EP1485889B1 - Method for following at least one object in a scene - Google Patents

Method for following at least one object in a scene Download PDF

Info

Publication number
EP1485889B1
EP1485889B1 EP03720139A EP03720139A EP1485889B1 EP 1485889 B1 EP1485889 B1 EP 1485889B1 EP 03720139 A EP03720139 A EP 03720139A EP 03720139 A EP03720139 A EP 03720139A EP 1485889 B1 EP1485889 B1 EP 1485889B1
Authority
EP
European Patent Office
Prior art keywords
scene
time
list
image
movement
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.)
Expired - Lifetime
Application number
EP03720139A
Other languages
German (de)
French (fr)
Other versions
EP1485889A1 (en
Inventor
Frank Rottmann
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.)
Robert Bosch GmbH
Original Assignee
Robert Bosch GmbH
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 Robert Bosch GmbH filed Critical Robert Bosch GmbH
Publication of EP1485889A1 publication Critical patent/EP1485889A1/en
Application granted granted Critical
Publication of EP1485889B1 publication Critical patent/EP1485889B1/en
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B13/00—Burglar, theft or intruder alarms
    • G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
    • G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
    • G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
    • G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
    • G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B13/00—Burglar, theft or intruder alarms
    • G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
    • G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
    • G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
    • G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
    • G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
    • G08B13/19604—Image analysis to detect motion of the intruder, e.g. by frame subtraction involving reference image or background adaptation with time to compensate for changing conditions, e.g. reference image update on detection of light level change
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B13/00—Burglar, theft or intruder alarms
    • G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
    • G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
    • G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
    • G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
    • G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
    • G08B13/19608—Tracking movement of a target, e.g. by detecting an object predefined as a target, using target direction and or velocity to predict its new position
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B13/00—Burglar, theft or intruder alarms
    • G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
    • G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
    • G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
    • G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
    • G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
    • G08B13/19613—Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion

