Michael Haag

Combination of Edge Element and Optical Flow Estimates for 3D-Model-Based Vehicle Tracking in Traffic Image Sequences

A model-based vehicle tracking system for the evaluation of inner-city traffic video sequences has been systematically tested on about 15 minutes of real world video data. Methodological improvements during preparatory test phases affected—among other changes—the combination of edge element and optical flow estimates in the measurement process and a more consequent exploitation of background knowledge. The explication of this knowledge in the form of models facilitates the evaluation of video data for different scenes by exchanging the scene-dependent models. An extensive series of experiments with a large test sample demonstrates that the current version of our system appears to have reached a relative optimum: further interactive tuning of tracking parameters does no longer promise to improve the overall system performance significantly. Even the incorporation of further knowledge regarding vehicle and scene geometry or illumination has to cope with an increasing level of interaction between different knowledge sources and system parameters. Our results indicate that model-based tracking of rigid objects in monocular image sequences may have to be reappraised more thoroughly than anticipated during the recent past.

INCREMENTAL RECOGNITION OF TRAFFIC SITUATIONS FROM VIDEO IMAGE SEQUENCES

Abstract Our image evaluation system XTRACK tracks multiple-vehicle-configurations in image sequences. The resulting geometric state descriptions are associated with fuzzy attributes and relations and thereby form the basis for incremental characterization of traffic situations from the point of view of selected road users or observers. Knowledge representation and inference is performed by means of fuzzy metric temporal logic in order to provide an in-depth analyzable transition from raw video data to conceptual descriptions of traffic situations.

Tracking of Occluded Vehicles in Traffic Scenes

Vehicles on downtown roads can be occluded by other vehicles or by stationary scene components such as traffic lights or road signs. After having recorded such a scene by a video camera, we noticed that the occlusion may disturb the detection and tracking of vehicles by previous versions of our computer vision approach. In this contribution we demonstrate how our image sequence analysis system can be improved by an explicit model-based recognition of 3D occlusion situations. Results obtained from real world image sequences recording gas station traffic as well as inner-city intersection traffic are presented.

Integration of Image Sequence Evaluation and Fuzzy Metric Temporal Logic Programming

Advanced image sequence evaluation systems generate a voluminous amount of quantitative data which is increasingly difficult to assess. The challenge consists in abstracting from and reasoning with these data in order to create a more intuitive access to image evaluation results.

Beginning a Transition from a Local to a More Global Point of View in Model-Based Vehicle Tracking

This contribution attempts to move beyond the status where single moving objects in video image sequences are tracked separately in the scene domain, based on individually adapted approaches and parameters. Instead, we investigate which performance can be achieved by a combination of approaches based on edge element orientation and on optical flow, applied to a variety of image sequences and vehicles. Five different image sequences of traffic scenes recorded under different conditions have been evaluated. Quantitative statements are provided about the success rates of the approach after evaluating over 5.500 full video-frames, i. e. more than 3 1/2 minutes of real-world video, using one single approach and a single parameter set. Remaining tracking failures are analyzed and classified.

Influence of an Explicitly Modelled 3D Scene on the Tracking of Partially Occluded Vehicles

Model-based tracking of vehicles in real world image sequences of traffic may fail due to different reasons. A careful analysis of failed tracking experiments brought to light that one of these phenomena consists in an incorrect match of parts of the vehicle model to image features belonging to other scene components. This effect appears in particular if the vehicle is occluded by either stationary scene components or by another moving vehicle. Although we got some encouraging results when we modeled the occluding scene components explicitly, we still encountered cases in which we did not succeed in tracking partially occluded vehicles properly. In this contribution we show some successful tracking results obtained from real world image sequences and discuss the cases of failures. Additionally, we present a framework for the conceptual characterization of occurring occlusions which became amenable to experimental analysis in the course of our investigations.

QUANTITATIVE COMPARISON BETWEEN TRAJECTORY ESTIMATES OBTAINED FROM A BINOCULAR CAMERA SETUP WITHIN A MOVING ROAD VEHICLE AND FROM THE OUTSIDE BY A STATIONARY MONOCULAR CAMERA

Abstract The image sequence evaluation system Xtrack detects, initializes, and tracks images of moving road vehicles in videosequences recorded at innercity traffic scenes by a stationary camera. An additional model-based vision system Ximage tracks road models in video sequences recorded from within a driving vehicle . Both systems produce the same kind of results (vehicle trajectories). We have established a link between Xtrack and Ximage which enables us to quantitatively compare trajectories of the same vehicle obtained independently by both systems for the same observed traffic situation.

Visualisation of Conceptual Descriptions Derived from Image Sequences

Synthetic image sequences are generated based on conceptual descriptions which have been extracted automatically by model- based tracking from video sequences recording vehicle manouvers in inner-city traffic scenes. A detailed comparison between original and synthesized image sequences offers clues as to which knowledge must be provided in addition to elementary conceptual descriptions in order to obtain acceptable agreement.

Zuordnung natürlichsprachlicher Begriffe zu Geschehen an einer Tankstelle

Das in diesem Beitrag vorgestellte Bildauswertungssystem berechnet begriffliche Beschreibungen fur automatisch aus Videobildfolgen ermittelte Trajektoriendaten. Dabei sind alle notwendigen Schritte von den Sensordaten uber geometrische bis hin zu begrifflichen Beschreibungen in einem einheitUchen System implementiert. Ausgehend von den theoretischen Uberlegungen in [Nagel 91] werden zulassige Aktionssequenzen beim Betanken eines Fahrzeuges weiterentwickelt und an — aus den Bilddaten ermittelte — Geschehensbeschreibungen gekoppelt. Damit stehen reichhaltigere begriffliche Beschreibungen zur Verfugung als bisher.

‘Begriffliche Rückkopplung’ zur Behandlung temporärer Verdeckungssituationen in der Bildfolgenauswertung von Straßenverkehrsszenen

Ein Bildfolgenauswertungssystem zur Verfolgung sich bewegender Objekte in Strasenverkehrsszenen und zur begrifflichen Charakterisierung ihrer Verkehrssituation wird um die Behandlung zeitweise vollstandig verdeckter Objekte erganzt. Typische Verkehrssituationen werden hierzu begrifflich modelliert und unter Ausnutzung von automatisch extrahierten geometrischen Verfolgungsergebnissen schritthaltend ausgepragt. Solch begriffliches Zusatzwissen gestattet die Erschliesung von Zusammenhangen, die (etwa aufgrund vollstandiger Verdeckungen) nicht explizit im Bild zu sehen sind. Begriffliches Wissen uber typische Objektbewegungen in bestimmten Verkehrssituationen wird wieder auf die geometrische Auswertungsebene ruckgekoppelt, um eine geometrische Zustandsschatzung auch wahrend Phasen vollstandiger Verdeckung plausibel fortschreiben zu konnen. Die so gewonnene rechnerinterne Reprasentation bildet den Ausgangspunkt fur eine naturlichsprachliche Beschreibung der in einer Bildfolge erfasten Geschehen.