Novel multiresolution approach for an adaptive structured light system

Abstract

3-D triangulation measurement systems with a fixed geometrical configuration have practical limitations that make them inappropriate for a wide variety of applications. The reason is that the ratio between the depth recovery error and the lateral extension is a constant that depends on the geometrical setup. Therefore, with a fixed triangulation setup, there is a tradeoff between field of view and depth resolution. As a consequence, measuring large areas with low depth recovery error necessitates the use of multiresolution techniques. In this work, we propose a multiresolution technique based on a camera-projector system previously calibrated and a second auxiliary camera that can move freely. The method consists of making first a measurement with a large field of view (coarse measurement). Afterwards, the geometrical configuration of the 3-D rig is changed to acquire a small field of view (fine measurement) that is referred to the original reference system and calibration parameters by means of the auxiliary camera. Using this method, a multiresolution reconstruction is possible without any optimization, registration, or recalibration process. Experimental results, which show a decrease of approximately one order of magnitude in the depth recovery error between fine and coarse measures, demonstrate the feasibility of the proposed method.