Definitions

  • the invention is based on a method for tracking at least one object in a scene.
  • the Japanese patent application JP 08 150125 discloses monitoring a sickroom with an imaging device, wherein an alarm is generated when an abnormal condition of a patient, such as falling, persists for a certain period of time.
  • the method according to the invention for tracking at least one object in a scene having the features of the independent patent claim has the advantage of generating a signaling at a specific point as a function of the dwell time of an object.
  • an atypical behavioral pattern is detected immediately in a supervised scene where typically moving objects are located. This can lead to faster alerting and thus better monitoring. Lingering objects that come to a standstill from a movement are detected and then lead to signaling.
  • the signaling which is generated in dependence on the count, causes an alarm. This means that when the object is idle for a predetermined time, an alarm is generated.
  • the predetermined time represents a threshold value for the counter reading.
  • the movement of a respective object is described by a list or a matrix, the image coordinates of the object being stored in the x-direction at different times for the respective images of a sequence of images, and the y-values in a second line vertical values, in the scene at these times.
  • the x and y values at a given time represent a motion vector, that is, in the direction in which the object is moving. If this motion vector is zero, then a standstill is detected and the counter is incremented.
  • the list can be managed dynamically or, if a predefined number of list places have been processed, totals can be transferred to a new list for this object. This list representation allows several objects to be observed simultaneously and tracked in this way.
  • the image generator is a camera that generates the image sequence.
  • the images are generated here at relatively large intervals of, for example, half a second in order to capture the corresponding movements meaningful.
  • Such motion vectors in a list can also be generated between images that are not directly consecutive, for example very slow movements. Thus, a large resolution of the movement is possible. This can be done dynamically, ie no movement is detected, then a counter can be started, but at the same time a comparison continues to take place up to a specific image sequence number, from which then finally to a standstill is detected and the count is ultimately relevant.
  • the list can then be reinitialized each time a movement is reinstated. If there are already values from a past list for the object that can be identified by various object recognition methods, these can be transferred to the new list. This also makes it possible to analyze a whole sequence of movements. This then represents a very efficient method in terms of storage resources.
  • a reference image is generated in order to recognize an object in a simple manner.
  • a so-called reference image which possibly contains only the background of the scene without tracked objects. This is basically not obtained in any object present from, for example, the penultimate picture.
  • This image can then be taken as a reference.
  • This method works especially for a few objects that are only briefly in the scene.
  • a reference image is generated in such a way that after the object positions have been determined in the current image, the reference image is taken over from the penultimate image into the remaining region. This is then called a local adaptation of the reference image.
  • the inventive method is used in particular in a video surveillance, which has at least one imager, a processor, a memory and also output means, with which then a signaling, such as an alarm, can take place.
  • a signaling such as an alarm
  • the signaling can also be used as a signal for other systems.
  • This video surveillance can be used in particular for monitoring a parking space.
  • FIG. 1 a block diagram of the video surveillance according to the invention
  • FIG. 2 a flow chart of the method according to the invention
  • FIG. 3 a first example of video surveillance
  • FIG. 4 a second example of video surveillance
  • FIG. 5 a third example of video surveillance.
  • Video sensors based on modern computer architectures are able to detect objects that are directed in the sequence of images of stationary cameras. Usually, the approach of exceeding predetermined thresholds with respect to a line distance or size is followed. This crossing is detected by changes in the image signal, for example by movement of objects such as persons, and a corresponding signaling or image recording takes place. In this way, for example, the task of alerting an unauthorized access is handled.
  • an object is recognized within a short time as a directed moving object and again traced through the scene, which is also referred to as tracking, until the object leaves the image again.
  • This will be in FIG. 3 shown.
  • a person enters a scene monitored by an imager, such as a video sensor or other camera such as a thermal imager. Since the evaluation of the video surveillance detects only movements, the object is not yet recognized in the first image sequence. Only at time 14, after a motion was detected, was this object identified. At time 15, the object that is moving continuously has left the observed scene.
  • FIG. 4 shows a second scenario.
  • a person enters the scene at time 14 and is recognized as a moving object at time 14.
  • the person comes to a halt from the movement, whereby now the inventive method starts a counter to monitor the residence time.
  • a signaling is generated. This can mean, for example, an alarm.
  • FIG. 5 shows a third scenario, further possibilities, which can occur with several objects.
  • An object 14 is recognized as being moved.
  • a split ie a separation of the object into two objects, is made here. This is detected by different motion vectors emanating from the same object.
  • Point 18 shows a so-called merge, ie a merger of two objects, which can then lead again to a split. This merge is caused by another object 19, which was recognized as a moving object 20.
  • the first object leaves to Time 21, the scene, while the second object at time 22 from the movement comes to a standstill.
  • FIG. 1 now shows a block diagram of a video surveillance according to the invention.
  • An imager here a surveillance camera 1
  • a memory 3 is connected to the processor 2 via a data input / output.
  • the processor 2 is connected to a control 4, which is connected on the one hand to a loudspeaker 5 and on the other hand to a display 6.
  • step 7 with the camera 1 and the processor 2, as in FIG. 3 represented, a moving object detected by the motion vector.
  • a list or matrix is set up, in which, for example, each column denotes a specific image in an image sequence, the image sequences being separated by time intervals, for example one second.
  • the list has two lines which define the motion vector in a plane. This is usually defined by the coordinates x and y, for example.
  • List 1 a motion vector with x value 123 and y value 12 at time 0 is detected.
  • time 99 that's 50 seconds, a standstill is counted, which has already been used at time 1, which corresponds to one second.
  • This is a list that is predefined, ie only allows new 100 entries. Therefore, when the time reaches 99, the list is reinitialized, and the second list that takes the value is passed.
  • the value 123 and 12 are entered again and at time 1 the standstill 0/0.
  • the object is tracked in step 8. If a standstill is detected in method step 9, specifically by the entries 0/0 in the list, the counter is started in method step 10. If this is not the case, then the object is tracked further with the list. If, however, the counter has been started in method step 10, then it is checked in method step 11 whether the threshold which is predetermined is reached. This threshold requires when they are reached in step 12 signaling. Die.Signalmaschine can be done here by means of the speaker 5 and the display 6, ie for example, the output of an alarm. If, however, the threshold is not reached and the object moves again, the process jumps back to step 8 and the object tracking is resumed.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Closed-Circuit Television Systems (AREA)
  • Image Analysis (AREA)

Abstract

Disclosed is a method for following at least one object in a scene, which is characterized by the fact that the time is counted during which an object is at a standstill once said object has stopped moving so as to generate a signaling according to the duration of the standstill position. The object is described by means of a list or a matrix in relation to the direction of movement and the time.

Description

Stand der TechnikState of the art

Die Erfindung geht aus von einem Verfahren zur Verfolgung wenigstens eines Objekts in einer Szene.The invention is based on a method for tracking at least one object in a scene.

Die japanische Offenlegungsschrift JP 08 150125 offenbart die Überwachung eines Krankenzimmers mit einer Bildaufnahmevorrichtung, wobei ein Alarm erzeugt wird, wenn ein abnormaler Zustand eines Patienten, beispielsweise ein Hinfallen, über einen bestimmten Zeitraum fortbesteht.The Japanese patent application JP 08 150125 discloses monitoring a sickroom with an imaging device, wherein an alarm is generated when an abnormal condition of a patient, such as falling, persists for a certain period of time.

