This is the author's version of an article that has been published in this journal. Changes were made to this version by the publisher prior to publication. The final version of record is available at http://dx.doi.org/10.1109/TIP.2018.2800289
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT)
REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT)
REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT)
REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT)
REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT)
𝜶. In fact, ‖𝜸𝑺𝒔 (𝒆 − 𝒛)‖𝟐𝟐 applies an extra denoising model alongside the denoising role of ‖𝜷𝑫′′ 𝒆‖𝟐𝟐 , on the marginal samples. The result of denoising and estimation after adding ‖𝜸𝑺𝒔 (𝒆 − 𝒛)‖𝟐𝟐 to the original TRM are shown in Figures 18.e and 18.f respectively. Until now, the proposed objective function has evolved to ‖𝜶𝑺(𝒆 − 𝒄)‖𝟐𝟐 + ‖𝜷𝑫′′ 𝒆‖𝟐𝟐 + ‖𝜸𝑺𝒔 (𝒆 − 𝒛)‖𝟐𝟐 in which, only ‖𝜷𝑫′′ 𝒆‖𝟐𝟐 contributes to the estimation of missing samples located at the central region of the corrupted signal (a group of 𝑵𝒈 samples). As mentioned earlier, ‖𝜷𝑫′′ 𝒆‖𝟐𝟐 contributes in the estimation process by enforcing the missing samples to follow the trend and the curve shape of the known samples of the signal. Accordingly, based on the direction of the existing samples
Copyright (c) 2018 IEEE. Personal use is permitted. For any other purposes, permission must be obtained from the IEEE by emailing
[email protected].
This is the author's version of an article that has been published in this journal. Changes were made to this version by the publisher prior to publication. The final version of record is available at http://dx.doi.org/10.1109/TIP.2018.2800289
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT)
REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT)
