Extraction and Analysis of Coherent Targets Using Quad-Pol ALOS-PALSAR L-band Data
Changcheng Wang(1), Mingsheng Liao(1), Yong Wang(2) and Jianya Gong(1)
(1) Wuhan University, 129 Luoyu Road, 430079 Wuhan, China
(2) East Carolina University, Greenville, NC 27858, United States
The extraction and analysis of coherent targets is an important task for applications of SAR/InSAR. Especially, the extraction of permanent scatterers based on their temporal stability in time-series of SAR images with a long period. Souyris etc. extracted coherent targets based on the interferometric correlation properties of the different sublook images. However, for the data from long wavelength SAR (such as L-band ALOS PALSAR), the sublook correlation can be quite high for the forested areas with tall and large trees. As the wavelength increases, attenuation from tree canopy decreases, and microwave energy can penetrate the canopy and hit the ground. The interactions of double-bounce trunk-ground interaction can be strong. Thus, it is difficult to separate urban man-made corner-reflector-type targets from large trees by using the sublook correlation. In this paper we develop a coherent targets extraction method that can overcome the un-separability by combining the sublook correlation and polarimetric signatures.
Due to presence of houses and buildings, backscattered signals from urban areas are the combination coherently of single-, double- and triple-bounce returns of various parts of the structures, but they seldom come from volume scattering of random media, such as vegetation. Conversely, for the scatterers in forested areas with tall and large trees, the contribution of volume scattering is significant. Therefore, we use Cloude-Pottier Eigenvalue/Eigenvector polarimetric decomposition method to extract the entropy (H), mean alpha angle (α) and anisotropy (A), and analyze the scattering properties from man-made structures versus scatterers in forest. Then, the polarimetric parameters are combined with the sublook correlation of each channel of SAR image to extract the coherent targets.
In our study, several L-band ALOS-PALSAR Quad-Pol images are chosen to evaluate the developed method. After the extraction of the coherent targets, the targets are then classified into three categories of scattering types: surface, volume and even bounce. Finally, their polarimetric characteristics are assessed. The results show that our method can effectively differentiate the scatterers of man-made structures from those of tall and large trees.
Keywords: coherent target, permanent scatterers interferometry (PSI), polarimetry, sublook correlation.
Keywords: ESA European
Space Agency - Agence spatiale europeenne,
observation de la terre, earth observation,
satellite remote sensing,
teledetection, geophysique, altimetrie, radar,
chimique atmospherique, geophysics, altimetry, radar,