Vorteile der ErfindungAdvantages of the invention

Das erfindungsgemäße Verfahren zur Verfolgung wenigstens eines Objekts in einer Szene mit den Merkmalen des unabhängigen Patentanspruchs hat den Vorteil, in Abhängigkeit von der Verweildauer eines Objekts an einem bestimmten Punkt eine Signalisierung zu erzeugen. Dadurch wird in einer überwachten Szene, in der sich üblicherweise sich bewegende Objekte befinden, ein atypisches Verhaltensmuster sofort erkannt. Dies kann zu einer schnelleren Alarmierung und damit besseren Überwachung führen. Dabei werden verweilende Objekte, die aus einer Bewegung heraus zum Stillstand kommen, erkannt und führen dann zur Signalisierung.The method according to the invention for tracking at least one object in a scene having the features of the independent patent claim has the advantage of generating a signaling at a specific point as a function of the dwell time of an object. As a result, an atypical behavioral pattern is detected immediately in a supervised scene where typically moving objects are located. This can lead to faster alerting and thus better monitoring. Lingering objects that come to a standstill from a movement are detected and then lead to signaling.

Besonders vorteilhaft ist, daß die Signalisierung, die in Abhängigkeit von dem Zählerstand erzeugt wird, einen Alarm hervorruft. Das bedeutet, daß wenn das Objekt für eine vorgegebene Zeit stillsteht, ein Alarm erzeugt wird. Die vorgegebene Zeit stellt einen Schwellwert für den Zählerstand dar.It is particularly advantageous that the signaling, which is generated in dependence on the count, causes an alarm. This means that when the object is idle for a predetermined time, an alarm is generated. The predetermined time represents a threshold value for the counter reading.

Die Bewegung eines jeweiligen Objekts wird dabei durch eine Liste oder eine Matrix beschrieben, wobei in einer Zeile die Bildkoordinaten des Objekts in x-Richtung zu verschiedenen Zeitpunkten für die jeweiligen Bilder einer Bildfolge abgelegt werden und in einer zweiten Zeile die y-Werte, also die vertikalen Werte, in der Szene zu diesen Zeitpunkten. Der x- und der y-Wert zu einem vorgegebenen Zeitpunkt stellt einen Bewegungsvektor dar, also in der Richtung, in der sich das Objekt bewegt. Ist dieser Bewegungsvektor Null, dann wird ein Stillstand erkannt und der Zähler inkrementiert. Die Liste kann dynamisch verwaltet werden oder, wenn eine vorgegebene Anzahl von Listenplätzen abgearbeitet ist, können Summenwerte in eine neue Liste für dieses Objekt überführt werden. Diese Listenrepräsentation macht es möglich, daß gleichzeitig mehrere Objekte beobachtet werden können und auf diese Weise verfolgt werden können. Dies wird dann von einem Prozessor der Videoüberwachung und im zugehörigen Speicher verwaltet. Als Bildgeber wirkt dabei eine Kamera, die die Bildfolge erzeugt. Die Bilder werden hier in relativ großen Abständen von beispielsweise einer halben Sekunde erzeugt, um die entsprechenden Bewegungen sinnvoll erfassen zu können. Solche Bewegungsvektoren in einer Liste können auch zwischen Bildern erzeugt werden, die nicht direkt aufeinander folgen, beispielsweise bei sehr langsamen Bewegungen. Damit ist eine große Auflösung der Bewegung möglich. Dies kann dynamisch erfolgen, d.h. wird keine Bewegung erkannt, dann kann ein Zähler gestartet werden, aber gleichzeitig weiterhin ein Vergleich stattfinden bis zu einer bestimmten Bildfolgenzahl, ab der dann endgültig auf einen Stillstand erkannt wird und der Zählerstand letztlich relevant ist.In this case, the movement of a respective object is described by a list or a matrix, the image coordinates of the object being stored in the x-direction at different times for the respective images of a sequence of images, and the y-values in a second line vertical values, in the scene at these times. The x and y values at a given time represent a motion vector, that is, in the direction in which the object is moving. If this motion vector is zero, then a standstill is detected and the counter is incremented. The list can be managed dynamically or, if a predefined number of list places have been processed, totals can be transferred to a new list for this object. This list representation allows several objects to be observed simultaneously and tracked in this way. This is then managed by a video surveillance processor and its associated memory. The image generator is a camera that generates the image sequence. The images are generated here at relatively large intervals of, for example, half a second in order to capture the corresponding movements meaningful. Such motion vectors in a list can also be generated between images that are not directly consecutive, for example very slow movements. Thus, a large resolution of the movement is possible. This can be done dynamically, ie no movement is detected, then a counter can be started, but at the same time a comparison continues to take place up to a specific image sequence number, from which then finally to a standstill is detected and the count is ultimately relevant.

Die Liste kann dann jeweils bei einem Neueinsetzen einer Bewegung neu initialisiert werden. Liegen für das Objekt, das durch verschiedene Objekterkennungsmethoden identifiziert werden kann, bereits Werte aus einer vergangenen Liste vor, können diese in die neue Liste überführt werden. Damit ist es auch möglich, einen ganzen Bewegungsablauf zu analysieren. Dies stellt dann eine sehr effiziente Methode im Hinblick auf die Speicherressourcen dar.The list can then be reinitialized each time a movement is reinstated. If there are already values from a past list for the object that can be identified by various object recognition methods, these can be transferred to the new list. This also makes it possible to analyze a whole sequence of movements. This then represents a very efficient method in terms of storage resources.

