JP3645404B2 - Method and apparatus for recognizing door from architectural drawing - Google Patents

Description

【００４５】
【表２】

【００４６】
「輪郭」とは「外輪郭」のことであり、表１に○印で示すように外周壁の外部に面している箇所のみ（図４の（ａ）に示す二重線の外側の線の部分）を意味するケース１の場合と、外部に接する壁の全体（図４の（ｂ）に太線で示す部分）を意味するケース２，３の場合とがある。
【００４７】
「骨格」とは、壁の総て（図４の（ｂ）（ｃ）に太線で示す部分の両方）を意味するケース１，２の場合と、外輪郭を除く壁（図４の（ｃ）に太線で示す家屋内の仕切り部分のみ）を意味するケース３の場合とがある。これらの定義において、特に断わらない場合はケース１の通常の意味として扱われる。
【００４８】
「外壁」とは表２に○印で示すように、外輪郭と同じく外周壁の外部に面している箇所のみ（図４の（ａ）に示す二重線の外側の線で示す部分）を意味するケース４の場合と、外部に接する壁の全体（図４の（ｂ）に太線で示す部分）を意味するケース５の場合とがある。
【００４９】
「内壁」とは表２に○印で示すように、ケース４，５とも外壁を除く壁（壁＝外壁＋内壁）であるが、ケース４の場合は図４の（ａ）に示す二重線の内側の線の部分と図４の（ｃ）に太線で示す部分であり、ケース５の場合は図４の（ｃ）に太線で示す部分である。これらの定義においても、特に断わらない場合はケース５の通常の意味として扱われる。
【００５０】
次に、図１に示した建築図面認識装置による建築図面（主に家屋やビル等の建築図面）認識の手順について、図５乃至図８のフロー図と、図２５乃至図３０によって説明する。上記図６〜図８において、各ステップを「Ｓ」で示している。また、この実施例では、認識する建築図面がポジ図面であるものとする。
【００５１】
図５は、建築図面の認識処理のメインルーチンを示すフローチャートである。図５の処理を開始すると、図１の画像読取部２にセットされた建築図面を読み取り、そのイメージ画像データを入力して、自動スキュー補正部５によって自動的にそのスキューを補正する。
【００５２】
そして、ステップＡのサブルーチンでドット数計測配列データ作成部６及び輪郭・骨格認識部７等によりその入力した画像データに対して壁の認識処理を実行する。この処理については後に詳述する。なお、イメージ画像データを符号化したデータを通信制御部３から入力するようにしてもよい。
【００５３】
その後、ドア認識部１３により、ステップＢとＣの処理で、水平壁から垂直壁へと順に個々の壁位置情報抽出を行なって、ステップＤ〜ステップＫの処理へ移行する。そして、ステップＢで壁の位置情報の抽出を終了すると、当該図面の解析を終了してステップＬに移行し、ドア認識結果を出力して処理を終了する。
ステップＣにおいても、それぞれの種類の壁（水平壁と垂直壁）を全てチェックし終わった（壁数エンド）後に、ステップＢへ戻る。
【００５４】
ステップＤでは、壁の位置情報である壁座標を基準にして認識範囲となる水平・垂直の両方向に単位長（９１ｃｍ＝半間又は１００ｃｍ＝１メートルに相当する縮尺寸法）の矩形エリア（領域）を壁の両側に設定する。
ステップＥでは、その対象エリア（指定された矩形エリア）の壁座標を基準とした連続する黒ドット数計測配列データを作成する。ここでの連続する黒ドット数とは、壁を基準とした黒ドットの連続性であるが、カスレなどを考慮しある程度のドットぬけを包括する条件付けをする必要がある。
【００５５】
そして、ステップＦでは、その黒ドット数計測配列データを順にチェックし、ドア半径候補としての許容範囲（例えば、単位長の１／４以上）に入っているときのみ、ステップＧに移行する。また、上記配列データの終了によって、次の壁データの抽出を行なうステップＣへ移行する。
【００５６】
ステップＧでは、ドア（開き戸，開き窓）の表現方法のチェックを行なって、形状抽出処理を行なう。そのドア半径候補（戸板部）の先端には、９０°扇型の円弧部分の立ち上がりが存在する。それを計測して半径と仮定し、対応するもう一方の円弧部分の立ち上がりを解析して、この９０°扇型の中心点／半径／ドア方向によるドア形状データを抽出する。
【００５７】
そして、ステップＨでは、ドア形状抽出ができたか否かを判断し、できた場合はステップＩへ、できなかった場合次のドア半径候補抽出のためにステップＦへ戻る。
ステップＩでは、ドア形状抽出データに基づいて、ドア認識候補の形状解析処理を行なう。つまり、ステップＧよりの中心点／半径／ドア方向のドア形状データによって、扇型の円弧部分のドット検索による形状解析を行なう。
【００５８】
全てのドアの表現方法を、この建築図面認識方法及び認識装置によって、完全に自動的に認識できるわけではなく、当認識装置の使用者次のいずれかを予め指定するものである。
【００５９】
ドアの表現方法には、主に以下の三つがある。
▲１▼形状輪郭で片方向に９０°扇型のもの
：図２８の（ａ）〜（ｄ）に示すような平面表示のもの
▲２▼形状輪郭で両方向に９０°扇型があるもの
：図２９の（ａ)，(ｂ）に示すような平面表示のもの
▲３▼形状輪郭で壁の中に戸板部分が包括されているもの
：図３０の（ａ)，(ｂ）に示すような平面表示のもの
【００６０】
そして、ステップＪでは、ドアの形状解析の結果がこれらのいずれかに該当する（ＯＫ）か否かを判断し、ＯＫであればステップＫへ、ＯＫでなければ（ＮＧであれば）次のドア半径候補抽出のためステップＦへ戻る。
ステップＫでは、上記、中心点／半径／ドア方向のドア形状データを「ドア」情報として認識し、その認識結果の情報を格納した後、ステップＦへ戻る。
【００６１】
以上のステップＤ〜ステップＫまでの処理を行なって認識した結果を、ステップＬによって出力して終了する。ステップＬによる処理では、図２７に示した画像データのドアの位置及びサイズ（図中に三角形で示している）を建築図面からのドアの認識結果としてメモリ４又は外部記憶装置１１に記憶、あるいは表示部９に表示する等の出力をする。
【００６２】
図２５は、壁の芯線を認識し、その結果から壁情報を抽出する処理に使用した入力サンプル元図面(イメージ画像)を、表示部９で再マッピングさせた画像データの表示例である。
図２６は、ドア認識の基準となる壁情報を表示した例を示す図である。
図２７は、その認識結果のドアを三角形で表示した例を示す図である。
【００６３】
ここで、上述したドア認識処理について、図３１から図３３を用いてさらに具体的に説明する。
例えば、図３１に示すような水平壁Ｗｈの位置情報（Ｗｖは垂直壁の一部）が抽出された場合、この水平壁に沿ってその両側に、水平・垂直の両方向に単位長（ｘｕ＝ｙｕ＝９１ｃｍ又は１００ｃｍに相当する縮尺寸法）の矩形エリアを、その全長に亘って連続して設定する。
【００６４】
そして、その各矩形エリア毎に、後述する壁認識処理の場合と同様にして、水平壁Ｗｈに沿う１ドット幅毎の垂直方向に連続する黒ドット数の計測配列データ（射影のヒストグラム）である垂直方向黒ドット数計測配列データを作成する。
その黒ドット数計測配列データのピーク値が単位長以下で単位長の１／４以上の許容範囲に入っていれば、ドアの半径候補（図３１にａ〜ｆで示す）と見なす。
【００６５】
建築図面におけるドア表示の代表的な例としては、図３２に示すように、半径ｒの円をその中心Ｐを通る水平線Ｈと垂直線Ｖによって４分割した開き角度９０度の１／４円をなす４種類の扇型のいずれかである。
そして、図２８の（ａ)，(ｂ）に示したように、その各扇型の一方の半径線（壁２１に垂直な半径線）が戸板２０の開き位置を示し、他方の半径線（壁２１に平行な半径線）が戸板２０の閉じ位置を示すが、同図の（ｃ)，(ｄ）に示すように戸板２０の閉じ位置の半径線を図示しない場合もある。
【００６６】
この戸板の幅を示す半径ｒは、一般に単位長（半間又は１メートル）以下で単位長の１／４（出窓の開き窓など）以上である。したがって、連続する黒ドット数計測配列データのピーク値が上記の許容範囲に入っている場合にドア半径候補と見なす。
【００６７】
次に、ドア認識候補の形状抽出を行なうが、そのため図３３に示すように、各ドア半径候補Ｒと見なした黒ドット数計測配列データの壁位置にある基点を中心点Ｐ、長さを半径ｒと仮定して、その両側に壁に沿ってｒ／３幅ずつｒまでの各区間Ｄ１〜Ｄ３およびＤ４〜Ｄ６で円弧画像の有無をパターンマッチング等によってチェックする。
その結果、ドア半径候補Ｒの左側に連続する円弧画像Ａが有った場合は右開きのドア、右側に連続する円弧画像Ａが有った場合は左開きのドア、いずれの側にも連続する円弧が無かった場合はドアではないと判別する。
【００６８】
例えば、図３１に示すａ，ｂ，ｅ，ｆのドア半径候補は左右いずれか一方の側に円弧が見つかるのでドアであると判別され、ｃ，ｄのドア半径候補は、左右何れの側にも円弧が見つからないのでドアではない（実際は垂直壁）と判別される。そして、ドアと判別した場合は、上述した９０°扇型の中心点Ｐ／半径ｒ／ドア方向（右開き／左開き）のドア形状データを抽出する。
【００６９】
また、複数のドア半径候補の長さ（半径ｒ）と距離および開き方向によって、図２８に示したような片開きドア（片開き窓も含む）だけでなく、図２９に示したような親子ドア（一対の戸板２０ａ，２０ｂの幅が異なる）を含む一対の戸板２０ａ，２０ｂが左右両側に開く両開きドア（両開き窓も含む）も判別することができる。
【００７０】
さらに、図３０に示したようにドア半径候補の戸板２０の部分が壁２１に包含されてしまっている場合でも、その一方の側の単位幅内に連続する円弧が有ることが確認されれば、ドアであると認識できる。
このようなドア認識処理を、水平方向および垂直方向の認識された全ての壁候補に対して行なう。
【００７１】
次に、図５のステップＡによる壁の位置の認識処理について説明する。
図６はその壁の位置を認識する処理のサブルーチンの内容を示すフロー図である。まず、ステップ３において、自動スキュー補正された画像データの全体を調査対象とする。そして、ステップ４において、ネスト変数は０（初期値：画像全体を対象にするという意味）である。「ネスト変数」は建築図面の解析範囲をトップダウンで絞り込む時の絞り込み段階を表す。ネスト変数の値が大きいほど解析が深くなっている（細かい部分まで進んでいる）ことを表わす。
【００７２】
ステップ５において、壁の位置の調査対象の領域を限定する処理を行なう。具体的には、この処理にいたる直前に調査領域の指示が示されていて、ここでは以降のステップ６〜９の処理のための準備（インタフェースの共通化）を行なうだけである。個々の調査対象領域の形は矩形図になる。最初はネスト変数が０なので図面全体を調査対象とする。
【００７３】
ステップ６において、水平方向の黒ドット数計測配列データを作成する。これは、画像データの垂直方向の１ドット幅毎に水平方向の黒ドット数を計数（計測）し、その各計数データを保持するものである。
次にステップ７において、ステップ６で作成した水平方向の黒ドット数計測配列データに基づいて、壁の抽出（認識）処理を行なう。その処理手順については後述する。
【００７４】
ステップ８では、ステップ６と同様に垂直方向の黒ドット数計測配列データを作成する。すなわち、画像データの水平方向の１ドット幅毎に垂直方向の黒ドット数を計数（計測）し、その各計数データを保持する。ステップ９では、その垂直方向の黒ドット数計測配列データに基づいて壁の抽出（認識）処理を行なう。
【００７５】
そして、ステップ１０において領域を分割する壁候補があるかどうかを判断する。ここで、水平あるいは垂直方向で１つでも領域を分割する壁候補があれば、ステップ１１へ、１つもそのような壁候補がなければ、ステップ１３へ進む。
例えば、図９の（ａ）に示すような調査対象領域Ｓａ内に領域を分割する壁候補Ｗｄが存在するかどうかを判断する。
ステップ１１では、ネスト変数を＋１して再設定する。これは、現在の解析領域の中から壁を認識し、その壁を使って新たに区切られた現在の領域内の小領域に解析範囲を限定する段階に入ることを表わす。
【００７６】
そして、ステップ１２において、その調査対象の領域の細分化を行なう。具体的には、ステップ７，ステップ９で認識した壁候補の芯線（中心線）で、例えば図９の（ａ）に示すように調査対象領域を壁候補Ｗ，Ｗｄの細線で示す芯線によって領域Ｓａ１，Ｓａ２に２分割する。さらに、その最初の細分化領域（例えば最も左上の領域）に調査対象の位置づけを行なう。
【００７７】
この新たに細分化された領域群の中での解析の順番には特別な順序が必要になる訳ではないが、例えば、領域開始位置のｘ，ｙ座標値の小さい順番に行なうことなどが考えられる。図９の（ｂ）は、壁候補によって細分化された各領域のネストＮo.とその解析処理順序の一例を示し、実線は壁候補の芯線（中心線）を、▲１▼，▲２▼，▲３▼はネストＮo.を、１〜８の小さい数字は処理順序をそれぞれ示している。
【００７８】
その後、図６のステップ５に戻って上述の処理を繰り返し行なう。
ステップ１０において、領域を分割すべき壁候補が１つもなかった場合は、ステップ１３に進んで同次ネスト領域（図９の（ｂ）で同じネストＮo.の領域）の残りがないかどうかを判断する。残りがある場合は、ステップ１５において同次ネストの次の領域に調査対象を進めてステップ５に戻る。
【００７９】
同次ネスト領域の残りがない場合は、ステップ１４へ進んでネスト変数が０であるかどうかを判断する。ネスト変数が０でない場合は、ステップ１６においてネスト変数を−１して１段階上位のネスト領域の処理に戻り、ステップ１３でその段階の残りのネスト領域があるかどうかを判断する。あればステップ１５で次のネスト領域に調査対象を進めてステップ５に戻る。
【００８０】
ステップ１４でネスト変数が０の場合は、ステップ１７へ進んで、ステップ７，９によって得られた各領域毎の壁候補の認識データをもとに各壁の位置及びサイズを確定し、その壁認識データをメモリ４に格納する。その格納方法については後述する。以上の処理の後、当該サブルーチンを終了する。
【００８１】
次に図７に基づいて、図６のステップ７及び９の壁の抽出（認識）処理の手順を説明するが、それに先立って、建築図面（家屋の間取り図）の画像データとその全領域から作成した水平及び垂直方向の黒ドット数計測配列データの具体例を図１０及び図１１に示す。これらの図において、（ａ）は建築図面の画像データ、（ｂ）は水平方向の黒ドット数計測配列データ、（ｃ）は垂直方向の黒ドット数計測配列データである。さらに、図１０の（ｃ）に示した垂直方向の黒ドット数計測配列データを拡大して図１２に示す。
【００８２】
図１０の建築図面では、壁のシンボルが壁の両面と芯線によって表わされており、図１１の建築図面では、壁のシンボルが壁の厚さ内の塗りつぶし（黒）によって表わされている。
この黒ドット数計測配列データにおいて、一番細い黒線の幅が１ドット幅であり、各黒線の長さが（ａ）に示す建築図面の画像データの１ドット幅毎の水平方向又は垂直方向の黒ドット数の計数値に相当する。
【００８３】
これらの図から明らかなように、建築図面を構成する線分の大部分(９０％以上）は水平方向又は垂直方向に描かれており、特に壁の部分で黒ドットの密度が高くなっている。そのため、水平方向及び垂直方向の黒ドット数計測配列データには、壁が存在する位置にピークが表われることになる。
【００８４】
図７のフローの処理を開始すると、まずステップ２１において、指定方向とクロスする方向（水平方向の黒ドット数計測配列データ作成の指定であれば垂直方向に、また垂直方向の指定であれば水平方向に）に、建設図面の基準の単位長に相当する所定間隔ごとに何回壁認識処理のループが可能かを確認する。
ここで、この単位長は一般の住宅の場合にはその最小壁間隔である半間あるいは１メートルであり、ここでは半間（９１ｃｍ）とする。
【００８５】
