[show abstract][hide abstract] ABSTRACT: Statistical shape models provide an important means in many applications in computer vision and computer graphics. However, the major problems are that the majority of these shape models require dense point-correspondences along all training shapes and that a large number of training shapes is needed in order to capture the full amount of intra-class shape variation. In this contribution, we focus on a statistical shape model that can be constructed from a set of training shapes without defining any point-correspondences. Additionally, we show how a local statistical shape model can make better use of the available shape information, greatly reducing the number of required training shapes. Finally, we present a new framework to fit this local statistical shape model without correspondences to range scans that represent incomplete parts of the trained shape class. The fitted model is then used to reproduce a natural-looking approximation of the complete shape.
[show abstract][hide abstract] ABSTRACT: The generic camera model considered in this work can be regarded as a mapping between image pixels and viewing rays. These rays are independent of each other which prohibits a standard parametric approach for calibration and modeling of these cameras. Spline surfaces are used here to calibrate and model generic imaging devices. This allows the utilization of sparse planar calibration boards and facilitates general forward projection as well as subpixel back projection. In contrast to other works the complete image area is to be calibrated, not only a part of it. This is done by adding further views of calibration patterns after an initial calibration step, which expands the calibrated region of the camera image. Results with two different imaging devices prove the general applicability of the proposed method and the comparison to an established parametric calibration procedure shows its superiority.
Asian Conference on Computer Vision (ACCV); 11/2012
[show abstract][hide abstract] ABSTRACT: Articulated structures like the human body have many degrees of freedom. This makes an evaluation of the configuration's likelihood very challenging. In this work we propose new linked hierarchical graphical models which are able to efficiently evaluate likelihoods of articulated structures by sharing visual primitives. Instead of evaluating all configurations of the human body separately we take advantage of the fact that different configurations of the human body share body parts, and body parts, in turn, share visual primitives. A hierarchical Markov random field is used to integrate the sharing of visual primitives in a probabilistic framework. We propose a scalable hierarchical representation of the human body and show that this representation is especially well suited for human gait analysis from a frontal camera perspective. Furthermore, the results of the evaluation on a gait dataset show that sharing primitives substantially accelerates the evaluation and that our hierarchical probabilistic framework is a robust method for scalable detection of the human body.
Computer methods and programs in biomedicine 02/2012; 106(2):104-13. · 1.14 Impact Factor
[show abstract][hide abstract] ABSTRACT: Cameras are a commonly used sensor in advanced driver assistance systems (ADAS). They serve to get vast amounts of information about a vehicle's environment. To accurately localize the measured data in relation to the own car, exact camera calibration is a prerequisite. This includes extrinsic as well as intrinsic parameters. While many works in the area of ADAS focus on extrinsic calibration, this work covers the intrinsic calibration. We use a generic camera model which regards the viewing ray of every pixel separately and can therefore be used to describe arbitrary imaging devices even with massive lens distortions. As the calibration procedure works for any camera, only one method has to be implemented, which simplifies the sensor calibration process. Former works have shown the applicability of generic camera models but do not cover important practical aspects which are subpixel ray determination and forward projection of arbitrary 3d points to the image plane. Furthermore, the calibration processes described so far are cumbersome and prone to inaccuracies. We propose to use spline surfaces to simplify the calibration procedure and implement general back and forward projection. The applicability of our approach is proved by showing calibration results for various real cameras.
[show abstract][hide abstract] ABSTRACT: Statistical shape models are one of the most powerful methods in medical image segmentation problems. However, if the task is to segment complex structures, they are often too constrained to capture the full amount of anatomical variation. This is due to the fact that the number of training samples is limited in general, because generating hand-segmented reference data is a tedious and time-consuming task. To circumvent this problem, we present a Locally Deformable Statistical Shape Model that is able to segment complex structures with only a few training samples at hand. This is achieved by allowing a unique solution in each contour point. Unlike previous approaches, trying to tackle this problem by partitioning the statistical model, we do not need predeﬁned segments. Smoothness constraints ensure that the local solution is restricted to the space of feasible shapes. Very promising results are obtained when we compare our new approach to a global ﬁtting approach.
Machine Learning in Medical Imaging - Second International Workshop, MLMI 2011, Held in Conjunction with MICCAI 2011, Toronto, Canada, September 18, 2011. Proceedings; 09/2011
[show abstract][hide abstract] ABSTRACT: Electric motors clearly constitute the most common drive principle in robotics and mechatronics. Smart materials, however, offer considerably higher power-to-mass ratios than electric motors. If mechanical energy instead of electrical energy can be distributed through a system, highly dynamic and efficient torque transmission elements based on smart materials, e.g. piezoceramics, can be used to transmit torque from an input to an output element. Just like electric motors, they can thus provide position, velocity, and force-torque control of the output element. This paper introduces machine components, called adaptronic couplers, which can transmit variable torques highly dynamically from an input element to an output element employing static and/or dynamic friction. In the long run, systems (e.g. robots) based on these machine components are envisaged to compete with systems based on classic drive principles - especially electric motors - w.r.t. dynamics and power-to-mass-ratio. Apart from the concept itself, this paper addresses different control approaches and discusses their influence on energy consumption and wear. Moreover, various experimental results proving the basic concept are presented.
