Nonlinear stochastic image modeling by means of multidimensional finite mixture distributions

Proceedings of SPIE - The International Society for Optical Engineering

Volumn 3304, Issue , 1998, Pages 96-107

  Maison, Benoît ^a   Vandendorpe, Luc ^a  

^a UNIVERSITÉ CATHOLIQUE DE LOUVAIN   (Belgium)

Author keywords

HOS; Image Modeling; Mixture Distributions; Nonlinear; Stochastic Modeling

Indexed keywords

ALGORITHMS; CORRELATION METHODS; IMAGE CODING; IMAGE COMPRESSION; IMAGE RECONSTRUCTION; IMAGE SEGMENTATION; MATHEMATICAL TRANSFORMATIONS; MEDICAL IMAGING; NONLINEAR SYSTEMS; PARAMETER ESTIMATION; RANDOM PROCESSES; SPECTRUM ANALYSIS; VECTORS;

IMAGE MODELING;

IMAGE PROCESSING;

EID: 0032390184     PISSN: 0277786X     EISSN: None     Source Type: Conference Proceeding    
DOI: 10.1117/12.304619     Document Type: Conference Paper

References (6)

1. Entropy coding of composite sources modeled by infinite gaussian mixture distributions
2. Maximum-likelihood parameter estimation for unsupervised stochastic model-based image segmentation
3. Time-varying subband image coding with efficient reduction of higher order redundancy
4. On the modeling of DCT and subband image data for compression
5. The EM algorithm and extensions