67

221

S.325-332

S.162-171

Visual scene decomposition into semantic entities is one of the major challenges when creating a reliable object grasping system. Recently, we introduced a bottom-up hierarchical clustering approach [1] which is able to segment objects and parts in a scene. In this paper, we introduce a transform from such a segmentation into a corresponding, hierarchical saliency function. In comprehensive experiments we demonstrate its ability to detect salient objects in a scene. Furthermore, this hierarchical saliency defines a most salient corresponding region (scale) for every point in an image. Based on this, an easy-to-use pick and place manipulation system was developed and tested exemplarily.

S.344-351

We present a fast method of detecting loop closure opportunities through the use of similarity measures on histograms extracted from 3-D LIDAR data. We avoid computationally expensive features and compute histograms over simple global statistics of the LIDAR scans. The resulting histograms encode sufficient information to detect spatially close scans with high precision and recall and can be computed at rates faster than data acquisition on modest consumer-grade hardware. Our approach is able to match previously established results in LIDAR loop closure detection with less computational overhead.

People detection surely is one of the hottest topics in Computer Vision. In this work we propose and evaluate the fusion of thermal images and images from the visible spectrum for the task of people detection. Our main goal is to reduce the false positive rate of the Implicit Shape Model (ISM) object detector, which is commonly used for people detection. We describe five possible methods to integrate the thermal data into the detection process at different processing steps. Those five methods are evaluated on several test sets we recorded. Their performance is compared to three baseline detection approaches. The test sets contain data from an indoor environment and from outdoor environments at days with different ambient temperatures. The data fusion methods decrease the false positive rate especially on the outdoor test sets.

Mobile robots with reconfigurable chassis are able to traverse unstructured outdoor environments with boulders or rubble, and overcome challenging structures in urban environments, like stairs or steps. Autonomously traversing rough terrain and such obstacles while ensuring the safety of the robot is a challenging task in mobile robotics. In this paper we introduce a two-phase motion planning algorithm for actively reconfigurable tracked robots. We first use the completeness of a graph search on a regular grid to quickly find an initial path in a low dimensional space, considering only the platform's operating limits instead of the complete state. We then take this initial path to focus the RRT* search in the continuous high-dimensional state space including the actuators of the robot. We do not rely on a detailed structure/terrain classification or use any predefined motion sequences. Hence, our planner can be applied to urban structures, like stairs, as well as rough unstructured environments. Simulation results prove our method to be effective in solving planning queries in such environments.

We developed a robust real time hand gesture based interaction system to effectively communicate with a mobile robot which can operate in an outdoor environment. The system enables the user to operate a mobile robot using hand gesture based commands. In particular the system offers direct on site interaction providing better perception of environment to the user. To overcome the illumination challenges in outdoors, the system operates on depth images. Processed depth images are given as input to a convolutional neural network which is trained to detect static hand gestures.

This paper presents an approach to exploit the richer information of sensor data provided by 3d laser rangefinders for the purpose of person tracking. Introduced is a method to adapt the observation model of a particle filter, to identify partial and full occlusions of a person, to determine the amount of occlusion behind an obstacle, and the occluding obstacle itself. This is done by tracing rays from positions near the person to the sensor and determining whether the ray hits an obstacle. The laser range data is represented using a voxel grid, which facilitates efficient retrieval and data reduction. As our experiments show, our proposed tracking approach is able to reliably keep track of a person in real-time, even when only partially visible, when moving in uneven terrain, or when the person passes closely another person of different size.

An active needle is proposed for the development of MRI guided percutaneous procedures. The needle uses internal laser heating, conducted via optical fibers, of a shape memory alloy (SMA) actuator to produce bending in the distal section of the needle. Active bending of the needle as it is inserted allows it to reach small targets while overcoming the effects of interactions with surrounding tissue, which can otherwise deflect the needle away from its ideal path. The active section is designed to bend preferentially in one direction under actuation, and is also made from SMA for its combination of MR and bio-compatibility and its superelastic bending properties. A prototype, with a size equivalent to standard 16G biopsy needle, exhibits significant bending with a tip rotation of more than 10°. A numerical analysis and experiments provide information concerning the required amount of heating and guidance for design of efficient optical heating systems.

