DocumentCode :
1158949
Title :
Late-Season Rural Land-Cover Estimation With Polarimetric-SAR Intensity Pixel Blocks and \\sigma -Tree-Structured Near-Neighbor Classifiers
Author :
Barnes, Christopher F. ; Burki, Jehanzeb
Author_Institution :
Georgia Inst. of Technol., GA
Volume :
44
Issue :
9
fYear :
2006
Firstpage :
2384
Lastpage :
2392
Abstract :
Synthetic aperture radar (SAR) image classification for late-season rural land-cover estimation is investigated. A novel tree-structured nearest neighbor-like classifier is applied to polarimetric SAR intensity image pixel blocks. The novel tree structure, called a sigma-tree, is generated by an ordered summation of unweighted template refinements. Computation and memory costs of a sigma-tree classifier grow linearly. The reduced costs of sigma-tree classifiers are obtained with the tradeoff of a guarantee of nearest neighbor mappings. Causal-anticausal refinement-template design methods, combined with causal multiple-stage search engine structures, are shown to yield sequential search decisions that are acceptably near-neighbor mappings. The performance of a sigma-tree classifier is demonstrated for rural land-cover estimation with detected polarimetric C-band AirSAR pixel data. Experiments are conducted on various polarization/pixel block size combinations to evaluate the relative utility of spatial-only, polarimetric-only, and combined spatial/polarimetric classifier inputs
Keywords :
geophysical techniques; image classification; radar polarimetry; synthetic aperture radar; vegetation; C-band AirSAR pixel data; causal multiple-stage search engine structures; causal-anticausal refinement-template design methods; classifier inputs; computation costs; late-season rural land-cover estimation; memory costs; nearest neighbor mappings; polarimetric-SAR intensity pixel blocks; search decisions; sigma-tree-structured near-neighbor classifiers; synthetic aperture radar image classification; Classification tree analysis; Computational efficiency; Costs; Design methodology; Image classification; Nearest neighbor searches; Pixel; Search engines; Synthetic aperture radar; Tree data structures; Additive successive refinements; direct sum; residual vector quantization; successive approximation; synthetic aperture radar (SAR) land-use classification;
fLanguage :
English
Journal_Title :
Geoscience and Remote Sensing, IEEE Transactions on
Publisher :
ieee
ISSN :
0196-2892
Type :
jour
DOI :
10.1109/TGRS.2006.875449
Filename :
1677748
Link To Document :
بازگشت