Structure-Oriented Spatio-Temporal Video Noise Estimation Conference Paper

abstract

• Noise can highly impact the performance of video processing algorithms. This paper proposes a new real-time spatio-temporal method for estimating the noise variance in video signals. The proposed algorithm selects 3D regions or cubes in the video signal with high intensity uniformity. The noise variance is estimated from the selected set of spatially, temporally and spatio-temporally intensity-uniform cubes using local variances calculated along homogeneous plains. The proposed algorithm works well for sequences with high structure and motion activity and outperforms other methods with a worst-case estimation error of 2 dB. It works well for highly noisy and non-noisy sequences

name of conference

• 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings

authors

• Ghrayeb, Ali

published proceedings

• 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings

author list (cited authors)

• M. Ghazal, .., A. Amer, .., & A. Ghrayeb.

publication date

• January 2006

published in

• 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings  Journal

keywords

• Additive White Noise
• Awgn
• Estimation Error
• Gaussian Noise
• High Performance Computing
• Motion Estimation
• Noise Robustness
• Signal Processing
• Video Compression
• Wavelet Domain

start page

• II

end page

• II

volume

• 2