JPH02236787A - State change detector - Google Patents

State change detector

Info

Publication number
JPH02236787A
JPH02236787A JP1058532A JP5853289A JPH02236787A JP H02236787 A JPH02236787 A JP H02236787A JP 1058532 A JP1058532 A JP 1058532A JP 5853289 A JP5853289 A JP 5853289A JP H02236787 A JPH02236787 A JP H02236787A
Authority
JP
Japan
Prior art keywords
state change
image
pictures
circuit
cumulative addition
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.)
Pending
Application number
JP1058532A
Other languages
Japanese (ja)
Inventor
Yasushi Yagi
康史 八木
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.)
Mitsubishi Electric Corp
Original Assignee
Mitsubishi Electric Corp
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 Mitsubishi Electric Corp filed Critical Mitsubishi Electric Corp
Priority to JP1058532A priority Critical patent/JPH02236787A/en
Publication of JPH02236787A publication Critical patent/JPH02236787A/en
Pending legal-status Critical Current

Links

Abstract

PURPOSE:To improve the reliability of state change detection by image-picking up a monitor object at prescribed time intervals, calculating a differential value between plural continuous image-picked up pictures for each picture element and executing the cumulative addition of this differential value for each picture element. CONSTITUTION:To an n-pieces of original pictures to be image-picked up and inputted at certain time intervals by an ITV camera 11, smoothing process is executed by a preprocessing circuit 13 for noise reduction. Next, in a differential circuit 16, difference is obtained between the continuous pictures out of the n-pieces of the pictures and the (n-1)-pieces of differential pictures are prepared. In a cumulative adder circuit 17, the total sum of the differential value for each picture is calculated to these (n-1)-pieces of the differential pictures. Then, a cumulative added picture is prepared and a state change area is extracted. Thus, even when liquid leaks from a pipe, etc., in a plant, etc., namely, even when a state change is limited in the small area, the state change can be detected with the satisfactory reliability.

Description

【発明の詳細な説明】 [産業上の利用分野コ この発明は形状、位置、状態が時間的な変化を呈する監
視対象物の変化の様子を検知するための状態変化検知装
置に関するものである。
DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] This invention relates to a state change detection device for detecting changes in a monitored object whose shape, position, and state change over time.

[従来の技術コ 従来、この種の状態変化検知装置においては、ITVカ
メラ等の撮像手段により撮像された基準画像と原画像と
の差分な取り、その差画像を用いて変化検知を行ってい
た。第4図は例えば特開昭62−147888号公報に
示された状態変化検知装置であり、図において(1)は
原画像フレームバッファ、〈2)は基準画像フレームバ
ッファ、(3)は差画{191算回路、〈4》は差画像
フレームバッファ、ク5)は投影回路、〈6〉は比較回
路、(7〉は判定回路である。
[Conventional technology] Conventionally, in this type of state change detection device, the difference between a reference image and an original image captured by an imaging means such as an ITV camera was taken, and the difference image was used to detect a change. . FIG. 4 shows a state change detection device disclosed in, for example, Japanese Unexamined Patent Publication No. 62-147888, in which (1) is an original image frame buffer, (2) is a reference image frame buffer, and (3) is a difference image {191 calculation circuit, <4> is a difference image frame buffer, 5) is a projection circuit, <6> is a comparison circuit, and (7> is a determination circuit.

変化検知のやり方としては、差画像計算回路(3〉で1
準られな差画像において、状態変化等による有効データ
とそれ以外の雑音等とを区別するために、投影回路(5
〉において、前記差画像の投影をX軸及びy軸方向に行
い、比較回路(6〉で各々の投影データに対して任意の
しきい値を設け、そのしきい値を超えるデータが存在す
れば、判定回路(7)により変化有りと判断していた。
The method for detecting changes is to use a difference image calculation circuit (1 in 3).
In an irregular difference image, a projection circuit (5
〉, the difference image is projected in the X-axis and y-axis directions, and a comparison circuit (6〉) sets an arbitrary threshold value for each projection data, and if there is data exceeding the threshold value, , the determination circuit (7) determined that there was a change.

