Title :
Novel criteria of uniqueness for signal reconstruction from phase
Author_Institution :
Inst. voor Perceptie Onderzoek, Eindhoven, Netherlands
fDate :
4/1/1991 12:00:00 AM
Abstract :
An approach for ascertaining whether a signal is uniquely determined by its Fourier transform phase is proposed. It is shown that uniqueness corresponds to the nonsingularity of a matrix which can be formed from the finite-length real sequence. The criterion of uniqueness for reconstructing a one-dimensional finite-length sequence from its phase is discussed. The criterion of uniqueness for reconstructing a multidimensional finite-length sequence is presented
Keywords :
Fourier transforms; matrix algebra; signal synthesis; Fourier transform phase; finite-length real sequence; matrix singularity; multidimensional finite-length sequence; one-dimensional finite-length sequence; signal reconstruction; uniqueness; Filters; Fourier transforms; Geophysical signal processing; Image processing; Image reconstruction; Polynomials; Signal processing; Signal processing algorithms; Signal reconstruction; Speech processing;
Journal_Title :
Signal Processing, IEEE Transactions on
