Electrical Engineering
      and Computer Sciences

Electrical Engineering and Computer Sciences

COLLEGE OF ENGINEERING

UC Berkeley

Improving the Accuracy of Differential-Based Optical Flow Algorithms

Roberto Manduchi

EECS Department
University of California, Berkeley
Technical Report No. UCB/CSD-93-776
November 1993

http://www.eecs.berkeley.edu/Pubs/TechRpts/1993/CSD-93-776.pdf

This work deals with the design of the low-level stages of any different differential-based optical flow algorithm. The problem is the accurate estimation of the spatio-temporal derivatives of the moving image, to be used in the solution of the gradient constraint equation. We claim that the "traditional" techniques for the partial derivatives estimation can be improved by i) exploiting all the information provided by both fields within a frame in the case of interlaced scanning systems, ii) adopting prolate spheroidal filters instead of gaussian filters in order to get rid of noise and aliasing, and iii) using larger size differentiators, designed by a weighted least squares technique.


BibTeX citation:

@techreport{Manduchi:CSD-93-776,
    Author = {Manduchi, Roberto},
    Title = {Improving the Accuracy of Differential-Based Optical Flow Algorithms},
    Institution = {EECS Department, University of California, Berkeley},
    Year = {1993},
    Month = {Nov},
    URL = {http://www.eecs.berkeley.edu/Pubs/TechRpts/1993/6293.html},
    Number = {UCB/CSD-93-776},
    Abstract = {This work deals with the design of the low-level stages of any different differential-based optical flow algorithm. The problem is the accurate estimation of the spatio-temporal derivatives of the moving image, to be used in the solution of the gradient constraint equation. We claim that the "traditional" techniques for the partial derivatives estimation can be improved by i) exploiting all the information provided by both fields within a frame in the case of interlaced scanning systems, ii) adopting prolate spheroidal filters instead of gaussian filters in order to get rid of noise and aliasing, and iii) using larger size differentiators, designed by a weighted least squares technique.}
}

EndNote citation:

%0 Report
%A Manduchi, Roberto
%T Improving the Accuracy of Differential-Based Optical Flow Algorithms
%I EECS Department, University of California, Berkeley
%D 1993
%@ UCB/CSD-93-776
%U http://www.eecs.berkeley.edu/Pubs/TechRpts/1993/6293.html
%F Manduchi:CSD-93-776