Shape From Focus by Total Variation

Usually, in Shape From Focus (SFF) methods, a single focus measure is applied on an image sequence to obtain an initial depth map and then an approximation technique is used to recover three-dimensional (3D) shape of an object. However, different focus measures perform differently in diverse conditions and it is hard to get accurate 3D shape based on a single focus measure. In this paper, we propose a nonlinear Total Variation (TV) based method for recovering 3D shape of an object by diffusing several initial depth maps obtained through different focus measures. Several experiments have been conducted using images of synthetic and real objects to evaluate the performance of the proposed method. Comparative analysis demonstrates of the effectiveness of the proposed approach.