Denoising real images using complex-valued wavelets


Stuart Barber & Guy P. Nason.
Barber & Nason (2003) have demonstrated that complex-valued wavelets can significantly outperform real-valued wavelets when denoising one-dimensional signals. We extend the complex multi-wavelet style thresholding methodology of Barber & Nason (2003) to the two dimensional case. Initial results indicate a promising improvement in the denoising of test image.

Some key words:
Complex-valued wavelets, denoising, image analysis.


Back to publications list