Martin Roser

Efficient large-scale stereo matching

In this paper we propose a novel approach to binocular stereo for fast matching of high-resolution images. Our approach builds a prior on the disparities by forming a triangulation on a set of support points which can be robustly matched, reducing the matching ambiguities of the remaining points. This allows for efficient exploitation of the disparity search space, yielding accurate dense reconstruction without the need for global optimization. Moreover, our method automatically determines the disparity range and can be easily parallelized. We demonstrate the effectiveness of our approach on the large-scale Middlebury benchmark, and show that state-of-the-art performance can be achieved with significant speedups. Computing the left and right disparity maps for a one Megapixel image pair takes about one second on a single CPU core.

Video-based raindrop detection for improved image registration

In this paper we present a novel approach to improved image registration in rainy weather situations. To this end, we perform monocular raindrop detection in single images based on a photometric raindrop model. Our method is capable of detecting raindrops precisely, even in front of complex backgrounds. The effectiveness is demonstrated by a significant increase in image registration accuracy which also allows for successful image restoration. Experiments on video sequences taken from within a moving vehicle prove the applicability to real-world scenarios.

Classification of weather situations on single color images

Present vision based driver assistance systems are designed to perform under good-natured weather conditions. However, limited visibility caused by heavy rain or fog strongly affects vision systems. To improve machine vision in bad weather situations, a reliable detection system is necessary as a ground base. We present an approach that is able to distinguish between multiple weather situations based on the classification of single monocular color images, without any additional assumptions or prior knowledge. The proposed image descriptor clearly outperforms existing descriptors for that task. Experimental results on real traffic images are characterized by high accuracy, efficiency, and versatility with respect to driver assistance systems.

Raindrop detection on car windshields using geometric-photometric environment construction and intensity-based correlation

Vision-based Driver Assistance Systems (DAS) are becoming pervasive in todays automotive industry. However, most of these systems are designed to perform in good weather conditions and they perform very poorly in adverse weather particularly in rain. A big problem related to rainy weather conditions that highly limits the performance of DAS is raindrops on car windshields. We present a novel approach that detects raindrops on a car windshield using only a single image from an in-vehicle camera and a standard interest point detector for pre-selection of raindrop candidates. The algorithm models the geometric shape of a raindrop on the car windshield, utilizes its photometric properties and establishes a relationship between raindrop and environment. The proposed algorithm outperforms existing machine vision-based approaches for the task of raindrop modeling and detection from an in-vehicle perspective. It functions very accurately and is robust in terms of imprecise positions of raindrop candidates. Its results can be further used for image restoration and vision enhancement and hence it is a valuable tool for DAS.

Realistic modeling of water droplets for monocular adherent raindrop recognition using Bézier curves

In this paper, we propose a novel raindrop shape model for the detection of view-disturbing, adherent raindrops on inclined surfaces. Whereas state-of-the-art techniques do not consider inclined surfaces because they assume the droplets as sphere sections with equal contact angles, our model incorporates cubic Bezier curves that provide a low dimensional and physically interpretable representation of a raindrop surface. The parameters are empirically deduced from numerous observations of different raindrop sizes and surface inclination angles. It can be easily integrated into a probabilistic framework for raindrop recognition, using geometrical optics to simulate the visual raindrop appearance. In comparison to a sphere section model, the proposed model yields an improved droplet surface accuracy up to three orders of magnitude.

Camera-based bidirectional reflectance measurement for road surface reflectivity classification

In this paper we propose a novel framework for road reflectivity classification in cluttered traffic scenarios by measuring the bidirectional reflectance distribution function of road surfaces from inside a moving vehicle. The predominant restrictions in our application are a strongly limited field of observations and a weakly defined illumination environment. To overcome these problems, we estimate the parameters of an extended Oren-Nayar model that considers the diffuse and specular behavior of real-world surfaces and extrapolate the surface reflectivity measurements to unobservable angle combinations. Model ambiguities are decreased by utilizing standardized as well as customized reflection characteristics. In contrast to existing approaches that require special measurement setups, our approach can be implemented in vision-based driver assistance systems using radiometrically uncalibrated gray value cameras and GPS information. The effectiveness of our approach is demonstrated by a successful classification of the road surface reflectance of expressway scenes with low error rates.

Numerische Modellierung des Rissfortschrittes in porösen CVI-CFC-Verbundwerkstoffen

Kurzfassung Zur Beschreibung der Rissausbreitung in spröden Materialien wurde in der Literatur eine Kombination von Spannungs- und Energiekriterien vorgeschlagen. Ausgehend von der inkrementellen Energiebilanz während des Rissfortschritts wird unter Einbeziehung der dort formulierten Kriterien ein Algorithmus zur Berechnung des Rissfortschritts hergeleitet. Dieser wird zur Simulation der Rissausbreitung in porösen Materialien mit folgenden Mikrostrukturen verwendet:1. Modellierung der typischen CVI-CFC-Mikrostruktur auf Basis von Querschnittsaufnahmen einer Probe mit anschließendem Vergleich der Resultate aus der Simulation und dem 4-Punkt-Biegeversuch,2. idealisiertes, poröses Material mit kreisförmigen Poren. Ausgehend von den Resultaten der durchgeführten Untersuchungen sind Aussagen über die Rissentwicklung in einem realen Material möglich.

Modellbasierte und positionsgenaue Erkennung von Regentropfen in Bildfolgen zur Verbesserung von videobasierten Fahrerassistenzfunktionen

Witterungsbedingte Effekte, insbesondere Regentropfen auf der Windschutzscheibe, fuhren zu einer starken Beeintrachtigung gegenwartiger videobasierter Fahrerassistenzfunktionen. In dieser Arbeit wird ein Verfahren zur Erkennung von Regentropfen auf der Windschutzscheibe vorgestellt, das auf einer physikalischen und signaltheoretischen Regentropfen-Modellierung sowie der Betrachtung des Bildentstehungsprozesses in realen optischen Systemen basiert.

Fatigue failure of titanium implants for mandibular reconstruction

Abstract In recent years mandibular reconstruction following resection surgery has become a standard medical treatment procedure. The mandible is bridged by a metallic implant which is bent into the required shape by the surgeon and screwed to the residual parts of the mandible. Awidely used product for this application is the titanium implant 2.4™ UniLock by Synthes GmbH & Co. KG, Umkirch, Germany. According to clinical experience, this kind of implant occasionally fails by apparently sudden fracture. In the present work, the reliability of this implant is investigated from a materials perspective. A series of fatigue tests was carried out using a jaw-simulator emulating the load distribution in an implant. Comparison of the experimental results to data reported in the literature for typical chewing forces and frequencies show that the safety margin of the implant is presently insufficient. As a consequence, a solution is proposed which increases the fatigue performance while still allowing easy adjustme...