Weiterhin ist es von Vorteil, daß ein Referenzbild erzeugt wird, um ein Objekt auf einfache Weise zu erkennen. Zur Generierung einer Zeit- und Objektmaske ist es notwendig, ein sogenanntes Referenzbild zu erhalten, das möglichst nur den Hintergrund der Szene ohne verfolgte Objekte enthält. Dies wird grundsätzlich bei keinem anwesenden Objekt aus beispielsweise dem vorletzten Bild gewonnen. Dieses Bild kann dann als Referenz übernommen werden. Dieses Verfahren funktioniert insbesondere für wenige Objekte, die sich auch nur kurzzeitig in der Szene befinden. Werden jedoch über größere Zeiträume viele Objekte verfolgt, dann wird ein Referenzbild in der Weise erzeugt, daß nach erfolgter Bestimmung der Objektpositionen im aktuellen Bild das Referenzbild in den übrigen Bereich aus dem vorletzten Bild übernommen wird. Dies wird dann als lokale Adaption des Referenzbildes bezeichnet.Furthermore, it is advantageous that a reference image is generated in order to recognize an object in a simple manner. In order to generate a time and object mask, it is necessary to obtain a so-called reference image, which possibly contains only the background of the scene without tracked objects. This is basically not obtained in any object present from, for example, the penultimate picture. This image can then be taken as a reference. This method works especially for a few objects that are only briefly in the scene. However, if many objects are tracked over relatively long periods of time, then a reference image is generated in such a way that after the object positions have been determined in the current image, the reference image is taken over from the penultimate image into the remaining region. This is then called a local adaptation of the reference image.

Das erfindungsgemäße Verfahren wird insbesondere in einer Videoüberwachung verwendet, die wenigstens einen Bildgeber, einen Prozessor, einen Speicher und auch Ausgabemittel aufweist, mit denen dann eine Signalisierung, beispielsweise ein Alarm, erfolgen kann. Die Signalisierung kann jedoch auch als Signal für andere Systeme verwendet werden. Diese Videoüberwachung kann insbesondere zur Überwachung eines Parkraumes verwendet werden.The inventive method is used in particular in a video surveillance, which has at least one imager, a processor, a memory and also output means, with which then a signaling, such as an alarm, can take place. However, the signaling can also be used as a signal for other systems. This video surveillance can be used in particular for monitoring a parking space.

Zeichnungdrawing

Ausführungsbeispiele der Erfindung sind in der Zeichnung dargestellt und werden in der nachfolgenden Beschreibung näher erläutert. Es zeigen Figur 1 ein Blockschaltbild der erfindungsgemäßen Videoüberwachung, Figur 2 ein Flußdiagramm des erfindungsgemäßen Verfahrens, Figur 3 ein erstes Beispiel einer Videoüberwachung, Figur 4 ein zweites Beispiel einer Videoüberwachung und Figur 5 ein drittes Beispiel einer Videoüberwachung.Embodiments of the invention are illustrated in the drawings and are explained in more detail in the following description. Show it FIG. 1 a block diagram of the video surveillance according to the invention, FIG. 2 a flow chart of the method according to the invention, FIG. 3 a first example of video surveillance, FIG. 4 a second example of video surveillance and FIG. 5 a third example of video surveillance.

Beschreibungdescription

Videosensoren auf der Basis moderner Rechnerarchitekturen sind in der Lage, Objekte, die sich gerichtet in Bildfolgen stationärer Kameras bewegen, zu detektieren. Üblicherweise wird der Ansatz des Überschreitens von vorgegebenen Schwellwerten bezüglich einer Streckendistanz oder -größe verfolgt. Dabei wird dieses Überschreiten durch Änderungen im Bildsignal, beispielsweise durch Bewegung von Objekten wie Personen, erfaßt, und eine entsprechende Signalisierung bzw. Bildaufzeichnung erfolgt. Auf diese Weise wird beispielsweise die Aufgabe der Alarmierung eines unerlaubten Zutritts abgewickelt.Video sensors based on modern computer architectures are able to detect objects that are directed in the sequence of images of stationary cameras. Usually, the approach of exceeding predetermined thresholds with respect to a line distance or size is followed. This crossing is detected by changes in the image signal, for example by movement of objects such as persons, and a corresponding signaling or image recording takes place. In this way, for example, the task of alerting an unauthorized access is handled.

