JP2004178358A - Method and equipment for guarding and watching event - Google Patents

Method and equipment for guarding and watching event Download PDF

Info

Publication number
JP2004178358A
JP2004178358A JP2002345156A JP2002345156A JP2004178358A JP 2004178358 A JP2004178358 A JP 2004178358A JP 2002345156 A JP2002345156 A JP 2002345156A JP 2002345156 A JP2002345156 A JP 2002345156A JP 2004178358 A JP2004178358 A JP 2004178358A
Authority
JP
Japan
Prior art keywords
people
security
data
inflow
flow
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
JP2002345156A
Other languages
Japanese (ja)
Inventor
Tomoaki Hirashima
倫明 平嶋
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Meidensha Corp
Meidensha Electric Manufacturing Co Ltd
Original Assignee
Meidensha Corp
Meidensha Electric Manufacturing Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Meidensha Corp, Meidensha Electric Manufacturing Co Ltd filed Critical Meidensha Corp
Priority to JP2002345156A priority Critical patent/JP2004178358A/en
Publication of JP2004178358A publication Critical patent/JP2004178358A/en
Withdrawn legal-status Critical Current

Links

Images

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method and equipment for guarding and watching an event, which quantitatively judge and implement traffic control, consider a passage route and space, control a flow of people, and draft a guarding plan by history data, actual observation data, forecast data, etc. without simply depending on an organizer's experience and gut feeling. <P>SOLUTION: The invention is characterized in that a camera is installed in a place closely related to a turnout in a target guard area such as an event site and a passageway in an arbitrarily set cycle during an arbitrarily set period, that a flow of people in the place is measured by processing a picture captured by the camera, and that the flow of people in each place and the turnout in the target guard area are systematically forecast from the forecast inflow and outflow data of people in the inflowing and outflowing place such as public transportation facilities directly related to the turnout. <P>COPYRIGHT: (C)2004,JPO

Description

【0001】
【発明の属する技術分野】
本発明は、イベント警備監視方法及びイベント警備監視装置に関する。具体的には、イベント等の行事において、画像処理を利用して、防犯乃至警備監視を行う装置及び方法に関する。
【0002】
【従来の技術】
現在、全国各地で国、県、市町村主催のお祭り、花火大会、パレードや、民間主催のイベント、コンサート、各種スポーツ大会等が春夏秋冬日夜を問わず数多く行われている。
これらの行事は、大勢の人が訪れ群衆が集まることとなり、主催者、警察、警備会社等(以下、主催者側と略す)の警備計画や努力にも関わらず、予想外の人出や出来事によりある特定の区域での群衆密度が安全な限界を超え、人身事故につながるケースがある、或いは発生する危険性を秘めていることは否定できない。
【0003】
主催者側はこれら危険性の回避のために、人出を予想し、警備計画を立て、警備本部を設置し、各要所への警備員を配置してイベント会場周辺の交通規制、歩道の確保を行い、更に人の流れ(流入出速度)を無線で連絡を取りながら制御調整している。
更に、狭い通路等の危険な区域には防犯カメラシステムを設置して、各地点の人出の状況を警備本部で監視したり、画像処理装置で画面上を通過する人間を検出し、人数の自動計測を行い人の流れの制御に使用したりしている場合もある。
【0004】
学会等の研究報告では、調査場所にカメラを設置して交差点や通路を撮影しその映像を画像処理することにより道路走行車両の検出と追跡計数や歩行者の検出追跡を行うアプローチが多く報告されている。
例えば、高速道路走行車両の検出については非特許文献1、歩行者の挙動検出についての研究報告は非特許文献2に記載されている。
【0005】
【非特許文献1】
小沢慎治、“ITS道路画像における認識と理解”、電子情報通信学会技術研究報告、PRMU98−91〜105,pp.99−104.1998.
【非特許文献2】
竹内、金子、五十嵐、佐藤、羽根、”ロバスト背景差分及び領域抽出に基づく歩行挙動の画像解析”,画像電子学会誌,vol.31,no.2,pp.193−201,2002.
【0006】
【発明が解決しようとする課題】
現在の画像処理装置を導入した防犯システムでは、警備対象区域の状況の監視や、人出の人数は把握可能であるが、下記の点が実現されていない。
【0007】
▲1▼任意に設定した、期間中、任意に設定した周期にて、イベント会場や通路等の警備対象区域の人出(人数、密度)に関連深い周辺地点にカメラを設置し、その地点における人流(流出入量)を画像処理にて計測するとともに、人出と直接関連する交通機関等の流出入地点における予想された人出のデータから各周辺地点(カメラ設置地点)の人流をシステマチックに予測し、通知すること
【0008】
▲2▼実測値から以降の予測値を再予測すること
▲3▼警備対象区域の人出の危険警報値を予め設定し、上記▲1▼の実測値、予測値がその値を超えた場合は、通知すること
▲4▼任意の時刻に任意の人流計測地点での通行規制の設定や解除を行った場合の、以降の人出と人流を予測処理して通知すること。
▲5▼上記▲3▼の予測警報発生時、任意に通行規制した場合の人出と人流を再予測し通知すること
▲6▼上記▲3▼の予測警報復帰時、任意に通行規制を解除した場合の人出と人流を再予測し通知すること
【0009】
▲7▼人出に間接的に影響すると考えられる過去の諸条件(日時、曜日、天気、気温、夏休み等の特殊な日、交通機関の到着時刻、開催時刻、閉会時刻等)、人出と直接関連する交通機関等からの流出入データ、過去の各流入出地点における人流の時系列データ(流出入量と通行規制の有無)と警備対象区域の人出(人数及び密度)の時系列データを履歴ファイルとして蓄積記憶させ、通知すること
▲8▼上記▲7▼の履歴ファイルから次回予測の参考データとして使用したいファイルを指定し、次回の諸条件を模擬入力することで、人出と直接関連する交通機関等からの流出入データを予測処理し、各流出入地点における時系列の人の流れ(流出入量)及び各警備対象区域における時系列の人出(人数及び密度)を予測処理し、次回警備計画の立案の基とすること
【0010】
以上から、現状のイベント運営おける交通規制、通行ルート、スペースの検討、人の流れの制御や、警備計画立案は主催者側の経験や勘に頼るところが大きい。
過去の人出や、各地点の人流に関する定量的なデータが少ないため、次回の警備計画立案時の予想が精度が良くない。
また、主催者の人事異動等で担当者が代わると感覚やノウハウが伝わりにくい。
混雑区域における群衆の密集に対する危険判断は、現場の警備員の及び監視カメラの映像からの状況把握によるものとなり、定量的な危険度の把握及びその値における現場状況の混雑度合いの感覚を掴むことが行えない。
【0011】
また、定量的な数値による自動通知がなされないため常に人間が全ての危険個所の混雑度合いを監視していなければならず、監視対象が多く、広範囲になるほど、負担が多く、また、見落としの危険性もある。
イベント会場への流入経路が多く複雑になると、実際の人出の状況から、以降の各地点での人流や交通規制の実施、解除を複数の地点で実施する場合もあるが、複雑になるため、全体の人流を把握しながら、各地点連動させて規制の制御を行う判断が難しい。
【0012】
【課題を解決するための手段】
上記課題を解決する本発明の請求項1に係るイベント警備監視方法は、任意に設定した期間中、任意に設定した周期にて、イベント会場や通路等の警備対象区域の人出に関連深い周辺地点にカメラを設置し、該カメラにより撮影された画像を画像処理することにより前記周辺地点における人流を計測するとともに、人出と直接関連する交通機関等の流出入地点における予想された人出の流出入データから各周辺地点の人流及び警備対象区域の人出をシステマチックに予測することを特徴とする。
【0013】
上記課題を解決する本発明の請求項2に係るイベント警備監視方法は、請求項1において、前記周辺地点の人流の実測値に基づいて以降の人流及び警備対象区域の人出の予測値を再予測することを特徴とする。
【0014】
上記課題を解決する本発明の請求項3に係るイベント警備監視方法は、請求項1又は2において、任意の時刻に任意の前記周辺地点での通行規制の設定又は解除を行った場合に、以降の人出と人流を予測することを特徴とする。
【0015】
上記課題を解決する本発明の請求項4に係るイベント警備監視装置は、任意に設定した期間中、任意に設定した周期にて、イベント会場や通路等の警備対象区域の人出に関連深い周辺地点にカメラを設置し、該カメラにより撮影された画像を画像処理することにより前記周辺地点における人流を計測するとともに、人出と直接関連する交通機関等の流出入地点における予想された人出の流出入データから各周辺地点の人流及び警備対象区域の人出をシステマチックに予測する予測処理部を設けたイベント警備監視装置において、警備対象区域に人出の危険警報値を予め設定し、人出の実測値、予測値が該危険警報値を超えた場合には警報を通知する警報判定部を設けたことを特徴とする。
【0016】
上記課題を解決する本発明の請求項5に係るイベント警備監視装置は、請求項4において、人出に間接的に影響すると考えられる過去の諸条件、前記流出入地点の流出入地点からの流出入データ、過去の各流入出地点における人流の時系列データと警備対象区域の人出の時系列データを履歴ファイルとして蓄積記憶させるデータ記憶部を設けたことを特徴とする。
【0017】
上記課題を解決する本発明の請求項6に係るイベント警備監視方法は、請求項5記載のイベント警備監視装置を使用し、上記履歴ファイルから次回予測の参考データとして使用したいファイルを指定し、次回の諸条件を模擬入力することで、前記流出入地点からの流出入データを予測処理し、前記流出入地点における時系列の人流及び各警備対象区域における時系列の人出を予測処理し、次回警備計画の立案の基礎とすることを特徴とする。
【0018】
【発明の実施の形態】
以下、本発明について図面に示す実施例を参照して詳細に説明するが、本発明においては、人出とは人数及び密度をいい、人流とは流出入量を意味するものとする。
本発明の一実施例に係るイベント警備監視装置を図3に示す。
図3に示すように、本実施例のイベント警備監視装置は、映像入力処理部20表示部30、表示処理部40、画像処理部50、計測処理部60、画像記録部70、操作入力部80、警報判定部90、予測処理部100及びデータ記憶部110よりなり、図1に示す警備本部に設置される。
主催者側は、イベント警備監視装置を利用して警備計画案を立案する。
警備計画案は、以下の通り、会場周辺警備地図の作成、会場警備系統図の作成及び人出の予測と通行規制計画の立案に従って作成される。
【0019】
(1)会場周辺警備地図の作成
過去の実績や、データをもとに、交通規制区域と通行路の区分、通行方向、関連交通機関も考慮した人の流出入地点、警部本部、各警備地点、カメラの設置地点、警備対象区域、イベント警備監視装置の設置場所等を明確化し、図1のような会場周辺警備地図を予め作成する。
システム的に人流計測及び予測を行うため、最寄りの交通機関等、警備対象区域A1〜Anの人出に関連する人の流入出地点を全て予め抽出して、カメラ10の設置地点P1〜Pnを予め決定する。
本実施例では、図1に示すように、警備対象区域としては、イベント会場A1、歩道橋A2、道路(歩行者天国)A3,A4の四箇所とし、カメラ10の設置地点としては、P1〜P13の13箇所とした。
尚、図1では、警備員を配置したカメラ監視地域(図中、斜線を入れて示す)も設置した。
【0020】
(2)会場警備系統図の作成
図1に示す会場周辺警備地図をもとに、図2のような警備対象区域(会場、通路等)A1〜An、通行方向、カメラの設置地点P1〜Pn、人出と直接関連する交通機関等の流出入地点B1〜B4の簡易な系統図である会場警備系統図を作成し、データ記憶部110に記憶させる。
本系統図は、イベント警備監視装置の表示部30に表示する系統監視画面として作成し、警備対象区域A1〜Anの面積、人出(人数、密度)、カメラの設置地点P1〜Pnでの人流(流出入量)等の計測値、通行規制の有無を表示することに用いる。
【0021】
(3)人出の予測と通行規制計画の立案(過去履歴データを使用しない場合)
人出の予測と通行規制計画は、人出データ一覧表の作成、予測条件、流出入データの予想入力、カメラの設置地点の人流、警備対象区域の人出の予測、予測結果の表示、警備対象区域の危険警報処理、通行規制の計画の策定よりなる。
▲1▼人出データ一覧表の作成
イベント警備監視装置の表示部30に人出に間接的に影響すると考えられる過去の諸条件(日時、曜日、天気、気温、夏体み等の特殊日、各交通機関の到着時刻、開催時刻、閉会時刻等)を抽出表示し、表1の人出データ一覧表(事前予測用)のフォーマットを作成し、データ記憶部110に記憶させる。
【0022】
【表1】

