We have developed an algorithm for segmenting fully polarimetric single look TerraSAR-X, multilook SIR-C and 7 band Landsat 5 imagery using neural nets. The algorithm uses a feedforward neural net with one hidden layer to segment different surface classes. The weights are refined through an iterative filtering process characteristic of a relaxation process. Features selected from studies of fully polarimetric complex single look TerraSAR-X data and multilook SIR-C data are used as input to the net. The seven bands from Landsat 5 data are used as input for the Landsat neural net. The Cloude-Pottier incoherent decomposition is used to investigate the physical basis of the polarimetric SAR data segmentation. The segmentation of a SIR-C ocean surface scene into four classes is presented. This segmentation algorithm could be a very useful tool for investigating complex polarimetric SAR phenomena.

Date Published: 21 May 2015
PDF: 16 pages
Proc. SPIE 9461, Radar Sensor Technology XIX; and Active and Passive Signatures VI, 94610P (21 May 2015); doi: 10.1117/12.2175837

Published in SPIE Proceedings Vol. 9461:
Radar Sensor Technology XIX; and Active and Passive Signatures VI
G. Charmaine Gilbreath; Kenneth I. Ranney; Armin Doerry; Chadwick Todd Hawley, Editor(s)