Erfindungsgemäß ist es nun möglich, Objekte vollständig durch eine Szene zu verfolgen. Ein Objekt wird beim Betreten der Szene innerhalb kurzer Zeit als gerichtet bewegtes Objekt erkannt und wieder durch die Szene verfolgt, was auch als Tracking bezeichnet wird, bis das Objekt das Bild wieder verläßt. Dies wird in Figur 3 dargestellt. Am Ort 13 betritt eine Person eine Szene, die durch einen Bildgeber, also beispielsweise einen Videosensor oder eine andere Kamera wie eine Wärmebildkamera, überwacht wird. Da die Auswertung der Videoüberwachung nur Bewegungen erkennt, wird in der ersten Bildfolge das Objekt noch nicht erkannt. Erst zum Zeitpunkt 14, nachdem eine Bewegung erkannt wurde, wurde dieses Objekt identifiziert. Zum Zeitpunkt 15 hat das Objekt, das sich fortlaufend bewegt, die beobachtete Szene verlassen.According to the invention, it is now possible to completely track objects through a scene. Upon entering the scene, an object is recognized within a short time as a directed moving object and again traced through the scene, which is also referred to as tracking, until the object leaves the image again. This will be in FIG. 3 shown. At location 13, a person enters a scene monitored by an imager, such as a video sensor or other camera such as a thermal imager. Since the evaluation of the video surveillance detects only movements, the object is not yet recognized in the first image sequence. Only at time 14, after a motion was detected, was this object identified. At time 15, the object that is moving continuously has left the observed scene.

Figur 4 zeigt ein zweites Szenario. Auch hier betritt zum Zeitpunkt 14 eine Person die Szene und wird zum Zeitpunkt 14 als ein sich bewegendes Objekt erkannt. Zum Zeitpunkt 16 kommt jedoch die Person aus der Bewegung zum Stillstand, wobei nun das erfindungsgemäße Verfahren einen Zähler startet, um die Verweildauer zu überwachen. In Abhängigkeit von der Verweildauer wird eine Signalisierung erzeugt. Dies kann beispielsweise einen Alarm bedeuten. FIG. 4 shows a second scenario. Here as well, a person enters the scene at time 14 and is recognized as a moving object at time 14. At time 16, however, the person comes to a halt from the movement, whereby now the inventive method starts a counter to monitor the residence time. Depending on the residence time, a signaling is generated. This can mean, for example, an alarm.

Figur 5 zeigt ein drittes Szenario, weitere Möglichkeiten, die bei mehreren Objekten auftreten können. Ein Objekt 14 wird als bewegt erkannt. Zum Zeitpunkt 17 wird hier ein Split, also ein Auftrennen des Objekts in zwei Objekte vorgenommen. Dies wird durch unterschiedliche Bewegungsvektoren, die von dem gleichen Objekt ausgehen, erkannt. Zeitpunkt 18 zeigt einen sogenannten Merge, also eine Verschmelzung von zwei Objekten, die dann wieder zu einem Split führen kann. Dieser Merge wird durch ein weiteres Objekt 19, das als bewegendes Objekt 20 erkannt wurde, hervorgerufen. Das ersten Objekt verläßt zum Zeitpunkt 21 die Szene, während das zweite Objekt zum Zeitpunkt 22 aus der Bewegung zum Stillstand kommt. FIG. 5 shows a third scenario, further possibilities, which can occur with several objects. An object 14 is recognized as being moved. At time 17, a split, ie a separation of the object into two objects, is made here. This is detected by different motion vectors emanating from the same object. Point 18 shows a so-called merge, ie a merger of two objects, which can then lead again to a split. This merge is caused by another object 19, which was recognized as a moving object 20. The first object leaves to Time 21, the scene, while the second object at time 22 from the movement comes to a standstill.

Figur 1 zeigt nun ein Blockschaltbild einer erfindungsgemäßen Videoüberwachung. Ein Bildgeber, hier eine Überwachungskamera 1, ist an einen Prozessor 2 angeschlossen. Über einen Datenein-/-ausgang ist ein Speicher 3 mit dem Prozessor 2 verbunden. Über eine Datenleitung ist der Prozessor 2 mit einer Ansteuerung 4 verbunden, die einerseits mit einem Lautsprecher 5 verbunden ist und andererseits mit einer Anzeige 6. FIG. 1 now shows a block diagram of a video surveillance according to the invention. An imager, here a surveillance camera 1, is connected to a processor 2. A memory 3 is connected to the processor 2 via a data input / output. Via a data line, the processor 2 is connected to a control 4, which is connected on the one hand to a loudspeaker 5 and on the other hand to a display 6.

