Robust automatic coarse registration of specular free-form surfaces

Many industrial applications of optical 3D sensors require the alignment (so-called “registration”) of different 3D data sets into one common coordinate system. This is necessary e. g. for variance analysis or to combine several partial measurements of an object into a complete 3D view. Smooth specular surfaces, however, have no prominent features like edges or corners. That makes registration difficult. We present a promising registration method based on Hough transformation. It is robust against small shape deviations, outliers, and noise, and it does not require user interaction.