Figure 2004178358
【0023】
▲2▼予測条件、流出入データの予想入力
天気予報、時刻表等の各種情報から、上記諸条件と指定した対象期間中の流出入地点B1〜B4の流出入データを予想し全て操作入力部80より入力し、データ記憶部110に記憶させる。
流出入地点としては、人出に直接関連する地点をいい、表1に示すように、例えば、B1は○○電鉄△駅(上り)、B2は○○電鉄△駅(下り)、B3は××バス○停留所(A方面)、B4は××バス○停留所(B方面)、B5は徒歩・自転車その他(合計)がある。
【0024】
▲3▼カメラの設置地点の人流、警備対象区域の人出の予測
入力されたB1〜B4の流出入データの流出入量の値から、各カメラの設置地点P1〜Pnにおける人流の時系列データ(流出入量と通行規制等)と警備対象区域A1〜Anの人出の時系列データ(人数及び密度)を予測演算するための演算式及び係数を作成し、予測処理部100に入力しておく。
演算式及び係数の一例を以下に示す。
【0025】
各時間毎のカメラ設置地点Pnの人の流入量をPnin、流出量をPnoutとすると、人出の流出入データBnin(流入量),Bnout(流出量)及び各係数を用いて下記の通り予測算出する。
ただしX11in,X11outは、P11地点の流出入量からバス停利用の流出入量分を除いた数値である。
anin:地点nに流入する遅延係数
anout:地点nから流出する遅延係数
cnin:地点nに流入する合流比率係数
cnout:地点nから流出する分流比率係数
各地点での出入り口広さ等から決まる最大人流量をPnMAX、そのときのanをanMAXとし、ある時刻におけるan、cnをそれぞれantn、cntnとするとantn=0(その方向に通行規制有りの場合) …(1)
antn=antnMAX(antnMAX≦1の場合) …(2)
antn=1としておく(antnMAX>1の場合) …(3)
cntn=一定で任意に設定 …(4)
【0026】
P1in=0 …(5)
P1out=a1out×c1out×B5out …(6)
P2in=0 …(7)
P2out=a2out×P4out …(8)
P3in=a3in×P5in …(9)
P3out=0 …(10)
P4in=0 …(11)
P4out=a4out(P6out+P7out+P8out) …(12)
P5in=a5in(P6in+P7in+P8in) …(13)
P5out=0 …(14)
【0027】
P6in=a6in×c6in×B5in …(15)
P6out=a6out×c6out×B5out …(16)
P7in=a7in×c7in×B5in …(17)
P7out=a7out×c7out×B5out …(18)
P8in=a8in(P9in+P10in+P11in) …(19)
P8out=a8out(P9out+P10out+P11out) …(20)
P9in=a9in×c9in×(B1in+B2in) …(21)
P9out=a9out×c9out×(B1out+B2out) …(22)
【0028】
P10in=a10in×c10in×(B1in+B2in) …(23)
P10out=a10out×c10out×(B1out+B2out) …(24)
P11in=a1in(P12in+P13in+X11in) …(25)
P11out=a11out(P12out+P13out+X11out) …(26)
P12in=a12in×B3in …(27)
P12out=a12out×B3out …(28)
【0029】
P13in=a13in×B4in …(29)
P13out=a13out×B4out …(30)
ただし、上記において
X11in=B5in−(P6in+P7in) …(31)
X11out=B5out−(P1out+P6out+P7out) …(32)
【0030】
各比率係数の関係は
c6in+c7in+c11in=1 徒歩その他で流入量B5inに対する地点nに流入する比率係数の総和は1 …(33)
c1out+c6out+c7out+c11out=1 徒歩その他で流入量B5outに対する地点nから流出する比率係数の総和は1 …(34)
c9in+c10in=1 電鉄駅からの流入量(B1in+B2in)に対する地点nに流入する比率係数の総和は1 …(35)
c9out+c10out=1 電鉄駅からの流出量(B1out+B2out)に対する地点nから流出する比率係数の総和は1 …(36)
【0031】
Anにおける人出をAnh 、その初期値(設定可)をAns 、密度をAnm とすると
Anm =Anh /Anの面積
A1h =Σ(P1in+P2in+P3in)+A1s …(37)
A2h =Σ(P2out+P3out+P4in+P5in)+A2s …(38)
A3h =Σ(P4out+P5out+P6in+P7in+P8in)+A3s …(39)
A4h =Σ(P8out+P9in+P10in+P11in)+A4s …(40)
【0032】
尚、各遅延係数は、上記各式から決まる。
例えば、上記(29)式a13inを求める。
(29)式からa13in=P13in÷B4in
a13intnMAX=P13inMAX÷B4intn
仮にB4intn=50,P13inMAX=100の場合
上記よりa13intnMAX=100÷50=2>1となるため
上記の(1)〜(3)式のうち(3)式が適応され
a13intn=1となる。
【0033】
上記数式と予め人間にて過去のデータ等から予想したPnMAXの各個を入力しておき、各遅延係数は予測処理部100が自動計算処理する。
また、合流比率係数は人間が予想若しくは過去のデータから想像して上記(33)〜(36)式を満たす値を、操作入力部80から設定入力する。
なお、(33)〜(36)式の関係は、事前に作成した会場警備系統図から各関係を人間が導き出して設定時に忘れないよう装置に入力保存しておく。
更に、上記演算式は会場警備系統図に合わせてその都度、人間が作成及び修正する。
【0034】
▲4▼予測結果の表示
予測処理部100は、データ記憶部110に入力された情報から、上記演算式にて演算した結果を表示処理部40に渡して、表示部30の人出のデータ一覧表(事前予測用)に表示する。
この予測値をベース予測値としてデータ記憶部110に保存しておく。
【0035】
▲5▼警備対象区域の危険警報処理
操作入力部80から各警備対象区域A1〜Anの人出密度(人/m3)の危険警報設定値(1段、2段)を入力し、警報判定部90は、予測値が危険警報設定値を越える場合は表示処理部40に通知し数値の色替え表示を行うことにより警報を通知する。
また、警報判定部90は、再予測値が危険警報設定値から下限に復帰した場合には表示処理部40に通知し数値を点滅表示して通知する。
【0036】
▲6▼通行規制の計画の策定
危険警報設定値を越えた時間帯の人出を安全にする(警報値よりも少なくする)ための通行規制の地点と時間帯を検討して、操作入力部80より通行規制の該当欄に通行規制の有無予定を入力する。
予測処理部100は、通行規制有りの場合は、該当の予測値は0値にして次回通行規制無しの時刻の各予測数値にもとの予測値を加算して、各予測値を再予測し表示処理部40により表示する。
但し、加算した値が、その時刻の最大値(PnMAX )を越える場合はその予測値は最大値として扱い、越えた分の値はそれ以降通行規制が無い時刻の予測値に加算する。
【0037】
通行規制の設定を解除した場合、予測処理部100において、予測値はベース予測値をもとにその逆の演算を行いもとの数値に戻す。
以降、予測処理部100により通行規制の模擬を繰り返し、危険警報設定値を越える時間帯が無く、かつ警備対象区域A1〜Anの人出が地域によって及び時間帯によってできるだけ平滑化される通行規制方法を模索した後、データ記憶部110に保存操作を行って、事前警備計画案とする。
更に予想以上に人出が多かった場合の対応にも備え、条件表の予測入力値(各交通機関からの流出入量)を変えて、複数の事前警備計画案を立案保存しておく。
【0038】
(4)人出の予測と通行規制計画の立案(過去履歴データを使用する場合)
過去実績があり過去履歴ファイルを使用する場合は、上記(3)項の▲2▼の処理に代えて下記処理が可能となり、その他の処理は上記と同様である。
即ち、人出データ一覧表(事前予測用)を表示して、次回予測に使用したい過去のファイルを複数選択し、諸条件を入力すると、予測処理部100は、過去の各交通機関からの流出入量の実測値データから決められた処理を行い予測データを表示する。
決められた処理とは、過去ファイルの中から同一時刻の流出入データを平均化、或いは時刻表が異なる場合は、今回時刻表のタイミングに合わせて加算、前後へ均一分配処理を言う。
【0039】
警備実施時には、主催者側はイベント警備監視方法として、警備本部の運用及びイベント警備監視装置の処理を以下の通り行う。
(1)警備本部の運用
警備本部は、イベント警備監視装置の表示部30に表示される、会場警備系統図及び予測実測データ一覧画面、各地点のカメラ映像を監視しながら各警備地点の警備員に通行規制、誘導等の警備指示を行う。
【0040】
(2)イベント警備監視装置の処理
イベント警備監視装置の処理は、以下の通り、映像の表示、画像処理による人流の計測、人流の表示、人出データ一覧表(実測用)への実測値の表示、実測値からの以降の予測、警備対象区域の危険警報処理、通行規制の検討及び画像の記憶及び再生よりなる。
【0041】
▲1▼映像の表示
各カメラ10から入力された映像信号を、映像入力処理部20により信号変換して表示処理部40にて表示処理し、表示部30に表示する。
即ち、映像入力処理部20は、NTSC方式等の入力されたアナログ映像信号をMPEGII形式等のデジタル信号に変換し、表示処理部40に送る。
表示処理部40はデジタル信号を表示部30に表示するための処理を行うと同時に、画像処理部50にデジタル信号を送信する。
【0042】
▲2▼画像処理による人流の計測
画像処理部50は入力された各画像から画像処理にて人物を抽出し、計測処理部60に出力した通行方向別の人数を計測し、設定した計測周期毎にデータ記憶部110にデータを保存する。
即ち、画像処理部50にて、エッジ抽出、モデルマッチング処理にて人間一人一人を認識する。計測誤差を最小にするために、カメラは撮影箇所に対してできるだけ真上から撮影できる位置に設置する。
計測処理部60にて、カメラ映像中に人間の進行方向に基準線を設け、それと垂直に二次元の方向軸を設定し、認識した人間の座標を計算し、動き検出処理にて人間進行方向(入か出か)を検出し、更に単位時間における人間の座標位置の変化を差分処理することで、基準軸を通過した人間の人数を検出する(例えば、特願2002−343587「動き方向検出方法及び装置」)。
【0043】
▲3▼人流の表示
表示処理部40は操作入力部80から選択された画面(映像、会場警備系統図、人出データ一覧表(実測用)等)及び実測値を表示部30に表示する。
▲4▼人出データ一覧表(実測用)への実測値の表示
複数の人出データ一覧表(事前予測用)ファイル(=事前計画警備案)の中から、今回実測用として使用したいファイルを操作入力部80より選択操作し、予測処理部100は選択されたファイルを人出データ一覧(実測用)ファイルとして新規に作成し表示する。
予測値に代わって実測値に表示更新される。
更に、予測処理部100は人流の実測値から、警備対象区域A1〜Anの人出のデータを演算し実測値として更新し、表示処理部40により表示更新する。
【0044】
▲5▼実測値からの以降を再予測
予測処理部100は実測値を更新処理した後、以降の予測値を全て演算して更新する。即ち、再予測する。
再予測演算処理は以下の通り。
実測値表示更新した時刻におけるPnを実測Pnt(n)、
事前に予測していたPnを予測Pnt(n)、
次回の更新すべき予測値を再予測Pnt(n+1)とすると、
通行規制が有りの場合は、再予測Pnt(n+m)=0
通行規制無しの場合は、再予測Pnt(n+m)=jn×予測Pnt(n)
jnは変動係数で実測Pnt(n)/予測Pnt(n)…(ア)
(但し実測Pnt(n)或いは予測Pnt(n)が0の場合はjn=1とする)
(ア)式の実測開始からの平均値をjnとする。
また、(3)項▲4▼のベース予測値もその度にjnを積算して更新しておく。
【0045】
▲6▼警備対象区域の危険警報処理
操作入力部80から入力設定された各警備対象区域A1〜Anの人出密度(人/m3)の危険警報設定値(1段、2段)に基づき、警報判定部90は実測値及び再予測した予測値が危険警報設定値を越える場合は表示処理部40に通知し数値の色替え表示を行うことにより警報を通知する。
警報判定部90は、再予測値が危険警報設定値から下限に復帰した場合には表示処理部40に通知し数値を点滅表示して通知する。
【0046】
▲7▼通行規制の検討
監視員は危険警報設定値を越えた時間帯の人出を安全にする(警報値よりも少なくする)ための通行規制の地点と時間帯を検討(いつどこを通行規制すべきか或いは解除できるか)をして、操作入力部80より通行規制の該当欄に通行規制の有無予定を入力する。
予測処理部100は、その都度入力された通行規制の条件を加味して、予測出力値を再予測し表示処理部40により表示する。
以降、監視員は、予測処理部100を利用して通行規制の模擬を繰り返し、危険警報設定値を越える時間帯が無く、かつ警備対象区域A1〜Anの人出が地域によって及び時間帯によってできるだけ平滑化される通行規制方法を模索し、最新の警備計画案として運用する。
【0047】
▲8▼画像の記憶及び再生
画像記録部70は、指定したカメラ10の画像を指定した期間、或いは指定したカメラ10の画像をサイクリックに指定した周期で指定した期間、記録保存し、指令に基づいて再生可能とする。
▲9▼計測終了時のデータ保存
計測期間が終了した時点で、各種データは履歴ファイルとしてデータ記憶部110に蓄積保存される。
【0048】
【発明の効果】
以上、実施例に基づいて具体的に説明したように、本発明によれば、現状のイベント運営おける交通規制、通行ルート、スペースの検討、人の流れの制御や、警備計画立案は主催者側の経験や勘のみに頼らずに、履歴データ、実測データ、予測データ等から定量的に判断し実施できる。
過去の人出や、各地点の人流に関する定量的なデータを使用して予測するため次回の警備計画立案時の予想が精度が良くなる。
また、主催者の人事異動等で担当者が代わっても履歴データを参照し、計画を立てることができる。
混雑区域における群衆の密集に対する危険判断は、現場の警備員の及び監視カメラの映像からの状況把握によるものだけでなく、定量的な危険度の把握及びその値における現場状況の混雑度合いの感覚を掴むことができ、事故等の危険性を事前に確実にシステマチックに抑えることが可能となる。
また、定量的な数値による自動通知がされるため常に人間が全ての危険個所の混雑度合いを監視していなくてもよくなり、監視対象が多く、広範囲になるほど、監視員の負担が軽くまた、見落としの危険性も無くなる。
イベント会場への流入経路が多く複雑になっても、全体の人流を把握しながら、各地点連動させて規制の制御を行う判断が容易となる。
【図面の簡単な説明】
【図1】本発明の一実施例に係る会場周辺警備地図である。
【図2】本発明の一実施例に係る会場警備系統図である。
【図3】本発明の一実施例に係るイベント警備監視装置を示すブロック図である。
【符号の説明】
10 カメラ
20 映像入力処理部
30 表示部
40 表示処理部
50 画像処理部
60 計測処理部
70 画像記録部
80 操作処理部
90 警報判定部
100 予測処理部
110 データ記憶部
P1〜Pn カメラの設置地点
A1〜An 警備対象区域[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an event security monitoring method and an event security monitoring device. More specifically, the present invention relates to an apparatus and a method for performing crime prevention or security monitoring using image processing in an event such as an event.
[0002]
[Prior art]
At present, there are many festivals, fireworks, and parades hosted by the national, prefectural, and municipalities throughout the country, as well as events, concerts, and various sports tournaments hosted by the private sector, regardless of spring, summer, autumn, winter, and night.
These events attract a large number of people and attract crowds, and despite the security plans and efforts of organizers, police, security companies, etc. (hereinafter abbreviated as the organizers), unexpected crowds and events occur. It cannot be ruled out that the density of the crowd in a certain area exceeds the safe limit and may lead to or cause the risk of personal injury.
[0003]
In order to avoid these dangers, the organizer anticipates crowds, makes a security plan, establishes a security headquarters, places guards at each key point, regulates traffic around the event venue, It secures and controls and adjusts the flow of people (inflow / outflow speed) while communicating wirelessly.