Beispielhaft ist hier nur eine Kamera 1 angegeben. Es können jedoch mehrere Kameras vorhanden sein, um mehrere Szenen gleichzeitig zu überwachen und von einem Prozessor 2 bedienen zu lassen. Weiterhin ist hier beispielhaft nur ein Lautsprecher 5 bzw. eine Anzeige 6 dargestellt, die zur Ausgabe eines Alarms dienen. Die Signalisierung, die vom Prozessor 2 zur Ansteuerung 4 übertragen wird, kann jedoch auch verwendet werden. Auf dem Prozessor 2 läuft das erfindungsgemäße Verfahren, das durch das in Figur 2 dargestellte Flußdiagramm nun erläutert wird. In Verfahrensschritt 7 wird mit der Kamera 1 und dem Prozessor 2, wie in Figur 3 dargestellt, ein sich bewegendes Objekt anhand des Bewegungsvektors erkannt. Dazu wird eine Liste oder Matrix aufgestellt, bei der beispielsweise jede Spalte ein bestimmtes Bild in einer Bildfolge bezeichnet, wobei die Bildfolgen durch Zeitintervalle, beispielsweise eine Sekunde, getrennt sind. Weiterhin hat die Liste zwei Zeilen, die den Bewegungsvektor in einer Ebene definieren. Dies wird beispielsweise durch die Koordinaten x und y üblicherweise definiert. Im Folgenden sind zwei Listen dargestellt, die dies illustrieren. In der Liste 1 wird ein Bewegungsvektor mit dem x-Wert 123 und dem y-Wert 12 zum Zeitpunkt 0 erkannt. Zum Zeitpunkt 99, das entspricht hier 50 Sekunden, wird ein Stillstand gezählt, der bereits beim Zeitpunkt 1, das entspricht einer Sekunde, eingesetzt hat. Dies ist eine Liste, die vordefiniert ist, d.h. nur neue 100 Einträge erlaubt. Daher wird die Liste dann, bei Erreichen des Zeitpunkts 99, neu initialisiert, und man geht zur zweiten Liste über, die den Wert übernimmt. Hier dargestellt, wird zum Zeitpunkt 0 wieder der Wert 123 und 12 eingetragen und zum Zeitpunkt 1 der Stillstand 0/0. Zum Zeitpunkt 2 jedoch werden nun die 50 Sekunden hinzuaddiert, und es wird neu weitergezählt. Liste 1: Zeitpkt 0 1 2 3 .... 99 X 123 1 2 0 .... 0 Y 12 0 3 0 ..... 0 0:00 0:01 0:02 0:03 0:50 Liste 1, Neuinitialisiert: Zeitpkt 0 1 2 3 X 126 0 0 0 Y 15 0 0 0 0:00 0:03 0:51 0:52 By way of example only one camera 1 is indicated here. However, multiple cameras may be present to simultaneously monitor and operate multiple scenes from a processor 2. Furthermore, only a loudspeaker 5 or a display 6 is shown here by way of example, which serve to output an alarm. However, the signaling transmitted from processor 2 to driver 4 may also be used. On the processor 2, the method according to the invention, which is performed by the in FIG. 2 illustrated flowchart will now be explained. In step 7, with the camera 1 and the processor 2, as in FIG. 3 represented, a moving object detected by the motion vector. For this purpose, a list or matrix is set up, in which, for example, each column denotes a specific image in an image sequence, the image sequences being separated by time intervals, for example one second. Furthermore, the list has two lines which define the motion vector in a plane. This is usually defined by the coordinates x and y, for example. The following are two lists that illustrate this. In List 1, a motion vector with x value 123 and y value 12 at time 0 is detected. At time 99, that's 50 seconds, a standstill is counted, which has already been used at time 1, which corresponds to one second. This is a list that is predefined, ie only allows new 100 entries. Therefore, when the time reaches 99, the list is reinitialized, and the second list that takes the value is passed. Presented here, at time 0, the value 123 and 12 are entered again and at time 1 the standstill 0/0. At time 2, however, the 50 seconds are added, and it is recounted. List 1: Zeitpkt 0 1 2 3 .... 99 X 123 1 2 0 .... 0 Y 12 0 3 0 ..... 0 00:00 12:01 12:02 12:03 12:50 Zeitpkt 0 1 2 3 X 126 0 0 0 Y 15 0 0 0 00:00 12:03 12:51 12:52

Durch diese Liste wird im Verfahrensschritt 8 das Objekt verfolgt. Wird in Verfahrensschritt 9 ein Stillstand erkannt, und zwar durch die Einträge 0/0 in der Liste, dann wird in Verfahrensschritt 10 der Zähler gestartet. Ist das nicht der Fall, dann wird das Objekt weiter mit der Liste verfolgt. Wurde jedoch in Verfahrensschritt 10 der Zähler gestartet, dann wird in Verfahrensschritt 11 überprüft, ob die Schwelle, die vorgegeben ist, erreicht wird. Diese Schwelle bedingt bei ihrem Erreichen in Verfahrensschritt 12 eine Signalisierung. Die.Signalisierung kann hier mittels des Lautsprechers 5 bzw. der Anzeige 6 erfolgen, also beispielsweise der Ausgabe eines Alarms. Wird jedoch die Schwelle nicht erreicht und das Objekt.bewegt sich wieder, dann wird zu Verfahrensschritt 8 zurückgesprungen und die Objektverfolgung wieder aufgenommen.Through this list, the object is tracked in step 8. If a standstill is detected in method step 9, specifically by the entries 0/0 in the list, the counter is started in method step 10. If this is not the case, then the object is tracked further with the list. If, however, the counter has been started in method step 10, then it is checked in method step 11 whether the threshold which is predetermined is reached. This threshold requires when they are reached in step 12 signaling. Die.Signalisierung can be done here by means of the speaker 5 and the display 6, ie for example, the output of an alarm. If, however, the threshold is not reached and the object moves again, the process jumps back to step 8 and the object tracking is resumed.