そして、水平方向の黒ドット数計測配列データに対しては垂直方向の画像幅より若干長い寸法を幅サイズＬとし、垂直方向の黒ドット数計測配列データに対しては水平方向の画像幅より若干長い寸法を幅サイズＬとして自動設定する。
また、半間サイズをｈとし、この値は予め図面の縮尺データ及び半間長を入力するか又は計算による自動算出などにより決定する。
この幅サイズＬと半間サイズをｈからＬ／ｈを算出して小数点以下は切上げた数値を壁認識処理の大ループの実行回数ｎとする。図１２にｈで示す範囲が１回の大ループでの処理範囲である。
【００８６】
次いで、ステップ２２で大ループの回数カウンタのカウント値ｉの初期設定（ｉ←１）を行なう。
そして、ステップ２３において回数カウンタのカウント値ｉが可能な大ループの実行回数ｎを超えた（ｉ＞ｎ）かどうかを判定する。超えていれば、当該領域の解析を終了する。超えていなければ、ステップ２４において当該ｉ番目の半間内の最初の解析処理として最高ピーク（図１２にＰで示す）に位置付け、その点をｘｐとする。このように半間毎に解析処理することにより、その中のピーク値が壁の一部である可能性が高いことになる。
【００８７】
次に、ステップ２５において壁の対象としての最初の条件であるピーク（極大値）の高さが半間（ｈ）以上かどうかを判定する。その結果、ピークの高さが半間未満の場合には壁の認識不明として、次の半間先の解析に移るためにステップ３４に進む。ピークの高さが半間以上ある時には次の解析ステップ２６に移行する。このステップ２６において、最高ピークの位置から左右両側（例えば、２Ｗ幅）を調べて、壁の厚みの範囲（両面の位置：Ｗ₁，Ｗ₂及び厚みＷｅ＝｜Ｗ₁−Ｗ₂｜）の絞り込みを行なう。ここでＷは壁の厚みの意味で、例えばＷmaxと同じ値で使用する。
【００８８】
この絞り込み方法としては、（最高ピーク値−min）* rate＋min以上の値を持つ最高ピーク位置の両端又は片側のピーク位置を、壁の両面の位置Ｗ₁，Ｗ₂又は最高ピーク位置ｘｐが壁の一方の面の位置Ｗ₁であるときの他方の面の位置Ｗ₂として絞り込む方法がある。
【００８９】
図１３はこの絞り込み処理の説明図であり、（ａ）は最高ピーク位置ｘｐの両側に壁の両面の位置Ｗ₁，Ｗ₂が存在する場合の例である。この場合は、最高ピーク位置ｘｐは壁の芯線位置の候補と推定される。
【００９０】
図１３の（ｂ）は最高ピーク位置ｘｐが壁の一方の面の位置Ｗ₁であり、その片側に他方の面の位置Ｗ₂が存在する場合の例である。この場合は、長い方のピーク位置が図２に二重線で示した外輪郭の外壁位置の候補で、それに近接する短い方のピークがその内壁位置の候補と推定し得る。
【００９１】
ここで、 rate は解析の前半(ネスト変数の値が小さい時)は小さめに、後半（ネスト変数の値が大きい時）では大きめにする(例えば、最初は rater＝０.７０とする)。min は図１３の（ｃ）に示すように現在注目している最大ピークＰの位置ｘｐの左右両側２Ｗに拡がった４Ｗ幅程度の幅内の黒ドット数計測データの最小値である。
【００９２】
このようにして、図７のステップ２６で絞り込んだ結果を基に、ステップ２７と２８で壁としての妥当性を確認する。
まずステップ２７においては、壁の厚みＷｅがその最大値 Ｗmax（例えば３０ｃｍ）を超えているか否かを判断し、超えている時は壁の認識不明として次の半間先の解析に移るためステップ３４に進む。超えていなければステップ２８へ進み、壁の厚みＷｅが最小値Ｗmin（例えば２.５ｃｍ）未満か否かを判断する。
【００９３】
その結果、壁の厚みＷｅが最小値 Ｗmin未満の場合はステップ３０へ進み、そうでない時はステップ２９に進む。ステップ３０においては壁以外のピーク（例えば、畳，窓，引戸など）を認識したものとしてその情報を得る。
ステップ２９においては、壁の候補となる領域から、壁としての条件を満たすかどうかを後述する２等分割探索法によって判定し、壁だと認識できたものについて、ステップ３１において当該壁の両面の位置Ｗ₁，Ｗ₂及び厚みＷｅなどの情報を退避（記憶）してステップ３２に進む。
【００９４】
ステップ２９で壁としての条件を満たさなければステップ３２に移行する。ステップ３２においては、壁の両面位置（座標）Ｗ₁及びＷ₂が共に現在処理を行なっている領域の内側かどうかを判定し、内側であればステップ３３に進む。そうでなければステップ３４に移行する。
【００９５】
ステップ３３では、現在の処理領域を細分化し、新しい細分領域を示す境界データとして、Ｗ₁，Ｗ₂，Ｗｅを退避（記憶）する。また、（Ｗ₁＋Ｗ₂）／２がその壁の芯線位置である。ステップ３４では、次の半間先の処理を行なうために大ループの回数カウンタのカウント値ｉを＋１してから、ステップ２３に戻って上述の処理を繰り返し行なう。
【００９６】
このようにして、対象となる黒ドット数計測配列データの一端（先頭要素）から半間毎に解析処理し、反対の端まで解析が終われば、今回の範囲の解析を終了する。水平方向と垂直方向の２方向のを黒ドット数計測配列データに対して別々に解析し、次回の解析範囲は、今回の解析で壁と認識できた範囲に限定する。
そのため、水平及び垂直方向の黒ドット数計測配列データを解析した結果を組み合わせて、総当たりの場合分けを行なう。
【００９７】
この実施例では、壁の芯線位置は隣の壁との間隔が半間単位の整数倍になるように配置されているとみなす。
そして、特徴的なピークが発見できる範囲までを解析データとして有効に使用する。逆に云えば、解析対象とするある範囲内に壁に相当する特徴が、水平方向及び垂直方向の両方合わせても一つも発見できなかったときは解析を終了する。１回の解析範囲は、最初は図面全体を対象にし、以後発見された壁で区切られた範囲に限定し直す方法をとって、壁のピークが発見されやすくし、且つ細部までの解析が容易になるようにする。
【００９８】
次に、図７のステップ２９において「壁としての条件を満たすか」を判断する２等分割探索法について、図８のフロー図によって説明する。
この図８に示すフローの処理を開始すると、まずステップ４１において、２等分割探索法の初期設定を行なう。
すなわち、壁分析のための領域分割要素数ｎを１にし、壁部か非壁部かが未確定である分割要素の配列要素の最も小さいＮo.を指すｋを０に、また、調査領域の分割配列要素として、Ｓ〔０〕，Ｅ〔０〕それぞれに入力のスタート及びエンド画像アドレスを代入して初期設定とする。
【００９９】
その後、ステップ４２に進み、調査領域の開始アドレスＳ〔ｋ〕と終了アドレスＥ〔ｋ〕が等しければステップ５３に移行し、等しくなければステップ４３に進む。ステップ４３においてはイメージ画像データの対象領域の分割処理を行ない、元の領域を小数以下の誤差を除いて２等分割する。
【０１００】
すなわち、分割する前半の領域のスタート及びエンド画像アドレスＳ１，Ｅ１を、Ｓ１＝Ｓ〔ｋ〕，Ｅ１＝(１／２)（Ｓ〔ｋ〕＋Ｅ〔ｋ〕）とし、後半の領域のスタート及びエンド画像アドレスＳ２，Ｅ２を、Ｓ２＝(１／２)（Ｓ〔ｋ〕＋Ｅ〔ｋ〕）＋１，Ｅ２＝Ｅ〔ｋ〕とする。
【０１０１】
それによって、例えば図１４に示すように、建設図面のイメージ画像データ中において、壁候補が存在する線（一点鎖線で示す）の元の領域幅を最初は１の位置で２等分割する。その後壁の有無を判別できるまで、順次図１４に示す位置２で２回目、３の位置で３回目、４の位置で４回目というように２等分割を繰り返して細分化した領域でステップ４４及び４５の壁調査を行なうようにする。
ステップ４４及び４５においては、２等分割したそれぞれの領域Ｐ１，Ｐ２が壁で満たされているかどうかを調査し、ステップ４６に進む。
【０１０２】
ここでは、指定された領域（スタートアドレスからエンドアドレスまで）の黒ドット数計測配列データを分析した結果、黒ドットのピーク（壁の厚みの広がりを保って）高さが、指定された領域の幅（長さ）と比較して次の▲１▼〜▲３▼の判断をする。
【０１０３】
▲１▼：５％以下のとき、 ｖ＝０：非壁部と判断
▲２▼：９５％以上のとき、ｖ＝１：壁部と判断
▲３▼：上記以外のとき、 ｖ＝２：どちらとも判断できない
これを図に示すと図１５に▲１▼，▲２▼，▲３▼で示すようになる。
【０１０４】
ステップ４６においては、ステップ４３で分割された前半の領域Ｐ１について、壁が存在するかどうか判断できない（ｖ＝２）場合はステップ４７に進み、そうでなければステップ４９に進む。ステップ４７においては、ステップ４３によって分割する前の領域範囲の格納配列要素Ｓ〔ｋ〕，Ｅ〔ｋ〕，ｖ〔ｋ〕を、ステップ４３で分割された後半の領域データ(スタート及びエンド画像アドレスＳ２，Ｅ２と判定結果ｖ２)で置き換える。
【０１０５】
ステップ４８においては、新しい格納配列要素Ｓ〔ｎ〕，Ｅ〔ｎ〕，ｖ〔ｎ〕として、ステップ４３で分割された前半の領域データ（スタート及びエンド画像アドレスＳ１，Ｅ１と判定結果ｖ１）を退避し、ステップ５１に移行する。
ステップ４９においては、ステップ４３によって分割する前の領域範囲の格納配列要素Ｓ〔ｋ〕，Ｅ〔ｋ〕，ｖ〔ｋ〕を、ステップ４３で分割された前半の領域データ(スタート及びエンド画像アドレスＳ１，Ｅ１と判定結果ｖ１)で置き換える。
【０１０６】
ステップ５０においては、新しい格納配列要素Ｓ〔ｎ〕，Ｅ〔ｎ〕，ｖ〔ｎ〕として、ステップ４３で分割された後半の領域データ（スタート及びエンド画像アドレスＳ２，Ｅ２と判定結果ｖ２）を退避し、ステップ５１に移行する。
ステップ５１では、領域アドレスの新しい格納配列要素を示せるように新しい格納配列要素Ｎo.を示す変数ｎを＋１してから、ステップ５２に進む。
【０１０７】
ステップ５２においては、壁部か非壁部かが未確定である分割要素の最も小さいＮo.を指すｋ要素内の分類コードｖが壁が存在するかどうか判らない内容（ｖ＝２）の場合は、ステップ４２に戻って更に細分割する処理を繰り返す。そうでない場合はステップ５３に進む。
ステップ５３では、壁が存在するかどうか判らない内容が一つ解決したとして、その指標ｋを＋１してからステップ５４に進む。
【０１０８】
ステップ５４においては、壁分析のための領域分割要素数ｎと、壁部か非壁部かが未確定である分割要素の配列要素の最も小さいＮo.を指すｋとが、等しくなっているかどうか判定し、等しければ壁認識のための分割処理が終了したと判断してステップ５５へ進む。等しくなければステップ５２へ戻る。
【０１０９】
ステップ５５においては、分割された領域アドレス・データ（配列）が上昇順に並ぶよう、スタートアドレス順（昇順）にソートを実行してステップ５６に移行する。すなわち、Ｓ〔０〕〜Ｓ〔ｎ−１〕，Ｅ〔０〕〜Ｅ〔ｎ−１〕，ｖ〔０〕〜ｖ〔ｎ−１〕のデータをＳ〔 〕をキーにして昇順にソートする。
【０１１０】
ステップ５６においては、壁部分及び非壁部分が連続している場合は、それぞれ縮退処理（一つの範囲で表現）して終了する。すなわち、連続したｖ〔 〕値が０又は１の状態の場合は、Ｓ〔 〕，Ｅ〔 〕データを圧縮する。
なお、この時に、壁が存在するかどうか判らない内容（ｖ＝２）の配列要素を含む場合は、その要素の前後の要素が壁を示している場合には壁データに変更し、また、非壁を示している場合には非壁データに変更して処理する。
【０１１１】
上述した二等分割探索処理による画像データ中の壁位置の分析例を図１６に示す。この図１６の（ａ）には壁のイメージ画像（斜線を施した部分）Ｗとその調査対象領域を破線で示しており、この調査領域は先に認識された壁候補の存在位置に沿って設定される。そして、Ｓ〔０〕＝０がその調査領域の最初のスタートアドレス、Ｅ〔０〕＝１５が最初のエンドアドレスである。４，５，７等の途中の数字は分割後の対象領域のスタート又はエンドアドレス（いずれも画像アドレス）である。
【０１１２】
図１６の（ｂ）には変数ｎ＝１〜１０の各調査段階における各対象領域のスタートアドレスＳ，エンドアドレスＥ，及び壁の有無に関する判断結果ｖとその確定状況、ソート状況、並びに縮退処理結果をｋの値と共に示している。そして、最終的には画像アドレス５〜１２に壁が存在することを認識している。
【０１１３】
次に、図１７によって簡単な建築図面の壁認識例を説明する。
この図１７には、ネスト変数（ｎｅｓｔ）と、領域分割状態と認識された実際の壁の状態とを示している。
まず、ネスト変数＝０で建築図面の全体を壁位置の調査対象として壁の抽出を行なう。その結果（Ａ）に実線で示すように建築図面の家屋部の輪郭と水平及び垂直方向の壁候補の位置を認識できたとする。しかし、その認識できた壁候補のうち実際の壁は（ａ）に示す部分だけであるが、それはまだ判らない。
【０１１４】
そこで次に、ネスト変数＝１にして、（Ａ）に示す認識できた壁候補の芯線で区切られた各矩形領域毎に調査対象領域を限定して壁の抽出を行なう。それによって（Ｂ）に▲１▼，▲２▼，▲３▼，▲４▼で示す４つの調査対象領域で新たに太線で示す壁候補が認識されると共に、先に認識された壁候補のうち、実際の壁は（ａ）に示された部分だけであったことが確認され、（ｂ）に示す壁の状態が認識される。
【０１１５】
さらに、ネスト変数＝２にして、（Ｂ）において新たな壁候補が認識された４つの領域▲１▼〜▲４▼をそれぞれその新たに認識された壁によって分割して、調査対象領域をさらに限定して壁の抽出を行なう。それによって、（Ｃ）に▲５▼，▲６▼で示す２つの調査対象領域で新たに太線で示す壁候補が認識され、（ｃ）に示す壁の状態が認識される。
【０１１６】
その後、ネスト変数＝３にして、（Ｃ）において新たな壁候補が認識された２つの領域▲５▼，▲６▼をそれぞれその新たに認識された壁によって分割して、調査対象領域をさらに限定して壁の抽出を行なう。その結果いずれの分割領域でも新たな壁候補を抽出できなにかった場合には、それによって壁位置の調査を終了し、（ｃ）に示す壁位置が最終的な壁認識結果であることが確定し、そのデータをメモリに格納する。
【０１１７】
このように、分割した各調査対象領域のいずれでも新たな壁候補が抽出されなくなるまで、調査対象領域を細分化して壁の抽出を行なう。それによって、小さな壁でも確実に認識することができ、且つ壁候補のうち実際には壁が存在する部分と存在しない部分とを正確に判別することができる。
【０１１８】
ここでさらに、前述した図６のフローチャートに従った具体的な建築図面の認識処理手順の例を、図１８乃至図２０によって説明する。図１８乃至図２０は一連の図であるが、図示の都合上３枚の図に分割して示している。これらの図におけるＳ３〜Ｓ１７は、図６のＳ３〜Ｓ１７の各ステップに対応している。また、各段階での領域分割図と実壁状態も図示している。
【０１１９】
以下の説明ではステップを「Ｓ」と略称する。図１８のＳ３で図面全体を調査対象とし、Ｓ４でネスト変数を０にする。Ｓ５で調査対象の限定を行なうがネスト変数が０なのでやはり図面全体を調査対象とする。Ｓ６〜Ｓ７で水平方向の黒ドット数計測配列データを作成して壁を抽出するが、輪郭以外の壁を発見できず、Ｓ８〜Ｓ９で垂直方向の黒ドット数計測配列データを作成して壁を抽出し、輪郭以外の壁を２か所に発見する。したがって、Ｓ１０でＹＥＳになり、Ｓ１１でネスト変数を１にし、Ｓ１２で領域の細分化（各壁の位置で）をしてＳ５へ戻り、調査対象領域を一番左の領域に限定する。
【０１２０】
そして、Ｓ６〜Ｓ７で壁を２か所に発見し、Ｓ８〜Ｓ９では壁を発見できなかったがＳ１０ではＹＥＳになり、Ｓ１１でネスト変数を２にして「入れ子処理」を、図１９のＳ１４でＮＯになり入れ子処理を終了するまで繰り返し実行し、左側の縦長の領域を新たに発見された２つの壁によって区切った３つの分割領域に対して、順次壁の抽出処理を行なう。
【０１２１】
この例ではそれによって新たな壁は発見されず、図１９のＳ１６でネスト変数を−１して１に戻し、真中の縦長の領域を調査対象領域として同様に壁の抽出を行なうが、この例では新たな壁候補は発見されない。
そこで、Ｓ１５，Ｓ５で右側の縦長の領域に調査対象領域を変更し、Ｓ６〜Ｓ７で壁を１ケ所発見する。
【０１２２】
