Duality of reconstruction and positioning from projective views
Published 19 November 2002
Sten Carlsson
Citations60
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
The problem of computing camera positions from m points in n views can be solved with the same algorithm as the problem of directly reconstructing n+4 points in m-4 views with similar results.
Abstract
Given multiple image data from a set of points in 3D, there are two fundamental questions that can be addressed: (1) What is the structure of the set of points in 3D? (2) What are the positions of the cameras relative to the points? In this paper, we show that for projective views and with structure- and position-defined modulo linear transformations, these problems are are dual in the sense that their solution arises from constraint equations where space point and camera positions occur in a reciprocal way. The problem of computing camera positions from m points in n views can be solved with the same algorithm as the problem of directly reconstructing n+4 points in m-4 views. This unifies different approaches for projective reconstruction: methods based on external calibration and direct methods exploiting constraints that exist between space and image invariants.
Keywords
Computer ScienceEngineering
Proceedings of the Royal Society of London. Series B, Biological sciencesThe interpretation of a moving retinal image
1,427 Citations1980Hugh Christopher Longuet-Higgins, K. Prazdny
It is shown that from a monocular view of a rigid, textured, curved surface it is possible, in principle, to determine the gradient of the surface at any point, and the motion of the eye relative to it, from the velocity field of the changing retinal image, and its first and second spatial derivatives.
Lecture notes in computer scienceCamera self-calibration: Theory and experiments
928 Citations1992O. D. Faugeras, Q.-T. Luong +1 more
It is shown, using experiments with noisy data, that it is possible to calibrate a camera just by pointing it at the environment, selecting points of interest and then tracking them in the image as the camera moves.
Lecture notes in computer scienceWhat can be seen in three dimensions with an uncalibrated stereo rig?
833 Citations1992Olivier Faugeras
This paper addresses the problem of determining the kind of three- dimensional reconstructions that can be obtained from a binocular stereo rig for which no three-dimensional metric calibration data is available, and shows that even in this case some very rich non-metric reconstructions of the environment can nonetheless be obtained.
Lecture notes in computer scienceEstimation of relative camera positions for uncalibrated cameras
539 Citations1992Richard Hartley
A non-iterative algorithm is given for determining the focal lengths of the two cameras, as well as their relative placement, assuming all other internal camera parameters to be known.
IEEE Transactions on Pattern Analysis and Machine IntelligenceAlgebraic functions for recognition
265 Citations1995Amnon Shashua
The trilinearity result is shown to be of much practical use in visual recognition by alignment-yielding a direct reprojection method that cuts through the computations of camera transformation, scene structure and epipolar geometry.
Matching constraints and the joint image
193 Citations2002Bill Triggs
The geometry of multi image perspective projection and the matching constraints that this induces on image measurements are studied and their complex algebraic interdependency is captured by quadratic structural simplicity constraints on the Grassmannian.
Lecture notes in computer scienceCanonic representations for the geometries of multiple projective views
160 Citations1994Q.-T. Luong, Thierry Viéville
It is shown how a special decomposition of general projection matrices, called canonic, enables us to build geometric descriptions for a system of cameras which are invariant with respect to a given group of transformations.
Lecture notes in computer scienceTrilinearity in visual recognition by alignment
158 Citations1994Amnon Shashua
The trilinearity result is shown to be of much practical use in visual recognition by alignment — yielding a direct method superior to the conventional epipolar line intersection method.
Trilinearity of three perspective views and its associated tensor
133 Citations2002Amnon Shashua, Michael Werman
It is shown that a permutation of the the trilinear coefficients produces three homography matrices (projective transformations of planes) of three distinct intrinsic planes, respectively, which yields the result that 3D invariants are recovered directly-simply by appropriate arrangement of the tensor's coefficients.
Extracting projective structure from single perspective views of 3D point sets
124 Citations2002C. Rothwell, David Forsyth +2 more
It is proved by example that although the general statement is true, invariants do exist for structured three-dimensional point sets and they can be used to compute projective structure.
The International Journal of Robotics ResearchRelative 3D Reconstruction Using Multiple Uncalibrated Images
111 Citations1995Roger Mohr, Long Quan +1 more
It is shown how relative 3D reconstruction from point correspondences of multiple uncalibrated images can be achieved through reference points, and robustness of the method and reconstruction precision are discussed.
Projective depth: A geometric invariant for 3D reconstruction from two perspective/orthographic views and for visual recognition
69 Citations2002Amnon Shashua
It is shown that with the correspondences coming from four non-coplanar points in the scene and the corresponding epipoles, it is possible to define and reconstruct a projective invariant that can be used later to reconstruct the projective or affine structure of the scene or directly to generate novel views of thescene.
Lecture notes in computer scienceInvariants of 6 points from 3 uncalibrated images
61 Citations1994Long Quan
There are three projective invariants of a set of six points in general position in space that cannot be recovered from one image, however an invariant relationship does exist between space invariants and image invariants.
Reconstruction from image sequences by means of relative depths
59 Citations2002Anders Heyden
Philosophical Transactions of the Royal Society of London Series A Physical and Engineering SciencesThe projective geometry of ambiguous surfaces
52 Citations1990Stephen J. Maybank
Image and Vision ComputingDepth computations from polyhedral images
46 Citations1992Gunnar Sparr
A method is developed for the computation of depth maps, modulo scale, from one single image of a polyhedral scene with only affine shape properties of the scene and image used, hence no metrical information.
Affine and Projective Structure from Motion
45 Citations1992Sabine Demey, Andrew Zisserman +1 more
Lecture notes in computer scienceThe double algebra: An effective tool for computing invariants in computer vision
43 Citations1994Stefan Carlsson
This paper shows how to compute linear invariants of general configurations points and lines observed in two images and polyhedral configurations observed in one image without reconstructing individual points and Lines.
Lecture notes in computer scienceA common framework for kinetic depth, reconstruction and motion for deformable objects
39 Citations1994Gunnar Sparr
In this paper, problems related to depth, reconstruction and motion from a pair of projective images are studied under weak assumptions, based on a novel interpretation of a theorem in projective geometry of M. Chasles and the notion of “affine shape”.
Lecture notes in computer scienceA case against epipolar geometry
30 Citations1994Andrew Zisserman, Stephen J. Maybank
It is demonstrated configurations for which 3D projective invariants can be computed from perspective stereo pairs, but epipolar geometry (and full projective structure) cannot, and the connection with camera calibration is investigated.
Lecture notes in computer scienceHow to use the cross ratio to compute projective invariants from two images
12 Citations1994Patrick Gros
A survey and clarification of the different invariant calculation methods are detailed in this extented technical report on applications of invariant theory to computer vision problems.
Image and Vision ComputingRelative positioning from model indexing
8 Citations1994Stefan Carlsson
Inverting this relationship between camera position and viewpoint invariants can compute camera position relative to a model of the environment using indexing methods, independent of camera calibration.