Geometric Invariance in Computer Vision

By Emanuele Trucco. 1995

Subfields: 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 Computer Vision and Pattern Recognition

Citations in the corpus

The dual-bootstrap iterative closest point algorithm with application to retinal image registration (2003, background): “Matches between two landmarks, one in each image, or between pairs of landmarks in each image are generated by computing and comparing invariants [8], [55].”
Learning the semantics of object–action relations by observation (2011, methodology): “Geometry-based approaches use a geometric description of a 3D object and match its projected shape against the image of the object (Mundy and Zisserman 1992; Mundy 2006).”
Active Contours (2000, methodology): “The most general linear approximation to perspective is known variously as paraperspective (Aloimonos, 1990) or weak perspective (Mundy and Zisserman, 1992) and can be particularly effective if separate approximations are constructed for different neighbourhoods of an image (Lawn and Cipolla, 1994).”
A search engine for 3D models (2003, background): “Other shape representations, such as Extended Gaussian Images [38], Spherical Attribute Images [27, 28], moments [68, 84], and wavelets [34], require a priori registration into a canonical coordinate system, which is difficult to achieve robustly.”
Actions sketch: a novel action representation (2005, methodology): “Using the affine camera model, the the world coordinates X are mapped to image coordinates x by xL = AX+TL and xR = CX+TR, where subscripts denote the left and right cameras [7].”
A statistical approach to 3d object detection applied to faces and cars (2000, background): “At the other extreme, others use a minimalist feature representation and emphasize geometric relationships [32][41][45][46][47].”
3D Object Modeling and Recognition Using Local Affine-Invariant Image Descriptors and Multi-View Spatial Constraints (2006, unknown): “…of generalized cylinders (Ponce et al., 1989; Liu et al., 1993)—admit invariants, general 3D shapes do not (Burns et al., 1993), which is the main reason why invariants have fallen out of favor after an intense ﬂurry of activity in the early 1990s (Mundy and Zisserman, 1992; Mundy et al., 1994).”
View-Invariant Representation and Recognition of Actions (2002, background): “affine projection model, which assumes that the depth of the 3-D trajectory of the action is small compared to the viewing distance (Mundy and Zisserman, 1992).”
Viewpoint Invariant Texture Description Using Fractal Analysis (2009, unknown): “A number of projective invariants (Mundy and Zisserman 1992) have been found, which are defined on sets of points and lines and planar curves (Weiss 1993), and they have been used for object recognition and calibration.”
Shape Quantization and Recognition with Randomized Trees (1997, background): “In general, the mathematical level of this work is advanced, borrowing ideas from projective, algebraic, and differential geometry (Mundy & Zisserman, 1992).”
Bayesian Model Estimation and Selection for Epipolar Geometry and Generic Manifold Fitting (2002, methodology): “…other approaches to vision such as the mechanical approach set out in Blake and Zisserman (1987) purely geometric approach (Faugeras and Papadopoulo, 1993; Mundy and Zisserman, 1992), or some combination of ortho-dox statistics and geometry (Hartley and Zisserman, Torr 2000; Kanatani, 1992, 1996).”
Shape distributions (2002, methodology): “The simplest approach of this type is to evaluate distances between feature vectors [Duda et al. 2001] in a multidimensional space where the axes encode global geometric properties, such as circularity, eccentricity, or algebraic moments [Prokop and Reeves 1992; Taubin and Cooper 1992].”