Abstract
This chapter proposes a new approach for learning invariant region descriptor operators through genetic programming. The approach relies on the synthesis of mathematical expressions that extract information derived from local image patches. These local features have been previously designed by human experts using traditional representations that have a clear and, preferably, mathematically wellfounded definition. We propose in this chapter that the mathematical principles that are used in the description of such local features could be evolved with the genetic programming paradigm to find a set of features that exhibit an optimal local description. Experimental results confirm the validity of our approach using a widely accepted testbed that is used for testing local descriptor algorithms. Furthermore, we provide results that illustrate the performance of our improved SIFT algorithm using an object recognition application for indoor and outdoor scenarios.
āThe theory of evolution by natural selection is an ecological theory ā founded on ecological observation by perhaps the greatest of all ecologists. It has been adopted by and brought up by the science of genetics, and ecologists, being modest people, are apt to forget their distinguished parenthood.ā
ā John L. Harper
āGoals are the fuel in the furnace of achievement.ā
ā Brian Tracy
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
Ā© 2016 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Olague, G. (2016). Evolutionary Synthesis of Feature Descriptor Operators with Genetic Programming. In: Evolutionary Computer Vision. Natural Computing Series. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-662-43693-6_9
Download citation
DOI: https://doi.org/10.1007/978-3-662-43693-6_9
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-662-43692-9
Online ISBN: 978-3-662-43693-6
eBook Packages: Computer ScienceComputer Science (R0)