[発明が解決しようとする課題] 従来の状態変化検知装置は以上のように構成されていた
ので、第5図<a)に示すように状態変化がある場合で
も、背景(釦の輝度と状態変化場所(9)の輝度の差が
小さい場合や、状態変化場所(9〉自身の輝度が小さい
場合、背景ク8》と状態変化場所〈9》のコシ1・ラス
1・のある差画像が得られず、第5図(b)及び(C)
に示すように、そのX軸、及びy軸投影データA.  
Bも非常に小さくなり、設定されたしきい値八B、 B
[+を越えるデータがなく、変化なしと判断される。ま
たしきい値Aa.  Bllを小さくすると、状態変化
場所<9〉だけでな《雑音等に対しても異常ありと判断
してしまう。さらに監視対象物が高速に運動するため,
状態変化が小領域である場合や、監視対象物が小さく状
態変化領域が小領域である場合等、スパーク状の雑音や
、背景物の振動やA/D変換時の量子化誤差による背景
物輪郭での線状雑音等との区別が困離であった。
[Problems to be Solved by the Invention] Since the conventional state change detection device is configured as described above, even when there is a state change as shown in FIG. If the difference in the brightness of the change location (9) is small, or if the brightness of the state change location (9) itself is small, the difference image with stiffness 1 and luster 1 between the background (8) and the state change location (9) will be Figure 5 (b) and (C)
As shown in the figure, the X-axis and y-axis projection data A.
B also becomes very small, and the set threshold 8B, B
[There is no data exceeding +, so it is determined that there is no change. Moreover, the threshold value Aa. If Bll is made small, it will be determined that there is an abnormality not only at the state change location <9> but also for noise, etc. Furthermore, since the monitored object moves at high speed,
When the state change is a small area, or when the monitored object is small and the state change area is a small area, spark-like noise, vibrations of the background object, and background object contours due to quantization errors during A/D conversion may occur. It was difficult to distinguish it from linear noise, etc.

この発明は上記のような問題点を解消するためになされ
たもので、ブラン1・等における配s、バルブ等からの
液体漏洩等、状態変化が小領域である場合や、監視対象
物が小さく状態変化領域が小領域である場合等も状態変
化を信頼性良く検知できる状態変化検知装置を提供する
ことを目的とする。
This invention was made in order to solve the above-mentioned problems, and it is used in cases where the state change is in a small area, such as liquid leakage from a valve, etc., or when the object to be monitored is small. It is an object of the present invention to provide a state change detection device that can reliably detect a state change even when the state change area is a small area.

[課題を解決するための手段] この発明に係わる状態変化検知装置は、監視対象物を所
定の時間間隔で撮像する撮像手段、二の撮像手段で撮像
された複数枚の連続画像間での差分値を各画素毎に計算
する計算手段、及び上記連続画像に対し、上記差分値を
上記各画素毎に累積加算する累積加算手段を備えたもの
である。
[Means for Solving the Problems] A state change detection device according to the present invention includes an imaging means for imaging an object to be monitored at predetermined time intervals, and a difference between a plurality of consecutive images taken by the second imaging means. The apparatus is equipped with a calculation means for calculating a value for each pixel, and a cumulative addition means for cumulatively adding the difference value for each pixel to the continuous images.

また、さらに累積加算手段より得られた累積加算画像か
ら背景を構成する輪郭線を除去する除去手段,及び輪郭
線除去後の累積加算画像を、監視対象物の移動方向と、
この移動方向に垂直な方向に投影し、各方向毎に所定の
しきい値を用いて状態変化画素を抽出する抽出手段を設
けてもよい。
Further, a removing means removes the contour line constituting the background from the cumulative addition image obtained by the cumulative addition means, and the cumulative addition image after the contour line removal is determined according to the moving direction of the monitored object.
An extraction means may be provided that projects in a direction perpendicular to this movement direction and extracts state-change pixels using a predetermined threshold value for each direction.

