WO2004055732A1 - Method for determining three-dimensional object contours by means of images taken synchronously - Google Patents

Method for determining three-dimensional object contours by means of images taken synchronously Download PDF

Info

Publication number
WO2004055732A1
WO2004055732A1 PCT/EP2003/014266 EP0314266W WO2004055732A1 WO 2004055732 A1 WO2004055732 A1 WO 2004055732A1 EP 0314266 W EP0314266 W EP 0314266W WO 2004055732 A1 WO2004055732 A1 WO 2004055732A1
Authority
WO
WIPO (PCT)
Prior art keywords
images
spatial
camera
object contours
curve
Prior art date
Application number
PCT/EP2003/014266
Other languages
German (de)
French (fr)
Inventor
Lars Krueger
Christian Woehler
Martin Wendler
Original Assignee
Pilz Gmbh & Co. Kg
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 Pilz Gmbh & Co. Kg filed Critical Pilz Gmbh & Co. Kg
Priority to AU2003294857A priority Critical patent/AU2003294857A1/en
Publication of WO2004055732A1 publication Critical patent/WO2004055732A1/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/149—Segmentation; Edge detection involving deformable models, e.g. active contour models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/12—Edge-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/50—Depth or shape recovery
    • G06T7/55—Depth or shape recovery from multiple images
    • G06T7/564—Depth or shape recovery from multiple images from contours
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/50—Depth or shape recovery
    • G06T7/55—Depth or shape recovery from multiple images
    • G06T7/593—Depth or shape recovery from multiple images from stereo images
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image
    • G06T2207/10012—Stereo images

Definitions

  • the present invention relates generally to a method for determining spatial object contours using synchronously recorded images.
  • the present invention relates to a method for determining the spatial coordinates of the contour of objects that are located in the common image field of several cameras.
  • a method which comprises a unit for determining contours from ultrasound images.
  • a first subunit for the determination of raw contours of an ultrasound image by means of a predetermined process (such as compensation, binarization / digitization or degeneration) and a second subunit for the dynamic determination of contours.
  • a predetermined process such as compensation, binarization / digitization or degeneration
  • a second subunit for the dynamic determination of contours.
  • an active model such as a snake model
  • the second subunit provides an exact contour of the ultrasound image.
  • Snake models are known, among others, from the publication by KR Castleman, Digital Image Processing, Prentice Hall, Englewood Cliffs, NJ, 1996, the content of which is incorporated herein by reference.
  • the method according to US-2002/0102023 AI does not work with sufficient accuracy and is not suitable for the processing of synchronously recorded images to determine spatial object contours of the images.
  • the method according to US-2002/0102023 AI does not provide satisfactory results for determining the spatial coordinates of the contour of objects that are in the common image field of several cameras.
  • Another specific object of the present invention is to provide a stereo method that is capable of determining the spatial coordinates of the contour of objects determine which are in the common image field of synchronously working camera pairs.
  • FIG. 2 shows the results of the segmentation circuit of the present invention, which is carried out on the output image of FIG. 1.
  • a method for determining spatial object contours using synchronously recorded images comprises the following steps:
  • K is an integer greater than or equal to 2.
  • the synchronous recording of a scene by means of stereo processes is well known from the prior art, as, inter alia, by U. Franke, I. Kutzbach in Fast Stereo based Object Detection for Stop & Go Traffic, Intelligent Vehicles Symposium, pp. 339-344, Tokyo, 1996 and by U. Franke, A. Joos. in Real-time Stereo Vision for Urban Traffic Scene Understanding IEEE Intelligent Vehicles Symposium, 2000.
  • the stereo recording of a dynamic scene makes it possible to obtain depth information that considerably simplifies object detection in a telematics application.
  • the contents of the above publications are hereby incorporated by reference.
  • a possible result of a synchronous recording of a scene with two cameras is shown in FIG. 1 of the present invention.
  • the images are segmented according to the invention, for. B. by binarization or multi-threshold methods that use color, grayscale, texture or other discrete information. Binarization requires less computing effort, whereas multi-threshold methods deliver an increased volume of information.
  • FIG. 2 The result of the segmentation of the two-camera image of FIG. 1 is shown in FIG. 2.
  • spatial coordinates for the correspondence points are determined on the basis of the corresponding points on the two-dimensional contours in the individual images.
  • a disparity taking into account the calibration data of the camera Pair (e.g. epipolar geometry) can be converted directly into spatial coordinates.
  • a spatial curve can be adapted to the set of correspondence points thus obtained by regression.
  • the three-dimensional height contours appear after segmentation, a contour analysis using smooth functions, e.g. B. B-splines, and after determining the disparity on corresponding contour points.
  • K is an integer that is greater than or equal to 2;
  • N the number of parameters of the spatial curve
  • I k the current image of the kth camera, h (x) the segmentation or generally the preprocessing; g ( ⁇ x ⁇ ) the transformation of the projection of the spatial curve into the value range of the preprocessing function h (x),
  • T fc the transformation from the world coordinate system into the coordinate system of camera k (eg transformation using a perforated camera model; w ( ⁇ x ⁇ ) the shape preservation function of the spatial curve ("penalty terms"), which describes the boundary conditions placed on the spatial curve. This term can also be used Be set to zero.
  • the present invention fulfills the objects set and provides an advantageous method which allows the spatial coordinates of the contour points which are in a topological relationship to be obtained.
  • the spatial curves are directly assigned to the segmented objects in the scene.
  • the method according to the invention requires little computing effort, since the correspondence only needs to be determined at the feature level (after segmentation). chen. This offers particular advantages in telematics applications.
  • the method according to the invention fundamentally provides a different functional principle than the "classic" correlation-based stereo image analysis, which means that it can be used in security-relevant detection systems, e.g. B. is suitable as a diversified approach together with "classic" stereo image analysis.

