A Faster High-Dimensional Bilateral Image Filtering By Efficient Utilization of Color Sparsness Offline

Abstract

Most of the image filtering algorithms involve nonlinear processing of high-dimensional data. Bilateral, joint bilateral, and non-local means filters are the examples of the same. Real-time implementation of high-dimensional filters has always been a research challenge. In this paper, we have proposed a solution utilizing color sparseness in an image which ensures a faster algorithm for generic high-dimensional filtering. The solution speeds up the filtering algorithm further by psycho-visual saliency lookup-table (LUT) evolved through iterations of batched image data processing. We use the proposed filtering algorithm with our LUT for edge-preserving smoothing and denoising of color images. The results demonstrate competitiveness of our proposed solution with existing fast bilateral algorithms in terms of both accuracy and performance.