Contents I. Introduction
When images are captured optically or electronically and processed, the original image signal is usually degraded. The degradation effect may be seen in the form of sensor noise, blur due to out of focus of camera, relative object camera motion, lens and film non linearity, impulse due to faulty storage and transmission media and many other factors. Noise is a high frequency component in an image, so we require low pass filter to denoise it. Noise removal using different types of low pass filtering techniques is an important area of research in digital image and signal processing. Denoising a corrupted image is known as image restoration. The objective of the restoration scheme is to recover the original image from the observed degraded image. Image restoration methods are used to model the degradation process and apply an approximately inverse process to the degraded image to recover the original image [1]. The effectiveness of such restoration techniques depends on the availability and completeness of knowledge about the impulse degradation process as well as on the structure of the filtering scheme. The linear filter is adequate in removing noise for a bandwidth limited additive noise like Gaussian noise from the corrupted images. However, it is observed in many situations that due to presence of impulse noise, linear filter produces more smoothing and poor performance [2]. To overcome these shortcomings, non-linear filters like median filter is used to denoise the Salt & Pepper noise from the corrupted image. In linear filter, the value of the output pixel is calculated using the weighted sum of input pixels, but non linear filter assigns a value to the output pixel which is directly based on the values of the pixel in the neighborhood. The median filter is a computationally efficient non linear filter for denoising impulse noise and also preserves edges. Different image restoration schemes have been proposed in the literature on digital image processing [3], [4]. Median filtering is a nonlinear signal processing technique developed by Tukey that is useful for noise suppression in images [5]. To improve the restoration performance several fast median filtering algorithms are described as in [6], [7], [8]. Different literatures are available for denoising Salt & Pepper noise by average filter as in [9], [10]. To denoise salt & Pepper noise from gray scale image by Rank order filtering, different efficient schemes are proposed in [11]–[14]. The outlier technique works on a principle in which the center pixel in a window is replaced by the average of its ‘k’ neighbors whose amplitudes are closest to the center pixel as discussed in [15]. Another variant of the outlier algorithm called the sigma filter has been suggested in [16]. Objective and subjective estimation of image restoration by different filtering techniques are reported as in [17], [18], [19].
