Adaptive algorithm for parameterization of feature extraction techniques in remote sensing image processing

Edmore Chikohora, Attlee Gamundani, Teressa Chikohora

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This study presents a novel adaptive heuristic algorithm named 'the GenApp' that is based on the Gabor Filter (GF) to generate useful solutions to optimization of parameter selection strategies for Feature Extraction Techniques (FET) in Remote Sensing Image (RSI) processing. The GenApp was evaluated against the existing algorithms namely, the Modified Gabor Filter (MGF) and the GF for efficiency and fitness of value. The triangulation approach was applied to compare the two sets of results and the results obtained from the simulations evidenced an improved fitness value of the GenApp. Although the GenApp posted higher complexity values of 0.63 and 0.03 points above the existing algorithms, the lapse is a direct result of the initialization processes introduced to the GenApp that screens less fit prospective solutions from the initial population data, which can be traded-off with the GenApp's overall performance as compared to the existing algorithms.

Original languageEnglish
Title of host publication2018 IST-Africa Week Conference, IST-Africa 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781905824601
Publication statusPublished - 20 Jul 2018
Externally publishedYes
Event2018 IST-Africa Week Conference, IST-Africa 2018 - Gaborone, Botswana
Duration: 9 May 201811 May 2018

Publication series

Name2018 IST-Africa Week Conference, IST-Africa 2018

Conference

Conference2018 IST-Africa Week Conference, IST-Africa 2018
Country/TerritoryBotswana
CityGaborone
Period9/05/1811/05/18

Keywords

  • Dimensionality reduction
  • Feature extraction
  • Genetic algorithms
  • Region of Interest
  • Remote sensing image processing

Cite this