Applicable Method to Specular Surface for Recovering Sign of Local Gaussian Curvature

Using more than three shading images gives more This paper proposes a new method to recover the sign of local Gaussian curvature from multiple shading images (more than three). The required information to recover the sign of Gaussian curvature is obtained by applying Principal Components Analysis (PCA) to the normalized irradiance measurements. The sign of the Gaussian curvature is recovered based on t i e relative orientation of measurements obtained on a local five point test pattern to those in the 2-D subspace, called the eigen plane. Using multiple shading images gives more correct and robust result and minimizes the effect of shadows by allowing a larger area of visible surface to be analyzed in comparison with the methods using the three shading images. Furthermore, it makes this method be applicable to the specular surface object where it is impossible for the previous method to recover the sign of the Gaussian curvature. On the other hand, since PCA removes a high degree of correlation between each image, this method can keep result high quality even when the light source directions are not widely dispersed.