[作用コ この発明における状態変化検知装置は、背景の輝変と状
態変化場所の輝度の差が小さい場合や、状態変化場所自
身の輝麿が小さい場合でも、連続画像の差分値を累積加
算することで、変化領域の拡大強調を行なうことができ
、状態変化場所と背景とを分離し、雑音と区別可能なコ
ントラストのある差画像が得られる。
[Operation] The state change detection device according to the present invention cumulatively adds the difference values of successive images even when the difference in brightness between the background brightness change and the state change location is small, or when the brightness of the state change location itself is small. By doing so, it is possible to enlarge and emphasize the change area, separate the state change location from the background, and obtain a difference image with contrast that can be distinguished from noise.

また、状態変化が小領域である場合でも、スパーク状の
雑音や、また背景物の振動やA/D変換時における量子
化誤差により現われる背景物の輪郭での線状雑音等を除
去し、さらに状態変化の特徴を用いて,変化領域の検知
を信頼性良く、自動的に行なう。
In addition, even when the state change is in a small area, it removes spark-like noise and linear noise at the contour of the background object that appears due to background object vibrations and quantization errors during A/D conversion. Automatically and reliably detect changing regions using characteristics of state changes.

[実施例コ 以下、この発明の一実施例による状態変化検知・装置を
図について説明する。
[Embodiment] Hereinafter, a state change detection/device according to an embodiment of the present invention will be explained with reference to the drawings.

第1図及び第2図は各々この発明の一実施例による状態
変化検知装置を示すブロック構成図である。
FIGS. 1 and 2 are block diagrams each showing a state change detection device according to an embodiment of the present invention.

図において〈l1)はITVカメラ、(l2)は画像入
力回路、《l3〉は雑音除去のための前処理回路、<1
4〉〈15)は連続した画像を順次蓄積する画像メモI
J 1、画像メモリ2、(l6)は差分回路、(l7)
は累積加算回路、〈1釦は累積加算画像メモリであり、
状態変化領域の抽出を行なう。さらに、〈l9)は2値
化回路、〈20〉は輪郭線検出回路、〈21)は輪郭線
近傍領域除去回路であり、(22)は状態変化方向投影
回路、〈23〉は領域判定回路であり、状態変化の有無
を自動的に判別する。
In the figure, <l1) is an ITV camera, (l2) is an image input circuit, <l3> is a preprocessing circuit for noise removal, and <1
4〉〈15) is an image memo I that stores consecutive images in sequence.
J 1, image memory 2, (l6) is a differential circuit, (l7)
is the cumulative addition circuit, <1 button is the cumulative addition image memory,
Extract the state change region. Furthermore, <19> is a binarization circuit, <20> is a contour line detection circuit, <21) is a contour line vicinity area removal circuit, (22) is a state change direction projection circuit, and <23> is a region determination circuit. The presence or absence of a state change is automatically determined.

上記のように構成された状態変化検知装置においては、
まず、第1図に示す構成により、状態変化領域を抽出す
る。即ち、 ITVカメラ(11)よりある時間間隔で
撮像入力されるn枚の原画像(第3図(Q〉)に対し、
前処理回路(13)にて雑音除去のための平滑化処理を
施す。なお、第3図(n)中、〈90〉は水漏れ、〈8
0)は配管を示している。次に差分回路(l6)にてn
枚の画像のうち連続した画像間毎に差分をとり、n−1
枚の差分画像を作成する(第3図(b)》。累積加算回
路・〈l7)にてこのn−1枚の差分画像にないし、各
画素毎での差分値の総和を計算し、累積加算画像を作成
、状態変化領域の抽出を行なう (第3図(C))。
In the state change detection device configured as above,
First, a state change area is extracted using the configuration shown in FIG. That is, for n original images (Fig. 3 (Q)) input from the ITV camera (11) at certain time intervals,
A preprocessing circuit (13) performs smoothing processing to remove noise. In addition, in Fig. 3(n), <90> indicates water leakage, and <8
0) indicates piping. Next, in the differential circuit (l6), n
Differences are taken between consecutive images among the images, and n-1
Create two differential images (Figure 3(b)).The cumulative addition circuit (l7) calculates the sum of the differential values for each pixel that is not present in the n-1 differential images, and An added image is created and the state change area is extracted (Figure 3 (C)).