Claims (8)

  1. Method for following at least one object in a scene, the at least one object being followed in the scene by means of an image generator (1), the image generator (1) generating a sequence of images of the scene, the at least one object being accorded a movement on the basis of successive images, a counter being started if a standstill of the previously moving object is detected, and signalling being generated on the basis of a counter reading, the movement of the at least one object being described, in terms of movement vectors and time, by a respective list.
  2. Method according to Claim 1, characterized in that the signalling causes an alarm.
  3. Method according to Claim 1, characterized in that the list is reinitialized after the start of a movement of the at least one object.
  4. Method according to one of the preceding claims, characterized in that a reference image for identifying the at least one object is generated.
  5. Method according to Claim 4, characterized in that the reference image is generated after the at least one object has been identified by adopting other areas of the scene from at least one preceding image.
  6. Method according to one of the preceding claims, characterized in that a time interval of at least half a second is provided between images.
  7. Method according to one of Claims 1 to 6, characterized in that a processor (2) is connected to the image generator (1), and in that output means (5, 6) are connected to the processor (2).
  8. Method according to Claim 7, characterized in that a parking space is monitored.
EP03720139A 2002-03-13 2003-03-06 Method for following at least one object in a scene Expired - Lifetime EP1485889B1 (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
DE10210926 2002-03-13
DE10210926A DE10210926A1 (en) 2002-03-13 2002-03-13 Device for tracking at least one object in a scene
PCT/DE2003/000700 WO2003079306A1 (en) 2002-03-13 2003-03-06 Method for following at least one object in a scene

Publications (2)

Publication Number Publication Date
EP1485889A1 EP1485889A1 (en) 2004-12-15
EP1485889B1 true EP1485889B1 (en) 2009-09-30

Family

ID=27815598

Family Applications (1)

Application Number Title Priority Date Filing Date
EP03720139A Expired - Lifetime EP1485889B1 (en) 2002-03-13 2003-03-06 Method for following at least one object in a scene

Country Status (4)

Country Link
US (1) US7738009B2 (en)
EP (1) EP1485889B1 (en)
DE (2) DE10210926A1 (en)
WO (1) WO2003079306A1 (en)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB0408208D0 (en) * 2004-04-13 2004-05-19 Globaleye Network Intelligence Area monitoring
JP4607797B2 (en) * 2006-03-06 2011-01-05 株式会社東芝 Behavior discrimination device, method and program
DE102007041893A1 (en) 2007-09-04 2009-03-05 Robert Bosch Gmbh Method for detecting and / or tracking moving objects in a surveillance scene with interferers, apparatus and computer program
DE102008001126A1 (en) * 2008-04-11 2009-10-15 Robert Bosch Gmbh Inventory module for a video surveillance system, procedure for inventorying a surveillance area and computer program
JP2010087598A (en) * 2008-09-29 2010-04-15 Fujifilm Corp Photographic apparatus, photographic control method and program therefor, image display apparatus, image display method and program therefor, and photographic system, control method therefor and program therefor
DE102009028212A1 (en) 2009-08-04 2011-02-10 Robert Bosch Gmbh Method for monitoring an area
DE102010003669B4 (en) 2010-04-07 2022-05-05 Robert Bosch Gmbh Method and device for locating people in a given area
DE102012200504A1 (en) 2012-01-13 2013-07-18 Robert Bosch Gmbh Analysis device for evaluating monitoring scene of e.g. airport for early detection of abnormal events, has output device outputting object list of moving objects, where each object satisfies spatial condition including specification
CN108024088B (en) * 2016-10-31 2020-07-03 杭州海康威视系统技术有限公司 Video polling method and device

