Invariant Features from Interest Point Groups

By Matthew A. Brown, David G. Lowe. British Machine Vision Conference, 2002

776 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/45cb5262b1fb9f149e8f4171e0b1e52d748e062d

How researchers see this paper

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

In Computer Vision and Pattern Recognition

Citations in the corpus

Class-Based Feature Matching Across Unrestricted Transformations (2008, background): “Several popular schemes [2], [3], [4], [15] work with socalled affine invariants.”
Multi-image matching using multi-scale oriented patches (2005, background): “Other researchers have developed feature descriptors that are invariant under affine transformations [7, 8, 9].”
Linear Time Maximally Stable Extremal Regions (2008, background): “Extraction of invariant regions has recently been the focus of intense study [1,2,3,4,5,6,7,8,9,10,11,12], supporting a wide variety of applications such as recognition, image retrieval, mosaicing, 3D reconstruction, tracking, robot navigation and more.”
Distinctive Image Features from Scale-Invariant Keypoints (2004, methodology): “However, recently Brown has developed a method (Brown and Lowe, 2002) for fitting a 3D quadratic function to the local sample points to determine the interpolated location of the maximum, and his experiments showed that this provides a substantial improvement to matching and stability.”
Recognising panoramas (2003, background): “Recently there has been great progress in the use of invariant features [12, 2, 4] for object recognition and matching.”
Distinctive Image Features from Scale-Invariant Keypoints Abstract by Matthijs Dorst Based on the paper by (2011, methodology): “Brown developed a method to determine the interpolated location of the maximum using a quadratic order Taylor expansion of the shifted scale-space function D(x, y, σ) [2] with D evaluated at the sample point.”
Distinctive Image Features from Scale-Invariant Keypoints (2004, background): “Carneiro and Jepson (2002) describe phase-based local features that represent the phas rather than the magnitude of local spatial frequencies, which is likely to provide improved invariance to illumination.”
Machine Learning for High-Speed Corner Detection (2006, background): “Recently, scale invariance has been extended to consider features which are invariant to affine transformations [14,16,17].”
Faster and Better: A Machine Learning Approach to Corner Detection (2008, neutral): “October 14, 2008 DRAFT 6”
CSIFT: A SIFT Descriptor with Color Invariant Characteristics (2006, methodology): “In order to localize the interest points, subpixel/sub-scale approximation is performed for the obtained extrema to achieve the maximum geometrical stability of the detected interest points [4].”
Notes on the OpenSURF Library (2009, methodology): “This is done by fitting a 3D quadratic as proposed by Brown [3].”
Evaluation of Interest Point Detectors and Feature Descriptors for Visual Tracking (2011, methodology): “The interest point locations are then refined to subpixel accuracy by fitting a parabola to the sample point and its immediate neighbors (Brown and Lowe 2002).”
KAZE Features (2012, methodology): “Finally, the position of the keypoint is estimated with sub-pixel accuracy using the method proposed in [20].”
SURF: Speeded Up Robust Features (2006, methodology): “The maxima of the determinant of the Hessian matrix are then interpolated in scale and image space with the method proposed by Brown et al. [27].”
Speeded-Up Robust Features (SURF) (2008, methodology): “The maxima of the determinant of the Hessian matrix are then interpolated in scale and image space with the method proposed by Brown et al. [5].”