Estimating the parameters of the generalized KA distribution by applying the expectation maximization algorithm

Theonymphi M. Melesanaki; Ioannis O. Vardiambasis; Melina P. Ioannidou; Evangelos A. Kokkinos; Andreas M. Maras

doi:10.1117/1.JRS.13.014518

20 February 2019 Estimating the parameters of the generalized KA distribution by applying the expectation maximization algorithm

Theonymphi M. Melesanaki, Ioannis O. Vardiambasis, Melina P. Ioannidou, Evangelos A. Kokkinos, Andreas M. Maras

Author Affiliations +

Journal of Applied Remote Sensing, Vol. 13, Issue 1, 014518 (February 2019). https://doi.org/10.1117/1.JRS.13.014518

Abstract

Generalization of the KA distribution is formulated by combining the class A and K distributions; the resulting distribution is termed as generalized KA distribution. It is obtained as a mixture of a generalized Rayleigh and a class A distribution with gamma-distributed mean intensity, and it may be used to describe clutter statistics. Its parameters are estimated by implementing the expectation maximization algorithm. The latter provides estimates in the framework of the maximum likelihood principle, and it is widely used when the data set is incomplete and/or of limited size. The numerical results show that the absolute relative error of the estimated parameters may be <8 % even in the case of 100 data samples, whereas it is reduced significantly as the size of the data set increases.

Citation Download Citation

Theonymphi M. Melesanaki, Ioannis O. Vardiambasis, Melina P. Ioannidou, Evangelos A. Kokkinos, and Andreas M. Maras "Estimating the parameters of the generalized KA distribution by applying the expectation maximization algorithm," Journal of Applied Remote Sensing 13(1), 014518 (20 February 2019). https://doi.org/10.1117/1.JRS.13.014518

Received: 15 October 2018; Accepted: 7 February 2019; Published: 20 February 2019

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

;

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE