Two-Stage Image Denoising by Principal Component Analysis with Local Pixel Grouping.

Information Communication Engineering

This paper presents an efficient image denoising scheme by using principal component analysis (PCA) with local pixel grouping (LPG). For a better preservation of image local structures, a pixel and its nearest neighbors are modeled as a vector variable, whose training samples are selected from the local window by using block matching-based LPG. Such an LPG procedure guarantees that only the sample blocks with similar contents are used in the local statistics calculation for PCA transform estimation, so that the image local features can be well preserved after coefficient shrinkage in the PCA domain to remove the noise. The LPG-PCA denoising procedure is iterated one more time to further improve the denoising performance, and the noise level is adaptively adjusted in the second stage. Experimental results on benchmark test images demonstrate that the LPG-PCA method achieves very competitive denoising performance, especially in image fine structure preservation, compared with state-of-the-art Denoising algorithms. In this paper, six different image filtering algorithms are compared based on their ability to reconstruct noise affected images. The purpose of these algorithms is to remove noise from a signal that might occur through the transmission of an image. A new algorithm, the Spatial Median Filter, is introduced and compared with the current image smoothing techniques. Experimental results demonstrate that the proposed algorithm is comparable to popular image smoothing algorithms. In addition, a modification to this algorithm is introduced to achieve more accurate reconstructions over other popular techniques. In this paper, an effective algorithm for noise removal in an image is obtained by using PCA (principal component analysis) with LPG (Local Pixel Grouping). This technique ensures the preservation of image local structure. Here the pixels and its neighbors are treated as vector variables whose training samples are selected from local windows using block matching based LPG. This ensures only the similar samples are selected for the PCA transformation so that the desired local characteristics are only preserved with considerable noise reduction. The LPG – PCA algorithm is performed twice to enhance the quality of an image. The first iteration would remove the noise considerably and the second iteration would preserve the image features like edges etc. The LPG-PCA algorithm will adaptively adjust the noise level of an image unlike WT (Wavelet Transformation). Several experimental results show the effectiveness of the proposed algorithm.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.