A fast and robust method of calculating RFM parameters for satellite imagery. Issue 12 (1st December 2016)

• Record Type: Journal Article
• Authors: Wu, Yingdan; Ming, Yang
• Is Part Of: Remote sensing letters. Volume 7:Issue 12(2016)
• Page Start: 1112
• Page End: 1120
• Publication Date: 2016-12-01
• Subjects: Remote sensing -- Periodicals; Remote sensing; Periodicals; 621.3678
• DOI: 10.1080/2150704X.2016.1219459
• Languages: English
• ISSNs: 2150-704X

Abstract

ABSTRACT: Since the normal matrix is often ill-conditioned when solving Rational Function Model (RFM) for the satellite remote sensing imagery, combining with matrix orthogonal decomposition, Levenberg–Marquardt algorithm and Compute Unified Device Architecture high-performance computing technique, a fast and robust method directly based on the error equation coefficient matrix is proposed. The method is analysed by different parameters of control point grid and compared with the common methods, namely L-curve based ridge estimation and Iteration by Correcting Characteristic Value. The experimental results show that RFM parameters derived by our method have higher fitting accuracy.