Abstract

The invention relates to a method for determining three-dimensional object contours by means of images taken synchronously, comprising the following method steps: synchronous recording of a scene with a number of cameras, segmentation of the images of the scene and determination of the object contours in the images in the two-dimensional space of the image, calculation of an initial parameterization of the three-dimensional curve of the object contours in the images and determination of an optimal parameterization of the there-dimensional snake function. The synchronous recording of an image in a stereo method can be carried out by means of two cameras.

Description

Verfahren zur Bestimmung räumlicher Obηektkonturen anhand von synchron aufgenommenen Bildern Method for determining spatial object contours based on synchronously recorded images
Die vorliegende Erfindung betrifft allgemein ein Verfahren zur Bestimmung räumlicher Objektkonturen anhand von synchron aufgenommenen Bildern. Darüber hinaus betrifft die vorliegende Erfindung ein Verfahren zur Bestimmung der räumlichen Koordinaten der Kontur von Objekten, die sich im gemeinsamen Bildfeld mehrerer Kameras befinden.The present invention relates generally to a method for determining spatial object contours using synchronously recorded images. In addition, the present invention relates to a method for determining the spatial coordinates of the contour of objects that are located in the common image field of several cameras.
Stand der TechnikState of the art
Aus der US-2002/0102023 AI ist ein Verfahren bekannt, das eine Einheit zur Bestimmung von Konturen aus Ultraschallbildern umfasst. Im Rahmen dieser Einheit werden eine erste Untereinheit zur Bestimmung von rohen Konturen eines Ultraschallbildes mittels eines vorgegebenen Vorgangs (wie beispielsweise Ausgleich, Binarisierung/Digitalisierung oder Degenerierung) und eine zweite Untereinheit zur dynamischen Ermittlung von Konturen, bereitgestellt. Die zweite Untereinheit liefert ausgehend von den rohen Konturen der ersten Untereinheit und mittels eines aktiven Modells, wie beispielsweise einem Snake-Modell, eine exakte Kontur des Ultraschallbildes. Snake-Modelle sind unter anderem aus der Veröffentlichung von K. R. Castleman, Digital Image Processing, Prentice Hall, Englewood Cliffs, NJ, 1996, bekannt, deren Inhalt durch die Bezugnahme hierin eingeschlossen ist.From US-2002/0102023 AI a method is known which comprises a unit for determining contours from ultrasound images. Within the scope of this unit, a first subunit for the determination of raw contours of an ultrasound image by means of a predetermined process (such as compensation, binarization / digitization or degeneration) and a second subunit for the dynamic determination of contours. Starting from the raw contours of the first subunit and by means of an active model, such as a snake model, the second subunit provides an exact contour of the ultrasound image. Snake models are known, among others, from the publication by KR Castleman, Digital Image Processing, Prentice Hall, Englewood Cliffs, NJ, 1996, the content of which is incorporated herein by reference.
Das Verfahren gemäß der US-2002/0102023 AI arbeitet jedoch nicht mit einer ausreichender Genauigkeit und ist nicht für die Verarbeitung von synchron aufgenommenen Bildern zur Bestimmung räumlicher Objektkonturen der Bilder geeignet. Darüber hinaus liefert das Verfahren gemäß der US-2002/0102023 AI keine zufriedenstellenden Ergebnisse zur Bestimmung der räumlichen Koordinaten der Kontur von Objekten, die sich im gemeinsamen Bildfeld mehrerer Kameras befinden.However, the method according to US-2002/0102023 AI does not work with sufficient accuracy and is not suitable for the processing of synchronously recorded images to determine spatial object contours of the images. In addition, the method according to US-2002/0102023 AI does not provide satisfactory results for determining the spatial coordinates of the contour of objects that are in the common image field of several cameras.
Daher besteht eine Aufgabe der vorliegenden Erfindung in der Bereitstellung eines Verfahrens, das die Nachteile aus dem Stand der Technik beseitigt und das für die Verarbeitung von synchron aufgenommenen Bildern zur Bestimmung von Objektkonturen geeignet ist.It is therefore an object of the present invention to provide a method which overcomes the disadvantages of the prior art and which is suitable for the processing of synchronously recorded images for determining object contours.