In addition, security camera systems are installed in dangerous areas such as narrow passages to monitor the status of crowds at each point at the security headquarters, and to detect people passing on the screen with an image processing device, In some cases, automatic measurement is used to control the flow of people.
[0004]
Research reports from academic societies and others report many approaches that detect and track road vehicles, count and track pedestrians by installing a camera at the survey location, photographing intersections and passages, and processing the images. ing.
For example, Non-Patent Document 1 describes detection of a vehicle traveling on a highway, and Non-Patent Document 2 describes a research report on detection of pedestrian behavior.
[0005]
[Non-patent document 1]
Shinji Ozawa, "Recognition and Understanding in ITS Road Images", IEICE Technical Report, PRMU 98-91-105, pp. 139-157. 99-104.1998.
[Non-patent document 2]
Takeuchi, Kaneko, Igarashi, Sato, Hane, "Image analysis of walking behavior based on robust background subtraction and region extraction," Journal of the Institute of Image Electronics Engineers of Japan, vol. 31, no. 2, pp. 193-201, 2002.
[0006]
[Problems to be solved by the invention]
Although the security system using the current image processing device can monitor the status of the security target area and grasp the number of people coming out, the following points have not been realized.
[0007]
(1) At an arbitrarily set period, at an arbitrarily set period, cameras are installed at peripheral points that are closely related to the number of people (people, density) in the security area such as event venues and passageways, and at that point Measures the flow of people (inflow and outflow) by image processing, and systematically calculates the flow of people at each peripheral point (camera installation point) from the data of the expected outflow at the inflow and outflow points of transportation, etc. directly related to the outflow. Predict and notify
[0008]
(2) Re-forecast future forecast values from actual measurement values
(3) Pre-set danger warning values for traffic in the security target area, and notify when the measured and predicted values in (1) above exceed those values.
{Circle around (4)} In the case of setting or canceling the traffic regulation at an arbitrary person flow measurement point at an arbitrary time, the subsequent traffic and the person flow are predicted and notified.
(5) When the prediction warning of (3) above occurs, re-predict and notify the number of people and the flow of people when traffic is arbitrarily restricted.
(6) When returning to the prediction warning in (3) above, re-predict and notify the number of people and the flow of traffic when the restriction on traffic is arbitrarily released.
[0009]
(7) Past conditions (day and time, day of the week, weather, temperature, special days such as summer vacation, arrival time of transportation, holding time, closing time, etc.) considered to indirectly affect the crowd, Inflow / outflow data from directly related transportation facilities, time-series data of past human flows at each inflow / outflow point (inflow / outflow and presence / absence of traffic restrictions), and time-series data of traffic (people and density) in security areas As a history file and notify
(8) Specify the file you want to use as the reference data for the next forecast from the history file of (7) above, and simulate input of the next conditions so that the inflow / outflow data from transportation facilities etc. directly related to the crowding can be obtained. Predicting and processing the time-series flow of people (inflow and outflow) at each inflow / outflow point and the time-series outflow (number and density) in each security target area to be the basis for the next security planning thing
[0010]
From the above, the traffic regulation, traffic route, space examination, flow control, and security planning in the current event management largely depend on the experience and intuition of the organizer.
Because there is little quantitative data on past traffic and the flow of people at each location, predictions at the next security planning are not accurate.
In addition, if the person in charge changes due to a change in the personnel of the organizer, etc., it is difficult to convey the sense and know-how.
The risk judgment for crowd crowding in the congested area is based on the situation grasped by the guards of the site and the images of the surveillance cameras, and it is necessary to grasp the quantitative risk level and to grasp the degree of congestion of the site situation at that value. Can not do.
[0011]
In addition, since automatic notification using quantitative numerical values is not performed, humans must constantly monitor the degree of congestion at all danger points, and the monitoring target is large, the wider the area, the greater the burden, and the risk of oversight There is also.
If the influx routes to the event venue are many and complicated, depending on the actual situation of traffic, there may be cases where the enforcement and cancellation of traffic flow and traffic control at each point will be performed at multiple points, but it will be complicated. However, it is difficult to judge how to control regulations in conjunction with each point while grasping the overall flow of people.
[0012]
[Means for Solving the Problems]
The event security monitoring method according to claim 1 of the present invention that solves the above-mentioned problem, provides a method for monitoring an area closely related to a crowd in a security target area such as an event hall or a passage at an arbitrarily set period during an arbitrarily set period. A camera is installed at a point, and image processing of an image taken by the camera is performed to measure the flow of people at the surrounding points, and the expected amount of traffic at an inflow / outflow point of transportation or the like directly related to the traffic. It is characterized by systematically predicting the flow of people at each peripheral point and the number of people in the security target area from inflow / outflow data.
[0013]
According to a second aspect of the present invention, there is provided an event security monitoring method according to the first aspect of the present invention, wherein the predicted value of the subsequent human flow and the number of people in a security target area is re-established based on the actual measured value of the human flow at the surrounding point. It is characterized by predicting.
[0014]
The event security monitoring method according to claim 3 of the present invention that solves the above-mentioned problem, according to claim 1 or 2, further comprising: setting or canceling a traffic regulation at any of the surrounding points at an arbitrary time; It is characterized by predicting the number of people and the number of people.
[0015]
The event security monitoring device according to claim 4 of the present invention which solves the above-mentioned problem, provides a peripheral area closely related to traffic in a security target area such as an event venue or a passage at an arbitrary set period during an arbitrarily set period. A camera is installed at a point, and image processing of an image taken by the camera is performed to measure the flow of people at the surrounding points, and the expected amount of traffic at an inflow / outflow point of transportation or the like directly related to the traffic. In an event security monitoring device equipped with a prediction processing unit that systematically predicts the flow of people at each peripheral point and the number of people in the security target area from the inflow / outflow data, a danger warning value for the number of people in the security target area is set in advance, An alarm judging unit for notifying an alarm when the actual measured value or the predicted value exceeds the danger alarm value is provided.
[0016]