そこで、図２０のＳ１０でＹＥＳになり、Ｓ１１でネスト変数を２にして「入れ子処理」を開始し、右側の縦長の領域を新たに発見された壁によって分割し、その各分割領の壁抽出処理を順次行なう。その結果、いずれの分割領域でも新たな壁は発見されず、Ｓ１３でＮＯになり入れ子処理を終了し、Ｓ１６でネスト変数を−１して１にするが、ネスト変数１の領域は残っていないので、さらにネスト変数を０に戻すが、その領域も残っていない。そのため、壁抽出の処理は完了したと判断し、Ｓ１７で抽出された壁候補の認識データにより各壁の位置及びサイズを確定し、そのデータをメモリに格納する。
【０１２３】
次に、上述のようにして認識した建築図面の輪郭及び骨格に相当する壁の位置及びサイズ等の情報（解析結果データ）を図１に示したメモリ４及び外部記憶装置１１の記憶媒体に格納する内容の一例について、図２１によって説明する。
図２１において、（Ａ）はネスト数、（Ｂ）は固有ネスト情報、（Ｃ）水平方向の壁情報、（Ｄ）は垂直方向の壁情報、（Ｅ）は水平な壁のモデル、（Ｆ）は垂直な壁のモデルを示す。
【０１２４】
ネスト数は、子ネストポインタの入れ子（親子関係）の深さを示し、壁が全然認識されなかった場合は、ネスト数＝０である。
固有ネスト情報は、ネストＮo.，ＮＥＸＴ兄弟ポインタ，子ネストポインタ，壁数（水平方向及び垂直方向），壁情報ポインタ（水平方向及び垂直方向）からなる。
【０１２５】
ＮＥＸＴ兄弟ポインタは、同時階層ネスト情報の次のアドレスを持つ。したがって、このポインタが示す場所の固有ネスト情報内のネストＮo.は、当該処理のものと同じ値である。
子ネストポインタは、一階層下の階層ネスト情報の先頭データのアドレスを持つ。したがって、このポインタが示す場所の固有ネスト情報内のネストＮo.は、当該処理のものに＋１した値である。
【０１２６】
（Ｂ）に示す固有ネスト情報中の水平方向の壁情報ポインタが示すアドレスを先頭アドレスとして、（Ｃ）に示す水平方向の壁情報が格納される。
その壁情報は、次の水平方向の壁情報の先頭アドレスの位置を示すＮＥＸＴポインタ、壁の始点座標（ｘ座標：ａ，ｙ座標：ｂ）、壁の横（ｘ方向）サイズ：ｃ、壁の縦（ｙ方向）サイズ：壁の厚みｄからなる。これらのａ〜ｄによって（Ｅ）に示す水平な壁のモデルを記憶し、またそれを再現することができる。
【０１２７】
同様に、固有ネスト情報中の垂直方向の壁情報ポインタが示すアドレスを先頭アドレスとして、（Ｄ）に示す垂直方向の壁情報が格納される。
その壁情報は、次の垂直方向の壁情報の先頭アドレスの位置を示すＮＥＸＴポインタ、壁の始点座標（ｘ座標：ｅ，ｙ座標：ｆ）、壁の縦（ｙ方向）サイズ：ｇ、壁の横（ｘ方向）サイズ：壁の厚みｈからなる。これらのｅ〜ｈによって（Ｆ）に示す垂直な壁のモデルを記憶し、またそれを再現することができる。
【０１２８】
ところで、実際の建築図面の画像データに対して、その図面全体を調査対象領域として水平方向及び垂直方向の黒ドット数計測配列データを作成した例を図１０及び図１１に示したが、その黒ドット数計測配列データに基づいて壁候補を認識した次の段階で、その図面の領域を認識した壁候補によって分割し、調査対象領域を限定した画像データに基づく水平及び垂直方向の黒ドット数計測配列データの作成例を、図２２乃至図２４に示す。
【０１２９】
図２３及び図２４は、図２２よりさらに調査対象領域を細分化した例である。
このようにして、新たな壁候補が発見されなくなるまで、調査対象領域を細分化して、その画像データによる水平及び垂直方向の黒ドット数計測配列データを作成し、壁の抽出を行なう。
【０１３０】
なお、この実施形態ではポジ画像の建築図面を認識対象としたので、その２値化した画像データの水平及び垂直方向の黒ドット数を計測（カウント）して黒ドット数計測配列データを作成したが、ネガ画像の建築図面を認識対象とする場合には、その２値化した画像データの水平及び垂直方向の白ドット数を計測（カウント）して白ドット数計測配列データを作成すれば、壁の認識を同様に行なうことができる。
【０１３１】
また、上述のようにして認識した建築図面の輪郭及び骨格に関する認識データは、ＣＡＤ用ベクトルデータに変換をすることができ、異機種間のＣＡＤデータの互換性を得ることができる。
【０１３２】
そして、図２８又は図２９に示したような平面表示のドアの場合、この実施形態では黒ドットからなる形状輪郭より戸板部分に該当する大きさの連続ドット数を抽出し、その連続ドット数をドア半径候補と仮定し妥当性をチェックしてドアの」認識を行なう。これをさらに、図面全体の建具認識処理につなげられることが考えられる。
【０１３３】
さらに、図３０に示したような平面表示のドアについて、その連続する黒ドットの先端部に注目し、９０°扇型の円弧部分の立ち上がりを計測することによって、壁に包括された戸板部分を抽出し、それによって多少の位置ずれや正確な９０°扇型の記述でなくても認識処理を行なうので、次に示すような利点がある。また、ドット構成が細く弱々しい場合には、当該領域の黒又は白ドットに注目した画像膨張を行なってから、連続ドット数の抽出を行なうことにより、ドアの認識が可能になる。
【０１３４】
（利点１）大きさを勘案した連続ドット数計測配列データを基に、２つの９０°扇型の円弧立ち上げ部分の計測と、その中の円弧記述の認識を行なうことによって、ドアの記述を確実に認識できる。
（利点２）定規などを使用せずにフリーハンドで記述された図面でも、水平及び垂直方向に限らず斜め直線も含めて直線の揺らぎを相殺すると、形状輪郭が水平及び垂直方向への正対性がよければ認識することができる。
【０１３５】
なお、手書き図面で建築図面の正対方向のスキュー補正が困難な場合、手書き図面を方眼紙に記述して、スキュー補正はその方眼紙のマス目を利用して行なうようにするとよい。
また、やむをえず利用者の手作業に委ねる場合、スキャナ読込時にできるだけスキュー補正が不要な正対する図面を作成するように注意を促すマニュアルを添えるとよい。
【０１３６】
【発明の効果】
以上説明してきたように、この発明による建築図面からドアを認識する方法及びドアを認識する装置によれば、従来正確な認識が困難であった線が途切れたり直線が多少傾いて記載された手書き図面や、青焼きなどの比較的コントラストが低い図面や、ノイズの多い図面あるいは古い建築図面など、記載状態や画質の悪い建築図面からでも、その輪郭・骨格を形成する壁の位置とともにドア（開き戸，開き窓）の形状及び種類を簡単に精度よく認識することができる。
【図面の簡単な説明】
【図１】この発明による建築図面からドアを認識する方法を実施する建築図面認識装置の一例の概略構成を示すブロック図である。
【図２】図１の表示部９における再マッピングされた認識結果の画像データの表示例を示す図である。
【図３】同じくスキュー補正された建築図面の入力画像データを認識結果の画像データと重ね合わせて表示した例を示す図である。
【図４】建築図面における外輪郭，外壁，内壁及び骨格の定義を説明するための図である。
【図５】図１に示した建築図面認識装置による建築図面認識処理のメインルーチンを示すフローチャートである。
【図６】図５におけるステップＡの壁の位置認識処理のサブルーチンの詳細を示すフロー図である。
【図７】図６におけるステッブ７及び９の壁の抽出（認識）処理のサブルーチンのフロー図である。
【図８】同じく壁の位置を認識する２等分探索法を実行処理するフロー図である。
【図９】調査対象領域内の壁候補とそれによる領域分割例及び壁候補によって細分化された領域群のネストＮｏ．とその解析処理順序の一例を示す説明図である。
【図１０】建築図面（家屋の間取り図）の画像データとその全領域から作成した水平及び垂直方向の黒ドット数計測配列データの具体例を示す図である。
【図１１】同じくその他の具体例を示す図である。
【図１２】図１０の（ｃ）に示した垂直方向の黒ドット数計測配列データを拡大して示す図である。
【図１３】図７のステップ２６におけるピークの両側又は片側に壁の両面を絞り込む処理の説明図である。
【図１４】図８のステップ４３における領域幅の分割処理の説明図である。
【図１５】図８のステップ４４，４５における対象領域の壁調査による判断の説明図である。
【図１６】図８に示した２等分割探索処理による壁のサンプル（壁候補）の分析例を示す説明図である。
【図１７】この発明による簡単な建築図面の壁認識例の説明図である。
【図１８】図５のフローチャートに従った具体的な建築図面の認識処理手順の例を示す説明図である。
【図１９】図１８の続きの説明図である。
【図２０】図１９の続きの説明図である。
【図２１】解析結果データのメモリへの格納内容の一例を示す説明図である。
【図２２】図１０に示した建築図面の調査対象領域を限定した画像データとそれに基づく水平及び垂直方向の黒ドット数計測配列データの例を示す図である。
【図２３】図２２より調査対象領域をさらに限定した画像データとそれに基づく水平及び垂直方向の黒ドット数計測配列データの例を示す図である。
【図２４】図２２より調査対象領域をさらに限定した他の部分の画像データとそれに基づく水平及び垂直方向の黒ドット数計測配列データの例を示す図である。
【図２５】図１の表示部９におけるマッピングされたドア認識のための建築図面の入力画像データの表示例を示す図である。
【図２６】同じくその入力画像データに対して家屋の外壁及び家屋内の仕切り壁の認識結果を表示した例を示す図である。
【図２７】同じく図２５の入力画像データに対して外壁及び家屋内の仕切り壁とドアの認識結果の画像データを重ね合わせて表示した例を示す図である。
【図２８】建築図面における片開きドアの平面表示の例を示す図である。
【図２９】建築図面における両開きドアの平面表示の例を示す図である。
【図３０】建築図面における片開きドアの戸板部分が壁に包括された平面表示の例を示す図である。
【図３１】水平壁に沿ってドア半径候補を探索する方法の説明図である。
【図３２】一般的なドアの平面表示の種類を説明するための説明図である。
【図３３】ドア半径候補の両側の単位長範囲内の円弧画像を探索する処理の説明図である。
【符号の説明】
１：全体制御部（ＣＰＵ） ２：画像読取部
３：通信制御部 ４：メモリ
５：自動スキュー補正部
６：ドット数計測配列データ作成部
７：輪郭・骨格認識部 ８：再マッピング制御部
９：表示部 １０：操作入力部
１１：外部記憶装置 １２：印刷装置
１３：ドア認識部 １４：バス[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a method for recognizing “doors (folding doors, casement windows)” in architectural drawings of houses such as individual houses, apartments, condominiums, inns, hotels, stores, and office buildings widely used in the construction industry. The present invention relates to a recognition device.
[0002]
[Prior art]
2. Description of the Related Art In recent years, it has been practiced to easily create architectural drawings of houses, buildings, etc. using a CAD system or the like, and to store the data for use in design changes or extension / renovation. However, the data of the architectural drawing is not compatible with the created system, and cannot be used due to the passage of time or change of contractor.
In addition, when renovating or remodeling a house, there are often only old architectural drawings drawn on paper, and it was necessary to redraw all the drawings for expansion and renovation, including the parts that were not changed.
[0003]
Therefore, it is also attempted to read architectural drawings drawn on paper and recognize and store them as data that can be processed by a computer, but there is no special method or device for that purpose, and architectural drawings are read with an image scanner, The image data is input to a personal computer or the like, and line segment recognition or pattern recognition is performed using a general graphic recognition function.
[0004]
In addition, even if the automatic recognition function is somewhat incomplete, the basic diagram such as straight lines and arcs can be converted by raster-vector conversion even if the automatic recognition function is somewhat incomplete. Some are designed to be automatically recognized.
[0005]
[Problems to be solved by the invention]
However, such a conventional drawing recognition device requires a high level of operation knowledge and is limited to those who are used to personal computers and specialized operators, and for those involved in the industry who frequently use architectural drawings, It was by no means easy to use.
[0006]
In addition, although basic diagrams such as straight lines and arcs can be automatically recognized, it is possible to individually recognize graphic symbols (such as walls and pillars) that are unique to architectural drawings consisting of combinations of basic diagrams. There wasn't. For this reason, when recognizing the recognized drawing, it has to be performed for each line segment, which requires a lot of work.
