Photometric Stereo

Shading on a surface provides strong cues for us to perceive the 3D shape of an object. The study of computationally inferring a 3D shape from shading cues originated with shape from shading (Horn in Shape from shading: a method for obtaining the shape of a smooth opaque object from one view. Technical report, 1970, [1], Ikeuchi and Horn in Artif Intell 17(1–3):141–184, 1981, [2]), which aims to estimate a 3D shape from a single image observed under a light source. The problem has attracted many researchers’ attentions since then because of its mathematically rich problem structure and important practical applications. Later, it was shown that, instead of a single image, the use of multiple images observed under different lighting conditions could alleviate the difficulty of the problem of 3D shape estimation from shading cues. This multiple-image approach is known as photometric stereo. This chapter describes the problem of photometric stereo and its solution methods.