次に第2図に示す構成により状態変化の判別を自動的に
行なう。即ち、n枚の画像より作成された累積加算画像
に対し、2値化回路〈19)により、適切なしきい値で
2値化処理を施し、時間的変化を呈する成分のみを抽出
する(第3図(d))。そして輪郭線検出回路(20》
にて別途、画像入力回路(l2〉より簿られる原画像を
微分処理し、さらに2値化処理して輪郭線画像(第3図
《e))を作成する。次に輪郭線近傍領域除去回路〈2
l》では、上記輪郭線画像(第3図〈e〉)により時間
変化成分画像(第3図(d))をマスクする。そして背
景を構成する輪郭線(81)が存在する画素については
、時間変化成分画像の値を零とし、背景環境の振動やA
/D変換時の量子化誤差による輪郭線近傍での雑音を除
去する《第3図《f))。次に状態変化方向投影回路(
22)では液体落下の方向性を利用し、落下方向に画像
を投影、所定のしきい値以上の領域を抽出し、状態変化
候補領域を抽出する(第3図クg》)。この状態変化候
補領域に対し、領域判別回路<23八二で落下方向と垂
直な方向に投影を行(・,所定のしまい値を用いて状態
変化画素を抽出、確定するく第3図(h))。
Next, the configuration shown in FIG. 2 automatically determines whether the state has changed. That is, the cumulative addition image created from n images is subjected to binarization processing using an appropriate threshold value by the binarization circuit (19), and only the components exhibiting temporal changes are extracted (the third Figure (d)). And the contour detection circuit (20)
Separately, the original image input from the image input circuit (12) is differentiated and then binarized to create a contour image (Fig. 3 (e)). Next, the contour line vicinity area removal circuit 〈2
1), the time-varying component image (FIG. 3(d)) is masked by the contour image (FIG. 3(e)). For pixels where there is a contour line (81) constituting the background, the value of the time-varying component image is set to zero, and vibrations in the background environment and A
Noise near the contour line due to quantization error during /D conversion is removed (Fig. 3 (f)). Next, the state change direction projection circuit (
In 22), an image is projected in the direction of the drop by utilizing the directionality of the liquid falling, and a region exceeding a predetermined threshold value is extracted, and a state change candidate region is extracted (Fig. 3, g). For this state change candidate region, a region discrimination circuit <2382 projects in a direction perpendicular to the falling direction (Fig. 3 (h )).

なお、上記実施例では状態変化の判別を第2図に示す構
成により、自動化したものを示したが、このような手段
を用いず、累積加算回路(17)にて1孕られた累積加
算画像を用いて状態変化を目視するようにしてもよい。
In the above embodiment, the state change determination was automated using the configuration shown in FIG. Changes in status may be visually observed using the .

[発明の効果] 以上のように、この発明によれば監視対象物を所定の時
間間隔で撮像する撮像手段、この撮像手段で撮像された
複数枚の連続画像間での差分値を各画素毎に計算する計
算手段、及び上記連続画像に対し、上記差分値を上記各
画素苺に累積加算する累積加算手段により状態変化検知
装置を構成したので、水漏れのような変化率の小さい漏
れ等もノイズと区別して検出でき、信頼性の高い状態変
化検知装置が辱られる効果がある。
[Effects of the Invention] As described above, according to the present invention, there is an imaging means for imaging an object to be monitored at predetermined time intervals, and a difference value between a plurality of consecutive images taken by the imaging means is calculated for each pixel. Since the state change detection device is configured with a calculation means that calculates the difference value and a cumulative addition means that cumulatively adds the difference value to each pixel of the continuous image, it is possible to detect leaks with a small rate of change such as water leaks. It can be detected separately from noise, and has the effect of humiliating highly reliable state change detection devices.

