Geodesic
Linear Projection-Based Noise Filtering Framework for Image Denoising
Computer Science and Information Systems, Tome 22 (2025) no. 1.

Voir la notice de l'article provenant de la source Computer Science and Information Systems website

Image denoising remains a fundamental challenge in digital image processing due to the inevitable presence of noise during image acquisition and transmission. While existing noise filtering methods predominantly focus on local spatial information, they often overlook crucial structural information from other perspectives, such as local manifold and global structures. To address this limitation, we propose a novel linear projection-based noise filtering (LPNF) framework grounded in linear projection learning theory. This framework innovatively learns a linear projection for noise filtering by incorporating multiple structural information sources - local spatial, local manifold, and global structures - through well-defined criteria. We present two specialized implementations of the LPNF framework: PCA-based LPNF (LPNF-PCA) and LPP-based LPNF (LPNF-LPP). The LPNF-PCA simultaneously leverages local spatial and global information, while LPNF-LPP integrates both local manifold and spatial information for enhanced denoising performance. Comprehensive experiments conducted on four standard test images with various noise types demonstrate that both LPNF-PCA and LPNF-LPP consistently outperform state-of-the-art denoising methods in terms of both quantitative.
Keywords: Image denoising, linear projection-based noise filtering framework, principle component analysis,locality preserving projection 
@article{CSIS_2025_22_1_a14,
     author = {Congyin Cao and Yangjun Deng and Menglong Yang and Xinghui Zhu},
     title = {Linear {Projection-Based} {Noise} {Filtering} {Framework} for {Image} {Denoising}},
     journal = {Computer Science and Information Systems},
     publisher = {mathdoc},
     volume = {22},
     number = {1},
     year = {2025},
     url = {http://geodesic.mathdoc.fr/item/CSIS_2025_22_1_a14/}
}
TY  - JOUR
AU  - Congyin Cao
AU  - Yangjun Deng
AU  - Menglong Yang
AU  - Xinghui Zhu
TI  - Linear Projection-Based Noise Filtering Framework for Image Denoising
JO  - Computer Science and Information Systems
PY  - 2025
VL  - 22
IS  - 1
PB  - mathdoc
UR  - http://geodesic.mathdoc.fr/item/CSIS_2025_22_1_a14/
ID  - CSIS_2025_22_1_a14
ER  - 
%0 Journal Article
%A Congyin Cao
%A Yangjun Deng
%A Menglong Yang
%A Xinghui Zhu
%T Linear Projection-Based Noise Filtering Framework for Image Denoising
%J Computer Science and Information Systems
%D 2025
%V 22
%N 1
%I mathdoc
%U http://geodesic.mathdoc.fr/item/CSIS_2025_22_1_a14/
%F CSIS_2025_22_1_a14
Congyin Cao; Yangjun Deng; Menglong Yang; Xinghui Zhu. Linear Projection-Based Noise Filtering Framework for Image Denoising. Computer Science and Information Systems, Tome 22 (2025) no. 1. http://geodesic.mathdoc.fr/item/CSIS_2025_22_1_a14/