Abstract
In this work, a framework is proposed to build topological models in mobile robotics, using an omnidirectional vision sensor as the only source of information. The model is structured hierarchically into three layers, from one high-level layer which permits a coarse estimation of the robot position to one low-level layer to refine this estimation efficiently. The algorithm is based on the use of clustering approaches to obtain compact topological models in the high-level layers, combined with global appearance techniques to represent robustly the omnidirectional scenes. Compared to the classical approaches based on the extraction and description of local features, global-appearance descriptors lead to models that can be interpreted and handled more intuitively. However, while local-feature techniques have been extensively studied in the literature, global-appearance ones require to be evaluated in detail to test their efficacy in map-building tasks. The proposed algorithms are tested with a set of publicly available panoramic images captured in realistic environments. The results show that global-appearance descriptors along with some specific clustering algorithms constitute a robust alternative to create a hierarchical representation of the environment.
Highlights
Introduction to Map Building Using VisionSensorsOver the last few years, mobile robots have become a crucial technology to solve a variety of tasks.Creating an optimal model of the environment where they are moving is one of the most important abilities they must be endowed with, so that they can estimate its position and navigate towards the target points
The robot is equipped with a catadioptric vision sensor which provides it with omnidirectional images from the environment
Owing mainly to the large quantity of information they can capture with only one snapshot (360◦ around the camera axis), omnidirectional vision systems have become a popular option in the implementation of mapping algorithms such as the SLAM method presented by Caruso et al [1], who prove that even at relatively low resolutions, the model built with omnidirectional images provides a good localization accuracy
Summary
Over the last few years, mobile robots have become a crucial technology to solve a variety of tasks. Realistic indoor environments are prone to the perceptual aliasing problem, that is, images which have been captured from different positions may have a similar appearance, which would lead to confusion about the position of the robot Taking these facts into account, a processing step is necessary to extract relevant and distinguishable information from the scenes which is able to cope with these events. The current work continues and expands the scope of the research presented in [19], where a comparative evaluation between some global description methods is carried out to obtain a low-level map and to create the topological relationships between the images of this level This evaluation is extended and generalized, and a method is proposed to create hierarchical maps which arrange the information more optimally, into several layers, which would permit an efficient localization subsequently.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