また、さらに累積加算手段より簿られな累積加算画像か
ら背景を構成する輪郭線を除去し、輪郭線除去後の累積
加算画像を、監視対象物の移動方向と、この移動方向に
垂直な方向に投影し、各方向毎に所定のしきい値を用い
て状態変化画素を抽出するようにしたので,水漏れ等、
高速で変化する状態変化を自動的に、信頼性良《検出で
き、従来のようにプラン1・監視員が常時モニタテレビ
を監視する必要がなく,監視員の負担を軽減できるとと
もに、プラント運転の監視性及び操作性、店頼性を向上
できる効果がある。
Further, the cumulative addition means removes the contour line forming the background from the recorded cumulative addition image, and the cumulative addition image after the contour line removal is divided into the moving direction of the monitored object and the direction perpendicular to this moving direction. By projecting the image and extracting state-changing pixels using a predetermined threshold value in each direction, water leakage, etc.
Changes in conditions that change at high speed can be detected automatically and reliably. Monitoring personnel do not need to constantly monitor the TV monitor as in the past, reducing the burden on the monitoring personnel and improving plant operation. This has the effect of improving monitoring performance, operability, and store reliability.

【図面の簡単な説明】[Brief explanation of drawings]

第1図及び第2図は各々、この発明の一実施例による状
態変化検知装置を示すブロック構成図、第3図(a)〜
(h)は各々この発明の一実施例による状態変化検知装
置の動作を説明する説明図、第4図は従来の状態変化検
知装置を示すブロック構成図、並びに第5図(aXb)
(c)は、各々従来の状態変化検知装置の動作を説明す
る説明図である。 図において、クl1)はITVカメラ、(l2〉は画像
入力回路、(14)(15)は画像メモリ、(l6)は
差分回路、(17〉は累積加算回路、〈18〉は累積加
算画像メモIJ,  (20)は輪郭線検出回路、ク2
1)は輪郭線近傍領域除去回路、(22)は状態変化方
向投影回路、(23〉は領域判定回路である。 なお、図中、同一符号は同一または相当部分を示す。
FIGS. 1 and 2 are block diagrams showing a state change detection device according to an embodiment of the present invention, and FIGS.
(h) is an explanatory diagram illustrating the operation of a state change detection device according to an embodiment of the present invention, FIG. 4 is a block diagram showing a conventional state change detection device, and FIG. 5 (aXb)
(c) is an explanatory diagram illustrating the operation of each conventional state change detection device. In the figure, (l1) is an ITV camera, (l2> is an image input circuit, (14) and (15) are image memories, (l6) is a difference circuit, (17> is a cumulative addition circuit, and <18> is a cumulative addition image. Memo IJ, (20) is the contour line detection circuit, Ku2
1) is a contour line vicinity area removal circuit, (22) is a state change direction projection circuit, and (23> is an area determination circuit. In the figures, the same reference numerals indicate the same or equivalent parts.

Claims (2)

【特許請求の範囲】[Claims] (1)監視対象物を所定の時間間隔で撮像する撮像手段
、この撮像手段で撮像された複数枚の連続画像間での差
分値を各画素毎に計算する計算手段、及び上記連続画像
に対し、上記差分値を上記各画素毎に累積加算する累積
加算手段を備えた状態変化検知装置。
(1) An imaging means for taking images of the object to be monitored at predetermined time intervals, a calculation means for calculating the difference value for each pixel between a plurality of consecutive images taken by this imaging means, and , a state change detection device comprising cumulative addition means for cumulatively adding the difference value for each pixel.
(2)累積加算手段より得られた累積加算画像から背景
を構成する輪郭線を除去する除去手段、及び輪郭線除去
後の累積加算画像を、監視対象物の移動方向と、この移
動方向に垂直な方向に投影し、各方向毎に所定のしきい
値を用いて状態変化画素を抽出する抽出手段を備えた請
求項1記載の状態変化検知装置。
(2) a removal means for removing the contour line constituting the background from the cumulative addition image obtained by the cumulative addition means; and a removal means for removing the contour line forming the background from the cumulative addition image obtained by the cumulative addition means; 2. The state change detection device according to claim 1, further comprising an extraction means for projecting the pixel in each direction and extracting the state change pixel using a predetermined threshold value for each direction.
JP1058532A 1989-03-10 1989-03-10 State change detector Pending JPH02236787A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP1058532A JPH02236787A (en) 1989-03-10 1989-03-10 State change detector