The event security monitoring device according to claim 5 of the present invention that solves the above-mentioned problem is characterized in that, in claim 4, the past conditions considered to indirectly affect the number of people, the outflow from the inflow / outflow point of the outflow / inflow point. There is provided a data storage unit for accumulating and storing, as a history file, incoming data, past time-series data of human flow at each inflow / outflow point, and time-series data of crowding in a security target area.
[0017]
An event security monitoring method according to claim 6 of the present invention for solving the above-mentioned problems uses the event security monitoring device according to claim 5, specifies a file to be used as reference data for the next prediction from the history file, and By simulating the various conditions, the inflow / outflow data from the inflow / outflow point is predicted and processed, and the time-series human flow at the inflow / outflow point and the time-series outflow in each security target area are predicted and processed. It is characterized as a basis for security planning.
[0018]
BEST MODE FOR CARRYING OUT THE INVENTION
BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, the present invention will be described in detail with reference to embodiments shown in the drawings, but in the present invention, outgoing means the number of people and density, and human flow means outflow and inflow.
FIG. 3 shows an event security monitoring device according to an embodiment of the present invention.
As shown in FIG. 3, the event security monitoring device according to the present embodiment includes a video input processing unit 20, a display unit 30, a display processing unit 40, an image processing unit 50, a measurement processing unit 60, an image recording unit 70, and an operation input unit 80. , An alarm determination unit 90, a prediction processing unit 100, and a data storage unit 110, and are installed in the security headquarters shown in FIG.
The organizer drafts a security plan using the event security monitoring device.
The security plan draft is created in accordance with the creation of a security map around the venue, the creation of a security diagram for the venue, the prediction of traffic, and the drafting of a traffic regulation plan, as described below.
[0019]
(1) Creating a security map around the venue
Based on past results and data, the classification of traffic-controlled areas and traffic routes, the direction of traffic, the inflow / outflow point of humans, taking into account the related transportation, the police headquarters, each security point, the camera installation point, and the area subject to security The location of the event security monitoring device is clarified, and a security map around the venue as shown in FIG. 1 is created in advance.
In order to systematically measure and predict the flow of people, all the inflow and outflow points of persons related to the traffic in the security target areas A1 to An, such as the nearest transportation, are extracted in advance, and the installation points P1 to Pn of the camera 10 are determined. Determine in advance.
In the present embodiment, as shown in FIG. 1, the areas to be guarded are the event venue A1, the pedestrian bridge A2, the roads (pedestrian heaven) A3 and A4, and the installation points of the camera 10 are P1 to P13. 13 locations.
In FIG. 1, a camera surveillance area where security guards are arranged (shown by hatching in the figure) is also installed.
[0020]
(2) Creating a security diagram for the venue
Based on the security map around the venue shown in FIG. 1, areas to be guarded (venues, passageways, etc.) A1 to An, traffic directions, camera installation points P1 to Pn as shown in FIG. A security system diagram for the venue, which is a simple system diagram of the inflow / outflow points B1 to B4, is created and stored in the data storage unit 110.
This system diagram is created as a system monitoring screen to be displayed on the display unit 30 of the event security monitoring device, and the area of the security target areas A1 to An, the number of people (number of people, density), and the flow of people at the camera installation points P1 to Pn. Used to display measured values such as (inflow and outflow) and the presence or absence of traffic restrictions.
[0021]
(3) Forecast of traffic and drafting of traffic regulation plans (when past history data is not used)
Forecasting of traffic and traffic regulation plans include creating a list of traffic data, predicting conditions, entering inflow and outflow data, predicting the flow of people at the camera location, predicting traffic in security areas, displaying prediction results, and guarding It consists of processing hazard warnings in the target area and formulating traffic regulation plans.
▲ 1 ▼ Creation of a list of heading data
Past conditions (special days such as date and time, day of the week, weather, temperature, summer appearance, etc., arrival time of each transportation, holding time, etc.) which are considered to indirectly affect the crowd on the display unit 30 of the event security monitoring device, The closing time, etc.) are extracted and displayed, the format of the heading data list (for advance prediction) in Table 1 is created, and stored in the data storage unit 110.
[0022]
[Table 1]
Figure 2004178358
[0023]
(2) Forecast conditions and forecast input of inflow / outflow data
From various information such as a weather forecast and a timetable, the above conditions and the inflow / outflow data at the outflow / inflow points B1 to B4 during the designated target period are predicted, all are input from the operation input unit 80, and are stored in the data storage unit 110. .
The inflow / outflow points refer to points directly related to the crowding. As shown in Table 1, for example, B1 is XX railway station (up), B2 is XX railway station (down), and B3 is × × Bus ○ stop (A direction), B4 is XX bus ○ stop (B direction), B5 is walking / bicycle and others (total).
[0024]
(3) Prediction of the number of people at the camera installation point and the number of people in the security area
Based on the values of the inflow / outflow data of the inflow / outflow data of B1 to B4, the time-series data (inflow and outflow and traffic regulation, etc.) of the human flow at the installation points P1 to Pn of each camera and the people in the security target areas A1 to An An arithmetic expression and a coefficient for predictive calculation of the output time-series data (number of people and density) are created and input to the prediction processing unit 100.
An example of an arithmetic expression and a coefficient is shown below.
[0025]
Assuming that the inflow amount of a person at the camera installation point Pn at each time is Pnin and the outflow amount is Pnout, prediction is made as follows using the outflow / inflow data Bnin (inflow amount), Bnout (outflow amount) of the crowd and each coefficient. calculate.
However, X11in and X11out are numerical values obtained by subtracting the inflow / outflow amount at the bus stop from the inflow / outflow amount at the point P11.
anin: delay coefficient flowing into point n
anout: delay coefficient flowing out from point n
cnin: Confluence ratio coefficient flowing into point n
cnout: Dividing ratio coefficient flowing out from point n
If the maximum human flow determined by the width of the entrance at each point is PnMAX, an at that time is anMAX, and ann and cn at a certain time are anntn and cntn, respectively, then antn = 0 (in the case of traffic regulation in that direction). … (1)
antn = antnMAX (when antnMAX ≦ 1) (2)
antn = 1 (when antnMAX> 1) (3)
cnnt = constant and arbitrarily set ... (4)
[0026]
P1in = 0 (5)
P1out = a1out × c1out × B5out (6)
P2in = 0 (7)
P2out = a2out × P4out (8)
P3in = a3in × P5in (9)
P3out = 0 (10)
P4in = 0 (11)
P4out = a4out (P6out + P7out + P8out) (12)
P5in = a5in (P6in + P7in + P8in) (13)
P5out = 0 (14)
[0027]
P6in = a6in × c6in × B5in (15)
P6out = a6out × c6out × B5out (16)
P7in = a7in × c7in × B5in (17)
P7out = a7out × c7out × B5out (18)
P8in = a8in (P9in + P10in + P11in) (19)
P8out = a8out (P9out + P10out + P11out) (20)
