6533b82ffe1ef96bd1295ae7

RESEARCH PRODUCT

Chaotic multiagent system approach for MRF-based image segmentation

Kamal Eddine MelkemiSebti FoufouMohamed Batouche

subject

Segmentation-based object categorizationbusiness.industryComputer scienceMulti-agent systemCrossoverComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISIONChaoticScale-space segmentationImage segmentationComputingMethodologies_ARTIFICIALINTELLIGENCENonlinear Sciences::Chaotic DynamicsComputer Science::Multiagent SystemsComputerSystemsOrganization_MISCELLANEOUSComputer Science::Computer Vision and Pattern RecognitionIterated conditional modesSegmentationComputer visionArtificial intelligencebusiness

description

In this paper, we propose a new chaotic approach for image segmentation based on multiagent system (MAS). We consider a set of segmentation agents organized around a coordinator agent. Each segmentation agent performs iterated conditional modes (ICM) starting from its own initial image created using a chaotic mapping. The coordinator agent diversifies the initial images using a crossover and a chaotic mutation operators. The efficiency of our chaotic MAS approach is shown through some experimental results.

https://doi.org/10.1109/ispa.2005.195421