An Algorithmic Analysis of Variational Models for Perceptual Local Contrast Enhancement

Abstract

Color cast cancellation and local contrast enhancement are very important problems in computer vision. In this paper we review the algorithm proposed by Palma-Amestoy et al. Source CodeThe code implements the algorithm presented by Palma-Amestoy et al. [7], with a precomputation of the polynomials. These polynomials are inputted from text files which are also provided.The reviewed source code and documentation for this algorithm are available from the web page of this article 1 . Compilation and usage instruction are included in the README.txt file of the archive. Supplementary MaterialAs supplementary material, a reference dataset is provided, and the Matlab code to generate the polynomial approximation, although it is not necessary to execute it to test the algorithm.

Hue-preserving perceptual contrast enhancement

“…In Figure 1, we compare Algorithm 1 to the equalization implementation of Gimp, the algorithm of Nikolova et al [10,22] with implementation [11], the method of Bertalmio et al [13] and the one of Ferradans et al [15] with entropic Michelson function. For the method of [13], we use the implementation [15]. For the comparison approaches, we used the default parameters of the papers.…”

“…On the Flower image, the approaches of [13,15] as well as the one of the Gimp changes the hue, thus, the background becomes purple. The proposed approach, as well as the one of [10] preserves the hue and the colorfulness.…”

“…whereν,ν ∈ {R, G, B}. [10], of Bertalmio et al [13], of Ferradans et al [15]. The proposed algorithm produces reliable results in all these cases.…”

“…The method was generalized by Palma-Amestoy et al [14] for a larger class of functionals. This last method was implemented in [15]. Provenzi et al [16] propose to work in the wavelet domain in order to reduce the computation time.…”

Fusion-based Variational Image Dehazing

“…The FFT technique proposed in [24] can also be applied here, delivering an O(n) complexity, which in our implementation requires 90 seconds to process a 1 MP image 1 . An algorithmic analysis of the minimization of the energy in [24], which lies at the core of the proposed technique, is available at [25]. Fig.…”