In recent years multiple simultaneous localization and mapping (SLAM) algorithms have been proposed, which address the challenges of 3D environments in combination with six degress of freedom in the robot position. Commonly, solutions based on scan-matching algorithms are applied. In contrast to these approaches, we propose to use a particle filter transferring the concept of the 2D Rao-Blackwellized particle filter SLAM to 3D. As filter input, 3D laser range data and odometry readings are obtained while the robot is in motion. The ground plane is estimated based on previously built map parts, thereby approaching the problem that not all degrees of freedom are covered by the odometry. To gain control of the high memory requirements for the particles' 3D map representations, we introduce a memory efficient search structure and adapt a technique to efficiently organize and share maps between particles. We evaluate our approach based on experimental results obtained by

Segmenting range data into semantic categories has become a more and more active field of research in robotics. In this paper, we advocate to view this task as a problem of fast, large-scale retrieval. Intuitively, given a dataset of millions of labeled scan points and their neighborhoods, we simply search for similar points in the datasets and use the labels of the retrieved ones to predict the labels of a novel point using some local prediction model such as majority vote or logistic regression. However, actually carrying this out requires highly efficient ways of (1) storing millions of scan points in memory and (2) quickly finding similar scan points to a target scan point. In this paper, we propose to address both issues by employing Weiss et al.'s recent spectral hashing. It represents each item in a database by a compact binary code that is constructed so that similar items will have similar binary code words. In turn, similar neighbors have codes within a small Hamming distance of the code for the query. Then, we learn a logistic regression model locally over all points with the same binary code word. Our experiments on real world 3D scans show that the resulting approach, called spectrally hashed logistic regression, can be ultra fast at prediction time and outperforms state-of-the art approaches such as logistic regression and nearest neighbor.

In this paper we present a method for navigating a multi-robot system through an environment while additionally maintaining a predefined set of constraints. Possible constraints are the requirement to keep up the direct line-of-sight between robots or to ensure that robots stay within a certain distance. Our approach is based on graph structures that model movements and constraints separately, in order to cover different robots and a large class of possible constraints. Additionally, the partition of movement and constraint graph allows us to use known graph algorithms like Steiner trees to solve the problem of finding a target configuration for the robots. We construct so called separated distance graphs from the Steiner tree and the underlying roadmap graph, which allow assembling valid navigation plans fast.

In this paper, we present a visual object tracker for mobile systems that is able to specialize to individual objects during tracking. The core of our method is a novel observation model and the way it is automatically adapted to a changing object and background appearance over time. The model is integrated into the well known Condensation algorithm (SIR filter) for statistical inference, and it consists of a boosted ensemble of simple threshold classifiers built upon center-surround Haar-like features, which the filter continuously updates based on the images perceived. We present optimizations and reasonable approximations to limit the computational costs. Thus, the final algorithms are capable of processing video input at real-time. To experimentally investigate the gain of adapting the observation model we compare two different approaches with a non-adapting version of our observation model: maintaining a single observation model for all particles, and maintaining individual observation models for each particle. In addition, experiments were conducted to compare system performances between the proposed algorithms and two other state of the art Condensation based tracking approaches.

To ensure the safety of people, it is important that mobile robots operating in populated environments are able to take the motions of humans in their vicinity into account. An especially demanding task in this respect is accompanying a person walking through an unknown and busy environment, because it requires the robot to stay close to his client and simultaneously prevent bumping into any passers-by. This paper presents a local navigation planning approach for collision avoidance, which aims at achieving this goal. The presented technique uses probabilistic roadmaps to plan collision-free paths to a given target location relative to the robot. A laser-based people tracking component is used to estimate the motions of humans in the robot's surrounding, and a potential field method is applied for predicting the humans' future trajectories based on this information. In addition to preventing collisions, the predictions enable us to choose appropriate target locations relative to the person being attended. We tested our method on real robots and in simulations. The experiments carried out in an office environment confirm that the integrated motion prediction actually improves the performance of the collision avoidance and the robot's ability to stay close to the client it accompanies.

