Title :
Application of Legendre Transform for Decorrelation of Random Variables
Author :
Sahasrabudhe, S.C. ; Kekre, H.B. ; Goyal, N.C.
Author_Institution :
Electrical Engineering Department, Indian Institute of Technology (I.I.T.), Bombay, Powai, Bombay 400 076, India
fDate :
4/1/1982 12:00:00 AM
Abstract :
Karhunen-Loeve (K-L) transform yields a set of uncorrelated random variables from a set of correlated random variables of a data vector. However, K-L transform is computationally not efficient and hence other fast transforms have been investigated for use in their place. Possible use of Legendre transforms in this application has been investigated.
Keywords :
Communication channels; Covariance matrix; Data compression; Decorrelation; Discrete transforms; Indexing; Polynomials; Quantization; Random variables; Sampling methods; Legendre transforms; covariance matrix; nonstationary process; stationary process;
Journal_Title :
Geoscience and Remote Sensing, IEEE Transactions on
DOI :
10.1109/TGRS.1982.350393
