Polsar scene classification based on fast approximate neareast neighbours search

Author

Chen, Lijun ; Yang, Wen ; Liu, Ying ; Sun, Hong

Author_Institution

Sch. of Electron. Inf., Wuhan Univ., Wuhan, China

fYear

2011

fDate

24-29 July 2011

Firstpage

2137

Lastpage

2140

Abstract

In this paper, we attempt to solve the efficiency problem of PolSAR scene classification with non-parametric classifier. We employ the tree-structure based search strategy to perform fast approximate nearest neighbour search by introducing the multiple randomized kd-tree and hierarchical kmeans-tree into ONBNN classifier. The experimental results on RadarSat-2 PolSAR dataset demonstrate that our method can greatly improve the efficiency with a guaranteed accuracy.

Keywords

data analysis; geophysical image processing; geophysical techniques; image classification; synthetic aperture radar; ONBNN classification system; PoLSAR scene classification method; RadarSat-2 PolSAR dataset; fast approximate nearest neighbour search strategy; hierarchical kmeans-tree method; nonparametric classification analysis; randomized kd-tree method; tree-structure based search strategy; Accuracy; Buildings; Classification algorithms; Kernel; Search problems; Support vector machines; Training; Scene classification; kd-tree; kmeans-tree; nearest neighbor search;

fLanguage

English

Publisher

ieee

Conference_Titel

Geoscience and Remote Sensing Symposium (IGARSS), 2011 IEEE International

Conference_Location

Vancouver, BC

ISSN

2153-6996

Print_ISBN

978-1-4577-1003-2

Type

conf

DOI

10.1109/IGARSS.2011.6049588

Filename

6049588