Persistent localization and life-long mapping in changing environments using the Frequency Map Enhancement

Krajník, Tomáš; Fentanes, Jaime Pulido; Hanheide, Marc; Duckett, Tom

doi:10.1109/iros.2016.7759671

Cited by 39 publications

(20 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To handle the long-term changes, the place graph should be continuously updated. This is similar to the life-long localization and mapping tasks (Krajník et al , 2016; Sons and Stiller, 2018). In this paper, we track the vehicle location by using location particles, at each step, the predicted pose of each particle will be matched with nearest places to obtain the posterior probability and further finish the location updating process.…”

Section: System Designsupporting

confidence: 71%

Map-less long-term localization in complex industrial environments

Liu

Qiao

Suo

et al. 2021

View full text Add to dashboard Cite

Purpose This paper aims to study the localization problem for autonomous industrial vehicles in the complex industrial environments. Aiming for practical applications, the pursuit is to build a map-less localization system which can be used in the presence of dynamic obstacles, short-term and long-term environment changes. Design/methodology/approach The proposed system contains four main modules, including long-term place graph updating, global localization and re-localization, location tracking and pose registration. The first two modules fully exploit the deep-learning based three-dimensional point cloud learning techniques to achieve the map-less global localization task in large-scale environment. The location tracking module implements the particle filter framework with a newly designed perception model to track the vehicle location during movements. Finally, the pose registration module uses visual information to exclude the influence of dynamic obstacles and short-term changes and further introduces point cloud registration network to estimate the accurate vehicle pose. Findings Comprehensive experiments in real industrial environments demonstrate the effectiveness, robustness and practical applicability of the map-less localization approach. Practical implications This paper provides comprehensive experiments in real industrial environments. Originality/value The system can be used in the practical automated industrial vehicles for long-term localization tasks. The dynamic objects, short-/long-term environment changes and hardware limitations of industrial vehicles are all considered in the system design. Thus, this work moves a big step toward achieving real implementations of the autonomous localization in practical industrial scenarios.

show abstract

Section: System Designsupporting

confidence: 71%

Map-less long-term localization in complex industrial environments

Liu

Qiao

Suo

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Meanwhile, there is significant progress in outdoor visual place recognition [12] (VPR) under changing perceptual conditions [13]- [15]. Given the object dynamics and with an assumption of robust localization, people have also explored the potential of actively tracking [16], [17], modeling and inferring indoor object dynamics [18]- [20]. Without explicitly modeling the indoor object dynamics, our method combine both visual and semantic features to learn the appearance and structure of the scene which enables a robust similarity-based localization under dynamics.…”

Section: Related Workmentioning

confidence: 99%

SeanNet: Semantic Understanding Network for Localization Under Object Dynamics

Li¹,

Du²,

Zeng³

et al. 2021

Preprint

View full text Add to dashboard Cite

We aim for domestic robots to operate indoor for long-term service. Under the object-level scene dynamics induced by human daily activities, a robot needs to robustly localize itself in the environment subject to scene uncertainties. Previous works have addressed visual-based localization in static environments, yet the object-level scene dynamics challenge existing methods on long-term deployment of the robot. This paper proposes SEmantic understANding Network (SeanNet) that enables robots to measure the similarity between two scenes on both visual and semantic aspects. We further develop a similarity-based localization method based on SeanNet for monitoring the progress of visual navigation tasks. In our experiments, we benchmarked SeanNet against baselines methods on scene similarity measures, as well as visual navigation performance once integrated with a visual navigator. We demonstrate that SeanNet outperforms all baseline methods, by robustly localizing the robot under object dynamics, thus reliably informing visual navigation about the task status.

show abstract

“…Dymczyk et al proposed scoring functions for measuring landmark utility so that the summarized map would be available for long-term localization [21]. In [22], the map was represented as 2D spatiotemporal occupancy grid, and they tried to construct a map that corresponds to the change in the real environment. The temporal map was constructed for each robot of a multi-robot system, and the map was updated by merging the map [23].…”

Section: Related Workmentioning

confidence: 99%

A Hybrid Map with Permanent 3D Wireframes and Temporal Line Segments toward Long-Term Visual Localization

Kaneko

Takahashi

Ito

et al. 2019

SICE Journal of Control, Measurement, and System Integration

View full text Add to dashboard Cite

This paper deals with map-construction problems for visual localization. Basically, the map is an aggregation of visual landmarks, and it is desirable that such landmarks exist densely and permanently for long-term localization. However, there are no landmarks with these attributes at the same time. In order to solve this problem, we propose a hybrid map with permanent landmarks and temporal landmarks. As a permanent landmark, we employ 3D wireframe which can be easily obtained from architectural CAD. For a temporal landmark, we use line segments which are visually detected in images captured by a camera. To handle these two types of landmarks on the same map, we develop two algorithms. One is to extract temporal line segments from images containing two mixed landmarks, and the other is to reconstruct them into the 3D wireframe map. We experimentally demonstrated that the proposed hybrid map outperformed the 3D wireframe map in terms of localization accuracy.

show abstract

Persistent localization and life-long mapping in changing environments using the Frequency Map Enhancement

Cited by 39 publications

References 17 publications

Map-less long-term localization in complex industrial environments

Map-less long-term localization in complex industrial environments

SeanNet: Semantic Understanding Network for Localization Under Object Dynamics

A Hybrid Map with Permanent 3D Wireframes and Temporal Line Segments toward Long-Term Visual Localization

Contact Info

Product

Resources

About