A robust structure-adaptive hybrid vector filter is proposed for digital color image restoration in this paper. At each pixel location, the image vector (i.e., pixel) is first classified into several different signal activity categories by applying a modified quadtree decomposition to luminance component (image) of the input color image. A weight-adaptive vector filtering operation with an optimal window is then activated to achieve the best tradeoff between noise suppression and detail preservation. Through extensive simulation experiments conducted using a wide range of test color images, the filter has demonstrated superior performance to that of a number of well known benchmark techniques, in terms of both standard objective measurements and perceived image quality, in suppressing several distinct types of noise commonly considered in color image restoration, including Gaussian noise, impulse noise, and mixed noise.
In this thesis, various noise types in digital color imaging are reviewed. Based on the human visual perception, a structure- adaptive filter named SAHVF is proposed for color image restoration. The filter includes a noise-adaptive structure activity classification, which provides a robust structure activity detection and partition for each pixel position. The results of the classification are utilized by a switch-based hybrid filtering structure, which is able to select an optimal subfiltering and processing window to provide the best tradeoff between noise suppression and visual quality. The proposed SAHVF filter has demonstrated a robust performance in suppressing different types of noise contamination.
For additive noise contamination, it can produce a reconstruction with smoother background and sharp texture, which is significantly better in perceived image quality.
For impulse noise corruption, the filter provides a comparable performance to the state-of-the-art filters in noise removal and structure preservation. Beside it robust performance, the filter demands a moderate computation cost for most restoration applications, especially for high-corrupted color images, the SAHVF can provide a better reconstruction, with less than half of the processing time than those classic vector filters.