Visual odometry for ground vehicle applications

By David Nistér, Oleg Naroditsky, J. Bergen. J. Field Robotics, 2006

734 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/c8965cc5c62a245593dbc679aebdf3338bb945fc

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

Visual-lidar odometry and mapping: low-drift, robust, and fast (2015, background): “With stereo cameras [3], [4], the baseline provides a reference to help determine scale of the motion.”
Detailed Real-Time Urban 3D Reconstruction from Video (2007, methodology): “While this approach works well for relatively short image sequences, it is diﬃcult to reliably estimate the camera motion over long distances using video only, especially when using monocular structure from motion [43].”
Active Neural Localization (2018, methodology): “, 1990) geometry-based visual odometry methods (Nistér et al., 2006) and most recently, learning-based visual odometry methods which learn to predict motion between consecutive frames using recurrent convolutional neural networks (Clark et al.”
Two years of Visual Odometry on the Mars Exploration Rovers (2007, methodology): “This implementation contains two motion estimation schemes: the stereo scheme, which is an iterative pose refinement scheme, and the monocular scheme, which was based on the framework of 5-point algorithm (Nistér, 2004) and an outlier rejection scheme (Nistér, 2005) and good results have been repo”
Practice makes perfect? Managing and leveraging visual experiences for lifelong navigation (2012, background): “3 illustrates the output of the VO, which may be saved as an experience.”
A Multi-State Constraint Kalman Filter for Vision-aided Inertial Navigation (2007, methodology): “One algorithm that, similarly to the method proposed in this paper, directly uses the landmark measurements for imposing constraints between multiple camera poses is presented in [16].”
Visual-inertial navigation, mapping and localization: A scalable real-time causal approach (2011, background): “5 N/A Stereo vision only (Nister et al., 2006) 1.”
Simultaneous Localization and Mapping: A Survey of Current Trends in Autonomous Driving (2017, neutral): “Nist´er et al. [71] proposed a selection method based on a preemptive RANSAC [72].”
Incremental smoothing and mapping (2008, background): “While accuracy of the presented system can further be improved by applying fixedlag smoothing [111], by using key frames to prevent drift [89], and by using landmarks to provide longer range constraints [138], in our case visual SLAM takes care of the long range consistency.”
1 year, 1000 km: The Oxford RobotCar dataset (2017, methodology): “…datasets focus primarily on the development of algorithmic competencies for autonomous driving: motion estimation as in Geiger et al. (2011) and Nistér et al. (2006), stereo reconstruction as in Geiger et al. (2010) and Hirschmuller (2005), pedestrian and vehicle detection as in Benenson et…”
The Transferability Approach: Crossing the Reality Gap in Evolutionary Robotics (2013, methodology): “Such methods only rely on a single camera [10], sometimes coupled with classic inertial sensors [45], and appear to be robust and simple to implement.”
An Introduction to Inertial and Visual Sensing (2007, methodology): “Alternatively corner features can be found using the Shi–Tomasi (Tomasi and Shi 1994) or Harris (Harris and Stephens 1988) detectors in individual frames and then correspondence is determined, generally involving an exhaustive comparison of all features between consecutive image frames (Ma et al. 20”
Real-time monocular visual odometry for on-road vehicles with 1-point RANSAC (2009, background): “In the first category are the works of [2]–[5].”
SKYLINE2GPS: Localization in urban canyons using omni-skylines (2010, background): “can be classified into ones that use a motion model [2], [4] and the approaches free of motion models [20], [26].”
Inverse Depth Parametrization for Monocular SLAM (2008, neutral): “Recently, work in the spirit of these methods, but with “sliding window” processing and refinement rather than global optimization, has produced impressive real-time “visual odometry” results when applied to stereo sequences in [21] and for monocular sequences in [20].”
Real-time stereo visual odometry for autonomous ground vehicles (2008, background): “I. INTRODUCTION Localization is a key capability for autonomous ground vehicles, and is typically performed using a combination of wheel odometry (from joint encoders) and inertial sensing (gyroscopes and accelerometers).”
A novel parametrization of the perspective-three-point problem for a direct computation of absolute camera position and orientation (2011, background): “Applications such as feature-point-based camera tracking [29] [21], structure from motion, and visual odometry [26] require dealing with hundreds or even thousands of noisy feature points and outliers in real-time, which requires computationally efficient methods.”
Made to measure: Bespoke landmarks for 24-hour, all-weather localisation with a camera (2016, methodology): “The pose optimisation is performed in a similar manner to [28][29][30].”