the processing target space. Even if there is, if the temperature of the heating element is slightly different from the temperature of a human, the heating element can be excluded from the extraction target, and the human candidate region can be more accurately extracted. it can.
[0134]
[Fourth embodiment]
In the third embodiment, the case where the human candidate area is directly extracted from the collected thermal image information has been described. However, in the fourth embodiment, the human candidate area is generated based on two thermal image information having different collection times. The present invention according to claims 4 and 8, wherein a form in which an overlap region between the two thermal image information of a human is extracted as a human candidate region using the combined image information, that is, a human region extraction method. An embodiment in which is applied will be described. The configuration of the environment control system 10 according to the fourth embodiment is the same as that of the environment control system 10 according to the first embodiment, and a description thereof will not be repeated.
[0135]
Hereinafter, the operation of the environmental control system 10 according to the fourth embodiment when executing the environmental control processing will be described. The environment control process according to the fourth embodiment is different from the environment control process (see FIG. 4) according to the first embodiment only in the human area detection process executed in step 110. Now, only the human region detection processing according to the fourth embodiment will be described with reference to FIGS. Here, FIG. 15 is a flowchart showing the flow of processing of a human area detection processing program executed by the CPU 22 of the PC 20A when performing the human area detection processing according to the fourth embodiment. It is stored in a predetermined area in advance. Steps in FIG. 15 that perform the same processing as in FIG. 6 are assigned the same step numbers as in FIG. 6, and descriptions thereof are omitted.
[0136]
In step 218 of FIG. 15, the composite image information is derived in the same manner as in step 208 of the human area detection process (FIG. 9) according to the second embodiment. Thus, in the example shown in FIG. 16, it is possible to obtain composite image information indicating the composite image GD. Note that the thermal image ID1, the thermal image ID2, the difference image SD, and the composite image GD in FIG.
[0137]
In the next step 220, a lower limit threshold value THL (here, a predetermined lower threshold value THL) for extracting an overlapping area between the synthetic image information derived in the above step 218 and two thermal image information collected at different timings of a human. Then, an area composed of pixels indicating values within a range from 60 ° C.) to an upper threshold THH (here, 74 ° C.) is extracted as a human candidate area. Thus, in the example shown in FIG. 16, the overlapping region of the human and the region W of the heating element having a temperature substantially similar to the temperature region of the human are extracted as the human candidate region.
[0138]
In the next step 222, of the human candidate regions extracted in the above step 220, only the regions located within a predetermined distance from the non-overlapping region extracted in the process of the above step 202 are extracted as human regions. Note that, as in the third embodiment, an area located within a predetermined distance from the non-overlapping area is in contact with an area located within the non-overlapping area (an area overlapping with the non-overlapping area) and a non-overlapping area. Region is included. Also here, the distance between the closest pixels between the non-overlapping area and the human candidate area is applied as the distance corresponding to the predetermined distance. As a result, in the example shown in FIG. 16, only the human overlap region is extracted as a region located within a predetermined distance from the non-overlap region B ′.
[0139]
Then, in step 222, a human region is extracted by combining the extracted region (human overlapping region) with the non-overlapping region extracted in step 202. Thus, in the example shown in FIG. 16, only the human region B is extracted with high accuracy.
[0140]
The processing of step 200 in the human area detection processing program according to the present embodiment is performed by the difference image deriving means of the invention described in claim 4, and the processing of step 202 is performed by the non-overlapping area extraction means of the invention described in claim 4. The processing of 218 is performed by the synthesized image deriving means of the invention described in claim 4, the processing of step 220 is performed by the candidate area extraction means of the invention described in claim 4, and the processing of step 222 is performed by the human area extraction processing of the invention described in claim 4. Each corresponds to a means.
[0141]
As described in detail above, in the environment control system according to the present embodiment, the composite image information for each of the divided regions is derived by adding each of the divided regions (pixels in the present embodiment) in the two thermal image information. Then, the combined image information, and a region that is a candidate for a human region is extracted by comparing each of the divided regions with a predetermined threshold for extracting an overlapping region between the two thermal image information of a human, Since an area corresponding to a human is extracted by combining an area located within a predetermined distance from the non-overlapping area of the extracted area and the non-overlapping area, a human area is simply extracted from the thermal image information. An area corresponding to a human can be extracted with higher accuracy than when a candidate area is extracted.