P9in = a9in × c9in × (B1in + B2in) (21)
P9out = a9out × c9out × (B1out + B2out) (22)
[0028]
P10in = a10in × c10in × (B1in + B2in) (23)
P10out = a10out × c10out × (B1out + B2out) (24)
P11in = a1in (P12in + P13in + X11in) (25)
P11out = a11out (P12out + P13out + X11out) (26)
P12in = a12in × B3in (27)
P12out = a12out × B3out (28)
[0029]
P13in = a13in × B4in (29)
P13out = a13out × B4out (30)
However, in the above
X11in = B5in- (P6in + P7in) (31)
X11out = B5out- (P1out + P6out + P7out) (32)
[0030]
The relationship of each ratio coefficient is
c6in + c7in + c11in = 1 The sum of the ratio coefficients flowing into the point n with respect to the inflow amount B5in when walking or the like is 1 (33)
c1out + c6out + c7out + c11out = 1 The sum total of the ratio coefficients flowing out of the point n with respect to the inflow amount B5out during walking or the like is 1 (34)
c9in + c10in = 1 The sum of the ratio coefficient flowing into the point n with respect to the inflow amount (B1in + B2in) from the railway station is 1 ... (35)
c9out + c10out = 1 The sum of the ratio coefficients flowing out from the point n with respect to the outflow amount (B1out + B2out) from the railway station is 1 ... (36)
[0031]
Assuming that Anh is a protruding person in An, an initial value (settable) is Ans, and a density is Anm
Anm = Area of Anh / An
A1h = Σ (P1in + P2in + P3in) + A1s (37)
A2h = Σ (P2out + P3out + P4in + P5in) + A2s (38)
A3h = Σ (P4out + P5out + P6in + P7in + P8in) + A3s (39)
A4h = Σ (P8out + P9in + P10in + P11in) + A4s (40)
[0032]
Note that each delay coefficient is determined from the above equations.
For example, the above equation (13) a13in is obtained.
From equation (29), a13in = P13in ÷ B4in
a13intnMAX = P13inMAX ÷ B4intn
If B4intn = 50, P13inMAX = 100
From the above, a13intnMAX = 100 ÷ 50 = 2> 1 because
Equation (3) among the above equations (1) to (3) is applied.
a13intn = 1.
[0033]
The formula and each of PnMAX predicted by humans from past data and the like are input in advance, and each delay coefficient is automatically calculated by the prediction processing unit 100.
Also, as the merging ratio coefficient, a value that satisfies the above equations (33) to (36) is set and input from the operation input unit 80 as imagined by humans based on expected or past data.
The relations of the equations (33) to (36) are input and stored in a device so that a human can derive each relation from the venue security system diagram created in advance and do not forget when setting.
In addition, the above formula is created and corrected by a human in each case according to the venue security system diagram.
[0034]
▲ 4 ▼ Display of prediction result
The prediction processing unit 100 passes the result calculated by the above-described arithmetic expression from the information input to the data storage unit 110 to the display processing unit 40, and generates a data list (for advance prediction) of the protruding data on the display unit 30. indicate.
This predicted value is stored in the data storage unit 110 as a base predicted value.
[0035]
▲ 5 ▼ Danger warning processing for security target areas
From the operation input unit 80, the danger alarm set values (one step, two steps) of the traffic density (person / m3) of each of the security target areas A1 to An are input. If the number exceeds the limit, the display processing unit 40 is notified and an alarm is notified by performing color change display of the numerical value.
When the re-predicted value returns from the danger alarm set value to the lower limit, the alarm determination unit 90 notifies the display processing unit 40 and notifies the display processing unit 40 by blinking the numerical value.
[0036]
(6) Formulation of traffic regulation plan
Examine the point and time zone of the traffic regulation to make the traffic in the time zone exceeding the danger alarm set value safe (less than the alarm value), and pass from the operation input unit 80 to the corresponding column of the traffic regulation. Enter the schedule of the regulation.
When there is traffic regulation, the prediction processing unit 100 sets the corresponding predicted value to 0, adds the original predicted value to each predicted value at the time when there is no traffic regulation next time, and re-predicts each predicted value. The display is performed by the display processing unit 40.
However, if the added value exceeds the maximum value (PnMAX) at that time, the predicted value is treated as the maximum value, and the excess value is added to the predicted value at a time when there is no traffic regulation thereafter.
[0037]
When the setting of the traffic regulation is released, the prediction processing unit 100 performs the reverse operation on the predicted value based on the base predicted value to return to the original numerical value.
Thereafter, the simulation of the traffic control is repeated by the prediction processing unit 100, and there is no time zone exceeding the danger alarm set value, and the traffic in the security target areas A1 to An is smoothed as much as possible by the region and the time zone. After searching for, a save operation is performed in the data storage unit 110 to obtain a preliminary security plan.
Further, in preparation for a case where the number of people is larger than expected, a plurality of preliminary security plans are prepared and changed by changing the predicted input values (the amount of inflow / outflow from each transportation) in the condition table.
[0038]
(4) Forecast of traffic and drafting of traffic regulation plan (when using past history data)
In the case of using the past history file with past results, the following process can be performed instead of the process (2) in the above item (3), and other processes are the same as above.
That is, when a list of traffic data (for advance prediction) is displayed, a plurality of past files to be used for the next prediction are selected, and various conditions are input, the prediction processing unit 100 causes the past outflow from each transportation means. A process determined from the actually measured input data is performed and predicted data is displayed.
The determined processing is to average outflow / inflow data at the same time from the past files, or, if the timetables are different, add in accordance with the timing of the current timetable, and perform uniform distribution processing before and after.
[0039]
When security is implemented, the organizer performs the operation of the security headquarters and the processing of the event security monitoring device as follows as an event security monitoring method.
(1) Operation of the Security Headquarters
The security headquarters monitors the security system diagram of the venue, the list of predicted actual measurement data, and the camera image of each point displayed on the display unit 30 of the event security monitoring device, and controls traffic at security points at each security point, and controls traffic guidance and guidance. Give security instructions.
[0040]
(2) Processing of the event security monitoring device
The process of the event security monitoring device is as follows: video display, measurement of human flow by image processing, display of human flow, display of actual measurement values in a human appearance data list (for actual measurement), and subsequent prediction from actual measurement values , Security warning processing of the security target area, examination of traffic regulation, and storage and reproduction of images.
[0041]
Display of video
The video signal input from each camera 10 is converted into a signal by the video input processing unit 20, displayed by the display processing unit 40, and displayed on the display unit 30.
That is, the video input processing unit 20 converts an analog video signal input in the NTSC format or the like into a digital signal in the MPEGII format or the like and sends the digital signal to the display processing unit 40.
The display processing unit 40 performs processing for displaying the digital signal on the display unit 30 and, at the same time, transmits the digital signal to the image processing unit 50.