Eine weitere spezielle Aufgabe der vorliegenden Erfindung besteht in der Bereitstellung eines Stereo-Verfahrens , das in der Lage ist, die räumlichen Koordinaten der Kontur von Objekten zu bestimmen, die sich im gemeinsamen Bildfeld von synchron arbeitenden Kamera-Paaren befinden.Another specific object of the present invention is to provide a stereo method that is capable of determining the spatial coordinates of the contour of objects determine which are in the common image field of synchronously working camera pairs.
Diese und weitere der nachstehenden Beschreibung zu entnehmenden Aufgaben werden von einem Verfahren zur Bestimmung räumlicher Objektkonturen anhand von synchron aufgenommenen Bildern nach den anliegenden Ansprüchen gelöst.These and other tasks to be found in the description below are achieved by a method for determining spatial object contours on the basis of synchronously recorded images according to the appended claims.
Weitere Merkmale und Vorteile der vorliegenden Erfindung sowie die Wirkungsweise verschiedener Ausführungsformen der vorliegenden Erfindung werden unten mit Bezug auf die begleitenden Zeichnungen beschrieben. Die begleitenden Zeichnungen veranschaulichen die vorliegende Erfindung und dienen zusammen mit der Beschreibung weiterhin dazu, die Grundsätze der Erfindung zu erklären und einem Fachmann auf dem betreffenden Gebiet zu ermöglichen, die Erfindung herzustellen und zu verwenden. Dabei zeigt:Further features and advantages of the present invention as well as the operation of various embodiments of the present invention are described below with reference to the accompanying drawings. The accompanying drawings illustrate the present invention and, together with the description, further serve to explain the principles of the invention and to enable any person skilled in the art to make and use the invention. It shows:
Fig. 1 das Ausgangsbild eines Kamerapaares, das das im erfindungsgemäßen Verfahren verarbeitete Bild aufzeichnet;1 shows the output image of a pair of cameras which records the image processed in the method according to the invention;
Fig. 2 die Ergebnisse des Segmentierungssch itts der vorliegenden Erfindung, der am Ausgangsbild der Fig. 1 durchgeführt wird.FIG. 2 shows the results of the segmentation circuit of the present invention, which is carried out on the output image of FIG. 1.
Fig. 3 die erfindungsgemäße Ermittlung der dreidimensionalen Höhenkonturen ausgehend von dem segmentierten Bild der Fig. 2. Beschreibung der bevorzugten Ausführungsformen der Erfindung3 the determination of the three-dimensional height contours according to the invention based on the segmented image of FIG. 2. Description of the preferred embodiments of the invention
Erfindungsgemäß wird ein Verfahren zur Bestimmung räumlicher Objektkonturen anhand von synchron aufgenommenen Bildern bereitgestellt, das folgende Schritte umfasst:According to the invention, a method for determining spatial object contours using synchronously recorded images is provided, which comprises the following steps:
1. Synchrone Aufnahme der Szene mit einer Vielzahl von Kameras. In der nachstehenden Beschreibung wird die Anzahl der Kameras mit K bezeichnet, wobei K eine ganze Zahl größer oder gleich 2 ist.1. Synchronous recording of the scene with a variety of cameras. In the description below, the number of cameras is denoted by K, where K is an integer greater than or equal to 2.
Die synchrone Aufnahme einer Szene mittels Stereoverf hren ist aus dem Stand der Technik wohl bekannt, wie unter anderem von U. Franke, I. Kutzbach in Fast Stereo based Object Detection for Stop&Go Traffic, Intelligent Vehicles Symposium, S. 339- 344, Tokio, 1996 und von U. Franke, A. Joos . in Real-time Stereo Vision for Urban Traffic Scene Understanding IEEE Intelligent Vehicles Symposium, 2000 beschrieben. Dabei wird durch die Stereoaufzeichnung einer dynamischen Szene möglich, Tiefeninformation zu gewinnen, die die Objekterfassung in einer Telema- tikanwendung beträchtlich vereinfacht. Die Inhalte der vorstehenden Veröffentlichungen werden hierin durch die in Bezugnahme eingeschlossen. Ein mögliches Ergebnis einer synchronen Aufnahme einer Szene mit zwei Kameras (Stereoverfahren) ist in der Fig. 1 der vorliegenden Erfindung dargestellt.The synchronous recording of a scene by means of stereo processes is well known from the prior art, as, inter alia, by U. Franke, I. Kutzbach in Fast Stereo based Object Detection for Stop & Go Traffic, Intelligent Vehicles Symposium, pp. 339-344, Tokyo, 1996 and by U. Franke, A. Joos. in Real-time Stereo Vision for Urban Traffic Scene Understanding IEEE Intelligent Vehicles Symposium, 2000. The stereo recording of a dynamic scene makes it possible to obtain depth information that considerably simplifies object detection in a telematics application. The contents of the above publications are hereby incorporated by reference. A possible result of a synchronous recording of a scene with two cameras (stereo method) is shown in FIG. 1 of the present invention.
