Abstract
In this paper we present the robust Cascaded RM-filter to remove the mixture of impulsive and multiplicative noise from corrupted images. The designed filter uses combined R- and M-estimators called RM-estimators. The capability of the impulsive noise removal in the image processing applications by using of the RM-KNN (MM-KNN, WM-KNN, ABSTM-KNN or MoodM-KNNN) are presented. The presented Cascaded RM-filter is the consequential connection of two filters. The first filter uses one of the proposed filters to provide impulsive noise rejection and detail preservation. The second one uses an M-filter to provide multiplicative noise suppression. We investigated various types of influence functions used in the M-estimator of designed filter. Extensive simulation results in reference images have demonstrated that the proposed filters consistently can outperform other nonlinear filters by balancing the tradeoff between noise suppression and detail preservation. The criterions used to compare performance were the PSNR and MAE. Finally, we also present the real-time implementation of proposed filter using DSP TMS320C6701 to demonstrate that it potentially provides a real-time solution to quality TV/video transmission.
Original language | English |
---|---|
Pages (from-to) | 1630-1641 |
Number of pages | 12 |
Journal | Proceedings of SPIE - The International Society for Optical Engineering |
Volume | 5150 III |
DOIs | |
State | Published - 2003 |
Event | Visual Communications and Image Processing 2003 - Lugano, Switzerland Duration: 8 Jul 2003 → 11 Jul 2003 |
Keywords
- DSP
- Rank-M filters
- Real imaging
- Robust image processing