ORB: An efficient alternative to SIFT or SURF

By Ethan Rublee, Vincent Rabaud, Kurt Konolige, Gary R. Bradski. 2011

10,845 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

Delving into Egocentric Actions (2015, positive): “We model the camera motion by matching sparse interest points between successive frames using ORB [30].”
Deformable Convolutional Networks (2017, background): “Notable examples include scale invariant feature transform (SIFT) [38] and ORB [44] (O for orientation).”
Lifelong Graph Learning (2020, critical): “Although many hand-crafted feature descriptors such as SIFT [27] and ORB [33] have been proposed decades ago, their performance is still unsatisfied for large view point changes.”
Deformable ConvNets V2: More Deformable, Better Results (2018, background): “Prior to the deep learning era, notable works include scale-invariant feature transform (SIFT) [29], oriented FAST and rotated BRIEF (ORB) [33], and deformable part-based models (DPM) [11].”
Taskonomy: Disentangling Task Transfer Learning (2018, background): “Keypoint algorithmns usually consist of two parts, both a keypoint detector and some local patch descriptor which is invariant across multiple images [24, 4, 32].”
Unsupervised Learning of Edges (2015, methodology): “We first fit a homography matrix between consecutive frames using ORB descriptor matches [32] (which are fast to compute).”
RIFT: Multi-Modal Image Matching Based on Radiation-Variation Insensitive Feature Transform (2019, methodology): “The ORB (Oriented FAST and Rotated BRIEF) [27] algorithm uses the features from accelerated segment test (FAST) [28] to extract feature points and utilises the directional binary robust independent elementary features (BRIEF) [29] algorithm for feature description.”
An Overview to Visual Odometry and Visual SLAM: Applications to Mobile Robotics (2015, methodology): “[41] proposed an RGB-D-SLAM method that uses SIFT, SURF [10] and ORB [97] feature descriptors in place of FAST features.”
Recent developments in large-scale tie-point matching (2016, background): “Several variants exist such as SURF [4], ORB [33] or DAISY [33].”
ORB-SLAM2: An Open-Source SLAM System for Monocular, Stereo, and RGB-D Cameras (2016, methodology): “The system uses the same ORB features [17] for tracking, mapping, and place recognition tasks.”
Transfer learning for image classification using VGG19: Caltech-101 image data set (2021, methodology): “ORB algorithm is a local feature extraction algorithm that is presented by Rublee et al. (2011).”
D2-Net: A Trainable CNN for Joint Description and Detection of Local Features (2019, methodology): “The most common approach to sparse feature extraction – the detect-then-describe approach – ﬁrst performs feature detection [7, 19, 30, 32, 34] and then extracts a feature descriptor [7, 9, 25, 30, 45] from a patch centered around each keypoint.”
A review of multimodal image matching: Methods and applications (2021, methodology): “(3) Learnable feature Before deep learning, many detectors use classical learning (training a classifier) to identify more reliable and matchable features before matching, such as FAST [179], ORB [120], and others [145,180,181].”
Visual Place Recognition: A Survey (2016, methodology): “SIFT features as in [76], CenSurE features as in FrameSLAM [2], or ORB features [137] as in ORB-SLAM [136].”
LoFTR: Detector-Free Local Feature Matching with Transformers (2021, background): “SIFT [26] and ORB [35] are arguably the most successful hand-crafted local features and are widely adopted in many 3D computer vision tasks.”
PL-SLAM: A Stereo SLAM System Through the Combination of Points and Line Segments (2017, methodology): “In this work we use the well-known ORB method [23] due to its great performance for keypoint detection, and the binary nature of the descriptor it provides, which allows for a fast, eﬃcient keypoint matching.”
RTAB‐Map as an open‐source lidar and visual simultaneous localization and mapping library for large‐scale and long‐term online operation (2018, methodology): “Like visual odometry, covariance is computed using the MAD approach (Rusu & Cousins, 2011) between 3D point correspondences.”