2. Segmentierung der Bilder und Ermittlung der Objektkonturen im zweidimensionalen Bildraum. Die Segmentierung der Bilder erfolgt erfindungsgemäß z. B. durch Binarisierung oder Mehrschwellen-Verfahren, die Farb- Graustufen-, Textur- oder andere diskrete Informationen verwenden. Binarisierung erfordert weniger Rechenaufwand, wohingegen Mehrschwellen-Verfahren ein erhöhtes Volumen an Information liefern.2. Segmentation of the images and determination of the object contours in the two-dimensional image space. The images are segmented according to the invention, for. B. by binarization or multi-threshold methods that use color, grayscale, texture or other discrete information. Binarization requires less computing effort, whereas multi-threshold methods deliver an increased volume of information.
Die Ermittlung der Objektkonturen im zweidimensionalen Bildraum erfolgt entsprechend der Binarisierung oder dem Mehrschwellen- Verfahren durch BCC-Verfahren (BCC *= binary connected compo- nent) bzw. CCC-Verfahren (CCC = colored connected co ponent) .The object contours in the two-dimensional image space are determined in accordance with the binarization or the multi-threshold method using the BCC method (BCC * = binary connected component) or CCC method (CCC = colored connected component).
Weitere Einzelheiten zur Segmentierung insbesondere zu den BCC- und CCC-Algorithmen werden bei E. Mandler, M. Oberländer in der Veröffentlichung One Pass Encoding of Connected Components in Multi-Valued Images . IEEE Int. Conf. on Pattern Recognition, S. 64-69, Atlantic City, 1990 aufgezeigt, deren Inhalt ebenfalls durch die in Bezugnahme hierin eingeschlossen wird.Further details on segmentation, in particular on the BCC and CCC algorithms, are given by E. Mandler, M. Oberländer in the publication One Pass Encoding of Connected Components in Multi-Valued Images. IEEE Int. Conf. on Pattern Recognition, pp. 64-69, Atlantic City, 1990, the contents of which are also incorporated by reference herein.
Das Ergebnis der Segmentierung des Zweikamera-Bildes der Fig. 1 ist in der Fig. 2 dargestellt.The result of the segmentation of the two-camera image of FIG. 1 is shown in FIG. 2.
3. Berechnung einer initialen räumlichen Kurve .3. Calculation of an initial spatial curve.
Bei diesem Berechnungsschritt werden ausgehend von den korrespondierenden Punkten auf den zweidimensionalen Konturen in den einzelnen Bildern räumliche Koordinaten für die Korrespondenzpunkte ermittelt.In this calculation step, spatial coordinates for the correspondence points are determined on the basis of the corresponding points on the two-dimensional contours in the individual images.
Im Spezialfall von K = 2 (Stereo-Bildanalyse) kann eine Disparität unter Berücksichtigung der Kalibrierdaten des Kamera- Paares (z. B. Epipolargeometrie) direkt in räumliche Koordinaten umgerechnet werden. Unter Ausnutzung der topologischen Nachbarschaftsbeziehungen in den Bildern kann durch Regression eine räumliche Kurve an die Menge der so erhaltenen Korrespondenzpunkte angepasst werden.In the special case of K = 2 (stereo image analysis), a disparity taking into account the calibration data of the camera Pair (e.g. epipolar geometry) can be converted directly into spatial coordinates. Using the topological neighborhood relationships in the images, a spatial curve can be adapted to the set of correspondence points thus obtained by regression.
Die Fig. 3 zeigt die Ermittlung der dreidimensionalen Höhenkonturen nach dem erfindungsgemäßen Schritt 3. im Spezialfall von K = 2 (Stereo-Bildanalyse). Darin erscheinen die dreidimensionalen Höhenkonturen nach einer Segmentierung, einer Konturanalyse durch glatte Funktionen, z. B. B-Splines, und nach der Bestimmung der Disparität auf korrespondierenden Konturpunkten. Im konkreten Beispiel der Fig. 3 ist Dkopf > Dbodβrι- 3 shows the determination of the three-dimensional height contours after step 3 according to the invention in the special case of K = 2 (stereo image analysis). The three-dimensional height contours appear after segmentation, a contour analysis using smooth functions, e.g. B. B-splines, and after determining the disparity on corresponding contour points. In the concrete example in FIG. 3, D head > D bodβrι-
4. Ermittlung einer optimalen Parametrisierung der räumlichen Kurve .4. Determination of an optimal parameterization of the spatial curve.
Nach einem Aspekt der vorliegenden Erfindung wird vorgeschlagen - unter Verwendung eines geeigneten Optimierungsverfahrens (z. B. Simplex, Gradientenabstieg) - das Minimum einer Zielfunktion der folgenden Form zu suchen:According to one aspect of the present invention it is proposed - using a suitable optimization method (e.g. simplex, gradient descent) - to search for the minimum of an objective function of the following form:
Figure imgf000008_0001
Figure imgf000008_0001
Hierbei bedeutetHere means
II ... || eine geeignet definierte Norm, z. B. die euklidische Norm; K die Anzahl der Kameras. Wie erläutert, ist K eine ganze Zahl, die größer gleich 2 ist;II ... || a suitably defined standard, e.g. B. the Euclidean norm; K the number of cameras. As explained, K is an integer that is greater than or equal to 2;
N die Anzahl der Parameter der räumlichen Kurve;N the number of parameters of the spatial curve;
Ik das aktuelle Bild der k-ten Kamera, h(x) die Segmentierung bzw. allgemein die Vorverarbeitung; g({x}) die Transformation der Projektion der räumlichen Kurve in den Wertebereich der Vorverarbeitungsfunktion h(x),I k the current image of the kth camera, h (x) the segmentation or generally the preprocessing; g ({x}) the transformation of the projection of the spatial curve into the value range of the preprocessing function h (x),
Pk die Projektionsmatrix von Kamera k,P k the projection matrix of camera k,
Tfc die Transformation vom Weltkoordinatensystem in das Koordinatensystem von Kamera k (z.B. Transformation mittels Lochkameramodell; w({x}) die Formerhaltungsfunktion der räumlichen Kurve ( "Strafterme" ) , die die an die räumliche Kurve gestellten Randbedingungen beschreibt. Dieser Term kann auch zu Null gesetzt werden.T fc the transformation from the world coordinate system into the coordinate system of camera k (eg transformation using a perforated camera model; w ({x}) the shape preservation function of the spatial curve ("penalty terms"), which describes the boundary conditions placed on the spatial curve. This term can also be used Be set to zero.
Beispiele für eine projektive Abbildung werden von 0. Faugeras . in der Veröffentlichung Three-Dimensional Computer Vision , MIT Press, 1993 geschildert, deren Inhalt durch die in Bezugnahme ebenfalls hierin eingeschlossen wird.Examples of a projective mapping are provided by 0. Faugeras. in the publication Three-Dimensional Computer Vision, MIT Press, 1993, the content of which is also incorporated herein by reference.
Die vorliegende Erfindung erfüllt die gestellten Aufgaben und stellt ein vorteilhaftes Verfahren zur Verfügung, das die Gewinnung räumlicher Koordinaten der Konturpunkte, die zueinander in topologischer Beziehung stehen, gestattet. Die räumlichen Kurven sind direkt den segmentierten Objekten in der Szene zugeordnet.The present invention fulfills the objects set and provides an advantageous method which allows the spatial coordinates of the contour points which are in a topological relationship to be obtained. The spatial curves are directly assigned to the segmented objects in the scene.
Darüber hinaus erfordert das erfindungsgemäße Verfahren einen geringen Rechenaufwand, da die Korrespondenzen nur auf Merkmalsebene (nach der Segmentierung) ermittelt zu werden brau- chen. Dieses bietet besondere Vorteile in Telematik-Anwen- dungen.In addition, the method according to the invention requires little computing effort, since the correspondence only needs to be determined at the feature level (after segmentation). chen. This offers particular advantages in telematics applications.
Das erfindungsgemäße Verfahren stellt grundlegend ein anderes Funktionsprinzip als die "klassische" korrelationsbasierte Stereo-Bildanalyse zur Verfügung, wodurch die Einsatzmöglichkeit in sicherheitstechnisch relevanten ErkennungsSystemen, z. B. als diversitärer Ansatz gemeinsam mit "klassischer" Stereo-Bildanalyse geeignet ist. The method according to the invention fundamentally provides a different functional principle than the "classic" correlation-based stereo image analysis, which means that it can be used in security-relevant detection systems, e.g. B. is suitable as a diversified approach together with "classic" stereo image analysis.

