1Department of Mathematics, Zakir Husain Delhi College, University of Delhi, Delhi, India
2Department of Computer Science, Maulana Azad National Urdu University, Hyderabad, India
3Department of Mathematics, Al-Falah University, Faridabad, Haryana, India
Due to the imperfection of image acquisition systems and transmission channels, noise degrades the quality of almost every acquired image. There are number of denoising techniques which are in existence over the last few years, originated from various disciplines. Wavelet transform has speedily become one of the image processing tool for variety of applications. Unlike standard wavelet-based method, in the present paper, we have proposed new technique using local neighbourhood and modified mean in wavelet packet domain. On comparing this method with other methods we have obtained better quality of the estimated signal at the cost of more computational time.
Wavelet packet transform, denoising, thresholding