Tracking extended targets is of central interest in mobile robotics as it is a prerequisite for interaction with the environment. There are significant differences between tracking punctiform targets and tracking extended targets. Most of the existing algorithms assume the targets to be punctiform, which is not always suitable. In recent years, two advanced algorithms for tracking extended targets have been developed by the authors-a switching algorithm and the SJPDAF. The switching algorithm uses the Kalman filter and an adapted version of the Viterbi algorithm, which includes certain geometrical characteristics of the problem. The SJPDAF combines the idea of particle filtering with the JPDAF. In this paper we present these two algorithms and compare them with respect to accuracy, speed and robustness in case of crossing targets. We show, that the more recently developed switching algorithm outperforms the SJPDAF regarding these criterions

In this paper we look at the problem of searching a human intruder in a closed environment with a small group of mobile robots. In this context motion models for the intruder play an important role for planning the coordination of the robots. Often, simple Brownian motion models are used for this purpose. However, the assumed completely random change of direction in each time step is very unrealistic. We present an improved Markovian motion model that takes the intended motion direction of a person into account in order to achieve a more realistic motion prediction. This model is then used to estimate a probability distribution of an intruder's location within the environment. We develop a greedy algorithm that employs this distribution to coordinate the search of the environment by a group of robots. Finally, we compare our algorithm to two simple search methods and evaluate its behavior in simulation experiments

Efficient exploration of unknown environments is a fundamental problem in mobile robotics. We present an approach to distributed multirobot mapping and exploration. Our system enables teams of robots to efficiently explore environments from different, unknown locations. In order to ensure consistency when combining their data into shared maps, the robots actively seek to verify their relative locations. Using shared maps, they coordinate their exploration strategies to maximize the efficiency of exploration. This system was evaluated under extremely realistic real-world conditions. An outside evaluation team found the system to be highly efficient and robust. The maps generated by our approach are consistently more accurate than those generated by manually measuring the locations and extensions of rooms and objects

In this paper we discuss the problem of monitoring and searching an indoor environment for an intruder with a group of mobile robots. We present a graph-based algorithm to coordinate a group of robots which takes the limitations and uncertainties of sensors into account and is able to find good coordination plans efficiently even for large environments. We analyze and compare the approach against other coordination strategies based on a new probabilistic framework that allows to evaluate the performance of any coordination strategy based on a probabilistic sensor model and a worst case behavior model for intruders. Using this framework we demonstrate the capabilities of the planning algorithm in several simulation experiments.

In this article we introduce a hierarchical Bayesian model to estimate a set of colors with a mobile robot. Estimating colors is particularly important if objects in an environment can only be distinguished by their color. Since the appearance of colors can change due to variations in the lighting condition, a robot needs to adapt its color model to such changes. We propose a two level Gaussian model in which the lighting conditions are estimated at the upper level using a switching Kalman filter. A hierarchical Bayesian technique learns Gaussian priors from data collected in other environments. Furthermore, since estimation of the color model depends on knowledge of the robot's location, we employ a Rao-Blackwellised particle filter to maintain a joint posterior over robot positions and lighting conditions. We evaluate the technique in the context of the RoboCup AIBO league, where a legged AIBO robot has to localize itself in an environment similar to a soccer field. Our experiments show that the robot can localize under different lighting conditions and adapt to changes in the lighting condition, for example, due to a light being turned on or off.

Tracking the activity of people in indoor environments has gained considerable attention in the robotics community over the last years. Most of the existing approaches are based on sensors, which allow to accurately determining the locations of people but do not provide means to distinguish between different persons. In this paper we propose a novel approach to tracking moving objects and their identity using noisy, sparse information collected by id-sensors such as infrared and ultrasound badge systems. The key idea of our approach is to use particle filters to estimate the locations of people on the Voronoi graph of the environment. By restricting particles to a graph, we make use of the inherent structure of indoor environments. The approach has two key advantages. First, it is by far more efficient and robust than unconstrained particle filters. Second, the Voronoi graph provides a natural discretization of human motion, which allows us to apply unsupervised learning techniques to derive typical motion patterns of the people in the environment. Experiments using a robot to collect ground-truth data indicate the superior performance of Voronoi tracking. Furthermore, we demonstrate that EM-based learning of behavior patterns increases the tracking performance and provides valuable information for high-level behavior recognition.

