Marian Himstedt

Large scale place recognition in 2D LIDAR scans using Geometrical Landmark Relations

The recognition of places that have already been visited is a fundamental requirement for a mobile robot. This particularly concerns the detection of loop closures while mapping environments as well as the global localization w.r.t. to a prior map. This paper introduces a novel solution to place recognition with 2D LIDAR scans. Existing approaches utilize descriptors covering the local appearance of discriminative features within a bag-of-words (BOW) framework accompanied with approximate geometric verification. Though limiting the set of potential matches their performance crucially drops for increasing number of scans making them less appropriate for large scale environments. We present Geometrical Landmark Relations (GLARE), which transform 2D laser scans into pose invariant histogram representations. Potential matches are found in sub-linear time using an efficient Approximate Nearest Neighbour (ANN) search. Experimental results obtained from publicly available datasets demonstrate that GLARE significantly outperforms state-of-the-art approaches in place recognition for large scale outdoor environments, while achieving similar results for indoor settings. Our Approach achieves recognition rates of 93% recall at 99% precision for a dataset covering a total path of about 6.5 km.

Geometry matters: Place recognition in 2D range scans using Geometrical Surface Relations

Place recognition is a fundamental requirement for mobile robots. It is particularly needed for detecting loop closures in SLAM and to enable self-localization for mobile robots given a prior map. The multitude of existing approaches rely on appearance based methods, e.g. the extraction of interest points in terms of local extrema. It can be observed that the availability of these features is highly environment specific and the limited descriptiveness causes a large number of false-positive matches. This paper utilizes a generic environment description based on normal surface primitives. The association of different places is done using Geometrical Surface Relations (GSR) of co-occurring primitives. Experimental results obtained from publicly available datasets demonstrate that GSR outperforms state-of-the-art approaches in place recognition for large scale outdoor as well as indoor environments.

Awakening history: Preparing a museum tour guide robot for augmenting exhibits

While the idea of tour guide robots has been addressed several times, applying a video-projector based augmented reality component to this scenario is rather new. We show requirements of the localization system of the robot and how they can be fulfilled, as well as a basic system for projection correction and its suitability for this scenario.

Online semantic mapping of logistic environments using RGB-D cameras

Automated guided vehicles require spatial representations of their working spaces in order to ensure safe navigation and carry out high-level tasks. Typically, these models are given by geometric maps. Even though these enable basic robotic navigation, they off-the-shelf lack the availability of task-dependent information required to provide services. This article presents a semantic mapping approach augmenting existing geometric representations. Our approach demonstrates the automatic annotation of map subspaces on the example of warehouse environments. The proposals of an object recognition system are integrated in a graph-based simultaneous localization and mapping framework and eventually propagated into a global map representation. Our system is experimentally evaluated in a typical warehouse consisting of common object classes expected for this type of environment. We discuss the novel achievements and motivate the contribution of semantic maps toward the operation of automated guided vehicles in the context of Industry 4.0.

What's around me: Towards non-negative matrix factorization based localization

This paper presents a new method for representing the local and global environment captured by 2D range scans using non-negative matrix factorization (NMF). Unlike other approaches, we do not predefine features or geometric primitives, but, in contrast, extract environment specific basis primitives from occupancy grid maps. Similar to feature based methods, these enable fast map matching and simultaneously allow to reconstruct the entire initial map. Our approach enables efficient localization on highly compact environment representations. The proposed approach is applied to a publicly available dataset. First promising results are presented motivating further investigation in the application of NMF for localization and mapping.

Some Room for GLVQ: Semantic Labeling of Occupancy Grid Maps

This paper aims at an approach for labeling places within a grid cell environment. For that we propose a method that is based on non-negative matrix factorization (NMF) to extract environment specific features from a given occupancy grid map. NMF also computes a description about where on the map these features need to be applied. We use this description after certain pre-processing steps as an input for generalized learning vector quantization (GLVQ) to achieve the classification or labeling of the grid cells. Our approach is evaluated on a standard data set from University of Freiburg, showing very promising results.

Towards robust vision-based self-localization of vehicles in dense urban environments

Self-localization of ground vehicles in densely populated urban environments poses a significant challenge. The presence of tall buildings in close proximity to traversable areas limits the use of GPS-based positioning techniques in such environments. This paper presents an approach to global localization on a hybrid metric-topological map using a monocular camera and wheel odometry. The global topology is built upon spatially separated reference places represented by local image features. In contrast to other approaches we employ a feature selection scheme ensuring a more discriminative representation of reference places while simultaneously rejecting a multitude of features caused by dynamic objects. Through fusion with additional local cues the reference places are assigned discrete map positions allowing metric localization within the map. The self-localization is carried out by associating observed visual features with those stored for each reference place. Comprehensive experiments in a dense urban environment covering a time difference of about 9 months are carried out. This demonstrates the robustness of our approach in environments subjected to high dynamic and environmental changes.

Semantic Monte-Carlo localization in changing environments using RGB-D cameras

The localization with respect to a prior map is a fundamental requirement for mobile robots. The commonly used adaptive monte carlo localization (AMCL) can be found on most of the mobile robots ranging from small cleaning robots to large AGVs. While achieving accurate pose estimates in static environments, this algorithm tends to fail in the presence of significant changes. Recently published extensions and alternatives to AMCL observe the environment over longer times while building complex spatio-temporal models. Our approach, in contrast, utilizes object recognition and prior semantic maps to enable robust localization. It exploits the fact that putative changes in the environment can be predicted based on prior semantic knowledge. Our system is experimentally evaluated in a warehouse environment being subject to frequent changes. This emphasizes the importance of our algorithm for challenging industrial applications.