Snake-Based SFS Recovery of Layered Manufacturing Surfaces

In this paper we study the application of snake-based shape-from-shading to surface quality inspection in Layered Manufacturing. A cross-section of a road surface is modeled using the concept of snake and recovered from its grayscale image, called process signature. An iterative SFS procedure approximates the shape of the snake-modeled surface cross-section with the one obtained from the signature. Two snake SFS techniques are introduced, featuring constant and adaptive spacing between snake elements. Surface recovery of 12 simulated Layered Manufacturing roads is recovered using both techniques, in order to evaluate the accuracy and robustness of the proposed methodology. Snake SFS method is compared to some existing techniques. It is shown that our approach outperforms these earlier techniques when processing images with specular reflectivity and self-shadowing.