This paper presents an approach to collision avoidance for mobile robots that uses local path planning within a map built from the robot's latest range measurements followed by a search for control commands to steer the robot towards the goal safely and efficiently. While the search for control commands allows robot dynamics to be taken into account and therefore allows for very smooth navigation behavior, planning is able to determine the optimal path towards the goal and minimizes the likelihood that the robot gets trapped in dead end situations. We introduce and compare several approaches to utilize the results of path planning in the search for good control commands. In extensive experiments, we show that our combined approach outperforms a collision avoidance that is purely based on local planning and we empirically investigate the pros-and-cons of the different search strategies for control commands and find one to significantly outperform the others.

The problem of generating maps with mobile robots has received considerable attention over the past years. However, most of the approaches assume that the environment is static during the data-acquisition phase. In this paper we consider the problem of creating maps with mobile robots in populated environments. Our approach uses a probabilistic method to track multiple people and to incorporate the results of the tracking technique into the mapping process. The resulting maps are more accurate since corrupted readings are treated accordingly during the matching phase and since the number of spurious objects in the resulting maps is reduced. Our approach has been implemented and tested on real robot systems in indoor and outdoor scenarios. We present several experiments illustrating the capabilities of our approach to generate accurate 2D and 3D maps.

One of the goals in the field of mobile robotics is the development of mobile platforms which operate in populated environments. For many tasks it is therefore highly desirable that a robot can determine the positions of the humans in its surrounding. We introduce sample-based joint probabilistic data association filters to track multiple moving objects with a mobile robot. Our technique uses the robot's sensors and a motion model of the objects being tracked. A Bayesian filtering technique is applied to adapt the tracking process to the number of objects in the sensor range of the robot. Our approach to tracking multiple moving objects has been implemented and tested on a real robot. We present experiments illustrating that our approach is able to robustly keep track of multiple persons even in situations in which people are temporarily occluded. The experiments furthermore show that the approach outperforms other techniques developed so far.

One of the goals in the field of mobile robotics is the development of mobile platforms which operate in populated environments and offer various services to humans. For many tasks it is highly desirable that a robot can determine the positions of the humans in its surrounding. In this paper we present a method for tracking multiple moving objects with a mobile robot. We introduce a sample-based variant of joint probabilistic data association filters to track features originating from individual objects and to solve the correspondence problem between the detected features and the filters. In contrast to standard methods, occlusions are handled explicitly during data association. The technique has been implemented and tested on a real robot. Experiments carried out in a typical office environment show that the method is able to track multiple persons even when the trajectories of two people are crossing each other.

This paper describes an interactive tour-guide robot, which was successfully exhibited in a Smithsonian museum. During its two weeks of operation, the robot interacted with thousands of people, traversing more than 44 km at speeds of up to 163 cm/sec. Our approach specifically addresses issues such as safe navigation in unmodified and dynamic environments, and short-term human-robot interaction. It uses learning pervasively at all levels of the software architecture.

Visualization is an important precondition for successful teleoperation of instructable mobile robots. Data connections with varying and limited bandwidth such as the Internet, however prohibit the continuous transmission of video signals. In this paper we propose a predictive simulation technique which is designed to permit the reliable visualization of the robot's actions over the Internet. It differs from previous approaches in that it includes an odometry and sensor simulation. This simulation of the robot allows the integration of a complete robot control system to reliably predict complex actions of the robot even if large transmission gaps of several seconds occur. We describe an application of the predictive simulation technique to navigation experiments with mobile robots. We present different experiments carried out with a real robot illustrating that the predictive simulation technique provides accurate visualizations of the robot's actions even if transmission gaps of more than ten seconds occur.