Lucas, B.D.[Bruce D.],
Generalized Image Matching by the Method of Differences,
Ph.D.Thesis (CS), 1985.
BibRef
8500
CMU-CS-TR-85-160, CMU CS Dept.
Optical Flow.
BibRef
Lucas, B.D., and
Kanade, T.,
Optical Navigation by the Method of Differences,
IJCAI85(981-984).
BibRef
8500
And:
DARPA84(272-281).
Use the intensity gradient in an iterative matching
scheme to find the correspondence between two images.
BibRef
Lucas, B.D., and
Kanade, T.,
An Iterative Image Registration Technique with an
Application to Stereo Vision,
DARPA81(121-130).
HTML Version.
BibRef
8100
And:
IJCAI81(674-679).
HTML Version.
Code, Registration.
WWW Version. Another version in Matlab.
WWW Version. Uses differences in intensity between the two images and the
local gradient of one image (both?) to compute the shift.
A registration problem, but very applicable to stereo.
For more generalization:
See also Shape and Motion from Image Streams: A Factorization Method Part 3 - Detection and Tracking of Point Features.
BibRef
Lucas, B.D.,
Automatic Generation of Depth maps from Stereo Images,
DARPA82(309-314).
Basically assumes L(x,y)=R(x+h(x,y),y) and find h that minimizes
the error in the mapping. Find local errors in a neighborhood of
each point (use constraint of "real world" - smoothly varying
disparities). Apply different smoothing windows and h can be
computed at each of these (using the old value to limit the new
possibilities). The computations are essentially smoothing
operations. Also a Lucas paper at the Vancouver IJCAI.
BibRef
8200
Li, Z.N.,
Hu, G.Z.,
Analysis of Disparity Gradient-Based Cooperative Stereo,
IP(5), No. 11, November 1996, pp. 1493-1506.
IEEE DOI Link
9611
BibRef
Baker, S.[Simon],
Matthews, I.[Iain],
Lucas-Kanade 20 Years On: A Unifying Framework,
IJCV(56), No. 3, February-March 2004, pp. 221-255.
WWW Version.
0402
BibRef
And:
Lucas-Kanade 20 Years On: A Unifying Framework: Part 1,
CMU-RI-TR-02-16, July 2002.
WWW Version.
0211
BibRef
And:
Lucas-Kanade 20 Years On,
CMU-RI2006, Project Description.
HTML Version.
Code, Tracking. Matlab code is available.
See also Generalized Image Matching by the Method of Differences.
See also Iterative Image Registration Technique with an Application to Stereo Vision, An.
BibRef
Baker, S.,
Gross, R.,
Matthews, I.,
Ishikawa, T.,
Lucas-Kanade 20 Years On: A Unifying Framework: Part 2,
CMU-RI-TR-03-01, February, 2003.
HTML Version.
0306
BibRef
Baker, S.,
Gross, R.,
Matthews, I.,
Lucas-Kanade 20 Years On: A Unifying Framework: Part 3,
CMU-RI-TR-03-35, November, 2003.
HTML Version.
0501
BibRef
Baker, S.,
Gross, R.,
Matthews, I.,
Lucas-Kanade 20 Years On: A Unifying Framework: Part 4,
CMU-RI-TR-04-14, February, 2004.
HTML Version.
0501
BibRef
Baker, S.,
Patil, R.,
Cheung, K.M.,
Matthews, I.,
Lucas-Kanade 20 Years On: Part 5,
CMU-RI-TR-04-64, November, 2004.
HTML Version.
0501
BibRef
Baker, S.,
Datta, A., and
Kanade, T.,
Parameterizing Homographies,
CMU-RI-TR-06-11, March, 2006.
HTML Version.
BibRef
0603
Zhang, H.S.[Hong-Sheng],
Negahdaripour, S.[Shahriar],
BC&GC-Based Dense Stereo By Belief Propagation,
CVS06(14).
IEEE DOI Link
0602
Brightness constancy and gradient constancy --
apply optical flow ideas to stereo.
See also Revised Definition of Optical Flow: Integration of Radiometric and Geometric Cues for Dynamic Scene Analysis.
BibRef
Twardowski, T.[Tomasz],
Cyganek, B.[Boguslaw],
Borgosz, J.[Jan],
Gradient Based Dense Stereo Matching,
ICIAR04(I: 721-728).
WWW Version.
0409
BibRef
Tardon-Garcia, L.J.[Lorenzo-Jose],
Portillo-Garcia, J.[Javier],
Alberola-Lopez, C.[Carlos],
Markov Random Fields and the Disparity Gradient Constraint Applied to Stereo
Correspondence,
ICIP99(III:901-905).
IEEE Abstract. IEEE Top Reference.
See also Hypothesis Testing for Coarse Region Estimation and Stable Point Determination Applied to Markovian Texture Segmentation.
BibRef
9900
Trucco, E.,
Roberto, V.,
Tinonin, S., and
Corbatto, M.,
SSD Disparity Estimation for Dynamic Stereo,
BMVC96(Motion-Based Reconstruction).
9608
Heriot-Watt University and University of Udine, Italy
BibRef
Chapter on Stereo: Three Dimensional Descriptions from Two or More Views, Binocular, Trinocular continues in
Line Segment Based Stereo Analysis .