Claims

Patentansprüche claims
1. Verfahren zur Bestimmung räumlicher Objektkonturen anhand von synchron aufgenommenen Bildern, das folgende Schritte u fasst: synchrone Aufnahme einer Szene mit einer Vielzahl von Kameras ;1. A method for determining spatial object contours on the basis of synchronously recorded images, which comprises the following steps: synchronous recording of a scene with a multiplicity of cameras;
Segmentierung der Bilder der Szene und Ermittlung der Objektkonturen der Bilder im zweidimensionalen Bildraum; Berechnung einer initialen räumlichen Kurve, die die Objektkonturen in den Bildern beschreibt; und Ermittlung einer optimalen Kurve.Segmentation of the images of the scene and determination of the object contours of the images in the two-dimensional image space; Calculation of an initial spatial curve that describes the object contours in the images; and determining an optimal curve.
2. Verfahren nach Anspruch 1 , wobei die synchrone Aufnahme einer Szene im Stereoverfahren mittels zwei Kameras erfolgt.2. The method according to claim 1, wherein the synchronous recording of a scene takes place in the stereo method by means of two cameras.
3. Verfahren nach Anspruch 1 oder 2 , wobei die Segmentierung der Bilder durch Binarisierung oder Mehrschwellen-Ver- fahren erfolgt.3. The method according to claim 1 or 2, wherein the segmentation of the images is carried out by binarization or multi-threshold methods.
4. Verfahren nach Anspruch 3, wobei die Ermittlung der Objektkonturen im zweidimensionalen Bildraum entsprechend der Binarisierung oder dem Mehrschwellen-Verfahren durch BCC-Verfahren bzw. CCC-Verfahren erfolgt.4. The method according to claim 3, wherein the determination of the object contours in the two-dimensional image space is carried out in accordance with the binarization or the multi-threshold method by BCC method or CCC method.
5. Verfahren nach einem oder mehreren der Ansprüche 1-4, wobei die Berechnung der initialen räumlichen Kurve, die die Objektkonturen in den Bildern beschreibt, von korrespondierenden Punkten auf den zweidimensionalen Konturen in den einzelnen Bildern ausgeht, wobei räumliche Koordinaten für die Korrespondenzpunkte ermittelt werden.5. The method according to one or more of claims 1-4, wherein the calculation of the initial spatial curve, which describes the object contours in the images, from corresponding points on the two-dimensional contours in the individual images, whereby spatial coordinates for the correspondence points are determined.
Verfahren nach einem oder mehreren der Ansprüche 2-5, wobei eine Disparität unter Berücksichtigung der Kalibrierdaten des Kamera-Paares, insbesondere der Epipolar- geometrie direkt in räumliche Koordinaten umgerechnet wird, um unter Ausnutzung der topologischen Nachbarschaftsbeziehungen in den Bildern durch Regression eine räumliche Kurve an die Menge der so erhaltenen Korrespondenzpunkte anzupassen.Method according to one or more of claims 2-5, wherein a disparity, taking into account the calibration data of the camera pair, in particular the epipolar geometry, is converted directly into spatial coordinates, in order to use a spatial curve by regression using the topological neighborhood relationships in the images adjust the amount of correspondence points thus obtained.
Verfahren nach einem oder mehreren der Ansprüche 1-6, wobei die Ermittlung der optimalen Parametrisierung der räumlichen Kurve die Bestimmung des Minimums der folgenden Zielfunktion umfasst:Method according to one or more of claims 1-6, wherein the determination of the optimal parameterization of the spatial curve comprises the determination of the minimum of the following objective function:
ζ g(PkTkxi 2 + w({xi})
Figure imgf000012_0001
ζ g (P k T k x i 2 + w ({x i })
Figure imgf000012_0001
worinwherein
|| ... || eine geeignet definierte Norm ist, insbesondere die euklidische Norm ;|| ... || is a suitably defined norm, especially the Euclidean norm;
K die Anzahl der Kameras darstellt, wobei K eine ganze Zahl größer oder gleich 2 ist;K represents the number of cameras, where K is an integer greater than or equal to 2;
N die Anzahl der Parameter der räumlichen Kurve;N the number of parameters of the spatial curve;
Ik das aktuelle Bild der k-ten Kamera, h(x) die Segmentierung bzw. allgemein die Vorverarbeitung; g({x}) die Transformation der Projektion der räumlichen Kurve in den Wertebereich der Vorverarbeitungsfunktion h(x),I k the current image of the kth camera, h (x) the segmentation or generally the preprocessing; g ({x}) the transformation of the projection of the spatial curve into the value range of the preprocessing function h (x),
Pj- die Projektionsmatrix von Kamera k,Pj- the projection matrix of camera k,
Tk die Transformation vom Weltkoordinatensystem in das Koordinatensystem von Kamera k; w({x}) die Formerhaltungsfunktion der räumlichen Kurve, die die an die räumliche Kurve gestellten Randbedingungen beschreibt. Diese Term kann auch zu Null gesetzt werden . T k the transformation from the world coordinate system into the coordinate system of camera k; w ({x}) the shape preservation function of the spatial curve, which describes the boundary conditions placed on the spatial curve. This term can also be set to zero.
1/31.3
Figure imgf000014_0001
Figure imgf000014_0001
Ausgangsbild linke Kamera Ausgangsbild rechte KameraLeft camera output image Right camera output image
Fig. 1 Fig. 1
2/32.3
Figure imgf000015_0001
linke Kamera Fig
Figure imgf000015_0002
Figure imgf000015_0001
left camera Fig
Figure imgf000015_0002
3/33.3
Figure imgf000016_0001
linke Kamera rechte Kamera
Figure imgf000016_0001
left camera right camera
FIG. 3 FIG. 3
PCT/EP2003/014266 2002-12-18 2003-12-16 Method for determining three-dimensional object contours by means of images taken synchronously WO2004055732A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
AU2003294857A AU2003294857A1 (en) 2002-12-18 2003-12-16 Method for determining three-dimensional object contours by means of images taken synchronously

