Assessment of the introduction of spatial stratification and manual training in automatic supervised image classification

Daniel Moraes, Pedro Benevides, Hugo Costa, Francisco Moreira, Mário Caetano

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

1 Citation (Scopus)
3 Downloads (Pure)

Abstract

The performance of supervised classification depends on the size and quality of the training data. Multiple studies have used reference datasets to extract training data automatically in an efficient way. However, automatic extraction might be inappropriate for some classes. Furthermore, classes can have distinct spectral characteristics across large areas. Thus, dividing the study area into subregions can be beneficial. This study proposes to assess the impact of the introduction of spatial stratification and manually collected training data on classification performance. Two classifications were conducted with the Random Forest classifier and multi-temporal Sentinel-2 data. The classifications’ performance was evaluated by accuracy metrics and visual inspection of the maps. The results indicate that introducing spatial stratification and manual training yielded a higher overall accuracy (66.7%) when compared to the accuracy of a benchmark classification (60.2%) conducted without stratification and with training data collected exclusively by automatic methods. Visual inspection of the maps also revealed some advantages of the novel approach, namely constraining some land cover classes to be present only within specific strata, which avoids commission errors of the class to spread freely across the map. Most of the classification improvements were observed in subregions with specific landscapes and spectral patterns, although these strata represent a small fraction of the study area, which might have contributed to the small increase in accuracy.
Original languageEnglish
Title of host publicationEarth Resources and Environmental Remote Sensing/GIS Applications XII
EditorsKarsten Schulz, Ulrich Michel, Konstantinos G. Nikolakopoulos
Place of PublicationWashington
PublisherSPIE-International Society for Optical Engineering
Chapter1186311
Number of pages8
Volume11863
ISBN (Electronic)9781510645714
ISBN (Print)9781510645707
DOIs
Publication statusPublished - 12 Sep 2021
EventEarth Resources and Environmental Remote Sensing/GIS Applications XII - Online Only, Spain
Duration: 13 Sep 202118 Sep 2021

Publication series

NamePROCEEDINGS OF SPIE
PublisherSPIE
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X

Conference

ConferenceEarth Resources and Environmental Remote Sensing/GIS Applications XII
Period13/09/2118/09/21

Keywords

  • Supervised classification
  • Random Forest
  • Spatial Stratification
  • Sentinel-2

Fingerprint

Dive into the research topics of 'Assessment of the introduction of spatial stratification and manual training in automatic supervised image classification'. Together they form a unique fingerprint.

Cite this