Open AccessImage Sequence Denoising with Motion Estimation in Color Image Sequences
Author(s) -
M. V. Sarode,
P. R. Deshmukh
Publication year - 2011
Publication title -
engineering, technology and applied science research/engineering, technology and applied science research
Language(s) - English
Resource type - Journals
eISSN - 2241-4487
pISSN - 1792-8036
DOI - 10.48084/etasr.54
Subject(s) - artificial intelligence , computer vision , video denoising , computer science , noise reduction , motion estimation , monochrome , impulse noise , noise (video) , color image , motion compensation , pattern recognition (psychology) , image processing , image (mathematics) , video processing , pixel , video tracking , multiview video coding
In this paper, we investigate the denoising of image sequences i.e. video, corrupted with Gaussian noise and Impulse noise. In relation to single image denoising techniques, denoising of sequences aims to utilize the temporal dimension. This approach gives faster algorithms and better output quality. This paper focuses on the removal of different types of noise introduced in image sequences while transferring through network systems and video acquisition. The approach introduced consists of motion estimation, motion compensation, and filtering of image sequences. Most of the estimation approaches proposed deal mainly with monochrome video. The most usual way to apply them in color image sequences is to process each color channel separately. In this paper, we also propose a simple, accompanying method to extract the moving objects. Our experimental results on synthetic and natural images verify our arguments. The proposed algorithm’s performance is experimentally compared with a previous method, demonstrating comparable results.