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE10259698.0 2002-12-18
DE10259698A DE10259698A1 (en) 2002-12-18 2002-12-18 Display area of an automotive night vision system

Publications (1)

Publication Number Publication Date
WO2004055732A1 true WO2004055732A1 (en) 2004-07-01

Family

ID=32477827

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2003/014266 WO2004055732A1 (en) 2002-12-18 2003-12-16 Method for determining three-dimensional object contours by means of images taken synchronously

Country Status (3)

Country Link
AU (1) AU2003294857A1 (en)
DE (1) DE10259698A1 (en)
WO (1) WO2004055732A1 (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2008014831A3 (en) * 2006-08-02 2008-04-03 Daimler Ag Method for observation of a person in an industrial environment
US7729511B2 (en) 2002-09-24 2010-06-01 Pilz Gmbh & Co. Kg Method and device for safeguarding a hazardous area
CN102402788A (en) * 2011-12-22 2012-04-04 华南理工大学 Method for segmenting three-dimensional ultrasonic image

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
ATE335958T1 (en) 2002-09-24 2006-09-15 Pilz Gmbh & Co Kg METHOD AND DEVICE FOR SECURING A HAZARDOUS AREA
DE102009050850B4 (en) 2009-10-19 2019-01-03 Pilz Gmbh & Co. Kg Method and device for monitoring protection and standard areas

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0626655A2 (en) * 1993-05-25 1994-11-30 Matsushita Electric Industrial Co., Ltd. Vehicle recognition apparatus

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0631250B1 (en) * 1993-06-21 2002-03-20 Nippon Telegraph And Telephone Corporation Method and apparatus for reconstructing three-dimensional objects
DE10037491A1 (en) * 2000-08-01 2002-02-14 Stryker Leibinger Gmbh & Co Kg Process for three-dimensional visualization of structures inside the body

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0626655A2 (en) * 1993-05-25 1994-11-30 Matsushita Electric Industrial Co., Ltd. Vehicle recognition apparatus

