A Study of Land Cover Classification Using ALOS/PALSAR Polarimetric Data

Inhalt:
We applied various polarimetric decomposition and classification algorithms to the ALOS/PALSAR polarimetry data and evaluated its capabilities of land-cover characteristics identification and its application. We used some known algorithms without changes, or modified for our study. The results show several algorithms work well for PALSAR data. For example, the unsupervised Wishart classification is found to be effective method to identify the land-cover characteristics. The features of PALSAR observation; high resolution, global coverage, and strategic data acquisition, enable us to investigate land-cover characteristics and its changes.