Visual Odometry and Mapping for Autonomous Flight Using an RGB-D Camera

By Albert S. Huang, Abraham Bachrach, Peter Henry, Michael Krainin, Daniel Maturana, Dieter Fox, Nicholas A. Roy. International Symposium of Robotics Research, 2011

725 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/1457f7dbd7294cea1e7745820c5ed634420f87a8

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

Embodied Question Answering in Photorealistic Environments With Point Cloud Perception (2019, background): “Secondly, these problems are typically approached with 2D perception (RGB frames) despite the widespread use of depth-sensing cameras (RGB-D) on actual robotic platforms [13–15].”
Dense visual SLAM for RGB-D cameras (2013, methodology): “Most state-of-the-art methods establish correspondences between sparsely selected visual features to estimate the camera motion [13]–[16].”
VoxNet: A 3D Convolutional Neural Network for real-time object recognition (2015, background): “Meanwhile, active range sensors such as LiDAR and RGBD cameras are an increasingly common choice of sensor for modern autonomous vehicles, including cars [1], quadrotors [2] and helicopters [3].”
Kintinuous: Spatially Extended KinectFusion (2012, methodology): “…have addressed the problem of spatially extended mapping with RGB-D cameras however they have taken more traditional approaches to localisation and scene estimation and therefore have required post-processing of the map to generate a representation suitable, for example, for navigation [7].”
Collaborative mapping of an earthquake‐damaged building via ground and aerial robots (2012, background): “Recent progress in the area of dense 3D mapping suggests that high-fidelity maps suitable for real search-and-rescue applications are achievable using laser and RGB-D sensors (Ellekilde et al., 2007; Huang et al., 2011).”
A benchmark for RGB-D visual odometry, 3D reconstruction and SLAM (2014, methodology): “We evaluate 5 different odometers; (i) DVO [13] which performs full frame alignment between pairs of RGB-D frames making use of a robust cost function to improve robustness to outliers; (ii) FOVIS [14] which uses FAST feature correspondences between pairs of RGB-D frames to resolve camera motion, re”
Camera-based navigation of a low-cost quadrocopter (2012, neutral): “For solving the SLAM problem on MAVs, different types of sensors such laser range scanners [5], monocular cameras [6, 7], stereo cameras [8] and RGB-D sensors [9] have been explored in the past.”
Dynamic obstacle avoidance for quadrotors with event cameras (2020, background): “The existing literature on obstacle avoidance for MAVs relies on standard cameras [in a monocular ( 9 – 11 ) or stereo configuration ( 12 – 15 )] or on depth cameras ( 16 – 18 ).”
Stereo vision based indoor/outdoor navigation for flying robots (2013, methodology): “Huang et al. [8] fuse visual odometry estimates from an RGB-D sensor with IMU measurements in an EKF for local navigation.”
State estimation for aggressive flight in GPS-denied environments using onboard sensing (2012, background): “In contrast, vehicles that are capable of flight using state estimates computed from onboard sensor data, are either confined to wide open areas without obstacles, or slow moving hovering vehicles such as quadrotors [2], [9].”
RTAB‐Map as an open‐source lidar and visual simultaneous localization and mapping library for large‐scale and long‐term online operation (2018, methodology): “Seven approaches have been integrated in RTAB‐Map: fast visual odometry (FOVIS) (Huang et al., 2011), Viso2 F IGURE 2 Block diagram of rgbd_odometry and stereo_odometry ROS nodes.”
PX4: A node-based multithreaded open source robotics framework for deeply embedded platforms (2015, background): “the average robotic application will also provide a companion computer, commonly running ROS on Linux [18][19][3].”
Vision-based state estimation for autonomous rotorcraft MAVs in complex environments (2013, background): “Stereo vision-based state estimation approaches for autonomous MAVs such as those proposed in [2, 9, 10] do not suffer from the problem scale drift as seen in monocular systems or limit the observable camera motion.”
A comparative analysis of SIFT, SURF, KAZE, AKAZE, ORB, and BRISK (2018, background): “Image matching is the cornerstone of many computer and machine vision applications specially: Robot Navigation, Pose Estimation [1], Visual Odometry [2], [3]; Visual Simultaneous Localization and Mapping [4], [5]; Object Detection, Object Tracking, Augmented Reality [6], Image Mosaicing, and Panoram”