[0042]
(2) Human flow measurement by image processing
The image processing unit 50 extracts a person from each of the input images by image processing, measures the number of persons in each traveling direction output to the measurement processing unit 60, and stores the data in the data storage unit 110 at each set measurement cycle. I do.
That is, the image processing unit 50 recognizes each person through edge extraction and model matching processing. In order to minimize the measurement error, the camera should be installed at a position where the camera can shoot from directly above the shooting location as much as possible.
In the measurement processing unit 60, a reference line is provided in the camera image in the traveling direction of the human, a two-dimensional direction axis is set perpendicular to the reference line, the coordinates of the recognized human are calculated, and the human traveling direction is detected by the motion detection processing. (Incoming or outgoing) is detected, and the number of humans passing through the reference axis is detected by performing a difference process on a change in the coordinate position of the human in a unit time (for example, Japanese Patent Application No. 2002-343587 “Motion direction detection”). Methods and Apparatus ").
[0043]
▲ 3 ▼ Display of people flow
The display processing unit 40 displays on the display unit 30 the screen (video, venue security system diagram, crowding data list (for actual measurement), and the like) selected from the operation input unit 80 and the actual measurement value.
▲ 4 ▼ Display of actual measurement value on the list of heading data (for actual measurement)
A file to be used for actual measurement this time is selected and operated from the operation input unit 80 from among a plurality of traffic data list (for advance prediction) files (= prior plan security plan), and the prediction processing unit 100 selects the selected file. Is newly created and displayed as a heading data list (for actual measurement) file.
The display is updated to the actually measured value instead of the predicted value.
Further, the prediction processing unit 100 calculates the data of the traffic in the security target areas A1 to An from the actually measured values of the people flow, updates the calculated data as the actually measured values, and updates the display by the display processing unit 40.
[0044]
(5) Re-estimate the future from the actual measurement
After updating the actually measured value, the prediction processing unit 100 calculates and updates all subsequent predicted values. That is, the prediction is performed again.
The re-prediction calculation process is as follows.
Pn at the time of updating the measured value display is actually measured Pnt (n),
The predicted Pn is predicted Pnt (n),
Assuming that the next predicted value to be updated is re-prediction Pnt (n + 1),
If there is a traffic restriction, the re-prediction Pnt (n + m) = 0
When there is no traffic regulation, re-prediction Pnt (n + m) = jn × prediction Pnt (n)
jn is a coefficient of variation, which is actually measured Pnt (n) / predicted Pnt (n) (a)
(However, jn = 1 when the measured Pnt (n) or the predicted Pnt (n) is 0)
The average value of the equation (a) from the start of actual measurement is jn.
Also, the base predicted value of item (4) (4) is updated by integrating jn each time.
[0045]
(6) Danger warning processing for security target areas
Based on the danger alarm set values (1st, 2nd) of the traffic density (person / m3) of each of the security target areas A1 to An set by the operation input unit 80, the alarm determination unit 90 determines the actual measurement value and the re-prediction. When the predicted value exceeds the danger warning set value, the display processing unit 40 is notified and an alarm is notified by performing color change display of the numerical value.
When the re-predicted value returns to the lower limit from the danger alarm set value, the alarm determination unit 90 notifies the display processing unit 40 and notifies the display processing unit 40 by blinking the numerical value.
[0046]
7) Examination of traffic regulations
The observer considers the point and time zone of the traffic regulation to secure the traffic during the time zone exceeding the danger alarm setting value (less than the alarm value) (when and where the traffic should be restricted or released) ), And inputs the presence / absence schedule of the traffic regulation from the operation input unit 80 in the corresponding column of the traffic regulation.
The prediction processing unit 100 re-predicts the predicted output value in consideration of the traffic regulation condition input each time and displays the predicted output value by the display processing unit 40.
Thereafter, the observer repeats the simulation of the traffic regulation using the prediction processing unit 100, and there is no time zone exceeding the danger alarm set value, and the number of persons in the security target areas A1 to An depends on the area and time zone as much as possible. Find a way to regulate traffic that will be smoothed and use it as the latest security plan.
[0047]
(8) Image storage and playback
The image recording unit 70 records and saves the image of the specified camera 10 for a specified period or the specified image of the camera 10 in a cyclically specified period, and makes it possible to reproduce the image based on a command.
(9) Saving data at the end of measurement
At the end of the measurement period, various data are accumulated and stored in the data storage unit 110 as a history file.
[0048]
【The invention's effect】
As described above, according to the present invention, according to the present invention, the traffic regulation, the examination of the route, the space, the control of the flow of people, and the planning of the security in the current event management are performed by the organizer. Without relying solely on the experience or intuition of the user, it is possible to make quantitative judgments based on history data, actual measurement data, prediction data, and the like.
Since the prediction is made using quantitative data on past traffic and the flow of people at each point, the prediction at the next security planning is improved.
Further, even if the person in charge changes due to a change in personnel of the organizer or the like, it is possible to make a plan by referring to the history data.
The risk judgment for crowd crowding in the congested area is based not only on grasping the situation from the images of security guards and surveillance cameras on the site, but also on the quantitative risk assessment and the perception of the degree of congestion of the site situation at that value. It can be grasped, and the danger of an accident or the like can be reliably and systematically suppressed in advance.
In addition, since automatic notification with quantitative numerical values is performed, humans do not always need to monitor the congestion degree of all dangerous places, the number of monitoring targets is large, the wider the area, the lighter the burden on the supervisor, There is no danger of oversight.
Even if the inflow routes to the event venue are many and complicated, it becomes easy to determine the control of the regulation in conjunction with each point while grasping the overall flow of people.
[Brief description of the drawings]
FIG. 1 is a security map around a venue according to an embodiment of the present invention.
FIG. 2 is a security diagram of a venue according to one embodiment of the present invention.
FIG. 3 is a block diagram showing an event security monitoring device according to one embodiment of the present invention.
[Explanation of symbols]
10 Camera
20 Video input processing unit
30 Display
40 Display processing unit
50 Image processing unit
60 Measurement processing unit
70 Image recording unit
80 Operation processing unit
90 Alarm judgment unit
100 prediction processing unit
110 Data storage unit
P1 to Pn Camera installation point
A1-An Security target area

