Repairing imperfect video enhancement algorithms using classification-based trained filters

Abstract

Multifarious image enhancement algorithms have been used in different applications specifically. Still, some algorithm and modules are imperfect for practical use. When the image enhancement modules have been fixed or combined by a series of algorithm, we need to repair it as a whole part without changing the inside. This report aim to find an algorithm based on trained filters to repair low-quality image enhancement modules. A brief review on basic image enhancement techniques and pixel classification methods will be presented, and the procedure of trained filter will be described step by step. The experiments and results comparisons for this algorithm will be described in detail.

Fast image enhancement based on color space fusion

“…The second group can be further divided into two approaches: the RGBbased approaches and the HSV-based approaches. One of them mainly concentrates on image filtering improvements [8][9][10][11][12] and the illumination analysis. [13][14][15][16] Meylan and Susstrunk 9 proposed an adaptive filter combined with the principal component analysis (PCA) method.…”

An Adaptive Diffusion Scheme for Image Restoration and Selective Smoothing

“…14,17 Among a wealth of the filters available for image denoising and restoration we mention adaptive filters, 3 convex filters, 19,20 coupled PDEs 11 and more recently trained filters paradigm. [21][22][23] An important aspect in anisotropic diffusion-based filters and adaptive smoothing schemes is that the feature detection is part of the problem and is built-in. For most anisotropic PDEs, this is done using the spatial gradient, giving an indication of discontinuities in the image.…”