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH04345396A (en) * 1991-05-23 1992-12-01 Takayama:Kk Traveling object tracing method
EP0567059B1 (en) * 1992-04-24 1998-12-02 Hitachi, Ltd. Object recognition system using image processing
JP2752335B2 (en) * 1994-09-27 1998-05-18 鐘紡株式会社 Patient monitoring device in hospital room
JPH1093954A (en) * 1996-09-13 1998-04-10 Oki Electric Ind Co Ltd Object detection method/system
US5966074A (en) * 1996-12-17 1999-10-12 Baxter; Keith M. Intruder alarm with trajectory display
EP0967584B1 (en) * 1998-04-30 2004-10-20 Texas Instruments Incorporated Automatic video monitoring system
US6570608B1 (en) * 1998-09-30 2003-05-27 Texas Instruments Incorporated System and method for detecting interactions of people and vehicles
US7436887B2 (en) * 2002-02-06 2008-10-14 Playtex Products, Inc. Method and apparatus for video frame sequence-based object tracking

Also Published As

Publication number Publication date
WO2003079306A1 (en) 2003-09-25
DE10210926A1 (en) 2003-10-16
US20050128298A1 (en) 2005-06-16
DE50311966D1 (en) 2009-11-12
EP1485889A1 (en) 2004-12-15
US7738009B2 (en) 2010-06-15

Similar Documents

Publication Publication Date Title
DE69024537T2 (en) Infrared surveillance system
DE112016005965B4 (en) DEVICE FOR MONITORING UNATTENDED OBJECTS, SYSTEM FOR MONITORING UNATTENDED OBJECTS THEREOF, AND METHOD FOR MONITORING UNATTENDED OBJECTS
DE69016074T2 (en) Method and device for temporal interpolation of images with corrected motion compensation.
DE69510252T2 (en) Method and device for obtaining a moving object using background subtraction
DE3852927T2 (en) Emergency monitoring system with infrared processing.
EP2034461B1 (en) Method for detecting and/or tracking moved objects in a monitoring zone with stoppers, device and computer program
DE4411179A1 (en) Image capture device
EP1485889B1 (en) Method for following at least one object in a scene
DE102005045077A1 (en) A method and apparatus for distributed analysis of images
EP1919197B1 (en) Controller for an electronic line scan camera
DE102009055127A1 (en) Apparatus and method for monitoring video objects
DE102015102276A1 (en) DISPLAY CONTROL DEVICE AND DISPLAY METHOD
EP0543148A1 (en) Method and device for detecting changes in a video image
DE3214254A1 (en) Method for detecting movements in video camera images
DE4314483A1 (en) Surveillance system
EP4102477B1 (en) Method and device for detecting at least one object unsupervised by a person, said object being left in a surveillance area based on a recorded image of the monitored area
DE102009051100B4 (en) Method for counting objects
EP0777864A1 (en) Image-evaluation system and process
WO2010057732A1 (en) Device for creating and/or processing an object signature, monitoring device, method and computer program
EP0333069B1 (en) Device and method for discriminating movement in a video signal
DE69314448T2 (en) Method for detecting the appearance of point objects in an image
DE10153484B4 (en) Video / audio system and evaluation method for video / audio data
WO2019121729A1 (en) Method for determining distance information from an image of a spatial region
DE19938845B4 (en) Video monitoring method
DE102018204768B4 (en) Image recognition device

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

17P Request for examination filed

Effective date: 20041013

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LI LU MC NL PT RO SE SI SK TR

17Q First examination report despatched

Effective date: 20050126

APBN Date of receipt of notice of appeal recorded

Free format text: ORIGINAL CODE: EPIDOSNNOA2E

APBR Date of receipt of statement of grounds of appeal recorded

Free format text: ORIGINAL CODE: EPIDOSNNOA3E

APAF Appeal reference modified

Free format text: ORIGINAL CODE: EPIDOSCREFNE

APBT Appeal procedure closed

Free format text: ORIGINAL CODE: EPIDOSNNOA9E

GRAP Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOSNIGR1

GRAS Grant fee paid

Free format text: ORIGINAL CODE: EPIDOSNIGR3

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): DE FR GB IT

REG Reference to a national code

Ref country code: GB

Ref legal event code: FG4D

Free format text: NOT ENGLISH

REF Corresponds to:

Ref document number: 50311966

Country of ref document: DE

Date of ref document: 20091112

Kind code of ref document: P

PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

26N No opposition filed

Effective date: 20100701

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: IT

Payment date: 20120327

Year of fee payment: 10

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: FR

Payment date: 20130329

Year of fee payment: 11

Ref country code: GB

Payment date: 20130318

Year of fee payment: 11

GBPC Gb: european patent ceased through non-payment of renewal fee

Effective date: 20140306

REG Reference to a national code

Ref country code: FR

Ref legal event code: ST

Effective date: 20141128

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: GB

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20140306

Ref country code: FR

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20140331

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: IT

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20140306

REG Reference to a national code

Ref country code: DE

Ref legal event code: R084

Ref document number: 50311966

Country of ref document: DE

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: DE

Payment date: 20220525

Year of fee payment: 20

REG Reference to a national code

Ref country code: DE

Ref legal event code: R071

Ref document number: 50311966

Country of ref document: DE