Show simple item record

dc.contributor.authorLiang, Jie
dc.contributor.authorZhou, Jun
dc.contributor.authorQian, Yuntao
dc.contributor.authorWen, Lian
dc.contributor.authorBai, Xiao
dc.contributor.authorGao, Yongsheng
dc.date.accessioned2017-08-04T12:30:53Z
dc.date.available2017-08-04T12:30:53Z
dc.date.issued2017
dc.identifier.issn0196-2892
dc.identifier.doi10.1109/TGRS.2016.2616489
dc.identifier.urihttp://hdl.handle.net/10072/340634
dc.description.abstractAbstract: Spectral-spatial processing has been increasingly explored in remote sensing hyperspectral image classification. While extensive studies have focused on developing methods to improve the classification accuracy, experimental setting and design for method evaluation have drawn little attention. In the scope of supervised classification, we find that traditional experimental designs for spectral processing are often improperly used in the spectral-spatial processing context, leading to unfair or biased performance evaluation. This is especially the case when training and testing samples are randomly drawn from the same image - a practice that has been commonly adopted in the experiments. Under such setting, the dependence caused by overlap between the training and testing samples may be artificially enhanced by some spatial information processing methods, such as spatial filtering and morphological operation. Such enhancement of dependence in return amplifies the classification accuracy, leading to an improper evaluation of spectral-spatial classification techniques. Therefore, the widely adopted pixel-based random sampling strategy is not always suitable to evaluate spectral-spatial classification algorithms, because it is difficult to determine whether the improvement of classification accuracy is caused by incorporating spatial information into classifier or by increasing the overlap between training and testing samples. To tackle this problem, we propose a novel controlled random sampling strategy for spectral-spatial methods. It can greatly reduce the overlap between training and testing samples and provides more objective and accurate evaluation.
dc.description.peerreviewedYes
dc.languageEnglish
dc.language.isoeng
dc.publisherInstitute of Electircal and Electronics Engineers
dc.relation.ispartofpagefrom862
dc.relation.ispartofpageto880
dc.relation.ispartofissue2
dc.relation.ispartofjournalIEEE Transactions on Geoscience and Remote Sensing
dc.relation.ispartofvolume55
dc.subject.fieldofresearchGeophysics
dc.subject.fieldofresearchGeomatic engineering
dc.subject.fieldofresearchcode3706
dc.subject.fieldofresearchcode4013
dc.titleOn the Sampling Strategy for Evaluation of Spectral-Spatial Methods in Hyperspectral Image Classification
dc.typeJournal article
dc.type.descriptionC1 - Articles
dc.type.codeC - Journal Articles
dc.description.versionAccepted Manuscript (AM)
gro.rights.copyright© 2016 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
gro.hasfulltextFull Text
gro.griffith.authorWen, Larry
gro.griffith.authorGao, Yongsheng
gro.griffith.authorZhou, Jun


Files in this item

This item appears in the following Collection(s)

  • Journal articles
    Contains articles published by Griffith authors in scholarly journals.

Show simple item record