Median Morphological Filter Design Based on the PSO Algorithm
Design of morphological filter greatly depends on morphological operations and structuring elements selection. A filter design method used median closing morphological operation is proposed to enhance the image denoising ability and the PSO algorithm is introduced for structural elements selecting. The method takes the peak value signal-to-noise ratio (PSNR) as the cost function and may adaptively build unit structuring elements with zero square matrix. Experimental results show the proposed method can effectively remove impulse noise from a noisy image, especially from a low signal-to-noise ratio (SNR) image; the noise reduction performance has obvious advantages than the other.
Y. L. Zhu and C. Huang, "Median Morphological Filter Design Based on the PSO Algorithm", Applied Mechanics and Materials, Vols. 128-129, pp. 181-184, 2012