[0007]
Furthermore, the architectural drawings of houses and buildings used in the construction industry are not uniform in thickness even if each line segment on the drawing looks as a continuous line of the same size to the human eye. The trajectory also fluctuates. Even if it should be a continuous line, it may be broken in some places. These imperfections are caused not only at the time of drawing, but also from aging of paper and errors at the time of reading. Therefore, for example, if the limit of the thickness that can be recognized as a line segment is too thin, it is recognized that the line segment is cut even if it is slightly faint. A thick line segment may be recognized as a rectangle.
[0008]
This is of course the quality of the original drawing, but there are many drawings with low contrast (so the black and white borders are not clear) as in the so-called blue printing drawing using photosensitive paper. Since this point appears on the surface, it is difficult to perform automatic recognition with high accuracy with a conventional drawing recognition apparatus, and it takes a lot of time to correct the correction, so it has hardly been put to practical use.
[0009]
By the way, in the architectural drawing of a house, a wall that divides the floor plan plays a central role in the drawing, and by recognizing which part in the drawing is the wall, it is possible to understand an approximate floor plan of the architectural drawing. Therefore, recognizing the position of the wall and its length in the drawing is the most important matter in recognizing the construction drawing.
[0010]
In addition, in order to recognize doors (folding doors, hinged windows) from this architectural drawing, a 90 ° fan shape with a radius equal to or less than the unit length in contact with the outer wall of the house and the partition wall in the house (circle is divided into four) Often, this is determined by the shape description of the four patterns.
Therefore, when recognizing doors from architectural drawings such as blueprints, the wall information in the drawing recognizes the corresponding wall from the unit-length rectangular area in contact with it, and the “wall” recognition and “ Whether or not the linkage of “door” recognition can be surely performed is an important key.
[0011]
However, the door recognition from the conventional drawings has the following problems.
(1) In the contour tracking method, if there are many spots or irregularities, it is recognized as a broken straight line or as a straight line divided into two or three, and cannot be recognized accurately.
(2) In the projected feature extraction method obtained from the frequency distribution of black dots, when the line is thin, when the horizontal and vertical straight lines are found, the feature cannot be extracted if the line is inclined even a little, and I couldn't recognize it.
For the reasons described above, the “90 ° fan shape” in the middle of the “door” recognition process could not be recognized.
[0012]
The present invention has been made in view of the above points, and even with architectural drawings drawn on paper with lines that are broken or slightly inclined due to blue printing or handwriting, image data obtained by reading the drawing, or its run length It is an object of the present invention to enable automatic recognition of a door (a hinged door, a hinged window) with high accuracy from encoded image data obtained by encoding the above.
[0013]
[Means for Solving the Problems]
In order to achieve the above object, the present invention provides a method for recognizing a door and an apparatus for recognizing a door from the following architectural drawings.
[0014]
The method for recognizing a door from an architectural drawing according to the present invention is based on position information of a wall as a partition of an outer wall of a house and a partition in an area composed of black or white dots of image image data obtained by reading an image of the architectural drawing. As a unit length rectangular area on both sides of the image, and the continuous dot number measurement array data by measuring the number of black or white dots continuous in the direction perpendicular to the wall in the image data in each of the limited rectangular areas Are created, the door radius candidates are extracted based on the created continuous dot count measurement array data, the presence or absence of an arc image within the unit length range on both sides of each door radius candidate is determined, and the arc image The shape and type of the door are recognized based on the position of the door and the length of the door radius candidate.
[0015]
In this method of recognizing a door from an architectural drawing, horizontal and vertical dot number measurement array data is created by measuring the number of horizontal or vertical black or white dots in the image data obtained by reading the architectural drawing image. The position information of the wall of the architectural drawing can be recognized based on the created dot count measurement arrangement data in both directions.
[0016]
The apparatus for recognizing a door from an architectural drawing according to the present invention includes an image data input means for inputting image image data obtained by reading an image of the architectural drawing, and black and white in the horizontal and vertical directions of the image image data inputted by the means. Or, a dot number measurement array data creating means for measuring the number of white dots to create horizontal and vertical dot number measurement array data, and the outer wall of the house based on each dot number measurement array data created by the means Contour / skeleton recognition means for recognizing position information of a wall as a partition in the house.
[0017]
Further, with reference to the position information of the wall recognized by the contour / skeleton recognition means, the area consisting of black or white dots of the image image data is limited to unit length rectangular areas on both sides of the wall. In the image data in each rectangular area, the continuous dot number measurement array data is created by measuring the number of continuous black or white dots in the direction perpendicular to the wall, and the created continuous dot number measurement array data Door radius candidates are extracted on the basis of them, the presence / absence of an arc image within a unit length range on both sides of each door radius candidate is determined, and the shape of the door is determined according to the position of the arc image and the length of the door radius candidate And door recognition means for recognizing the type.
[0018]
The image data input means may be an image reading means for reading an image of an architectural drawing and inputting the image image data.
