Abstract
Many real word images are contaminated by noise. The noise not only degrades image quality but may also hinder further processing operations. Noise reduction techniques aim to both improve image quality and to aid further image processing. Spatial noise reduction techniques based on the discrete wavelet transform have been widely researched. This paper considers an undecimated shift invariant filter bank that has been used to decompose the image into components. The basic filters are derived from a biorthogonal wavelet basis. Reconstruction is obtained by a simple summation of the image components. A new thresholding scheme, which is obtained from Bayesian estimator theory, is used. The threshold parameters for each component are dependent on the noise level and are selected using a preliminary training procedure. The cost function utilised for the training is a weighted version of the mean square error which is designed to reflect human perception. The method compares favourably with other wavelet based noise reduction techniques and demonstrates significant noise reduction and visual quality enhancement.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
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.