Community Research and Development Information Service - CORDIS

Abstract

The main approach of GCP is to be clearly different from its closer environment. This approach should be computed and analysed to establish a measurement of its reliability. The concept of footprint is introduced for characterising this computational aptitude. The footprint is computed using a correlated function. It is applied to the central window frame of the candidate sub-image to correlate the GCP and every window frame extracted from this sub-image. So we obtain an auto-correlated matrix result, on which we applied some criteria on the basis of establishing the GCP reliability. These criteria are the variance, Euler characteristic and footprint value. In addition, a numerical criterion of direct comparisons about the two GCPs reliability that, excelling the criteria previously quoted, represent zones to correct very closely. This criterion is based on the necessity of having a quantity of GCPs not excessive on database and, simultaneously, to be sure that these are the best candidates for the process of geometrical correction. In this way, we divide the selection system in two stages:
-1. Sieve; gets rid of those that can lead into error on the localisation.
-2. Priority to the adjustment reliability of different GCPs with similar act zones. .

Additional information

Authors: CRUZ-MARTINEZ M, Universidad de Almeria (ES);LOPEZ-RAYA J-T, Universidad de Almeria (ES);BARON-MARTINEZ J, Universidad de Almeria (ES);CARMONA-MORENO C, European Commission, Joint Research Centre, Space Applications Institute, Ispra (IT);MORENO-RUIZ J-A, European Commission, Joint Research Centre, Space Applications Institute, Ispra (IT)
Bibliographic Reference: An oral report given at: EOS/SPIE - Symposium on Remote Sensing. Held at: Barcelona,(ES) 23-27 September 2000
Record Number: 200214300 / Last updated on: 2002-01-28
Category: PUBLICATION
Original language: en
Available languages: en