[0019]
Alternatively, the image data input means may be image data input means for inputting encoded image data obtained by encoding a run length of image image data obtained by reading an image of an architectural drawing. The number measurement array data creating means measures the number of black or white dots in the horizontal direction and the vertical direction of the original image image data from the encoded image data input by the image data input means, and calculates the number of dots in the horizontal direction and the vertical direction. Any means for creating measurement array data may be used.
[0020]
In this case, the image data input means may be image data receiving means for receiving and inputting the encoded image data by communication.
Furthermore, in the apparatus for recognizing a door from these architectural drawings, a display means for displaying a recognition result by the door recognition means may be provided.
The display means can display the result of recognition by the door recognition means superimposed on the image data input by the image data input means.
[0021]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be specifically described below with reference to the drawings.
FIG. 1 is a block diagram showing a schematic configuration of an example of an architectural drawing recognition apparatus (including an apparatus for recognizing a door from an architectural drawing) that implements a method for recognizing a door (a hinged door, an open window) from an architectural drawing according to the present invention. Yes, a hardware configuration and a software processing function by a microcomputer are shown mixedly.
[0022]
This apparatus includes an overall control unit 1, an image reading unit 2, a communication control unit 3, a memory 4, an automatic skew correction unit 5, a dot number measurement array data creation unit 6, a contour / skeleton recognition unit 7, a remapping control unit 8, It comprises a display unit 9, an operation input unit 10, an external storage device 11, a printing device 12, a door recognition unit 13, and a bus 14 for connecting them.
Note that an interface unit necessary between these units (or devices) and the bus 14 is not shown.
[0023]
The overall control unit 1 is a microcomputer (consisting of a CPU, ROM, RAM, etc., but is abbreviated as “CPU” as a representative) that controls the operation and function of the entire building drawing recognition apparatus, and an automatic skew correction unit. 5 and the functions of the dot number measurement array data creation unit 6, the contour / skeleton recognition unit 7, the remapping control unit 8, and the door recognition unit 13 according to the present invention can also be realized by software processing of the CPU.
[0024]
The image reading unit 2 is an image data input unit that scans a set architectural drawing, reads the image, and inputs image image data. The image reading unit 2 includes a scanning optical system, an image sensor such as a CCD, and a drive circuit thereof. This is an image scanner. Also included is a circuit that binarizes the read image data at a predetermined resolution to produce white dot and black dot image data.
[0025]
The communication control unit 3 is an image data receiving unit that receives and inputs encoded image data in which image image data or its run length is encoded by communication from the outside instead of capturing image data from the image reading unit 2. At the same time, architectural drawing data recognized by this device can be transmitted to an external device. Specifically, it includes a FAX modem and personal computer communication control means.
[0026]
The memory 4 includes image data read by the image reading unit 2, image image data or encoded image data received by the communication control unit 3, image data skew-corrected by the automatic skew correction unit 5, and dot number measurement array The dot count measurement array data created by the data creation unit 6, the contour and skeleton recognition results recognized by the contour / skeleton recognition unit 7, the shape and type data of the door recognized by the door recognition unit 13, and the remapping This is a large-capacity RAM or a hard disk memory for storing image data and the like remapped by the control unit 8.
[0027]
The automatic skew correction unit 5 is for adjusting the angle of the image data stored in the memory 4 to correct the horizontal and vertical line segment directions so as to coincide with the horizontal and vertical reference directions of the apparatus. The automatic skew correction technique can be used.
The image data corrected by the automatic skew correction unit 5 is stored in the memory 4 again.
[0028]
The dot number measurement array data creation unit 6 performs the automatic skew correction on the image image data or the encoded image data stored in the memory 4 and the image data remapped by the remapping control unit 8 described later. The image data is limited to two directions, the horizontal and vertical directions, and the number of black or white dots is measured (counted) in units of dot widths respectively. As a result, the horizontal and vertical dot number measurement array data (histogram) is obtained. This is dot number measurement array data creation means for creating and storing in the memory 4.
[0029]
If the read construction drawing is a positive drawing (a drawing whose brightness is lower than that of the ground), the number of black dots is measured, and if it is a negative drawing (a drawing whose brightness is higher than that of the ground), Measure the number of white dots.
[0030]
The outline / skeleton recognition unit 7 recognizes the outline and skeleton of the architectural drawing based on the horizontal and vertical dot number measurement array data created by the dot number measurement array data creation unit 6, and in particular, the outer wall of the house and the house This is a contour / skeleton recognition means for extracting information such as the position, length, thickness, type, etc. of the wall that is an indoor partition, and details thereof will be described later.
[0031]
In the reference dot number measurement array data, if wall recognition (extraction) is difficult or uncertain, a request for resetting the horizontal and vertical dot number measurement array data creation range in the drawing is sent to the overall control unit 1. The transmission range and the creation range are changed, and the dot number measurement array data creation unit 6 is made to create dot number measurement array data again.
[0032]
The remapping control unit 8 limits the range of the architectural drawing by the part recognized as the wall by the contour / skeleton recognition unit 7, re-maps the part in consideration of other recognition information, Image data for causing the dot number measurement array data creation unit 6 to create the dot number measurement array data again for each limited range is created. In this case, noise of the document or reading noise in the image reading unit 2 is also generated. Create the removed image data. This data is also stored in the memory 4.
[0033]
The door recognizing unit 13 extracts unit-length rectangular regions (areas) in both the horizontal and vertical directions based on the position information (recognition processing result) of the horizontal and vertical walls recognized by the contour / skeleton recognizing unit 7. A process for recognizing a door (a hinged door, a hinged window) is performed. Details will be described later.
[0034]
The display unit 9 is input from the image reading unit 2 or the communication control unit 3, and has been corrected by the automatic skew correction unit 5. The image data of the construction drawing and the horizontal and vertical directions created by the dot number measurement array data creation unit 6 are displayed. Black or white dot number measurement array data, wall data recognized by the outline / skeleton recognition unit 7, door data of the door recognized by the door recognition unit 13, image data remapped by the remapping control unit 8, etc. For example, a CRT or a liquid crystal display.
[0035]
2 and 3 show examples of the display state of the screen 9a of the display unit 9. FIG. 2 shows the outline and skeleton (wall) data of the architectural drawing recognized by the contour / skeleton recognition unit 7. It is an example of a display of the image data (recognition result) re-mapped by the mapping control part 8. FIG. In this display example, the double solid line indicates both surfaces of the wall, and the thin line indicates the core line of the wall and its extension line.
[0036]
FIG. 3 shows the image data of the architectural drawing (input image image data) skew-corrected by the automatic skew correction unit 5 and the wall image data and door image data, which are the remapped recognition results, at the same time. The example displayed together is shown. In this case, the image data of the architectural drawing that has been skew-corrected is displayed in halftone so that both can be easily identified (in FIG. 3, for convenience of illustration), the wall and the recognition result are displayed. The door image data is displayed as a solid line.
[0037]
Further, when the display unit 9 is a color display device, the distinguishability can be improved by changing and displaying the colors of both images. For example, the skew-corrected input image image data is light blue, and the recognition result image data is displayed in orange or green, so that the operator can easily identify the recognition result portion.
[0038]
Alternatively, the screen 9a of the display unit 9 can be divided and the skew-corrected input image image data and the recognition result image data can be displayed in comparison with the divided screens.
Alternatively, it is also possible to selectively display the input image image data subjected to skew correction and the image data of the recognition result remapped on the same screen. In that case, display selection means (keys or the like) may be provided in the operation input unit 10 described later.
[0039]
Further, the display unit 9 also displays a screen for the operator to confirm the recognition result. That is, the remapped image data of the recognition result is displayed, and a display such as “Are you sure about this recognition result? (YES / NO)” is performed. Thereby, the operator can confirm whether the wall and the door are correctly recognized.
[0040]
The operation input unit 10 is for inputting various operation instructions, function selection commands, editing data, and the like, and is a keyboard, a mouse, a touch panel, or the like.
The operation input unit 10 also has a function as display selection means, and can change the display state of the display unit 9 to a display state desired by the operator. For example, the input image data of the architectural drawing whose skew has been corrected and the image data of the recognition result that has been remapped can be superimposed and displayed by key operation, or only one of them can be selected and displayed.
[0041]
Further, the operation input unit 10 has an input means such as a key for inputting “YES” or “NO” information in response to the above display of “Are you sure about this recognition result? (YES / NO)”. . If “YES” is selected, the recognition process is terminated. If “NO” is selected, the process proceeds to a re-recognition process or a correction process. Thereby, the operator can confirm the content of the recognition result, and can confirm it.
[0042]
The external storage device 11 receives input image data, black dot number measurement array data created by the dot number measurement array data creation unit 6, wall data and door data recognized by the contour / skeleton recognition unit 7, and remapping control. This is a storage device for storing image data or the like as a recognition result remapped by the unit 8 in a storage medium such as a floppy disk (FD) or a magneto-optical disk (OMD) that can be taken out to the outside.
The printing device 12 is a printer or plotter that prints or draws the various data described above on paper.
[0043]
Here, the definitions of “contour” and “skeleton” and “outer wall” and “inner wall” in architectural drawings will be described with reference to Table 1, Table 2, and FIG.
[0044]
[Table 1]

[0045]
[Table 2]

[0046]
“Outline” means “outer contour”, and only the portion facing the outside of the outer peripheral wall as indicated by a circle in Table 1 (the outer line of the double line shown in FIG. 4A). The case 1 means a portion of (2) and the cases 2 and 3 mean the entire wall in contact with the outside (the portion indicated by a thick line in FIG. 4B).
[0047]
The “skeleton” means the case of cases 1 and 2 meaning all the walls (both shown by thick lines in FIGS. 4B and 4C) and the wall excluding the outer contour ((c) of FIG. There is a case 3 in which only the partition portion in the house indicated by a thick line in FIG. In these definitions, unless otherwise specified, it is treated as the normal meaning of case 1.
[0048]
“Outer wall” means only the portion facing the outside of the outer peripheral wall as indicated by a circle in Table 2 (the portion indicated by the outer line of the double line shown in FIG. 4A). And the case 4 which means the entire wall in contact with the outside (the portion indicated by the bold line in FIG. 4B).
[0049]
The “inner wall” is a wall excluding the outer wall (wall = outer wall + inner wall) for both cases 4 and 5, as indicated by a circle in Table 2, but in the case of case 4, the double wall shown in FIG. The part inside the line and the part indicated by a thick line in FIG. 4C, and in the case of the case 5, the part indicated by the thick line in FIG. 4C. Even in these definitions, unless otherwise specified, it is treated as the normal meaning of case 5.
[0050]
Next, a procedure for recognizing architectural drawings (mainly architectural drawings of houses, buildings, etc.) by the architectural drawing recognition apparatus shown in FIG. 1 will be described with reference to the flowcharts of FIGS. 5 to 8 and FIGS. 25 to 30. In FIG. 6 to FIG. 8, each step is indicated by “S”. In this embodiment, it is assumed that the architectural drawing to be recognized is a positive drawing.
[0051]
FIG. 5 is a flowchart showing a main routine of the architectural drawing recognition process. When the processing in FIG. 5 is started, the architectural drawing set in the image reading unit 2 in FIG. 1 is read, the image image data is input, and the skew is automatically corrected by the automatic skew correction unit 5.
[0052]
Then, in the subroutine of Step A, wall recognition processing is executed on the input image data by the dot number measurement array data creation unit 6, the contour / skeleton recognition unit 7, and the like. This process will be described in detail later. In addition, data obtained by encoding image image data may be input from the communication control unit 3.
[0053]
Thereafter, the door recognizing unit 13 extracts individual wall position information in order from the horizontal wall to the vertical wall in the processes of Steps B and C, and proceeds to the processes of Step D to Step K. When the extraction of the wall position information is finished in step B, the analysis of the drawing is finished, the process proceeds to step L, the door recognition result is output, and the process is finished.
Also in step C, after all the types of walls (horizontal wall and vertical wall) have been checked (wall number end), the process returns to step B.
[0054]
In step D, a rectangular area (region) having a unit length (scale size corresponding to 91 cm = half or 100 cm = 1 meter) in both the horizontal and vertical directions as a recognition range with reference to wall coordinates which are wall position information. Set on both sides of the wall.
In Step E, continuous black dot number measurement array data is created based on the wall coordinates of the target area (designated rectangular area). Here, the number of continuous black dots is the continuity of black dots with reference to the wall, but it is necessary to make a condition that includes a certain amount of dot skipping in consideration of blurring and the like.
[0055]
Then, in step F, the black dot count measurement array data is checked in order, and the process proceeds to step G only when it is within an allowable range as a door radius candidate (for example, ¼ or more of the unit length). Further, when the arrangement data is completed, the process proceeds to Step C where the next wall data is extracted.
[0056]
In step G, the method of extracting the shape is performed by checking the expression method of the door (the hinged door, the hinged window). At the tip of the door radius candidate (door plate portion), there is a rise of a 90 ° fan-shaped arc portion. This is measured and assumed to be a radius, and the rising of the other corresponding arc portion is analyzed to extract door shape data based on the center point / radius / door direction of the 90 ° fan shape.
