3-D Error Estimation Based on Binocular Stereo Vision Using Fringe Projection

Abstract

The accuracy of optical 3-D shape measuring system using fringe projection is associated with many factors and usually derived from calibrated reference objects. But such kind of result is a systematic assessment which is useless when considering typical complex surfaces where measurement errors tend to be different due to variances of measuring distance, object reflectivity and curve surface. In this paper, we present a method based on phase error and binocular stereo vision in order to estimate 3-D measurement error. The relationship between 3-D measurement error and system parameters is described at first according to classical binocular stereo vision and photogrammetric principles. An algorithm of phase error calculation is proposed which is focused on captured wavelength and the 3-D measurement error is then generated. We present results derived from practical experiments which demonstrate the acceptance of this method.