| Paper: | MA-P3.4 |
| Session: | Image and Video Denoising |
| Time: | Monday, September 17, 09:50 - 12:30 |
| Presentation: |
Poster
|
| Title: |
IMAGE DENOISING WITH DIRECTIONAL BASES |
| Authors: |
Heechan Park; University of Warwick | | |
| | Graham Martin; University of Warwick | | |
| | Zhen Yao; University of Warwick | | |
| Abstract: |
Directional information is an important feature of images, and is exploited in many image processing applications. This paper presents an empirical study of image denoising based on directional patterns. From the literature, the directional Wavelet transform, such as the Curvelet, has proven effective for denoising purposes. We introduce the Multiresolution Fourier Transform (MFT) facilitated with a multi-directional filter. From a statistical viewpoint, Independent Component Analysis (ICA) is drawing increasing attention. A combined solution is presented that benefits from the efficiency of the MFT and the data adaptiveness of ICA. Experimental results are very encouraging, even compared with other recent directional transforms such as the Curvelet. |