[0057]
In step H, it is determined whether or not the door shape has been extracted. If it has been determined, the process returns to step I. If not, the process returns to step F to extract the next door radius candidate.
In step I, door recognition candidate shape analysis processing is performed based on the door shape extraction data. In other words, the shape analysis is performed by dot search of the fan-shaped arc portion based on the center shape / radius / door direction door shape data from step G.
[0058]
Not all door expression methods can be recognized completely automatically by this architectural drawing recognizing method and recognizing device, and one of the following users of the recognizing device is designated in advance.
[0059]
There are mainly the following three ways to express the door.
(1) 90 ° fan shape in one direction with shape contour
: Planar display as shown in (a) to (d) of FIG.
(2) Shaped contour with 90 ° fan shape in both directions
: Planar display as shown in FIGS. 29 (a) and 29 (b)
(3) A door with a shape outline that includes a door plate in the wall
: Planar display as shown in FIGS. 30 (a) and 30 (b)
[0060]
In step J, it is determined whether or not the result of the door shape analysis corresponds to one of these (OK). If OK, the process proceeds to step K. Return to Step F to extract door radius candidates.
In step K, the door shape data in the center point / radius / door direction is recognized as “door” information, information on the recognition result is stored, and the process returns to step F.
[0061]
The result recognized by performing the processes from step D to step K is output in step L and the process ends. In the process of Step L, the door position and size (shown by triangles in the figure) of the image data shown in FIG. 27 are stored in the memory 4 or the external storage device 11 as the door recognition result from the architectural drawing, or An output such as display on the display unit 9 is performed.
[0062]
FIG. 25 is a display example of image data obtained by re-mapping the input sample source drawing (image image) used for the process of recognizing the core of the wall and extracting the wall information from the result, on the display unit 9.
FIG. 26 is a diagram illustrating an example in which wall information serving as a reference for door recognition is displayed.
FIG. 27 is a diagram showing an example in which the recognition result door is displayed in a triangle.
[0063]
Here, the door recognition process described above will be described more specifically with reference to FIGS.
For example, when the position information (Wv is a part of the vertical wall) of the horizontal wall Wh as shown in FIG. 31 is extracted, the unit length (xu = A rectangular area with a scale size corresponding to yu = 91 cm or 100 cm) is set continuously over its entire length.
[0064]
Then, for each rectangular area, measurement array data (projection histogram) of the number of black dots continuous in the vertical direction for each dot width along the horizontal wall Wh, as in the case of the wall recognition process described later. Create vertical black dot count measurement array data.
If the peak value of the black dot number measurement array data is less than the unit length and falls within the allowable range of 1/4 or more of the unit length, it is regarded as a door radius candidate (indicated by a to f in FIG. 31).
[0065]
As a typical example of the door display in the architectural drawing, as shown in FIG. 32, a circle with a radius of r is divided into four circles by a horizontal line H and a vertical line V passing through the center P, and a quarter circle with an opening angle of 90 degrees. One of the four types of fan shapes.
And as shown to (a), (b) of FIG. 28, one radial line (radius line perpendicular | vertical to the wall 21) of each fan type | mold shows the opening position of the doorplate 20, and the other radial line ( The radius line parallel to the wall 21 indicates the closed position of the door plate 20, but the radial line at the closed position of the door plate 20 may not be shown as shown in FIGS.
[0066]
The radius r indicating the width of the doorboard is generally not more than a unit length (half or 1 meter) and not less than ¼ of the unit length (such as a bay window opening window). Accordingly, when the peak value of the continuous black dot number measurement array data is within the above-described allowable range, it is regarded as a door radius candidate.
[0067]
Next, the shape of the door recognition candidate is extracted. Therefore, as shown in FIG. 33, the base point at the wall position of the black dot number measurement array data regarded as each door radius candidate R is the center point P, and the length is set. Assuming that the radius is r, the presence or absence of an arc image is checked by pattern matching or the like in each of the sections D1 to D3 and D4 to D6 up to r / 3 widths along the wall on both sides.
As a result, when there is a continuous arc image A on the left side of the door radius candidate R, the door opens to the right, and when there is a continuous arc image A on the right side, the door opens to the left side. If there is no arc to be made, it is determined that it is not a door.
[0068]
For example, the a, b, e, and f door radius candidates shown in FIG. 31 are determined to be doors because an arc is found on either the left or right side, and the c or d door radius candidates are on either side. Since no arc is found, it is determined that it is not a door (actually a vertical wall). If the door is determined to be a door, the door shape data of the above-described 90 ° fan-shaped center point P / radius r / door direction (right open / left open) is extracted.
[0069]
Further, depending on the length (radius r), distance and opening direction of a plurality of door radius candidates, not only a single door as shown in FIG. 28 (including a single door) but also a parent and child as shown in FIG. A double door (including a double door) in which the pair of door plates 20a and 20b including the door (the widths of the pair of door plates 20a and 20b are different) can be discriminated.
[0070]
Furthermore, even if the portion of the door plate 20 that is a door radius candidate is included in the wall 21 as shown in FIG. 30, if it is confirmed that there is a continuous arc within the unit width on one side thereof. Can be recognized as a door.
Such door recognition processing is performed for all recognized wall candidates in the horizontal and vertical directions.
[0071]
Next, the wall position recognition processing in step A of FIG. 5 will be described.
FIG. 6 is a flowchart showing the contents of a subroutine of processing for recognizing the wall position. First, in step 3, the entire image data that has been subjected to automatic skew correction is set as an investigation target. In step 4, the nesting variable is 0 (initial value: meaning that the entire image is targeted). “Nested variable” represents a refinement stage when narrowing down the analysis range of architectural drawings from the top down. The larger the value of the nested variable, the deeper the analysis (the more advanced the details).
[0072]
In step 5, processing for limiting the region to be investigated for the wall position is performed. Specifically, an instruction for the investigation area is shown immediately before this processing, and only preparation for the processing of the subsequent steps 6 to 9 (commonization of the interface) is performed here. The shape of each investigation target area is a rectangular diagram. Since the nesting variable is 0 at first, the entire drawing is examined.
[0073]
In step 6, horizontal black dot number measurement array data is created. This counts (measures) the number of black dots in the horizontal direction for each dot width in the vertical direction of the image data, and holds each count data.
Next, in step 7, wall extraction (recognition) processing is performed based on the horizontal black dot number measurement array data created in step 6. The processing procedure will be described later.
[0074]
In step 8, as in step 6, black dot count measurement array data in the vertical direction is created. That is, the number of black dots in the vertical direction is counted (measured) for each dot width in the horizontal direction of the image data, and each count data is held. In step 9, wall extraction (recognition) processing is performed based on the black dot count measurement array data in the vertical direction.
[0075]
In step 10, it is determined whether there is a wall candidate for dividing the region. If there is at least one wall candidate that divides the region in the horizontal or vertical direction, the process proceeds to step 11, and if there is no such wall candidate, the process proceeds to step 13.
For example, it is determined whether there is a wall candidate Wd that divides the region in the investigation target region Sa as shown in FIG.
In step 11, the nested variable is incremented by 1 and reset. This indicates that a wall is recognized from the current analysis region, and the analysis range is limited to a small region within the current region newly delimited using the wall.
[0076]
In step 12, the area to be investigated is subdivided. Specifically, the wall candidate core lines (center lines) recognized in step 7 and step 9 are areas indicated by the core lines indicated by the thin lines of the wall candidates W and Wd, for example, as shown in FIG. Divide into Sa1 and Sa2. Further, the investigation target is positioned in the first subdivided area (for example, the upper left area).
[0077]
The order of analysis in the newly subdivided area group does not require a special order. For example, it may be performed in the order of small x and y coordinate values of the area start position. It is done. FIG. 9B shows an example of the nest No. of each region subdivided by the wall candidate and the analysis processing order thereof, and the solid line indicates the core line (center line) of the wall candidate, and (1) and (2) , {Circle over (3)} indicate a nest No., and small numbers 1 to 8 indicate the processing order.
[0078]
Thereafter, the process returns to step 5 in FIG.
In step 10, if there is no wall candidate to divide the area, the process proceeds to step 13 to check whether there is any remaining homogeneous nest area (area of the same nest No. in FIG. 9B). to decide. If there is a remainder, the investigation target is advanced to the next area of the homogeneous nest in step 15 and the process returns to step 5.
[0079]
If there is no remaining homogeneous nest area, the process proceeds to step 14 to determine whether the nested variable is 0 or not. If the nesting variable is not 0, the nesting variable is decremented by -1 in step 16 and the process returns to the processing of the nesting area one level higher. In step 13, it is determined whether there is a remaining nesting area in that stage. If there is, the investigation object is advanced to the next nested area in step 15 and the process returns to step 5.
[0080]
If the nesting variable is 0 in step 14, the process proceeds to step 17 to determine the position and size of each wall based on the recognition data of the wall candidate for each area obtained in steps 7 and 9, and the wall The recognition data is stored in the memory 4. The storage method will be described later. After the above processing, the subroutine ends.
[0081]
Next, based on FIG. 7, the procedure of the wall extraction (recognition) processing in steps 7 and 9 of FIG. 6 will be described. Prior to that, from the image data of the architectural drawing (house floor plan) and its entire area, Specific examples of the created horizontal and vertical black dot number measurement array data are shown in FIGS. In these drawings, (a) is architectural drawing image data, (b) is horizontal black dot number measurement array data, and (c) is vertical black dot number measurement array data. Further, the vertical black dot number measurement array data shown in FIG. 10C is enlarged and shown in FIG.
[0082]
In the architectural drawing of FIG. 10, the symbol of the wall is represented by both sides of the wall and the core wire, and in the architectural drawing of FIG. 11, the symbol of the wall is represented by a fill (black) within the thickness of the wall. .
In this black dot count measurement array data, the width of the thinnest black line is one dot width, and the length of each black line is horizontal or vertical for each dot width of the architectural drawing image data shown in (a). This corresponds to the count value of the number of black dots in the direction.
[0083]
As is clear from these figures, most of the line segments (90% or more) constituting the architectural drawing are drawn in the horizontal direction or the vertical direction, and the density of black dots is particularly high in the wall portion. . For this reason, in the black dot count measurement array data in the horizontal direction and the vertical direction, a peak appears at the position where the wall exists.
[0084]
When the processing of the flow of FIG. 7 is started, first, in step 21, the direction crossing the specified direction (vertical direction is specified if black dot number measurement array data creation in the horizontal direction is specified, and horizontal direction is specified if the vertical direction is specified. In the direction), it is confirmed how many times the wall recognition processing loop is possible at predetermined intervals corresponding to the standard unit length of the construction drawing.
Here, in the case of a general house, this unit length is a half or 1 meter which is the minimum wall distance, and here it is a half (91 cm).
[0085]
For the horizontal black dot count measurement array data, the width L is set to a size slightly longer than the vertical image width, and for the vertical black dot count measurement array data, it is slightly larger than the horizontal image width. A long dimension is automatically set as a width size L.
In addition, the half size is set to h, and this value is determined by inputting scale data and half length of the drawing in advance or by automatic calculation by calculation.
The width size L and the half size are calculated from h to L / h, and the numerical value rounded up after the decimal point is set as the number n of executions of the large loop of the wall recognition processing. A range indicated by h in FIG. 12 is a processing range in one large loop.
[0086]
Next, in step 22, the count value i of the large loop number counter is initialized (i ← 1).
In step 23, it is determined whether or not the count value i of the number counter exceeds the number n of possible large loop executions (i> n). If so, the analysis of the area is terminated. If not, in step 24, it is positioned at the highest peak (indicated by P in FIG. 12) as the first analysis processing within the i-th half, and that point is designated xp. In this way, by performing the analysis process every half, the possibility that the peak value therein is a part of the wall is high.
[0087]
Next, in step 25, it is determined whether or not the height of the peak (maximum value), which is the first condition as the object of the wall, is equal to or greater than half (h). As a result, when the height of the peak is less than half, it is determined that the wall is unknown, and the process proceeds to step 34 in order to proceed to the next half-half ahead analysis. When the peak height is half or more, the process proceeds to the next analysis step 26. In this step 26, both the left and right sides (for example, 2 W width) are examined from the position of the highest peak, and the range of wall thickness (both positions: W ₁ , W ₂ And thickness We = | W ₁ -W ₂ |) Is narrowed down. Here, W means the wall thickness, and is used with the same value as Wmax, for example.
[0088]
As a method of narrowing down, the peak position on both ends or one side of the maximum peak position having a value of (maximum peak value−min) * rate + min or more is set to the position W on both sides of the wall. ₁ , W ₂ Or the highest peak position xp is the position W on one side of the wall. ₁ The position W of the other surface when ₂ There is a way to narrow down.
[0089]
FIG. 13 is an explanatory diagram of the narrowing-down process. FIG. 13A shows the positions W on both sides of the wall on both sides of the maximum peak position xp. ₁ , W ₂ This is an example in the case where exists. In this case, the highest peak position xp is estimated as a candidate for the wall core line position.