Claims (6)

任意に設定した期間中、任意に設定した周期にて、イベント会場や通路等の警備対象区域の人出に関連深い周辺地点にカメラを設置し、該カメラにより撮影された画像を画像処理することにより前記周辺地点における人流を計測するとともに、人出と直接関連する交通機関等の流出入地点における予想された人出の流出入データから各周辺地点の人流及び警備対象区域の人出をシステマチックに予測することを特徴とするイベント警備監視方法。During a arbitrarily set period, at an arbitrarily set period, install a camera at a peripheral point deeply related to the crowd in the security area such as an event venue or a passage, and perform image processing on the image taken by the camera The system measures the flow of people at the surrounding points and systematically detects the flow of people at each surrounding point and the number of people in the security area from the data on the flow of traffic expected at the inflow and outflow points of transportation and the like directly related to the traffic. An event security monitoring method characterized by predicting in advance. 請求項1記載のイベント警備監視方法において、前記周辺地点の人流の実測値に基づいて以降の人流及び警備対象区域の人出の予測値を再予測することを特徴とするイベント警備監視方法。2. The event security monitoring method according to claim 1, further comprising re-estimating a predicted value of a subsequent human flow and a number of people in a security target area based on an actually measured value of the human flow at the surrounding point. 請求項1又は2記載のイベント警備監視方法において、任意の時刻に任意の前記周辺地点での通行規制の設定又は解除を行った場合に、以降の人出と人流を予測することを特徴とするイベント警備監視方法。3. The event security monitoring method according to claim 1, wherein when traffic regulation at any of the surrounding points is set or canceled at an arbitrary time, a subsequent traffic and a traffic flow are predicted. Event security monitoring method. 任意に設定した期間中、任意に設定した周期にて、イベント会場や通路等の警備対象区域の人出に関連深い周辺地点にカメラを設置し、該カメラにより撮影された画像を画像処理することにより前記周辺地点における人流を計測するとともに、人出と直接関連する交通機関等の流出入地点における予想された人出の流出入データから各周辺地点の人流及び警備対象区域の人出をシステマチックに予測する予測処理部を設けたイベント警備監視装置において、警備対象区域に人出の危険警報値を予め設定し、人出の実測値、予測値が該危険警報値を超えた場合には警報を通知する警報判定部を設けたことを特徴とするイベント警備監視装置。During a arbitrarily set period, at an arbitrarily set period, install a camera at a peripheral point deeply related to the crowd in the security area such as an event venue or a passage, and perform image processing on the image taken by the camera The system measures the flow of people at the surrounding points and systematically detects the flow of people at each surrounding point and the number of people in the security area from the data on the flow of traffic expected at the inflow and outflow points of transportation and the like directly related to the traffic. In an event security monitoring device provided with a prediction processing unit for predicting, a danger warning value of a crowd is set in advance in a guard target area, and an alarm is output if the actual measurement value or the predicted value of a crowd exceeds the danger warning value. The event security monitoring device further comprising an alarm determination unit for notifying the user. 請求項4記載のイベント警備監視装置において、人出に間接的に影響すると考えられる過去の諸条件、前記流出入地点からの人出の流出入データ、過去の各流入出地点における人流の時系列データと警備対象区域の人出の時系列データを履歴ファイルとして蓄積記憶させるデータ記憶部を設けたことを特徴とするイベント警備監視装置。5. The event security monitoring device according to claim 4, wherein past conditions considered to indirectly affect the number of people, data on inflow / outflow of people from the inflow / outflow points, and time series of past people at each inflow / outflow point. An event security monitoring device comprising a data storage unit for storing and storing data and time-series data of people in a security target area as a history file. 請求項5記載のイベント警備監視装置を使用し、上記履歴ファイルから次回予測の参考データとして使用したいファイルを指定し、次回の諸条件を模擬入力することで、前記流出入地点からの人出の流出入データを予測処理し、各流出入地点における時系列の人流及び各警備対象区域における時系列の人出を予測処理し、次回警備計画の立案の基礎とすることを特徴とするイベント警備監視方法。Using the event security monitoring device according to claim 5, by specifying a file to be used as reference data for the next prediction from the history file and simulating input of various conditions for the next time, the number of people from the inflow / outflow point can be reduced. Event security monitoring that predicts inflow / outflow data, predicts time-series human flows at each inflow / outflow point and time-series crowding in each security target area, and forms the basis for planning the next security plan Method.
JP2002345156A 2002-11-28 2002-11-28 Method and equipment for guarding and watching event Withdrawn JP2004178358A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2002345156A JP2004178358A (en) 2002-11-28 2002-11-28 Method and equipment for guarding and watching event

