Resumen de A nonlocal variational modelfor retinex image decomposition

Frank W. Hammond Espinosa, Catalina Sbert Juan

• Unlike the human vision system, which is able to recognize color under different illuminationconditions, the information captured by a digital camera highly depends on the light reflectedby the objects in the scene. The Retinex theory assumes that a digital image can be decomposedinto a luminance component and a reflectance component. In this work, we propose a variationalmodel to solve the ill-posed inverse problem of estimating luminance and reflectance from agiven observation. We promote spatial smoothness for the illuminance in a suitable Sobolevspace and introduce nonlocal regularization, which takes advantage of image self-similarities,for the reflectance. A theoretical analysis of the resulting energy functional is provided