[0090]
FIG. 13B shows that the highest peak position xp is the position W of one surface of the wall. ₁ The position W of the other surface on one side ₂ This is an example in the case where exists. In this case, it can be estimated that the longer peak position is a candidate for the outer wall position of the outer contour indicated by a double line in FIG. 2, and the shorter peak adjacent thereto is a candidate for the inner wall position.
[0091]
Here, rate is set to be smaller in the first half of the analysis (when the value of the nested variable is small) and larger in the second half (when the value of the nested variable is large) (for example, initially rater = 0.70). As shown in FIG. 13 (c), min is the minimum value of the black dot number measurement data within the width of about 4W widened to the left and right sides 2W of the position xp of the maximum peak P of interest.
[0092]
In this manner, the validity of the wall is confirmed in steps 27 and 28 based on the result narrowed down in step 26 of FIG.
First, in step 27, it is determined whether or not the wall thickness We exceeds the maximum value Wmax (for example, 30 cm). If it exceeds, the wall is unknown and the process proceeds to the next halfway analysis. Proceed to If not, the process proceeds to step 28 to determine whether or not the wall thickness We is less than the minimum value Wmin (for example, 2.5 cm).
[0093]
As a result, if the wall thickness We is less than the minimum value Wmin, the process proceeds to step 30; otherwise, the process proceeds to step 29. In step 30, the information is obtained on the assumption that peaks other than walls (for example, tatami mats, windows, sliding doors, etc.) are recognized.
In step 29, whether or not a wall condition is satisfied is determined from a candidate region of the wall by a bisection search method described later, and what can be recognized as a wall is determined on both sides of the wall in step 31. Position W ₁ , W ₂ Then, information such as the thickness We is saved (stored), and the process proceeds to Step 32.
[0094]
If the condition as a wall is not satisfied in step 29, the process proceeds to step 32. In step 32, the double-sided position (coordinates) W of the wall ₁ And W ₂ Are both inside the currently processed area, and if inside, proceed to step 33. Otherwise, go to step 34.
[0095]
In step 33, the current processing area is subdivided and the boundary data indicating the new subdivision area is defined as W ₁ , W ₂ , We are saved (stored). Also, (W ₁ + W ₂ ) / 2 is the core line position of the wall. In step 34, the count value i of the large loop number counter is incremented by 1 in order to perform the next halfway ahead process, and then the process returns to step 23 to repeat the above-described process.
[0096]
In this way, analysis processing is performed every half from one end (first element) of the target black dot number measurement array data, and when the analysis is completed up to the opposite end, the analysis of the current range is ended. The horizontal direction and the vertical direction are analyzed separately for the black dot count measurement array data, and the next analysis range is limited to a range that can be recognized as a wall in the current analysis.
Therefore, the round robin cases are divided by combining the results of analyzing the horizontal and vertical black dot count measurement array data.
[0097]
In this embodiment, it is considered that the core positions of the walls are arranged so that the interval between the adjacent walls is an integral multiple of a half-half unit.
The range up to which a characteristic peak can be found is effectively used as analysis data. In other words, if no feature corresponding to the wall within a certain range to be analyzed is found in both the horizontal direction and the vertical direction, the analysis is terminated. The analysis range for one time is the entire drawing at first, and then the method is limited to the range delimited by the walls that have been discovered, making it easy to find the peaks of the walls and to analyze the details in detail. To be.
[0098]
Next, the bisection search method for determining whether “the condition as a wall is satisfied” in step 29 of FIG. 7 will be described with reference to the flowchart of FIG.
When the processing of the flow shown in FIG. 8 is started, first, in step 41, an initial setting of the bisection search method is performed.
That is, the number n of area dividing elements for wall analysis is set to 1, k indicating the smallest No. of the array elements of the dividing elements whose wall or non-wall is undetermined is set to 0, and As the divided array elements, the input start and end image addresses are assigned to S [0] and E [0], respectively, to be initialized.
[0099]
Thereafter, the process proceeds to step 42. If the start address S [k] and the end address E [k] of the investigation area are equal, the process proceeds to step 53, and if not equal, the process proceeds to step 43. In step 43, the target region of the image image data is divided to divide the original region into two equal parts, excluding the decimal error.
[0100]
That is, the start and end image addresses S1 and E1 of the first half area to be divided are S1 = S [k], E1 = (1/2) (S [k] + E [k]), and the start and end of the second half area are set. The end image addresses S2 and E2 are set as S2 = (1/2) (S [k] + E [k]) + 1, E2 = E [k].
[0101]
Thereby, as shown in FIG. 14, for example, in the image image data of the construction drawing, the original region width of the line where the wall candidate exists (indicated by the alternate long and short dash line) is initially divided into two equal parts at the position of 1. Thereafter, until the presence or absence of the wall can be discriminated, step 44 and step 44 are performed in the subdivided region by repeating the second equal division such as the second time at the position 2 shown in FIG. 14, the third time at the position 3, and the fourth time at the position 4. Do 45 wall surveys.
In steps 44 and 45, it is checked whether or not the regions P1 and P2 divided into two equal parts are filled with walls, and the process proceeds to step 46.
[0102]
Here, as a result of analyzing the black dot count measurement array data in the specified area (from the start address to the end address), the height of the black dot peak (maintaining the spread of the wall thickness) is the height of the specified area. Compared with the width (length), the following determinations (1) to (3) are made.
[0103]
(1): When 5% or less, v = 0: non-walled
(2): When 95% or more, v = 1: determined as a wall
(3): Other than the above, v = 2: Neither can be judged
This is shown by (1), (2), and (3) in FIG.
[0104]
In step 46, if it is not possible to determine whether or not a wall exists in the first half region P1 divided in step 43 (v = 2), the process proceeds to step 47. Otherwise, the process proceeds to step 49. In step 47, the storage array elements S [k], E [k], v [k] in the area range before dividing in step 43 are used as the latter half area data (start and end image addresses divided in step 43). Replace with S2, E2 and determination result v2).
[0105]
In step 48, the first half area data (start and end image addresses S1, E1 and determination result v1) divided in step 43 are used as new storage array elements S [n], E [n], v [n]. Evacuate and proceed to step 51.
In step 49, the storage array elements S [k], E [k], v [k] in the area range before dividing in step 43 are used as the first half area data (start and end image addresses divided in step 43). Replace with S1, E1 and determination result v1).
[0106]
In step 50, the second half area data (start and end image addresses S2, E2 and determination result v2) divided in step 43 are used as new storage array elements S [n], E [n], v [n]. Evacuate and proceed to step 51.
In step 51, the variable n indicating the new storage array element No. is incremented by 1 so as to indicate the new storage array element of the area address, and then the process proceeds to step 52.
[0107]
In step 52, if the classification code v in the k element indicating the smallest No. of the divided elements whose wall or non-wall is undetermined is a content that does not know whether a wall exists (v = 2) Returns to step 42 and repeats the subdivision process. Otherwise, go to step 53.
In step 53, assuming that one content that does not know whether a wall exists is solved, the index k is incremented by 1, and then the process proceeds to step 54.
[0108]
In step 54, whether or not the number n of area dividing elements for wall analysis is equal to k indicating the smallest No. of the array elements of the dividing elements for which the wall portion or the non-wall portion is undetermined. If it is determined that they are equal, it is determined that the dividing process for wall recognition has been completed, and the process proceeds to step 55. If not equal, return to step 52.
[0109]
In step 55, sorting is performed in the order of start addresses (ascending order) so that the divided area address data (array) are arranged in ascending order, and the process proceeds to step 56. That is, the data of S [0] to S [n-1], E [0] to E [n-1], v [0] to v [n-1] is sorted in ascending order using S [] as a key. To do.
[0110]
In step 56, when the wall portion and the non-wall portion are continuous, the respective reduction processes (represented by one range) are performed, and the process ends. That is, when the continuous v [] values are 0 or 1, the S [], E [] data are compressed.
At this time, when an array element having contents (v = 2) that do not know whether a wall exists is included, if the elements before and after the element indicate a wall, change to wall data, If it indicates a non-wall, it is changed to non-wall data and processed.
[0111]
FIG. 16 shows an analysis example of the wall position in the image data by the above-described bisection search process. In FIG. 16A, a wall image (shaded portion) W and its investigation target area are indicated by broken lines, and this investigation area is along the previously recognized position of the wall candidate. Is set. S [0] = 0 is the first start address of the investigation area, and E [0] = 15 is the first end address. Numbers in the middle of 4, 5, 7, etc. are start or end addresses (both are image addresses) of the target area after division.
[0112]
FIG. 16B shows the determination result v regarding the start address S, the end address E, and the presence / absence of the wall in each investigation stage of the variable n = 1 to 10 and its determination status, sorting status, and degeneration processing. The results are shown together with the value of k. Finally, it is recognized that a wall exists at the image addresses 5 to 12.
[0113]
Next, a wall recognition example of a simple architectural drawing will be described with reference to FIG.
FIG. 17 shows a nested variable (nest) and an actual wall state recognized as a region division state.
First, with the nesting variable = 0, the entire architectural drawing is extracted as a wall position investigation target. As a result, it is assumed that the outline of the house part of the architectural drawing and the positions of the horizontal and vertical wall candidates can be recognized as indicated by a solid line in FIG. However, of the recognized wall candidates, the actual wall is only the part shown in FIG.
[0114]
Therefore, the nest variable = 1 is set, and the wall is extracted by limiting the investigation target area for each rectangular area divided by the core lines of the recognized wall candidates shown in FIG. As a result, a new wall candidate indicated by a thick line is recognized in the four investigation target areas indicated by (1), (2), (3), and (4) in (B), and the previously recognized wall candidates Of these, it is confirmed that the actual wall was only the part shown in (a), and the state of the wall shown in (b) is recognized.
[0115]
Further, with the nesting variable = 2, the four regions {circle around (1)} to {circle around (4)} in which the new wall candidates are recognized in (B) are divided by the newly recognized walls, respectively, and the investigation target region is further divided. Limited wall extraction. As a result, new wall candidates indicated by bold lines are recognized in the two investigation target areas indicated by (5) and (6) in (C), and the wall state shown in (c) is recognized.
[0116]
After that, the nested variable = 3 is set, and the two regions (5) and (6) in which the new wall candidate is recognized in (C) are divided by the newly recognized walls, respectively, and the investigation target region is further divided. Limited wall extraction. As a result, if a new wall candidate could not be extracted in any of the divided areas, the wall position investigation is ended thereby, and it is determined that the wall position shown in (c) is the final wall recognition result. The data is stored in the memory.
[0117]
In this manner, the wall is extracted by subdividing the investigation target area until no new wall candidate is extracted in any of the divided investigation target areas. As a result, even a small wall can be reliably recognized, and a portion where a wall actually exists and a portion where a wall actually does not exist can be accurately determined.
[0118]
Here, an example of a concrete architectural drawing recognition processing procedure according to the flowchart of FIG. 6 described above will be described with reference to FIGS. 18 to 20 are a series of drawings, but are divided into three drawings for convenience of illustration. S3 to S17 in these drawings correspond to the steps S3 to S17 in FIG. In addition, an area division diagram and a real wall state at each stage are also illustrated.
[0119]
In the following description, step is abbreviated as “S”. In S3 of FIG. 18, the entire drawing is set as the object of investigation, and the nested variable is set to 0 in S4. Although the investigation target is limited in S5, since the nested variable is 0, the entire drawing is also the investigation target. In S6 to S7, a horizontal black dot number measurement array data is created and a wall is extracted, but a wall other than the outline cannot be found, and in S8 to S9, a vertical black dot number measurement array data is created and a wall is created. Is extracted and two walls other than the contour are found. Therefore, the answer is YES in S10, the nesting variable is set to 1 in S11, the area is subdivided (at the position of each wall) in S12, the process returns to S5, and the investigation target area is limited to the leftmost area.
[0120]
Then, in S6 to S7, two walls were found. In S8 to S9, the wall was not found, but in S10, the answer is YES. In S11, the nested variable is set to 2 and “nesting processing” is performed. The process is repeatedly executed until the nesting process is completed, and wall extraction processing is sequentially performed on three divided areas obtained by dividing the vertically long area on the left side by two newly found walls.
[0121]
In this example, no new wall is found, and the nesting variable is set to -1 and returned to 1 in S16 of FIG. 19, and the wall is extracted in the same manner with the middle vertically long region as the investigation target region. So no new wall candidates are found.
Therefore, in S15 and S5, the investigation target area is changed to the vertically long area on the right side, and one wall is found in S6 to S7.
[0122]
Therefore, the answer is YES in S10 of FIG. 20, the nested variable is set to 2 in S11 and “nesting processing” is started, and the vertically long area on the right side is divided by the newly found wall, and the wall extraction of each divided area is performed. Processing is performed sequentially. As a result, no new wall is found in any of the divided regions, and NO is determined in S13 and the nesting process is terminated. In S16, the nested variable is set to -1 to 1, but the region of the nested variable 1 does not remain. Therefore, the nested variable is returned to 0, but the area does not remain. Therefore, it is determined that the wall extraction process has been completed, the position and size of each wall are determined based on the wall candidate recognition data extracted in S17, and the data is stored in the memory.
