Extracting semantic indoor maps from occupancy grids

Abstract

The primary challenge for any autonomous system operating in realistic, rather unconstrained scenarios is to manage the complexity and uncertainty of the real world. While it is unclear how exactly humans and other higher animals master these problems, it seems evident, that abstraction plays an important role. The use of abstract concepts allows to define the system behavior on higher levels. In this paper we focus on the semantic mapping of indoor environments. We propose a method to extract an abstracted floor plan from typical grid maps using Bayesian reasoning. The result of this procedure is a probabilistic generative model of the environment defined over abstract concepts. It is well suited for higher-level reasoning and communication purposes. We demonstrate the effectiveness of the approach using real-world data.

Sensor based adaptive metric-topological cell decomposition method for semantic annotation of structured environments

“…Global approaches on the other hand derive semantic information via place labeling. For instance, [4] builds a topological map based on the notion of connectivity and adjacency, [5] explores an occupancy grid map through the Voronoi graph, [6] bases a topological map on the connectivity between acquired images, [7] provides a topological graph of the environment based on the concept of virtual door, and [8] employs a series of kernels based on Markov chain Monte Carlo for semantic labeling.…”

“…The common reliance on line extraction [9], [8], [10], [11], [12] indicates the importance of orientation in semantic mapping. However, we cannot readily reuse these results, due to clutter and discontinuities in the physical structures.…”

Not seeing is also believing: Combining object and metric spatial information

“…These hybrid representations have been successfully applied in tasks such as navigation ( [3]). A related field that has been growing recently is semantic mapping (e.g., [4], [5], [6], [7]), where typically the focus is to endow topological regions of space with semantic attributes, such as in the task of place classification. Topological and semantic information is typically extracted from metric layers (occupancy grids).…”

“…No object observations have been made, so the object location distribution is uniform over the feasible range. The free cells in the middle of the location posterior distribution (top right) indicate that it is highly unlikely that any object can occupy those cells (which correspond to x ∈ [3,7]). This makes the posterior distribution multimodal.…”

Modeling of a Large Structured Environment

“…Many researchers have been contributing to this particular aspect, from different points of view. Some tried to model the environment through the topology of open space in geometric map, like [6], where they employed a series of kernels for semantic labeling of regions. Some others like [5], [14] and [9] proposed spatial maps, enhanced conceptually by object recognition in regions of the map.…”