Robotics and Automation (ICRA), 2011 IEEE International Conference on; 06/2011
[show abstract][hide abstract] ABSTRACT: Rotary encoders are the most common sensors to measure angles in mechatronic and robotic components, e.g., servo motors or robot joints. Especially, if encoders are not mounted on the motor shaft but on the output side of geared motors, high encoder resolutions are required. Resolution requirements may increase further if velocities are to be derived from the encoder signals, e.g., for motion control purposes. To avoid noise amplification problems when estimating angular rates from encoder signals, angular velocity sensors may be employed instead. However, a significant drawback of angular rate sensors - particularly of cheap MEMS gyroscopes - is their drift. Both the requirements on encoder resolution and the drift of the angular rate signal can be reduced significantly by the sensor fusion approach presented in this paper. The approach utilizes the encoder to eliminate the drift of the angular rate signal and integrates the resulting signal to obtain estimates of the angle. Apart from a detailed description of the approach and a theoretical analysis of its disturbance characteristics, various simulation and experimental results are presented demonstrat- ing its effectiveness. the angle and the angular velocity. If encoders are used to provide angular velocity and angular acceleration signals, filters or estimators are generally required to reduce the noise in the velocity and acceleration signals. Regarding signals with low velocities, no encoder counts may occur during several sampling intervals. In this case, velocity estimation becomes especially delicate and sophisticated methods are required to provide meaningful estimates. Needless to say that increasing the encoder resolution is a viable solution in most cases but it comes at a non negligible cost. Therefore, direct angular rate measurement has become an attractive alternative with the advent of MEMS-based angular rate sensors. In contrast to expensive fiber optic gyroscopes (2), these sensors are affordable for most robotics applications. However, direct angular rate sensing with MEMS gyroscopes has a significant drawback: MEMS gyroscopes are not only affected by uncorrelated random noise but also by slowly time-varying disturbances denoted as drift. The characteris- tics of this random process have been analyzed in (3). The combination of low-resolution encoders with inexpen- sive (MEMS) angular rate sensors to compensate the major disadvantage of either sensor may be a viable alternative to high-resolution encoders. The angular rate sensor can be utilized to virtually increase encoder resolution while the encoder can be employed to eliminate the drift of the angular rate sensor. This approach may be implemented using model-based or model-free approaches. Considering industrial applicability, model-free approaches that can be applied with minimum setup and tuning efforts are generally more desirable than model-based approaches that require in- depth system knowledge. Therefore this paper introduces a model-free drift and offset compensator (DOC) employing FIR/IIR filtering tech- niques, which also lends itself to implementation in hardware (e.g. DSPs and FPGAs). Section II reviews related work. In Section III the concept of the DOC is presented and in Section IV its characteristics are analyzed. Section V shows the effectiveness of the approach in terms of encoder resolution enhancement and drift reduction with various simulations and experimental results. Section VI addresses further work and concludes the paper.
2011 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2011, San Francisco, CA, USA, September 25-30, 2011; 01/2011
[show abstract][hide abstract] ABSTRACT: One of the main tasks of intelligent vehicles is the extraction of information from the vehicle's surroundings and the understanding of the extracted information. The understanding of the environment allows the vehicle to drive autonomously or to support the driver in difficult or dangerous situations. In this paper we propose a vision-based hierarchical interpretation approach. First, we consider one single physical camera as a set of virtual sensors, where each virtual sensor gathers a type of 3d information. Then, the 3d information of this set is converted to high-level information that allows further reasoning. The interpretation is based on a hierarchical scene representation, where objects are recognized using nonparametric belief propagation. To demonstrate this approach we adopted the scene understanding to a parking spot finding application and show that it is real-time applicable and reliable even for multiple camera (on-board) systems.
[show abstract][hide abstract] ABSTRACT: Worldwide, ageing societies are bringing challenges for independent living and healthcare. Health-enabling technologies for pervasive healthcare and sensor-enhanced health information systems offer new opportunities for care. In order to identify, implement and assess such new information and communication technologies (ICT) the 'Lower Saxony Research Network Design of Environments for Ageing' (GAL) has been launched in 2008 as interdisciplinary research project. In this publication, we inform about the goals and structure of GAL, including first outcomes, as well as to discuss the potentials and possible barriers of such highly interdisciplinary research projects in the field of health-enabling technologies for pervasive healthcare. Although GAL's high interdisciplinarity at the beginning slowed down the speed of research progress, we can now work on problems, which can hardly be solved by one or few disciplines alone. Interdisciplinary research projects on ICT in ageing societies are needed and recommended.
