Open AccessImage restoration via DOST and total variation regularisation
Author(s) -
Bini A.A.
Publication year - 2019
Publication title -
iet image processing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.401
H-Index - 45
eISSN - 1751-9667
pISSN - 1751-9659
DOI - 10.1049/iet-ipr.2018.5504
Subject(s) - image restoration , wavelet transform , poisson distribution , computer science , image (mathematics) , noise (video) , artificial intelligence , shot noise , algorithm , wavelet , pattern recognition (psychology) , image processing , mathematics , computer vision , statistics , telecommunications , detector
In this study, the author proposes a total variational (TV) driven image restoration usingdiscrete orthogonal Stockwell transform (DOST) when the noise (in the image) isan outcome of a Poisson process. Stockwell transform or S‐transform (ST) is wellknown for its efficiency in resolving spatio‐frequency components with highaccuracy compared with many other transforms such as short‐term Fouriertransform, wavelet transform etc. This property of ST makes it more suitable formany image processing applications such as image restoration and imageinpainting. By deriving the objective function and constraints of theoptimisation problem (image restoration problem) based on the ST coefficients,the model becomes more robust in terms of preserving high resolution in thespatio‐frequency domain. Images are modelled as an outcome of a Poisson processin many medical and telescopic imaging applications. The Poisson noisecorruption is mainly due to the lack of a sufficient number of photons toreconstruct the data. In this study, corrupted images are restored due to thePoisson process (by which the data is formed) using the DOST under a non‐localTV framework. The model is analysed and compared with the state‐of‐the‐artPoisson noise removal methods using visual and statistical measures.