[0142]
Further, in the environment control system according to the present embodiment, using the parameters obtained based on the region corresponding to the human extracted as described above and the thermal image information used when extracting the region, A comfort index indicating the comfort in the space to be processed is derived, and the environment in the space is controlled based on the comfort index. Control of thermal environment can be performed.
[0143]
In the present embodiment, a case has been described in which a non-overlapping region is extracted from thermal image information collected later (thermal image ID2 in the example shown in FIG. 16), but the present invention is not limited to this. Instead, a non-overlapping area (non-overlapping area A 'in the example shown in FIG. 16) may be extracted from the previously collected thermal image information (the thermal image ID1 in the example shown in FIG. 16). A non-overlapping area (non-overlapping areas A ′ and B ′ in the example shown in FIG. 16) may be extracted from the thermal image information. In these cases, the same effects as in the present embodiment can be obtained.
[0144]
Furthermore, in the present embodiment, a case has been described where a predetermined threshold is applied as a threshold used for extracting a human candidate region. However, the present invention is not limited to this. The threshold value derived in step 204 of the human region detection process (FIG. 6) according to the embodiment or the threshold value derived in step 204 ′ of the human region detection process (FIG. 9) according to the second embodiment is applied. You can also. In this case, since these thresholds are based on the actual human temperature according to the environmental conditions such as the temperature and the season in the processing target space, the heating element having a temperature substantially equal to the human temperature in the processing target space. Even if there is, if the temperature of the heating element is slightly different from the temperature of a human, the heating element can be excluded from the extraction target, and the human candidate region can be more accurately extracted. it can.
[0145]
Further, in each of the above embodiments, a case has been described in which the comfortable region has a PMV value in the range from +0.5 to -0.5. However, the present invention is not limited to this. , The range from +0.6 to -0.6, the range from +0.4 to -0.4, and the like, the value indicating the comfort zone is appropriately changed according to the conditions of the space to be environmentally controlled. It goes without saying that you can do it.
[0146]
In each of the above embodiments, in the air conditioning control process (see step 122 in FIG. 4), all three parameters of the temperature of the air blown from the air conditioner 44, the humidity of the air, and the air volume of the air are adjusted. However, the present invention is not limited to this, and one or two of these three parameters may be adjusted. In this case, since the number of adjustment targets is reduced as compared with the above embodiments, the time required for the comfort to change by the adjustment becomes longer, but the control for the adjustment is simplified. be able to.
[0147]
Further, in each of the above embodiments, the case where the environmental control process is executed by the PC 20A capable of collecting thermal image information via the recording device 16 has been described. However, the present invention is not limited to this. It is needless to say that the environment control process can be executed by the PC 20B that can collect thermal image information without going through the recording device 16. In this case, the same effects as in the above embodiments can be obtained.
[0148]
Further, in each of the above embodiments, the case where the information processing apparatus and the information processing method of the present invention are applied to an environment control apparatus that controls a thermal environment has been described. However, the present invention is not limited to this. As shown in FIG. 17, the present invention can be applied to a disaster prevention / facility maintenance / management system using the disaster prevention / facility maintenance / management device 50.
[0149]
In this case, when a human area is detected by the human area detection processing, the PC 20A according to each of the above-described embodiments outputs a signal indicating that a human exists in the processing target space and a signal indicating the human's existing position in the processing target space. -Output to the facility maintenance management device 50. In response, the disaster prevention / facility maintenance / management device 50 assumes that when this time is a time zone when no humans exist (midnight, holidays, etc.), the suspicious individual may have entered the processing target space. Judgment is made and actions such as confirming the inside of the space, issuing an alarm, and notifying the security company are performed.
[0150]
Note that such a system usually collects raw images in a processing target space using a video camera. In this case, since individuals existing in the space are specified, privacy is not considered. There was a problem. However, in a system to which the information processing apparatus according to the present invention is applied, since a system can be constructed without using a raw image, such a privacy problem does not occur.
