Abstract

This paper presents a novel 3D reconstruction method based on smartphone device. By the usage of an extra light source and the smartphone camera, an uncalibrated photometric stereo (UPS) system can be constructed. To solve the surface normal accurately, the total variation minimization and energy minimization methods are used in the proposed approach. The total variation minimization method is used to remove image noise without loss of local gradient. A delta function with zero entropy is used to characterize the distribution of albedo with the assumption that albedos are constant for all points belonging to the reconstructed surface. By minimizing the entropy of recovered albedos, unknowns in the generalized bas relief (GBR) model can be solved. Experiments on public datasets and real objects are implemented to demonstrate the precision and feasibility of the proposed photometric stereo algorithm.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call