Noise Analysis of MAIA System and Possible Noise Suppression
This paper is devoted to the noise analysis and noise suppression in a system for double station observation of the meteors now known as MAIA (Meteor Automatic Imager and Analyzer). The noise analysis is based on acquisition of testing video sequences in different light conditions and their further statistical evaluation. The main goal is to find a suitable noise model and subsequently determine if the noise is signal dependent or not. Noise and image model in the wavelet domain should be based on Gaussian mixture model (GMM) or Generalized Laplacian Model (GLM) and the model parameters should be estimated by moment method. Furthermore, noise should be modeled by GMM or GLM also in the space domain. GMM and GLM allow to model various types of probability density functions. Finally the advanced denoising algorithm using Bayesian estimator is applied and its performance is verified.
Document typePeer reviewed
Document versionFinal PDF
SourceRadioengineering. 2011, vol. 20, č. 1, s. 110-117. ISSN 1210-2512
- 2011/1