Title:
|
FITNESS-AWARE OPERATORS FOR EVOLUTIONARY POLYGONAL APPROXIMATION |
Author(s):
|
José L. Guerrero, Antonio Berlanga, José M. Molina |
ISBN:
|
978-989-8533-14-2 |
Editors:
|
Hans Weghorn and Pedro Isaías |
Year:
|
2012 |
Edition:
|
Single |
Keywords:
|
Polygonal approximation, segmentation, genetic algorithm, chromosome representation, transformation operators |
Type:
|
Full Paper |
First Page:
|
283 |
Last Page:
|
290 |
Language:
|
English |
Cover:
|
|
Full Contents:
|
click to dowload
|
Paper Abstract:
|
Polygonal approximation is the decomposition of a given curve into a set of segments minimizing the representation error and/or the number of such segments used. This domain has been approached with a wide variety of methods, including evolutionary approaches based on genetic algorithms, according to the codification of a chromosome determining whether each of the individual points in the curve is considered dominant and computing at each generation the fitness values according to that chromosome. This work introduces a novel chromosome representation and a set of related operators in order to codify partial fitness information into the chromosome such that the transformation operators only apply partial fitness updates over the chromosome information to update the fitness values of the individuals. The proposal is tested over a set of traditional figures (chromosome, semicircle and leaf), along with modified versions of them according to an introduced method to scale the problems difficulty. Results show that the introduced method is able to improve the performance significantly, reducing the computational time down to a 40% of the one obtained by previous works without statistically significant degradation over the quality of the obtained results. |
|
|
|
|