FN Archimer Export Format
PT J
TI Fusion of textural statistics using a similarity measure: application to texture recognition and segmentation
BT 
AF KAROUI, I
   FABLET, Ronan
   BOUCHER, J
   PIECZYNSKI, W
   AUGUSTIN, Jean-Marie
AS 1:1;2:2;3:1;4:3;5:2;
FF 1:;2:PDG-DOP-DCB-STH-LASAA;3:;4:;5:PDG-DOP-DCB-NSE-AS;
C1 Telecom Bretagne, Inst Telecom, Dept SC, FRE CNRS Lab, Brest, France.
   Ifremer, Ctr Brest, F-29280 Plouzane, France.
   Inst Natl Telecommun, Dept CITI, CNRS UMR 5157, F-91011 Evry, France.
C2 TELECOM BRETAGNE, FRANCE
   IFREMER, FRANCE
   INST NATL TELECOMMUN, FRANCE
   UEB, FRANCE
SI BREST
SE PDG-DOP-DCB-STH-LASAA
   PDG-DOP-DCB-NSE-AS
IN WOS Ifremer jusqu'en 2018
   copubli-france
   copubli-univ-france
IF 1.367
TC 15
UR https://archimer.ifremer.fr/doc/2008/publication-4546.pdf
LA English
DT Article
DE ;MRF based texture segmentation;Texture recognition;Feature fusion and selection;Non parametric feature statistics
AB Features computed as statistics (e.g. histograms) of local filter responses have been reported as the most powerful descriptors for texture classification and segmentation. The selection of the filter banks remains however a crucial issue, as well as determining a relevant combination of these descriptors. To cope with selection and fusion issues, we propose a novel approach relying on the definition of the texture-based similarity measure as a weighted sum of the Kullback-Leibler measures between empirical feature statistics. Within a supervised framework, the weighting factors are estimated according to the maximization of a margin-based criterion. This weighting scheme can also be considered as a filter selection method: texture filter response distributions are ranked according to the associated weighting factors so that the problem of selecting a subset of filters reduces to picking the first features only. An application of this similarity measure to texture recognition is reported. We also investigate its use for texture segmentation within a Bayesian Markov Random Field (MRF)-based framework. Experiments carried out on Brodatz textures and sonar images show that the proposed weighting method improves the classification and the segmentation rates while relying on a parsimonious texture representation.
PY 2008
PD SEP
SO Pattern Analysis & Applications
SN 1433-7541
PU Springer
VL 11
IS 3-4
UT 000258579900016
BP 425
EP 434
DI 10.1007/s10044-008-0108-z
ID 4546
ER

EF