[0151]
Furthermore, in each of the above embodiments, the case where the PMV value is applied as the comfort index of the present invention has been described. However, the present invention is not limited to this. For example, the present invention was developed by Gagg et al. The air conditioner is adopted by the Japan Society for Air Conditioning), and it is also possible to adopt a mode in which ET * (ET star: new effective temperature), which is a thermal index based on a thermal equilibrium equation including perspiration, is applied. In this case, the same effects as those of the above embodiments can be obtained.
[0152]
【The invention's effect】
According to the information processing apparatus according to the first aspect and the information processing method according to the fifth aspect, a space to be processed is regarded as a two-dimensional plane, and the temperature of each predetermined divided region of the two-dimensional plane is determined. The difference image information for each of the divided regions is derived by subtraction for each of the divided regions in the two temporally different thermal image information collected by the collection unit that collects the thermal image information indicating, and zero in the difference image information is calculated. The region constituted by the divided regions having values outside the predetermined range is extracted as a non-overlapping region between the two thermal image information of a human, and the temperature indicated by the thermal image information corresponding to the non-overlapping region is extracted. A threshold for extracting an area having the same temperature as the non-overlapping area is derived based on the non-overlapping area, and a comparison is made between each of the two thermal image information and the threshold for each of the divided areas. Since the extracted region corresponding to the human in the thermal image information, an area of the human portion from the thermal image can be extracted with high accuracy, the effect is obtained that.
[0153]
According to the information processing apparatus described in claim 2 and the information processing method described in claim 6, based on the temperature indicated by the thermal image information corresponding to the non-overlapping area, the two thermal image information of a human being Deriving an overlap region extraction threshold for extracting an overlap region between the two thermal image information, and deriving composite image information for each of the divided regions by addition or multiplication for each of the divided regions in the two thermal image information, A region corresponding to a human by extracting the overlapping region between the two thermal image information pieces of the human by comparing each of the divided regions with the overlapping region extraction threshold value and the overlapping region extraction threshold, and combining the non-overlapping region and the overlapping region. Is extracted, so that an effect that a region corresponding to a human can be extracted with higher accuracy than in the case of simply extracting from thermal image information can be obtained.
[0154]
Further, according to the information processing apparatus of the third aspect and the information processing method of the seventh aspect, a space to be processed is regarded as a two-dimensional plane, and a predetermined divided area of the two-dimensional plane is determined. The difference image information for each of the divided regions is derived by subtraction for each of the divided regions in the two temporally different thermal image information collected by the collection unit that collects the thermal image information indicating the temperature of the difference image information. A region constituted by divided regions having values outside a predetermined range including zero is extracted as a non-overlapping region between the two thermal image information of a human, and any one of the two thermal image information and the human temperature By comparing each of the divided regions with a predetermined threshold value to extract a corresponding region, a candidate region of a human region in the thermal image information is extracted, and among the extracted regions, Since by extracting only a region located from the overlapping area within the predetermined distance as a region corresponding to a human can be extracted from the thermal image region of a human part with high accuracy, the effect is obtained that.
[0155]
Further, according to the information processing apparatus according to the fourth aspect and the information processing method according to the eighth aspect, synthetic image information for each of the divided areas is derived by adding or multiplying the two thermal image information for each of the divided areas. Then, the composite image information, and a region that is a candidate of a human region is extracted by comparing each of the divided regions with a predetermined threshold for extracting an overlapping region between the two thermal image information of a human, Since an area corresponding to a human is extracted by combining an area located within a predetermined distance from the non-overlapping area of the extracted area with the non-overlapping area, a human area is simply extracted from the thermal image information. As compared with the case where a candidate region is extracted, an effect that a region corresponding to a human can be extracted with higher accuracy can be obtained.
[0156]
Furthermore, according to the environment control device of the ninth aspect, the region corresponding to the person extracted by the human region extraction means of the information processing device according to the present invention, and the thermal image information used when extracting the region Since a comfort index indicating the comfort in the space to be processed is derived using the parameters obtained based on the and based on the comfort index, the environment in the space is controlled based on the comfort index. In this case, it is possible to appropriately control the thermal environment according to the actual condition in the space.
