Abstract:
The article studies the regularization method used in the problems of the compression signals by means of the discrete orthogonal conversions. The regularization parameters for the various discrete orthogonal conversions are founded. The matrices for the spectral component selection are determined, the errors are evaluated. Estimations of errors of restoration of the compressed signal coded by means of orthogonal transformations with regularization and without regularization are received. From the given estimations follows, that the signal restored with application of the regularization method is closer to an initial signal rather than a signal restored without application of the regularization method.
Fulltext:
PDF file (366 kB)