High quality mesostructure acquisition using specularities

Abstract

We propose a technique for cheap and efficient acquisition of mesostructure normal maps from specularities, which only requires a simple LCD monitor and a digital camera. Coded illumination enables us to capture subtle surface details with only a handful of images. In addition, our method can deal with heterogeneous surfaces, and high albedo materials. We are able to recover highly detailed mesostructures, which was previously only possible with an expensive hardware setup.

Shadow multiplexing for real-time silhouette extraction

“…First of all, it often requires an expensive setup because many cameras are required to obtain a detailed model of the actor. Alternatively, cameras can sometimes be replaced by controllable light sources for reconstructing 3D shapes, as shown in miscellaneous shape acquisition methods [23,9,13,26]. Similarly, in this work we will try to replace as many cameras as possible by much cheaper light sources.…”

Gloss and Normal Map Acquisition of Mesostructures Using Gray Codes

“…The majority of these methods requires a specialized hardware setup [6][7][8][9][10], have long acquisition/processing times [11,12] or are not able to scan specular surfaces [2,3]. In our work we will use a slightly adapted version of the Gray code based approach of Francken et al [1], employing a screen-camera setup as acquistion setup. Because of the use of a planar illuminant and Gray codes, fine-scale specular surface geometry can accurately be measured using only up to 40 input images.…”

“…As in the normal map acquisition technique [1], we display stepwise refining vertical and horizontal Gray code patterns. We also display each pattern's complement in order to robustly separate diffuse from specular reflection.…”

“…In our work, we start from an existing Gray code based mesostructure acquisition system [1] and show that glossiness information can easily be extracted from the already available data necessary for shape reconstruction. The original method only has to be slightly modified by replacing the polarization based specular-diffuse separation with the use of pattern complements instead.…”

“…We therefore employ a setup consisting of inexpensive and accessible components, including a regular computer screen and a digital still camera. We extend the Gray code based normal map acquisition approach of Francken et al [1] which utilizes a similar setup. The quality of the original method is retained and without requiring any extra input data we are able to extract per pixel glossiness information.…”