[0123]
Next, information (analysis result data) such as the position and size of the wall corresponding to the outline and skeleton of the architectural drawing recognized as described above is stored in the storage medium of the memory 4 and the external storage device 11 shown in FIG. An example of the contents to be described will be described with reference to FIG.
In FIG. 21, (A) is the number of nests, (B) is specific nest information, (C) horizontal wall information, (D) is vertical wall information, (E) is a horizontal wall model, (F) ) Indicates a vertical wall model.
[0124]
The number of nestings indicates the depth of nesting (parent-child relationship) of child nesting pointers. When no wall is recognized, the number of nestings = 0.
The unique nest information includes a nest No., a NEXT sibling pointer, a child nest pointer, the number of walls (horizontal and vertical directions), and a wall information pointer (horizontal and vertical directions).
[0125]
The NEXT sibling pointer has the next address of the simultaneous hierarchy nesting information. Therefore, the nest No. in the unique nest information at the location indicated by this pointer is the same value as that of the process.
The child nesting pointer has the address of the top data of hierarchical nesting information one level below. Therefore, the nest No. in the specific nest information at the location indicated by the pointer is a value obtained by adding 1 to the value of the process.
[0126]
The horizontal wall information shown in (C) is stored with the address indicated by the horizontal wall information pointer in the unique nest information shown in (B) as the head address.
The wall information includes a NEXT pointer indicating the position of the start address of the next horizontal wall information, wall start point coordinates (x coordinates: a, y coordinates: b), wall horizontal (x direction) size: c, wall Vertical (y direction) size: It consists of wall thickness d. By these a to d, the model of the horizontal wall shown in (E) can be stored and reproduced.
[0127]
Similarly, the vertical wall information shown in (D) is stored with the address indicated by the vertical wall information pointer in the unique nest information as the head address.
The wall information includes a NEXT pointer indicating the position of the start address of the next vertical wall information, wall start point coordinates (x coordinates: e, y coordinates: f), wall vertical (y direction) size: g, wall Horizontal (x direction) size: wall thickness h. By these e to h, the model of the vertical wall shown in (F) can be stored and reproduced.
[0128]
By the way, an example of creating black dot count measurement array data in the horizontal direction and the vertical direction with the entire drawing as an investigation target area is shown in FIG. 10 and FIG. In the next stage of recognizing wall candidates based on the dot count measurement array data, the area of the drawing is divided by the recognized wall candidates, and the number of black dots in the horizontal and vertical directions is measured based on the image data in which the investigation target area is limited. Examples of creating sequence data are shown in FIGS.
[0129]
FIG. 23 and FIG. 24 are examples in which the investigation target area is further subdivided from FIG.
In this way, until no new wall candidate is found, the investigation target area is subdivided, horizontal and vertical black dot number measurement array data based on the image data is created, and the walls are extracted.
[0130]
In this embodiment, since the architectural drawing of the positive image is a recognition target, the number of black dots in the horizontal and vertical directions of the binarized image data is measured (counted) to create black dot number measurement array data. However, when the architectural drawing of the negative image is to be recognized, if the white dot count measurement array data is created by measuring (counting) the number of white dots in the horizontal and vertical directions of the binarized image data, Wall recognition can be performed in the same way.
[0131]
In addition, recognition data regarding the outline and skeleton of the architectural drawing recognized as described above can be converted into CAD vector data, and compatibility of CAD data between different models can be obtained.
[0132]
Then, in the case of a flat display door as shown in FIG. 28 or FIG. 29, in this embodiment, the number of continuous dots having a size corresponding to the door plate portion is extracted from the shape outline made of black dots, and the number of continuous dots is calculated. Assuming that the door radius is a candidate, check its validity and recognize the door. It is conceivable that this can be further linked to the joinery recognition process for the entire drawing.
[0133]
Further, with respect to the flat display door as shown in FIG. 30, paying attention to the leading end of the continuous black dots, and measuring the rising edge of the 90 ° fan-shaped arc part, the door plate part included in the wall is Since the extraction process is performed and the recognition process is performed without a slight misalignment or an accurate 90 ° fan-shaped description, the following advantages are obtained. If the dot structure is thin and weak, the door can be recognized by extracting the number of consecutive dots after performing image expansion focusing on black or white dots in the area.
[0134]
(Advantage 1) Based on the continuous dot number measurement arrangement data taking the size into consideration, the description of the door is made by measuring two 90 ° fan-shaped arc rising portions and recognizing the arc description therein. Recognized surely.
(Advantage 2) Even in drawings that are described in freehand without using a ruler or the like, if the fluctuations of straight lines including not only horizontal and vertical directions but also diagonal straight lines are offset, the shape contours are aligned in the horizontal and vertical directions. You can recognize it if it ’s good.
[0135]
When it is difficult to correct the skew of the architectural drawing in the front direction in the handwritten drawing, the handwritten drawing may be described on a graph paper, and the skew correction may be performed using the grid of the graph paper.
In addition, if it is unavoidable to leave it to the user's manual work, it is advisable to attach a manual that urges attention to create a confronting drawing that does not require skew correction as much as possible when reading the scanner.
[0136]
【The invention's effect】
As described above, according to the method of recognizing a door from an architectural drawing according to the present invention and the apparatus for recognizing a door, a handwriting that has been described in which a line that has been difficult to accurately recognize is interrupted or a straight line is slightly inclined is described. Even from drawings with relatively low contrast, such as blueprints, noisy drawings, old architectural drawings, and other architectural drawings with poor description and image quality, the doors (open doors) together with the position of the walls that form the outline and skeleton , Open window) shape and type can be easily and accurately recognized.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a schematic configuration of an example of an architectural drawing recognition apparatus that implements a method for recognizing a door from an architectural drawing according to the present invention.
2 is a diagram illustrating a display example of image data of a recognition result that is remapped on the display unit 9 of FIG. 1;
FIG. 3 is a diagram showing an example in which input image data of an architectural drawing that has also been skew-corrected is displayed superimposed on image data of a recognition result.
FIG. 4 is a diagram for explaining definitions of an outer contour, an outer wall, an inner wall, and a skeleton in an architectural drawing.
FIG. 5 is a flowchart showing a main routine of architectural drawing recognition processing by the architectural drawing recognition apparatus shown in FIG. 1;
6 is a flowchart showing details of a subroutine for wall position recognition processing in step A in FIG. 5; FIG.
FIG. 7 is a flowchart of a subroutine of wall extraction (recognition) processing of steps 7 and 9 in FIG. 6;
FIG. 8 is a flowchart for executing a bisection search method for recognizing a wall position.
FIG. 9 shows a wall candidate in the investigation target region, a region division example based on the wall candidate, and a nesting No. of a region group subdivided by the wall candidate. It is explanatory drawing which shows an example of the analysis processing order.
FIG. 10 is a diagram showing a specific example of image data of an architectural drawing (house floor plan) and black dot number measurement array data in the horizontal and vertical directions created from the entire area thereof.
FIG. 11 is a diagram showing another specific example.
12 is an enlarged view of the vertical black dot number measurement array data shown in FIG. 10C. FIG.
13 is an explanatory diagram of processing for narrowing both sides of a wall to both sides or one side of a peak in step 26 of FIG. 7;
14 is an explanatory diagram of region width division processing in step 43 of FIG. 8;
FIG. 15 is an explanatory diagram of determination by wall survey of the target area in steps 44 and 45 in FIG. 8;
16 is an explanatory diagram showing an example of analysis of a wall sample (wall candidate) by the two-division search process shown in FIG. 8. FIG.
FIG. 17 is an explanatory diagram of a wall recognition example of a simple architectural drawing according to the present invention.
FIG. 18 is an explanatory diagram showing an example of a concrete architectural drawing recognition processing procedure according to the flowchart of FIG. 5;
FIG. 19 is an explanatory diagram continued from FIG. 18;
FIG. 20 is an explanatory diagram continued from FIG. 19;
FIG. 21 is an explanatory diagram showing an example of the contents of analysis result data stored in a memory.
22 is a diagram showing an example of image data in which a survey target area of the architectural drawing shown in FIG. 10 is limited and horizontal and vertical black dot number measurement array data based on the image data.
FIG. 23 is a diagram showing an example of image data further limiting the investigation target region from FIG. 22 and horizontal and vertical black dot number measurement array data based on the image data.
FIG. 24 is a diagram showing an example of image data of another part in which the investigation target region is further limited from FIG. 22 and horizontal and vertical black dot number measurement array data based on the image data.
25 is a diagram showing a display example of input image data of an architectural drawing for recognition of a mapped door on the display unit 9 of FIG.
FIG. 26 is a diagram showing an example in which recognition results of the outer wall of the house and the partition wall in the house are displayed for the input image data.
27 is a diagram showing an example in which image data of recognition results of the outer wall and the partition wall in the house and the door are superimposed on the input image data of FIG. 25 and displayed.
FIG. 28 is a diagram showing an example of a planar display of a single door in an architectural drawing.
FIG. 29 is a diagram showing an example of a planar display of a double door in an architectural drawing.
FIG. 30 is a diagram showing an example of a flat display in which a door plate portion of a single door in an architectural drawing is included in a wall.
FIG. 31 is an explanatory diagram of a method for searching for a door radius candidate along a horizontal wall.
FIG. 32 is an explanatory diagram for describing types of general door flat display;
FIG. 33 is an explanatory diagram of a process of searching for an arc image within a unit length range on both sides of a door radius candidate.
[Explanation of symbols]
1: Overall control unit (CPU) 2: Image reading unit
3: Communication control unit 4: Memory
5: Automatic skew correction unit
6: Dot number measurement array data creation unit
7: Outline / skeleton recognition unit 8: Re-mapping control unit
9: Display unit 10: Operation input unit
11: External storage device 12: Printing device
13: Door recognition unit 14: Bus

Claims (8)

The area consisting of black or white dots of the image image data obtained by reading the image of the architectural drawing is limited to the unit-length rectangular areas on both sides based on the positional information of the outer wall of the house and the wall as a partition in the house. The continuous dot number measurement array is created by measuring the number of black or white dots that are continuous in the direction perpendicular to the wall in the image data within each limited rectangular area, and creating the continuous dot number measurement array data. Candidate door radii are extracted based on the data, the presence / absence of an arc image in the unit length range on both sides of each door radius candidate is determined, and the door is determined based on the position of the arc image and the length of the door radius candidate. A method for recognizing a door from an architectural drawing, characterized by recognizing the shape and type of the door. In the method of recognizing a door from the architectural drawing of Claim 1,
The number of black or white dots in the horizontal and vertical directions of the image image data obtained by reading the image of the architectural drawing is measured to create the horizontal and vertical dot number measurement array data, and the created dot number measurement array data in both directions. A method of recognizing a door from an architectural drawing, comprising recognizing positional information of the wall of the architectural drawing based on the architectural drawing. Image data input means for inputting image image data obtained by reading an image of an architectural drawing;
Dot number measurement array data creating means for measuring the number of black or white dots in the horizontal and vertical directions of the image image data input by the means to create dot number measurement array data in the horizontal and vertical directions;
Contour / skeleton recognition means for recognizing positional information of the outer wall of the house and the wall as a partition in the house based on each dot count measurement array data created by the means;
With reference to the position information of the wall recognized by the means, the area consisting of black or white dots of the image image data is limited to unit-length rectangular areas on both sides of the wall, and each of the limited rectangular areas In the image image data, the number of continuous black or white dots in the direction perpendicular to the wall is measured to create a continuous dot number measurement array data, and a door radius candidate based on the generated continuous dot number measurement array data Is extracted, the presence / absence of an arc image within the unit length range on both sides of each door radius candidate is determined, and the shape and type of the door are recognized based on the position of the arc image and the length of the door radius candidate. Door recognition means;
The apparatus which recognizes a door from the architectural drawing characterized by providing. 4. The apparatus for recognizing a door from an architectural drawing according to claim 3, wherein the image data input means is an image reading means for reading an image of the architectural drawing and inputting the image image data. The image data input means is image data input means for inputting encoded image data obtained by encoding a run length of image image data obtained by reading an image of an architectural drawing, and the dot number measurement array data creation means is the image It is a means for measuring the number of black or white dots in the horizontal direction and the vertical direction of the original image data from the encoded image data input by the data input means, and creating dot number measurement array data in the horizontal direction and the vertical direction. The apparatus which recognizes a door from the architectural drawing of Claim 3 characterized by the above-mentioned . 6. The apparatus for recognizing a door from an architectural drawing according to claim 5, wherein the image data input means is image data receiving means for receiving and inputting the encoded image data by communication. The apparatus for recognizing a door from the architectural drawing according to any one of claims 3 to 6, wherein a display means for displaying a recognition result by the door recognizing means is provided, and the door is recognized from the architectural drawing. apparatus. 8. The door is recognized from an architectural drawing according to claim 7 , wherein the display means is a means for displaying the recognition result by the door recognition means in a superimposed manner with the image image data input by the image data input means. apparatus.

Images