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
BASCLE B ET AL: "Stereo matching, reconstruction and refinement of 3D curves using deformable contours", COMPUTER VISION, 1993. PROCEEDINGS., FOURTH INTERNATIONAL CONFERENCE ON BERLIN, GERMANY 11-14 MAY 1993, LOS ALAMITOS, CA, USA,IEEE COMPUTER SOCI, 11 May 1993 (1993-05-11), pages 421 - 430, XP010128594, ISBN: 0-8186-3870-2 *
BRINT A T ET AL: "Stereo Matching Of Curves By Least Deformation", INTERNATIONAL WORKSHOP ON INTELLIGENT ROBOTS AND SYSTEMS'89, 4 September 1989 (1989-09-04), pages 163 - 170, XP010257781 *
MANDLER E ET AL: "ONE-PASS ENCODING OF CONNECTED COMPONENTS IN MULTI-VALUED IMAGES", PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION. ATLANTIC CITY, JUNE 16 - 21, 1990, LOS ALAMITOS: IEEE COMP. SOC. PRESS, US, vol. 2 CONF. 10, 16 June 1990 (1990-06-16), pages 64 - 69, XP000166482, ISBN: 0-8186-2062-5 *
OLIVERAS A ET AL: "Stereo image analysis using connected operators", ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, 1997. ICASSP-97., 1997 IEEE INTERNATIONAL CONFERENCE ON MUNICH, GERMANY 21-24 APRIL 1997, LOS ALAMITOS, CA, USA,IEEE COMPUT. SOC, US, 21 April 1997 (1997-04-21), pages 3169 - 3172, XP010225830, ISBN: 0-8186-7919-0 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7729511B2 (en) 2002-09-24 2010-06-01 Pilz Gmbh & Co. Kg Method and device for safeguarding a hazardous area
WO2008014831A3 (en) * 2006-08-02 2008-04-03 Daimler Ag Method for observation of a person in an industrial environment
US8154590B2 (en) 2006-08-02 2012-04-10 Pilz Gmbh & Co. Kg Method for observation of a person in an industrial environment
CN102402788A (en) * 2011-12-22 2012-04-04 华南理工大学 Method for segmenting three-dimensional ultrasonic image

Also Published As

Publication number Publication date
DE10259698A1 (en) 2004-07-08
AU2003294857A1 (en) 2004-07-09

Similar Documents

Publication Publication Date Title
CN110008915B (en) System and method for estimating dense human body posture based on mask-RCNN
DE69922183T2 (en) FACE RECOGNITION FROM VIDEO IMAGES
DE69932619T2 (en) METHOD AND SYSTEM FOR RECORDING AND REPRESENTING THREE-DIMENSIONAL GEOMETRY, COLOR AND SHADOWING OF ANIMATED OBJECTS
DE102007045835B4 (en) Method and device for displaying a virtual object in a real environment
DE69817879T2 (en) DETECTION OF DISTORMED IMAGES BY FAST FOLDING AND IMAGE CAPTURE OF MARKANT POINTS AND VOLUME INFORMATION
EP1316057B1 (en) Evaluation of edge direction information
EP2206089A1 (en) Method and device for illustrating a virtual object in a real environment
DE102005008131A1 (en) Object e.g. road sign, detecting method for use with e.g. driver assistance system, involves determining position and movement of relevant pixels using filter and combining relevant pixels to objects under given terms and conditions
EP0474304B1 (en) Circuit arrangement for recognizing a human face
DE102019114622A1 (en) DETECTION AND PLANAR REPRESENTATION OF THREE-DIMENSIONAL DRIVING TRACKS IN A STREET SCENE
DE602004012341T2 (en) Method and system for providing a volume rendering of a three-dimensional object
DE60207367T2 (en) Method and apparatus for texture generation for a 3-D facial model
EP0886833B1 (en) Process for capturing and recording individuals facial data and the use of such data for determining or confirming the identity of persons
DE102006045828B4 (en) Method and device for recognizing a face and a face recognition module
WO2004055732A1 (en) Method for determining three-dimensional object contours by means of images taken synchronously
DE102004026782A1 (en) Method and apparatus for computer-aided motion estimation in at least two temporally successive digital images, computer-readable storage medium and computer program element
DE10145608B4 (en) Model-based object classification and target recognition
DE10358770A1 (en) Digital object recognition method for recognizing spatial object contours in digital images taken with stereo cameras, whereby scenes are segmented and initial spatial curves are determined prior to determination of optimum curves
EP1282070A1 (en) Hierarchical adaptation of an image model
DE102022120332A1 (en) 3D OBJECT RECOGNITION PROCESS USING THE SYNERGY OF HETEROGENE SENSORS FOR AUTONOMOUS DRIVING
WO2010028718A2 (en) Method for image processing stereo images
DE20220762U1 (en) 3D object contour determination method in which synchronously recorded images from multiple cameras are processed to generate a 3D snake function with optimum parameterization
EP1756748B1 (en) Method for the classification of an object by means of a stereo camera
WO2020164671A1 (en) Modular inpainting method
DE19509373C1 (en) Object based coding of image sequences

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BW BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE EG ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): BW GH GM KE LS MW MZ SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LU MC NL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG

DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application
122 Ep: pct application non-entry in european phase
NENP Non-entry into the national phase

Ref country code: JP

WWW Wipo information: withdrawn in national office

Country of ref document: JP