3D shape, color and specular estimation using image scanner with multiple illuminations

In this study, by using the image scanner equipped with multiple illuminations, we proposes a method to estimate 3D shape, color and specular reflections of an object. The proposed method consists of two parts. The first process is the initial shape estimation. This method uses several scanned images taken by illuminations located on the circumference, and estimate object shape by the photometric stereo using average images reduced the specular reflection. The second process is the iterative estimation of specular components, color and shape. In the experiments, we show shape reconstruction results using synthetic images, and we confirm the effectiveness of our proposed methods.