Adaptive Color Image Filtering Based on Center-Weighted Vector Directional Filters

TLDR: By varying the center weight, the proposed CWVDF framework can provide the smoothing characteristics ranging from an identity operation to that of the BVDF, which removes impulses and outliers from the image while simultaneously preserving the structural information.

Abstract: This paper presents a new filtering approach for impulsive noise removal in color images. Incorporating the nonnegative integer weight corresponding to the central sample into the structure of the basic vector directional filter (BVDF), the proposed framework constitutes a class of center-weighted vector directional filters (CWVDF). It can be easily observed that the CWVDF filters are computationally efficient and extend design flexibility of the standard BVDF scheme. By varying the center weight, the proposed CWVDF framework can provide the smoothing characteristics ranging from an identity operation to that of the BVDF. Therefore, design characteristics relate to the CWVDF, which removes impulses and outliers from the image while simultaneously preserving the structural information. To adaptively determine the optimal value of the center weight, two adaptive approaches based on the angular thresholds are provided. Both techniques achieve excellent results in terms of the commonly used objective image quality criteria and significantly outperform standard multichannel filtering algorithms.