GPGM-SLAM: a Robust SLAM System for Unstructured Planetary Environments with Gaussian Process Gradient Maps

Abstract

Simultaneous Localization and Mapping (SLAM) techniques play a key role towards longterm autonomy of mobile robots due to the ability to correct localization errors and produce consistent maps of an environment over time. Contrarily to urban or man-made environments, where the presence of unique objects and structures offer unique cues for localization, the apperance of unstructured natural environments is often ambiguous and self-similar, hindering the performances of loop closure detection. In this paper, we present an approach to improve the robustness of place recognition in the context of a submap-based stereo SLAM based on Gaussian Process Gradient Maps (GPGMaps). GPGMaps embed a continuous representation of the gradients of the local terrain elevation by means of Gaussian Process regression and Structured Kernel Interpolation, given solely noisy elevation measurements. We leverage the image-like structure of GPGMaps to detect loop closures using traditional visual features and Bag of Words. GPGMap matching is performed as an SE(2) alignment to establish loop closure constraints within a pose graph. We evaluate the proposed pipeline on a variety of datasets recorded on Mt. Etna, Sicily and in the Morocco desert, respectively Moon-and Mars-like environments, and we compare the localization performances with state-of-the-art approaches for visual SLAM and visual loop closure detection.

Resiliency in Space Autonomy: a Review

“…In contrast, various Simultaneous localization and mapping (SLAM) frameworks are established for the structured environment. A few SLAM frameworks embark on the challenge of unstructured, uncontrolled and unknown natural environments [41]. However, in general, SLAM-based exploration methods are computationally expensive and require a high configuration onboard computers to localize and store the previous information on the map accurately.…”

Simultaneous Localization and Mapping (SLAM) and Data Fusion in Unmanned Aerial Vehicles: Recent Advances and Challenges

“…Considering that m landmarks obey Gaussian distribution mean and variance, a new landmark l N+1 adds a variable to the collection of N + 1 landmarks [114,161,162]. For m landmarks and d sensor data, since there are usually far more landmarks than sensors, M N. The non-zero elements in the S matrix correspond to the association between the sensor data and the landmark [163]. If there is the same observation made by two sensors, it is referred to as co-visibility.…”

GTOC 11: Results found at Beijing Institute of Technology and China Academy of Space Technology

“…The main difference between relative and absolute localization is that the geographic position, such as latitude and longitude, can not be determined in the relative localization method. The most widely used method for relative localization in terrestrial scenes is the visual odometry (VO) [32]- [34], which provides rapid positioning and high success rates. However, VO primarily estimates the local 6-DoF poses of the vehicles through interframe matching, rather than in the world coordinate system.…”