[Brief description of the drawings]
FIG. 1 is a schematic diagram for explaining a method of extracting a non-overlapping region of a human part according to an embodiment.
FIG. 2 is a schematic block diagram illustrating an overall configuration of an environment control system 10 according to the embodiment.
FIG. 3 is a block diagram showing a configuration of an electric system of a PC 20A according to the embodiment.
FIG. 4 is a flowchart showing a processing flow of an environment control processing program according to the embodiment.
FIG. 5 is a schematic diagram showing an example of a fixed parameter setting screen.
FIG. 6 is a flowchart illustrating a flow of processing of a human area detection processing program according to the first embodiment.
FIG. 7 is an explanatory diagram for explaining a processing procedure of a human area detection process according to the first embodiment;
FIG. 8 is a graph showing a relationship between a PMV value and a dissatisfied person ratio and a comfortable area, for explaining an environmental control process according to the embodiment.
FIG. 9 is a flowchart illustrating a flow of processing of a human area detection processing program according to the second embodiment.
FIG. 10 is an explanatory diagram provided for describing a processing procedure of a human area detection process according to the second embodiment.
FIG. 11 is a diagram showing an example of a thermal image indicated by thermal image information actually obtained during the execution of the environmental control process by the environmental control system 10 according to the second embodiment.
FIG. 12 is a diagram illustrating an example of a binarized image represented by image information actually obtained during execution of an environmental control process by an environmental control system 10 according to a second embodiment.
FIG. 13 is a flowchart illustrating a flow of processing of a human area detection processing program according to a third embodiment.
FIG. 14 is an explanatory diagram for explaining a processing procedure of a human area detection process according to the third embodiment;
FIG. 15 is a flowchart illustrating a flow of processing of a human area detection processing program according to a fourth embodiment.
FIG. 16 is an explanatory diagram provided for describing a processing procedure of a human area detection process according to the fourth embodiment.
FIG. 17 is a block diagram showing another application example of the information processing apparatus and the information processing method of the present invention.
FIG. 18 is a schematic diagram for explaining a conventional technique.
[Explanation of symbols]
10 Environment control system
12 Infrared camera (collection means)
18A, 18B capture board
20A, 20B Personal computer
22 CPU

Claims (9)

Collection means for capturing the interior of the space to be processed as a two-dimensional plane and collecting thermal image information indicating the temperature of each predetermined divided region of the two-dimensional plane;
Difference image deriving means for deriving difference image information for each of the divided regions by subtraction for each of the divided regions in the two different thermal image information temporally collected by the collecting unit,
A non-overlapping region extracting means for extracting a region constituted by divided regions having values outside a predetermined range including zero in the difference image information as a non-overlapping region between the two thermal image information pieces of a human;
Based on the temperature indicated by the thermal image information corresponding to the non-overlapping region, a threshold deriving unit that derives a threshold for extracting a region indicating the same temperature as the non-overlapping region,
A human region extracting unit that extracts a region corresponding to the human in the thermal image information by comparing any one of the two thermal image information with the threshold for each of the divided regions,
Information processing device provided with. The threshold deriving unit sets the overlap region extraction threshold for extracting an overlap region between the two thermal image information of the human based on the temperature indicated by the thermal image information corresponding to the non-overlap region. Derived in place of
Composite image deriving means for deriving composite image information for each of the divided regions by addition or multiplication for each of the divided regions in the two thermal image information,
Overlapping area extraction means for extracting an overlapping area between the two thermal image information pieces of the human by comparing the composite image information and the overlapping area extraction threshold for each of the divided areas,
Further comprising
The human region extracting unit may extract a region corresponding to the person by combining the non-overlapping region extracted by the non-overlapping region extracting unit and the overlapping region extracted by the overlapping region extracting unit. Item 10. The information processing device according to Item 1. Collection means for capturing the interior of the space to be processed as a two-dimensional plane and collecting thermal image information indicating the temperature of each predetermined divided region of the two-dimensional plane;
Difference image deriving means for deriving difference image information for each of the divided regions by subtraction for each of the divided regions in the two different thermal image information temporally collected by the collecting unit,
A non-overlapping region extracting means for extracting a region constituted by divided regions having values outside a predetermined range including zero in the difference image information as a non-overlapping region between the two thermal image information pieces of a human;
