Real-time visual odometry from dense RGB-D images

By Frank Steinbrücker, Jürgen Sturm, Daniel Cremers. 2011

413 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

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

RGB-D datasets using microsoft kinect or similar sensors: a survey (2016, background): “Six research publications about evaluating ego-motion estimation and SLAM over TUM Benchmark dataset are [21, 38, 45, 84, 86, 87].”
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”
Visual SLAM algorithms: a survey from 2010 to 2016 (2017, methodology): “On the other hand, phtometric consistency based camera motion tracking methods have been proposed [22, 23, 64].”
Dense visual SLAM for RGB-D cameras (2013, methodology): “In recent years, dense tracking and mapping methods have appeared whose performance is comparable to feature-based methods [3], [4].”
Range sensor and silhouette fusion for high-quality 3D Scanning (2015, background): “[27] discuss an ap2015 IEEE International Conference on Robotics and Automation (ICRA) Washington State Convention Center Seattle, Washington, May 26-30, 2015”
Large-Scale Multi-resolution Surface Reconstruction from RGB-D Sequences (2013, background): “In contrast, recently upcoming dense methods aim at aligning the full image by minimizing the photometric error over all pixels [3, 13, 15, 18], thereby exploiting the available image information better than featurebased methods.”
BA-Net: Dense Bundle Adjustment Network (2018, methodology): “(2018) implement the direct method (Steinbruecker et al., 2011) as a differentiable component to compute camera motion after scene depth is estimated by the method in Zhou et al.”
NICP: Dense normal based point cloud registration (2015, methodology): “DVO reaches the highest frame rate, since it processes a substantial smaller quantity of data.”
Robust odometry estimation for RGB-D cameras (2013, methodology): “The advantage of these methods is that they give precise motion estimates and are fast enough to run in real-time on a single CPU core.”
Evaluating Egomotion and Structure-from-Motion Approaches Using the TUM RGB-D Benchmark (2012, background): “[7] (from our group) developed and analyzed an energy-based approach to egomotion estimation based on image warping.”
Robust real-time visual odometry for dense RGB-D mapping (2013, methodology): “Less commonly an image warping method is used in the parameterisation of a camera transformation as used by Audras et al. [4] and Steinbruecker et al. [8].”
Colored Point Cloud Registration Revisited (2017, methodology): “A photometric objective EI is formulated in terms of squared differences of intensities [35, 25]:”
Nonparametric Calibration for Depth Sensors (2015, background): “Depth sensors like Microsoft Kinect or ASUS Xtion are major technological achievements and gave a new impulse to a wide range of 3D applications such as SLAM [1, 10, 4], superresolution mapping [7], object recognition and many others.”
Learning Depth from Monocular Videos Using Direct Methods (2017, methodology): “We instead propose the incorporation of a Direct Visual Odometry (DVO) [27] pose predictor into our framework as: (i) it requires no learnable parameters, (ii) establishes a direct relationship between the input dense depth map and the output pose prediction, and (iii) it is derived from the same im”