Automatic appearance‐based loop detection from three‐dimensional laser data using the normal distributions transform

Abstract

We propose a new approach to appearance-based loop detection for mobile robots, using three-dimensional (3D) laser scans. Loop detection is an important problem in the simultaneous localization and mapping (SLAM) domain, and, because it can be seen as the problem of recognizing previously visited places, it is an example of the data association problem. Without a flat-floor assumption, two-dimensional laser-based approaches are bound to fail in many cases. Two of the problems with 3D approaches that we address in this paper are how to handle the greatly increased amount of data and how to efficiently obtain invariance to 3D rotations. We present a compact representation of 3D point clouds that is still discriminative enough to detect loop closures without false positives (i.e., detecting loop closure where there is none). A low false-positive rate is very important because wrong data association could have disastrous consequences in a SLAM algorithm. Our approach uses only the appearance of 3D point clouds to detect loops and requires no pose information. We exploit the normal distributions transform surface representation to create feature histograms based on surface orientation and smoothness. The surface shape histograms compress the input data by two to three orders of magnitude. Because of the high compression rate, the histograms can be matched efficiently to compare the appearance of two scans. Rotation invariance is achieved by aligning scans with respect to dominant surface orientations. We also propose to use expectation maximization to fit Magnusson et al.: Automatic Appearance-Based Loop Detection from 3D Laser Data Using NDT • 893a gamma mixture model to the output similarity measures in order to automatically determine the threshold that separates scans at loop closures from nonoverlapping ones. We discuss the problem of determining ground truth in the context of loop detection and the difficulties in comparing the results of the few available methods based on range information. Furthermore, we present quantitative performance evaluations using three realworld data sets, one of which is highly self-similar, showing that the proposed method achieves high recall rates (percentage of correctly identified loop closures) at low falsepositive rates in environments with different characteristics. C 2009 Wiley Periodicals, Inc.

SegMatch: Segment based place recognition in 3D point clouds

“…Using global descriptors of the local point cloud for place recognition is also proposed [13][14][15]. Rohling et al [13] propose to describe each local point cloud with a 1D histogram of point heights, assuming that the sensor keeps a constant height above the ground.…”

“…Finally, Iterative Closest Point (ICP) is used for computing the relative pose between point clouds. In another approach, Magnusson et al [15] split the cloud into overlapping grids and compute shape properties (spherical, linear, and several type of planar) of each cell and combine them into a matrix of surface shape histograms. Similar to other works, these descriptors are compared for recognizing places.…”

TGF‐β signaling is dynamically regulated during the alveolarization of rodent and human lungs

“…Alternative scan representations used for pairwise alignment include the normalized distributions transform (NDT) (Magnusson, Lilienthal, Andreasson, & Duckett, 2007), which provides a reduced representation of the scan by discretizing the scan volume. Although this scan representation was later extended to provide loop closure detection (Magnusson, Andreasson, Nüchter, & Lilienthal, 2009), these approaches were designed for localization in underground mines, where the wall structure can be exploited. Similarly, the use of planar surface patches (Pathak et al, 2010) for 3D mapping of urban disaster sites still utilized some structure inherent in the environment.…”

Semantic-assisted 3D normal distributions transform for scan registration in environments with limited structure

“…In the point-to-distribution (P2D) variant, the registration of a new scan then becomes a problem of fitting its points to the distribution, and is solved as a least-squares problem. Several proposals have been made regarding the segmentation of space [2], the representation of orientation [7], its use in loop closure [8], and the use of features instead of the full point cloud [9]. Important extensions include the distribution to distribution registration (D2D-NDT) presented in [10], the fusion of NDT with occupancy maps [11]- [14] its use with Monte-Carlo localization [15], [16], and the integration of color information in the registration [17].…”