Optimizing classification performance in an object-based very-high-resolution land use-land cover urban application

Stefanos Georganos, Tais Grippa, Sabine Vanhuysse, Moritz Lennert, Michal Shimoni, Eléonore Wolff

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

Abstract

This study evaluates the impact of three Feature Selection (FS) algorithms in an Object Based Image Analysis (OBIA) framework for Very-High-Resolution (VHR) Land Use-Land Cover (LULC) classification. The three selected FS algorithms, Correlation Based Selection (CFS), Mean Decrease in Accuracy (MDA) and Random Forest (RF) based Recursive Feature Elimination (RFE), were tested on Support Vector Machine (SVM), K-Nearest Neighbor, and Random Forest (RF) classifiers. The results demonstrate that the accuracy of SVM and KNN classifiers are the most sensitive to FS. The RF appeared to be more robust to high dimensionality, although a significant increase in accuracy was found by using the RFE method. In terms of classification accuracy, SVM performed the best using FS, followed by RF and KNN. Finally, only a small number of features is needed to achieve the highest performance using each classifier. This study emphasizes the benefits of rigorous FS for maximizing performance, as well as for minimizing model complexity and interpretation.

Original languageEnglish
Title of host publicationRemote Sensing Technologies and Applications in Urban Environments II
EditorsThilo Erbertseder, Ying Zhang, Nektarios Chrysoulakis
PublisherSociety of Photo-Optical Instrumentation Engineers
ISBN (Electronic)9781510613263
DOIs
Publication statusPublished - 2017
EventRemote Sensing Technologies and Applications in Urban Environments II 2017 - Warsaw, Poland
Duration: 11 Sept 201713 Sept 2017

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume10431
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X

Conference

ConferenceRemote Sensing Technologies and Applications in Urban Environments II 2017
Country/TerritoryPoland
CityWarsaw
Period11/09/1713/09/17

Keywords

  • feature selection
  • object based image analysis
  • random forest
  • supervised classification
  • support vector machine

Fingerprint

Dive into the research topics of 'Optimizing classification performance in an object-based very-high-resolution land use-land cover urban application'. Together they form a unique fingerprint.

Cite this