Authors: Žunić, Joviša 
Acketa, Dragan
Title: Least squares fitting of digital polynomial segments
Journal: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume: 1176
First page: 17
Last page: 23
Conference: 6th International Workshop on Discrete Geometry for Computer Imagery, DGCI 1996; Lyon; France; 13 November 1996 through 15 November 1996
Issue Date: 1-Jan-1996
ISBN: 978-3-540-62005-1
ISSN: 0302-9743
DOI: 10.1007/3-540-62005-2_2
Abstract: 
It is proved that digital polynomial segments and their least squares polynomial fits are in one-to-one correspondence. This enables an efficient representation of digital polynomial segments by n+3 parameters, under the condition that an upper bound, say n, for the degrees of the digitized polynomials is assumed. One of such representations is (x 1, m, an, an−1,…, a 0), where x 1 and m are the x-coordinate of the left endpoint and the number of digital points, respectively, while a n, a n−1,..., a 0 are the coefficients of the least squares polynomial fit Y=a nXn+an− 1Xn−1+ ...+a0, for a given digital polynomial segment.
Keywords: Coding | Computer vision | Digital polynomial segment | Image processing | Least squares fitting
Publisher: Springer Link

Show full item record

SCOPUSTM   
Citations

1
checked on Nov 19, 2024

Page view(s)

9
checked on Nov 19, 2024

Google ScholarTM

Check

Altmetric

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.