Abstract:
It is shown that the 2D indexation of a communication channel 1D noise with the help of the 2D Hadamard transform makes it possible to represent the distribution of this noise in the restored image space in the form of a 2D distribution of its variance. The width of noise variance distribution depends on the extent of preservation of correlation between elements of communication channel 1D noise. It is shown that the degree of 2D spatial localization of noise in this case exceeds the degree of noise localization in the case when 1D Hadamard transform is used.