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2002345156A JP2004178358A (en) 2002-11-28 2002-11-28 Method and equipment for guarding and watching event

Publications (1)

Publication Number Publication Date
JP2004178358A true JP2004178358A (en) 2004-06-24

Family

ID=32706406

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2002345156A Withdrawn JP2004178358A (en) 2002-11-28 2002-11-28 Method and equipment for guarding and watching event

Country Status (1)

Country Link
JP (1) JP2004178358A (en)

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009036653A (en) * 2007-08-02 2009-02-19 Pioneer Electronic Corp Device, method and program for creating driving plan, and recording medium
CN102509451A (en) * 2011-10-17 2012-06-20 北京世纪高通科技有限公司 Method and device for obtaining information of traffic incident
WO2014199675A1 (en) * 2013-06-12 2014-12-18 株式会社東芝 Regional-index computation device, regional-index computation method, and recording medium
JP2015026136A (en) * 2013-07-24 2015-02-05 株式会社ナビタイムジャパン Information processing system, information processing device, information processing method, and information processing program
CN104933867A (en) * 2015-06-17 2015-09-23 苏州大学 Road information real-time acquisition device and method based on traffic monitoring video
JP2015207041A (en) * 2014-04-17 2015-11-19 株式会社ニコン Congestion degree estimation server and congestion estimation system
WO2016013298A1 (en) * 2014-07-25 2016-01-28 日本電気株式会社 Image processing apparatus, monitor system, image processing method, and program
JP6062124B1 (en) * 2016-03-03 2017-01-18 三菱電機株式会社 Congestion prediction device and congestion prediction method
WO2017115586A1 (en) * 2015-12-28 2017-07-06 日本電気株式会社 Monitoring device, control method, and program
WO2017169068A1 (en) * 2016-03-30 2017-10-05 株式会社日立製作所 Train disembarking passenger number prediction system, congestion visualization and evaluation system, and riding capacity calculation system
JP2018173985A (en) * 2018-06-27 2018-11-08 株式会社ニコン Congestion degree estimation system and electric apparatus
US10733853B2 (en) 2015-04-20 2020-08-04 Nec Corporation Crowd guiding device, crowd guiding system, crowd guiding method, and storage medium
US10922632B2 (en) 2018-03-26 2021-02-16 Hitachi Solutions, Ltd. People flow prediction device
JP2022034358A (en) * 2020-08-18 2022-03-03 東日本旅客鉄道株式会社 Congestion prediction device, congestion prediction system, congestion prediction method, and congestion prediction program

Cited By (32)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2009036653A (en) * 2007-08-02 2009-02-19 Pioneer Electronic Corp Device, method and program for creating driving plan, and recording medium
CN102509451A (en) * 2011-10-17 2012-06-20 北京世纪高通科技有限公司 Method and device for obtaining information of traffic incident
WO2014199675A1 (en) * 2013-06-12 2014-12-18 株式会社東芝 Regional-index computation device, regional-index computation method, and recording medium
JP2014241088A (en) * 2013-06-12 2014-12-25 株式会社東芝 Local index calculation system, local index calculation method and local index calculation program
JP2015026136A (en) * 2013-07-24 2015-02-05 株式会社ナビタイムジャパン Information processing system, information processing device, information processing method, and information processing program
JP2015207041A (en) * 2014-04-17 2015-11-19 株式会社ニコン Congestion degree estimation server and congestion estimation system
US10628684B2 (en) 2014-07-25 2020-04-21 Nec Corporation Image processing apparatus, monitoring system, image processing method, and program
US10628685B2 (en) 2014-07-25 2020-04-21 Nec Corporation Image processing apparatus, monitoring system, image processing method, and program
JPWO2016013298A1 (en) * 2014-07-25 2017-06-01 日本電気株式会社 Image processing apparatus, monitoring system, image processing method, and program
WO2016013298A1 (en) * 2014-07-25 2016-01-28 日本電気株式会社 Image processing apparatus, monitor system, image processing method, and program
US11373408B2 (en) 2014-07-25 2022-06-28 Nec Corporation Image processing apparatus, monitoring system, image processing method, and program
JP7314974B2 (en) 2014-07-25 2023-07-26 日本電気株式会社 VIDEO MONITORING SYSTEM, VIDEO MONITORING METHOD, AND PROGRAM
US10699130B2 (en) 2014-07-25 2020-06-30 Nec Corporation Image processing apparatus, monitoring system, image processing method, and program
JP2020074512A (en) * 2014-07-25 2020-05-14 日本電気株式会社 Image processing apparatus, monitoring system, image processing method, and program
US10614317B2 (en) 2014-07-25 2020-04-07 Nec Corporation Image processing apparatus, monitoring system, image processing method, and program
JP2021182754A (en) * 2014-07-25 2021-11-25 日本電気株式会社 Monitoring system, method for monitoring, and program
US10733853B2 (en) 2015-04-20 2020-08-04 Nec Corporation Crowd guiding device, crowd guiding system, crowd guiding method, and storage medium
CN104933867A (en) * 2015-06-17 2015-09-23 苏州大学 Road information real-time acquisition device and method based on traffic monitoring video
CN104933867B (en) * 2015-06-17 2017-05-24 苏州大学 Road information real-time acquisition method based on traffic monitoring video
WO2017115586A1 (en) * 2015-12-28 2017-07-06 日本電気株式会社 Monitoring device, control method, and program
JP7095282B2 (en) 2015-12-28 2022-07-05 日本電気株式会社 Monitoring equipment, control methods, programs, and imaging equipment
JP2022105203A (en) * 2015-12-28 2022-07-12 日本電気株式会社 Information processing device, control method and program
JPWO2017115586A1 (en) * 2015-12-28 2018-10-18 日本電気株式会社 Monitoring device, control method, and program
JP7338739B2 (en) 2015-12-28 2023-09-05 日本電気株式会社 Information processing device, control method, and program
JP6062124B1 (en) * 2016-03-03 2017-01-18 三菱電機株式会社 Congestion prediction device and congestion prediction method
CN108701334A (en) * 2016-03-03 2018-10-23 三菱电机株式会社 Crowded prediction meanss and crowded prediction technique
WO2017149720A1 (en) * 2016-03-03 2017-09-08 三菱電機株式会社 Congestion forecasting device and congestion forecasting method
WO2017169068A1 (en) * 2016-03-30 2017-10-05 株式会社日立製作所 Train disembarking passenger number prediction system, congestion visualization and evaluation system, and riding capacity calculation system
US10922632B2 (en) 2018-03-26 2021-02-16 Hitachi Solutions, Ltd. People flow prediction device
JP2018173985A (en) * 2018-06-27 2018-11-08 株式会社ニコン Congestion degree estimation system and electric apparatus
JP2022034358A (en) * 2020-08-18 2022-03-03 東日本旅客鉄道株式会社 Congestion prediction device, congestion prediction system, congestion prediction method, and congestion prediction program
JP7175072B2 (en) 2020-08-18 2022-11-18 東日本旅客鉄道株式会社 Congestion prediction system, congestion prediction method and congestion prediction program

Similar Documents

Publication Publication Date Title
JP2004178358A (en) Method and equipment for guarding and watching event
Southworth Regional evacuation modeling: A state-of-the-art review
CN109074726B (en) People flow evaluation system and people flow control detection method
CN106652483A (en) Method for arranging traffic information detection points in local highway network by utilizing detection device
CN105117683B (en) Detection and early warning method for dense crowd in public place
JPH056500A (en) Moving body and equipment control system
Karmakar et al. A smart priority-based traffic control system for emergency vehicles
JP2015207041A (en) Congestion degree estimation server and congestion estimation system
EP4060642A1 (en) Method and system of predictive traffic flow and of traffic light control
KR20140134457A (en) Safety Accident Management System Of Bridge
Li et al. Risk-based crowd massing early warning approach for public places: A case study in China
CN114399726B (en) Method and system for intelligently monitoring passenger flow and early warning in real time
JP2021097397A (en) Degree of risk determination program and system
Vanajakshi Estimation and prediction of travel time from loop detector data for intelligent transportation systems applications
CN114898889A (en) Design method of aggregative risk control model based on big data
Lakshmi et al. Traffic response system based on data mining and internet of things (IoT) for preventing accidents
Wei et al. Integrated urban evacuation planning framework for responding to human-caused disasters over a surface transportation network
JP6835363B2 (en) Event safety management system
Parr Analysis and modeling of manual traffic control for signalized intersections
Рубель et al. Problems with effective traffic management
KR102614856B1 (en) System and method for predicting risk of crowd turbulence
JP2018173985A (en) Congestion degree estimation system and electric apparatus
CN111309954B (en) Scenic spot shopping guide behavior identification system
CN117690077A (en) People flow monitoring method, system, equipment and storage medium
Bonderup et al. Preventing drowning accidents using thermal cameras

Legal Events

Date Code Title Description
A300 Withdrawal of application because of no request for examination

Free format text: JAPANESE INTERMEDIATE CODE: A300

Effective date: 20060207