International Journal of Advanced Research in Computer and Communication Engineering

A monthly peer-reviewed online and print journal

ISSN Online 2278-1021
ISSN Print 2319-5940

Since 2012

Abstract:  Accurate estimation of Gaussian noise level is of fundamental interest in a wide variety of vision and image processing applications as it is critical to the processing techniques that follow. In this work, a new effective noise level estimation method is approaches on the basis of the study of singular values of noise-corrupted images. Two novel aspects of this project address the major challenges in noise estimation:1)how to infer the noise level according to image singular values out of SVD space .2)to add new noise to image to be estimated and analyses the change of singular values in order to determine the content related parameter in the model , so that the proposed scheme is adaptive to visual signals, thereby enabling a wider application scope of the proposed scheme. In this work example of algorithms noise estimates include motion estimation, denoising, super-resolution, shape-from-shading, and feature extraction.  Noise level estimation is useful for many computer vision and other image processing algorithms that require knowing the noise level. It the fast noise estimation algorithm using a Gaussian filters in order to estimate the amount noise, images are spilt into a number of blocks and smooth blocks are selected. SVD is a basic tool for signal processing and analysis for long, but it is explored for noise estimation in images. The analysis and experiment results demonstrate that the proposed algorithm can reliably infer noise levels and show robust behavior over a wide range of visual content and noise conditions, and that is outperforms relevant existing methods.

Keywords: SVD, AWGN, MAD, ACF.

PDF | DOI: 10.17148/IJARCCE.2018.7531