An Adaptive Hybrid Filtering for Removing Impulse Noise in Color Images

  • Authors:

  • Xuan Guo;Baoping Guo;Tao Hu;Ou Yang

  • Affiliations:

  • College of Optoelectronics Science and Engineering, Huazhong University of Science and Technology, Wuhan, China 430074;Institute of Optoelectronics, Shenzhen University, Shenzhen, China 518060;College of Optoelectronics Science and Engineering, Huazhong University of Science and Technology, Wuhan, China 430074;College of Optoelectronics Science and Engineering, Huazhong University of Science and Technology, Wuhan, China 430074

  • Venue:
  • ISNN 2009 Proceedings of the 6th International Symposium on Neural Networks: Advances in Neural Networks - Part III
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

An adaptive hybrid filter combining a group of sigma vector median filters with different thresholds with a filter based on neuro-fuzzy system is proposed for color image processing. The first subunit of the proposed filter is six sigma vector median filters, their outputs are used as optimum initial points to input the second subunit constituted by a simple Sugeno-type neuro-fuzzy system, and then the optimized result is obtained from the output of the second subunit. The parameters of the neuro-fuzzy model are automatically tuned and fixed by a learning method based on genetic algorithm. The results have indicated that the design of the proposed hybrid filter has met the requirement of removing impulse noise and preserving details. The proposed filter performs better than other filters.