Multi-objective evolutionary for synthetic aperture radar image segmentation with non-local means denoising

  • Authors:

  • Yangyang Li;Ying Wei;Yang Wang;Licheng Jiao

  • Affiliations:

  • Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education of China, Xidian University, Xi'an, China 710071;Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education of China, Xidian University, Xi'an, China 710071;Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education of China, Xidian University, Xi'an, China 710071;Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education of China, Xidian University, Xi'an, China 710071

  • Venue:
  • Natural Computing: an international journal
  • Year:
  • 2014

Quantified Score

Hi-index 0.00

Visualization

Abstract

Synthetic aperture radar (SAR) image segmentation is an important problem of the realm of image segmentation. In this study, a novel SAR image segmentation algorithm using a multi-objective evolutionary algorithm based on decomposition with non-local means denoising (MISD) is proposed. The novelty of MISD lies in the following issues: (1) an effective multi-objective method with decomposition to solve SAR image segmentation; (2) in order to denoise the SAR images and retain the details, we employ non-local means to remove the noise. The multi-objective decomposition method makes MISD have lower computational complexity. In order to evaluate the performance of the new method, we compared the results with three other popular segmentation approaches on four simulated and two real SAR images. In our experiments, the new method can always find better results, which means MISD is a promising SAR image segmentation method.