Multilevel Thresholding based on Fuzzy C Partition and Gravitational Search Algorithm

• Autor(a) principal: Gupta, Chhavi
• Data de Publicação: 2014
• Outros Autores: Jain, Sanjeev
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: INFOCOMP: Jornal de Ciência da Computação
• Texto Completo: https://infocomp.dcc.ufla.br/index.php/infocomp/article/view/3

Entropy based image thresholding methods are widely adopted for multilevel image segmentation. Bilevel thresholding partitions an image into two classes, whereas multilevel thresholding partitions an image into multiple classes depending upon thresholding level . The automatic selection of optimal threshold is often treated as an optimization problem. This paper contributes to novel thresholding method, that is based on entropy of fuzzy c partition and gravitational search algorithm (GSA). Experiments have been evaluated on the different test images and results were assessed by entropy, stability, computation time and peak signal to noise ratio (PSNR). The analysis of results conveys that the GSA outperform particle swarm optimization (PSO).