Informatics for Health and Social Care 12/2010; 35(3-4):92-103. · 1.27 Impact Factor
[show abstract][hide abstract] ABSTRACT: This paper introduces a generic framework for sensor-based robot motion control. The key contribution is the introduction
of an adaptive selection matrix for sensor-based hybrid switched-system control. The overall control system consists of multiple
sensors and open- and closed-loop controllers, in-between which the adaptive selection matrix can switch discretely in order
to supply command variables for low-level controllers of robotic manipulators. How control signals are chosen, is specified
by Manipulation Primitives, which constitute the interface to higher-level applications. This programming paradigm is formally specified in order to
establish the possibility of executing sensor-guided and sensor-guarded motion commands simultaneously and in a very open way, such that any kind and any number of sensors can be
addressed. A further key feature of this generic approach is, that the control structure can be directly mapped to a corresponding
software architecture. The resulting control system is freely scalable depending on the performance requirements of the desired
[show abstract][hide abstract] ABSTRACT: In this contribution we present a uniform notation for any kind of kinematic structure ranging from serial robots and parallel
robots to hybrid kinematic structures as well as from multi-finger grippers to the kinematics of locomotive or humanoid robots.
If kinematic structures contain passive joints, they often are of spherical or cardan nature. To describe these types of joints,
a new notation based on the well-known Denavit-Hartenberg notation is presented. Additionally, closed kinematic chains and
structures with more than one chain attached to a robot basis can be denoted by our graph based representation. Our goal is
to provide the robot community with a unified description of kinematic structures - as it has been done by the classical DH-parameter
notation for serial robots - in order to support the development for exchangeable programming tools and ideas.
[show abstract][hide abstract] ABSTRACT: Force guided assembly is attractive but the implementation is challenging when uncertainties are present. In these cases, contact models that guide the assembly process are attractive. This article elaborates the usage of static contact models, which map displacements to force-torque vectors of particular contact situations (force-torque map). This model, a map of discrete points, contains force and torque values that correspond to position errors. The inversion of this map, using forces and torques measured from an assembly attempt, yields correction movements in order to accomplish the assembly iteratively. A hypothesis stating the dependency of the model's quality on its injectivity is investigated. This aspect is studied thoroughly in so-called redundancy maps, which reveal regions of considerable ambiguity of the model. Experimental results are presented, which validate the hypothesis about the dependency of the convergence of the assembly process on the ambiguity of the initial position. In addition to the peg-in-hole problem, which has become a standard scenario to validate force guided assembly, the scope of this article also covers force guided assembly of more complex parts. Here, the analysis gives evidence to believe that it is unlikely that the implementation convergences acceptably, which is validated by experimental results.
Intelligent Robots and Systems (IROS), 2010 IEEE/RSJ International Conference on; 11/2010
[show abstract][hide abstract] ABSTRACT: This contribution aims to present a complete process chain starting from the initial specification of assembly tasks via assembly
sequence planning, all the way through task planning and finally task execution. It demonstrates how robot programs can be
generated automatically from CAD data. For assembly sequence planning, a new assembly planning system has been developed and
evaluated with several representative assembly groups. As an interface to robot control systems, manipulation primitives have
been introduced. Manipulation primitive nets are an appropriate way to deal with uncertainties occurring during assembly task
execution. This contribution introduces a new approach, demonstrating how to generate manipulation primitive nets automatically.
Firstly, CAD data are segmented into surface primitives. A contact formation graph based on topological contacts between such
surface primitives is generated afterwards. Based on these contact formation graphs, manipulation primitive nets can be successfully
derived. The concept for automated robot programming presented in this contribution is supported by real experiments of actual
[show abstract][hide abstract] ABSTRACT: We present a model-driven approach to the segmentation of nasal cavity and paranasal sinus boundaries. Based on computed tomography
data of a patients head, our approach aims to extract the border that separates the structures of interest from the rest of
the head. This three-dimensional region information is useful in many clinical applications, e.g. diagnosis, surgical simulation,
surgical planning and robot assisted surgery. The desired boundary can be made up of bone, mucosa or air what makes the segmentation
process very difficult and brings traditional segmentation approaches, like e.g. region growing, to their limits. Motivated
by the work of Tsai et al.  and Leventon et al. , we therefore show how a parametric level-set model can be generated from hand-segmented nasal cavity and paranasal sinus
data that gives us the ability to transform the complex segmentation problem into a finite-dimensional one. On this basis,
we propose a processing chain for the automated segmentation of the endonasal structures that incorporates the model information
and operates without any user interaction. Promising results are obtained by evaluating our approach on two-dimensional data
slices of 50 patients with very diverse paranasal sinuses.
[show abstract][hide abstract] ABSTRACT: Complex mechatronic systems with a number of independently driven elements generally employ an equal number of drive elements. Commonly, these systems are actuated by electric motors and hence electric energy is transmitted through them. In various systems, however, mechanical energy may be distributed through the system instead of electrical energy and may be used directly to drive output elements. Using electric servo motors and servo drives, highly dynamic torque, velocity, and position control of output elements can be achieved easily. However, there is no machine component that can transmit a variable torque from an input shaft to an output element and enable highly dynamic torque, velocity, and position control of the output element. Adaptronic couplers are novel machine components which can close this gap. This paper introduces the concept of adaptronic couplers and addresses potential applications in mechatronics/robotics. Various experimental results allow a preliminary evaluation of the viability of the approach as well as a discussion of constructional and control issues still to be addressed.
Advanced Intelligent Mechatronics (AIM), 2010 IEEE/ASME International Conference on; 08/2010