A Mobile Data Acquisition Platform for Photometric Stereo

The photometric stereo method (PSM) is a well known technique in the computer vision field that recovers 3D shape from a set of images of a still scene, using a single fixed-pose camera. Every image in the set is taken while a change in the direction of the light source occurs. Commonly, the acquisition of PMS imagery implies using rigid platforms with a static configuration of light sources. In this paper, we explore the use of a mobile platform as an alternative way to obtain PSM imagery. Using a 5 DOF robotic arm, a light source (placed over the end effector of the robot) is moved around an object so as to observe constraints concerning the elevation and azimuth angle of the illumination vector. These constraints are based on work by Drbohlav and Chantler, where a theoretical study about an optimal lightning configuration is introduced. Experiments with objects showing varying reflectance properties demonstrate the feasibility of the mobile platform for the purposes of gathering PSM imagery.