compsense.problems.probMissingPixels

class compsense.problems.probMissingPixels(fill_ratio=0.6, sigma=0.0055242717280199029, wavelet='db2', undecimated=False, wavelet_levels=None, noseed=False)

RandomMask example: Wavelet basis, masked Photographer. probMissingPixels creates a problem structure. The generated signal consists of the 256 by 256 grayscale ‘photographer’ image. A random binary mask is applied to the signal creating ~40% missing pixels and a ormally distributed noise with standard deviation SIGMA = 0.0055 is added to the final signal.

Examples

>>> P = probMissingPixels()   # Creates the default problem.

Parameters :

fill_ratio : float, optional (default=0.6)

Ratio of non zero (1) values in the mask.

sigma : float, optional (default=sqrt(2)/256)

Standard deviation of the additive noise.

wavelet : str, optional (default=’db2’)

Wavelet to use as saprsifying signal basis. If None, no sprasifying basis is used (dirac operator).

undecimated : bool, optional (default=False)

Use undecimated wavelet transform

wavelet_levels : int, optional (default=None)

Number of scaling levels used in the wavelet transform. If None, maximum possible number is used

noseed : bool, optional (default=False)

When True, the initialization of the random number generators is suppressed

Attributes

A	Response of the problem
B	Base matrix
M	Sampling matrix
b	Observation vector
name	Name of the problem
signal	Signal (Not in sparsifying basis)
signal_shape	Shape of the signal
x0	Solution to problem