A candidate region of a human region in the thermal image information is extracted by comparing each of the two thermal image information with a predetermined threshold value for extracting a region corresponding to a human temperature for each of the divided regions. Candidate area extracting means for
A human region extracting unit for extracting only a region located within a predetermined distance from the non-overlapping region extracted by the non-overlapping region extracting unit as a region corresponding to the human, of the regions extracted by the candidate region extracting unit; When,
Information processing device provided with. Further comprising a composite image deriving means for deriving composite image information for each of the divided regions by addition or multiplication for each of the divided regions in the two thermal image information,
The candidate area extracting means compares the synthesized image information with a threshold value for extracting an overlapping area between the two thermal image information of the human for each of the divided areas, and the candidate of the human area is obtained. Extract the area that becomes
The human region extracting unit is configured to include, among the regions extracted by the candidate region extracting unit, a region located within a predetermined distance from the non-overlapping region extracted by the non-overlapping region extracting unit, and the non-overlapping region. The information processing apparatus according to claim 3, wherein an area corresponding to the human is extracted by combining. Capture the space to be processed as a two-dimensional plane, and collect thermal image information indicating the temperature of each predetermined divided region of the two-dimensional plane at predetermined time intervals,
Deriving difference image information for each of the divided regions by subtraction for each of the divided regions in the collected two thermal image information different in time,
Extract the region constituted by the divided regions having a value outside the predetermined range including zero in the difference image information as a non-overlapping region between the two thermal image information of a human,
Based on the temperature indicated by the thermal image information corresponding to the non-overlapping region, to derive a threshold for extracting a region showing the same temperature as the non-overlapping region,
An information processing method for extracting a region corresponding to the human in the thermal image information by comparing one of the two thermal image information with the threshold value for each of the divided regions. Based on the temperature indicated by the thermal image information corresponding to the non-overlapping region, an overlapping region extraction threshold for extracting an overlapping region between the two thermal image information of the human is derived instead of the threshold. Deriving composite image information for each of the divided regions by addition or multiplication for each of the divided regions in the two thermal image information,
Extracting the overlapping area between the two thermal image information of the human by comparing the composite image information and the overlapping area extraction threshold for each of the divided areas,
6. The information processing method according to claim 5, wherein an area corresponding to the human is extracted by combining the non-overlapping area and the overlapping area. Capture the space to be processed as a two-dimensional plane, and collect thermal image information indicating the temperature of each predetermined divided region of the two-dimensional plane at predetermined time intervals,
Deriving difference image information for each of the divided regions by subtraction for each of the divided regions in the collected two thermal image information different in time,
Extract the region constituted by the divided regions having a value outside the predetermined range including zero in the difference image information as a non-overlapping region between the two thermal image information of a human,
A candidate region of a human region in the thermal image information is extracted by comparing each of the two thermal image information with a predetermined threshold value for extracting a region corresponding to a human temperature for each of the divided regions. And
An information processing method for extracting only a region located within a predetermined distance from the non-overlapping region as a region corresponding to the person among the extracted regions. Deriving composite image information for each of the divided regions by addition or multiplication for each of the divided regions in the two thermal image information,
The synthetic image information and a region that is a candidate for the human region is extracted by comparing each of the divided regions with a predetermined threshold to extract an overlapping region between the two thermal image information of the human,
8. The information processing method according to claim 7, wherein an area corresponding to the human is extracted by combining an area located within a predetermined distance from the non-overlapping area with the non-overlapping area among the extracted areas. The information processing apparatus according to any one of claims 1 to 4, an area corresponding to the person extracted by the human area extraction unit of the information processing apparatus, and an area used for extracting the area. Comfort index deriving means for deriving a comfort index indicating comfort in the space to be processed using a parameter obtained based on the thermal image information,
Environment control means for controlling an environment in the space based on the comfort index,
Environment control device equipped with.

Images