Machine Learning for High-Speed Corner Detection

By Edward Rosten, Tom Drummond. European Conference on Computer Vision, 2006

5,055 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/e0408181bccb7e3754dd5e6785ec47d8beb8b6bd

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

FREAK: Fast Retina Keypoint (2012, background): “Rostenand and Drummond proposed in [25] the FAST criterion for corner detection, improved by Mair et al . in [18] with their AGAST detector.”
Evaluation of local detectors and descriptors for fast feature matching (2012, background): “The largest number of keypoints are extracted by multiscale FAST detector and the least number of regions is provided by MSER.”
Large-Scale Visual Odometry for Rough Terrain (2007, methodology): “They also describe IMU measurements of gravity normal and yaw angle changes: gj = hg(Cj) (5) ∆ψj−1,j = h∆ψ(Cj−1, Cj) (6) 1 For 512x384 images: FAST 8ms, Harris 9ms, CenSurE 15ms, SIFT 138ms.”
BRISK: Binary Robust invariant scalable keypoints (2011, background): “The more recent FAST criterion [14] for keypoint detection has become increasingly popular in state-of-the-art methods with hard real-time constraints, with AGAST [10] extending this work for improved performance.”
Bags of Binary Words for Fast Place Recognition in Image Sequences (2012, methodology): “To overcome this problem, in this work we use FAST keypoints [10] and the state-of-the-art BRIEF descriptors [9].”
Advanced closed loop visual navigation for micro aerial vehicles (2014, methodology): “14a: While tasks like feature detection – FAST is one of the fastest performing feature detectors (Rosten and Drummond, 2006) – were already optimized and independent of the map-size, run-time increases for the process of re-projection of landmarks with growing map size.”
EKLT: Asynchronous Photometric Feature Tracking Using Events and Frames (2018, methodology): “Then, a standard feature detector and tracker (FAST [25] and KLT [35], respectively) were used on these event images to track the feature’s location.”
An Overview to Visual Odometry and Visual SLAM: Applications to Mobile Robotics (2015, methodology): “Features From Accelerated Segment Test (FAST) [96] is a corner detector based on the SUSAN method [107] in which a circle with a circumference of 16 pixels is placed around the center pixel.”
OpenVSLAM: A Versatile Visual SLAM Framework (2019, methodology): “In this approach, the FAST algorithm [12, 13] is used for keypoint detection.”
KinectFaceDB: A Kinect Database for Face Recognition (2014, background): “D map of indoor scene/object by camera tracking using sparse features (e.g., SIFT [33] and FAST [78]) and optimize the 3-”
BRIEF: Computing a Local Binary Descriptor Very Fast (2012, methodology): “In the case of BRIEF, any fast detector such as CenSurE [34] or FAST [35], [36] can be used.”
Omnidirectional 3D reconstruction in augmented Manhattan worlds (2014, methodology): “We employ the FAST corner detector [38] in combination with the BRIEF descriptor [39], which empirically led to the best results for our images which suffer from blur at the image boundaries and incidental noise due to small scratches in the mirror surface.”
A review of 3D reconstruction techniques in civil engineering and their applications (2018, background): “FAST [36] is a corner detector faster than Harris corner detector, and it identifies a pixel as a corner if the number of pixels in the circle consisting of sixteen pixels around the pixel meets the threshold.”
Deep Learning vs. Traditional Computer Vision (2019, methodology): “…CV tasks: • Scale Invariant Feature Transform (SIFT) [14] • Speeded Up Robust Features (SURF) [15] • Features from Accelerated Segment Test (FAST) [16] • Hough transforms [17] • Geometric hashing [18] Feature descriptors such as SIFT and SURF are generally combined with traditional machine…”