Steffen Abraham

Localization accuracy of region detectors

In this paper, a comparison of five state of the art region detectors is presented with regard to localization accuracy in position and region shape. Based on carefully estimated ground truth homographies, correspondences between frames are assigned using geometrical region overlap. Significant differences between detectors exist, depending on the type of images. Also, it is shown that localization accuracy linearly depends on region scale for some detectors, which may thus be used as a pre-selection criterion for the removal of error-prone regions. The presented results serve as a supplement to existing comparative studies, and can be used to facilitate the selection of an appropriate detector for a specific target application. When descriptor distance is used as assignment criterion instead of region overlap, a different set of correspondences results with lower accuracy. Set differences (and thus localization accuracy) are directly related to the density of regions in a local neighborhood. Based on the latter, a novel measure for the identification of error-prone regions - shape uniqueness - is introduced. In contrast to existing methods that are based on the descriptor distance of region correspondences, the new measure is pre-computed on each image individually. Thus, the complexity of the subsequent matching task can be significantly reduced.

On performance analysis of optical flow algorithms

Literally thousands of articles on optical flow algorithms have been published in the past thirty years. Only a small subset of the suggested algorithms have been analyzed with respect to their performance. These evaluations were based on black-box tests, mainly yielding information on the average accuracy on test-sequences with ground truth. No theoretically sound justification exists on why this approach meaningfully and/or exhaustively describes the properties of optical flow algorithms. In practice, design choices are often made based on unmotivated criteria or by trial and error. This article is a position paper questioning current methods in performance analysis. Without empirical results, we discuss more rigorous and theoretically sound approaches which could enable scientists and engineers alike to make sufficiently motivated design choices for a given motion estimation task.

Calibration Errors in Structure from Motion

In this paper we investigate the relation between camera calibration and structure from motion. A method is presented to analyze the effect of systematic errors and uncertainty in camera calibration on 3D-reconstruction and motion parameters. In two simple examples from stereo with lateral and forward motion the approach is demonstrated. The approach can easily be extended to more complex situations and used for planning and online diagnostics in calibration and structure from motion.

Estimating the mutual orientation in a multi-camera system with a non overlapping field of view

Multi-camera systems offer some advantages over classical systems like stereo or monocular camera systems. A multi-camera system with a non-overlapping field of view, able to cover a wide area, might prove superior e. g. in a mapping scenario where less time is needed to cover the entire area. Approaches to determine the parameters of the mutual orientation from common motions exist for more than 30 years. Most work presented in the past neglected or ignored the influence different motion characteristics have on the parameter estimation process. However, for critical motions a subset of the parameters of the mutual orientation can not be determined or only very inaccurate. In this paper we present a strategy and assessment scheme to allow a successful estimation of as many parameters as possible even for critical motions. Furthermore, the proposed approach is validated by experiments.

Affines Matching planarer Punktmengen mittles Normalisierung über diskrete Momente

Der vorliegende Beitrag stellt ein Matchingverfahren fur ungeordnete planare Punktmengen bezuglich affiner Abbildungen vor. Verwendet werden diskrete Momente als globale Merkmale. Mit Hilfe einer Normalisierung wird eine affine Abbildung zwischen den Punktmengen bestimmt, ohne das Punktkorrespondenzen bekannt sind. Aufbauend auf dieser Startlosung wird mit zwei iterativen Verfahren demonstriert, das eine Bestimmung der Punktkorrespondenzen und der Abbildungsparameter moglich ist. Fur gestaltahnliche Punktmengen ist auf Basis der Normalisierung mit diskreten Momenten damit eine einfache und schnelle Losung von Matching- oder Mustererkennungsaufgaben moglich.

A Comparison of Region Detectors for Tracking

In this work, the performance of five popular region detectors is compared in the context of tracking. Firstly, conventional nearest-neighbor matching based on the similarity of region descriptors is used to assemble trajectories from unique region-to-region correspondences. Based on carefully estimated homographies between planar object surfaces in neighboring frames of an image sequence, both their localization accuracy and length, as well as the percentage of successfully tracked regions is evaluated and compared. The evaluation results serve as a supplement to existing studies and facilitate the selection of appropriate detectors suited to the requirements of a specific application. Secondly, a novel tracking method is presented, which integrates for each region all potential matches into directed multi-edge graphs. From these, trajectories are extracted using Dijkstras algorithm. It is shown, that the resulting localization error is significantly lower than with nearest-neighbor matching while at the same time, the percentage of tracked regions is increased.

Erweiterte diskrete Relaxationsalgorithmen zum Subgraph-Isomorphie-Problem

Diskrete Relaxationsalgorithmen konnen die Suche nach Subgraph-Isomorphien (SGI), z.B. im Rahmen einer modellbasierten Objekterkennung, unterstutzen. Der vorliegende Artikel zeigt Moglichkeiten einer Erweiterung bekannter Relaxationsverfahren. Vorgestellt wird eine Reihe notwendiger Bedingungen, die eine Zuordnung eines Knotens des Modell-Graphen zu einem Knoten eines Szenen-Graphen, als Bestandteil einer Subgraph-Isomorphie erfullen mus. In Analogie zu bekannten Relaxationsalgorithmen werden uber diese Bedingungen unzulassige Knotenzuordnungen ermittelt, die kein Bestandteil einer Subgraph-Isomorphie sind.