ABSTRACT
This chapter surveys filtering methods suitable for restoring the desired microarray image information from the corresponding noisy measurements. To facilitate the discussion on microarray image filtering and restoration, Section 7.2 describes the problem of image formation and noise modeling, which are crucial for simulating the noise observed in microarray images and studying the effect of the filters on both the noise and the desired image features. Section 7.3 surveys various noise filtering methods that have been shown to be effective in suppressing the noise in microarray images. These methods are based on data averaging (e.g., Gaussian, bilateral, anisotropic diffusion, and nonlocal mean filters), order-statistics (e.g., median, weighted median, and combination filters), mathematical morphology (e.g., erosion, dilation, opening, and closing filters), and wavelets (e.g., hard and soft thresholding).
