Application of a conditioned classification methodology for optical and sar data in the mars-stat project: A case study
Use and integration MARS held data sets in a data analysis environment for support of the interpretation of earth resource satellite (ERS) synthetic aperture radar (SAR) time series for area estimation is presented. Also included are methods for fusion of ERS data with other spatial and non-spatial data, in a knowledge steered, or conditioned, classification procedure. Ideas are presented that will ultimately lead to the optimization and future automation of this classification methodology, highlighting a number of shortcomings that might be resolved with future data fusion techniques (eg segmentation of multi-source image data sets). A case study, Driffield 1996, proves that by the use of image logic both SAR and optical images can be integrated, leading to the timely advance of area estimation and increased reliability both with respect to initial estimates and increased accuracy throughout the growing season.
Bibliographic Reference: Paper presented: Fusion of Earth Data, Sophia Antipolis (FR), January 28-30, 1998
Availability: Available from (1) as Paper EN 41147 ORA
Record Number: 199810307 / Last updated on: 1998-03-09
Original language: en
Available languages: en