Salt-and-pepper noise removal by median-type noise detectors and detail-preserving regularization

• Published in: IEEE transactions on image processing Vol. 14; no. 10; pp. 1479 - 1485
• Main Authors: Chan, R.H., Chung-Wa Ho, Nikolova, M.
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01-10-2005; Institute of Electrical and Electronics Engineers; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Adaptive filters; Adaptive median filter; Algorithms; Applied sciences; Artificial Intelligence; Computer Simulation; Detection, estimation, filtering, equalization, prediction; Detectors; edge-preserving regularization; Exact sciences and technology; Image Enhancement - methods; Image Interpretation, Computer-Assisted - methods; Image processing; Image restoration; impulse noise; Information filtering; Information filters; Information Storage and Retrieval - methods; Information, signal and communications theory; Models, Statistical; Noise level; Noise reduction; Nonlinear filters; Pattern Recognition, Automated - methods; Phase noise; Pixel; Signal and communications theory; Signal processing; Signal, noise; Stochastic Processes; Telecommunications and information theory; Regularization method; Pulse noise; Image processing; Noise reduction; Adaptive median filter; Non linear filter; edge-preserving regularization; impulse noise; Adaptive filter; Image restoration; Median filter; Noise level
• ISSN: 1057-7149; 1941-0042
• DOI: 10.1109/TIP.2005.852196

This paper proposes a two-phase scheme for removing salt-and-pepper impulse noise. In the first phase, an adaptive median filter is used to identify pixels which are likely to be contaminated by noise (noise candidates). In the second phase, the image is restored using a specialized regularization method that applies only to those selected noise candidates. In terms of edge preservation and noise suppression, our restored images show a significant improvement compared to those restored by using just nonlinear filters or regularization methods only. Our scheme can remove salt-and-pepper-noise with a noise level as high as 90%.