Title :
Forest Spatial Structure Enhancing Non-Gaussian Texture in Airborne L-Band Polsar Images
Author_Institution :
Inst. of Space & Astronaut. Sci. (ISAS), Japan Aerosp. Exploration Agency (JAXA), Sagamihara
Abstract :
In this paper, results of texture analyses over coniferous forests using airborne L-band POLSAR data are shown. Interestingly, field experiments have disclosed existence of patch-like sparse areas in a forest division where low order parameters of the K-distribution are calculated in the texture analyses. In order to validate that the observed non-Gaussian texture is related to existence of the patch-like sparse forests, the two-component decomposition method is applied. It is successfully revealed that there is spotted high contribution of ground (double-bounce or direct) scattering in the highly textured forest division. Furthermore, difference of the ground contribution between polarization channels could encourage our understanding of polarimetric variation of the forest texture.
Keywords :
geophysical techniques; image enhancement; polarimetry; polarisation; radar cross-sections; remote sensing by radar; synthetic aperture radar; texture; vegetation; Japan; K-distribution; Polarimetry Synthetic Aperture Radar; SAR image; Tomakomai coniferous forest; airborne L-band POLSAR image; forest spatial structure enhancement; non-Gaussian texture; patch-like sparse forest; polarization; radar cross section; radar echoe; texture analysis; two-component decomposition method; Image analysis; Image resolution; Image texture analysis; L-band; Polarization; Radar cross section; Radar imaging; Radar scattering; Remote sensing; Spatial resolution; K-distribution; Non-Gaussian texture; airborne SAR; polarimetric texture; two-component decomposition;
Conference_Titel :
Geoscience and Remote Sensing Symposium, 2008. IGARSS 2008. IEEE International
Conference_Location :
Boston, MA
Print_ISBN :
978-1-4244-2807-6
Electronic_ISBN :
978-1-4244-2808-3
DOI :
10.1109/IGARSS.2008.4779073