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP1058532A JPH02236787A (en) 1989-03-10 1989-03-10 State change detector

Publications (1)

Publication Number Publication Date
JPH02236787A true JPH02236787A (en) 1990-09-19

Family

ID=13087043

Family Applications (1)

Application Number Title Priority Date Filing Date
JP1058532A Pending JPH02236787A (en) 1989-03-10 1989-03-10 State change detector

Country Status (1)

Country Link
JP (1) JPH02236787A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006031636A (en) * 2004-07-21 2006-02-02 Sony Corp Detection device and method for both eyes
JP2009133085A (en) * 2007-11-29 2009-06-18 Central Japan Railway Co Crack checking device for tunnel lining
JP2009194530A (en) * 2008-02-13 2009-08-27 Casio Comput Co Ltd Imaging apparatus, stroboscopic image generation method, and program
EP2998905A1 (en) 2014-09-17 2016-03-23 Panasonic Intellectual Property Management Co., Ltd. Substance detection device

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006031636A (en) * 2004-07-21 2006-02-02 Sony Corp Detection device and method for both eyes
JP2009133085A (en) * 2007-11-29 2009-06-18 Central Japan Railway Co Crack checking device for tunnel lining
JP2009194530A (en) * 2008-02-13 2009-08-27 Casio Comput Co Ltd Imaging apparatus, stroboscopic image generation method, and program
JP4513869B2 (en) * 2008-02-13 2010-07-28 カシオ計算機株式会社 Imaging apparatus, strobe image generation method, and program
US7995102B2 (en) 2008-02-13 2011-08-09 Casio Computer Co., Ltd. Imaging apparatus for generating stroboscopic image
EP2998905A1 (en) 2014-09-17 2016-03-23 Panasonic Intellectual Property Management Co., Ltd. Substance detection device
US9787907B2 (en) 2014-09-17 2017-10-10 Panasonic Intellectual Property Management Co., Ltd. Substance detection device

Similar Documents

Publication Publication Date Title
JP5644097B2 (en) Image processing apparatus, image processing method, and program
JPH0323860B2 (en)
US10853949B2 (en) Image processing device
JP2008009849A (en) Person tracking device
CN108921840A (en) Display screen peripheral circuit detection method, device, electronic equipment and storage medium
CN109030499B (en) Device and method suitable for continuous online detection of target defects and preventing repeated counting of defect number
Verma et al. Analysis of moving object detection and tracking in video surveillance system
JPH02236787A (en) State change detector
KR101395666B1 (en) Surveillance apparatus and method using change of video image
JPH07249128A (en) Picture processor for vehicle
JP2001043383A (en) Image monitoring system
JP3571628B2 (en) Image processing device
CN113936242B (en) Video image interference detection method, system, device and medium
JPH04345396A (en) Traveling object tracing method
JPH0514891A (en) Image monitor device
JP4998955B2 (en) Collapse detection system and method
JPH04311187A (en) Method for extracting change area of image to be monitored
JPH02236788A (en) State change detector
JP6348020B2 (en) Image processing apparatus, image processing method, and inspection method using the same
JPH04261290A (en) Moving object tracing device
KR102004724B1 (en) Method for detecting sperm and record media recorded program for realizing the same
JP3138381B2 (en) Monitoring device
WO2023121012A1 (en) Method and system of analyzing motion based on feature tracking
JPH04200084A (en) Image monitor device
CN112597800B (en) Method and system for detecting sitting-up actions of students in recording and broadcasting system