Selective Median Switching Filter for Noise Suppression inMicrostructure Images of Material

International Journal of Image Processing

P.S. Hiremath¹  Anita Sadashivappa¹ 

[1]$$

关键词: Pre-processing;    SMSF;    Median Filter;    MSE;    PSNR;    Correlation Coefficient;    Microstructure.;   

DOI  :  

来源: Computer Science Journals

【 摘 要 】

The image pre-processing is very critical and important task in any digital image analysis system.The eventual success and failure of image analysis depends on the performance of preprocessingtechniques applied on the image to be analyzed. In digital images, different noisetypes are noticed and to attenuate each type of noise, different pre-processing methods havebeen proposed in literature. The main focus of this paper is on pre-processing the microstructureimages. Among many types of noise, impulse noise is the one which is generally noticed inmicrostructure images. This paper is to present a novel, efficient and suitable pre-processingmethod for negotiating the impulse noise that is generally present in microstructure images.Through this paper, a new filtering method, selective median switching filter (SMSF) has beenproposed. The proposed method is compared with filtering methods those belong to median filterfamily for their efficiency in negotiating with impulse noise. The efficiency of proposed method iscompared with other methods by computing the three image quality assessment methods,namely, mean square error (MSE), peak signal-to-noise ratio (PSNR) and correlation coefficient.The experimental results confirm that the proposed SMSF method is efficient in handling theimpulse noise present in microstructure images of material. Also, the proposed SMSF method isquite efficient in preserving the edge information in images.

【 授权许可】

Unknown