Abstract
This paper describes some tools for adding and removing film grain. The film grain is represented using an additive signal-dependent model. The approach adopted for artificial grain synthesis avoids subjectivity and an assumption of Gaussianity. The grain within a user-defined plain area is analysed and the synthesis routine generates grain with matching spatial structure having the same probability distribution function as the original. The grain reduction method is based on manipulation of the coefficients achieved using a bi-orthogonal undecimated wavelet decomposition and is extremely advantageous for real-time implementation. The scheme for modifying the coefficient is derived from Bayesian estimation and approximates a range of optimal non-linear functions. Training to deduce parameter values is conducted by contaminating several nominally noise-free images with various realisations of grain noise. Using real and synthetically generated grain noise demonstrated an improvement of objective and visual qualities of the image. The ability of the technique to adapt with respect to image and noise characteristics is also clear.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Journal of Visual Communication and Image Representation
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
