オリビエ・クリング,肥塚隆,秋山 健二,ダニエル・リー,小林幸雄
Relative Order Determination in Ambigus Moire Pictures
Surface Curvatures Computation in Moire Pictures
Abstract:The research performed during this internship has been concerned with two main problems with the moire range acquisition system. They are : ambiguous moire pictures interpretation and surface curvatures computation from moire 3D data.
The current moire system can only take pictures of simple objects with no great depth variations. Especially when there are occluding edges in the object, the extracted contours will give incorrect relative orders leading to a distorted or incorrect wire frame model reconstruction. An algorithm has been developed an algorithm to detect inconsistencies in moire patterns using a graph representation of the pattern. That same representation is also used to compute more efficiently the relative orders of the moire fringes. An algorithm has been implemented to correct the inconsistencies of the pattern
(such as occluding edges) : it uses a special kind of pixels called "wrong points" which
are to be found in high density on the location of pattern inconsistencies. Providing the noise is not too important, it is shown that this algorithm allows a very precise detection of the location of the occluding edge and an accurate computation of relative orders.
The last part of this report deals with the surface curvatures computation. we present a novel surface curvature computation scheme that directly computes the surface curvatures (principal curvatures, Gaussian and mean curvatures) from the equidistance contours without any explicit computations or implicit estimates of partial derivatives. We show how the special nature of the equidistance contours, specifically, the dense information of the surface curves in the 2D contour plane, turns into an advantage for the computation of the surface curvatures. The approach is based on using simple geometric construction to obtain the normal sections and using osculating circles to obtain normal curvatures. It is also general and can be extended to any dense range image data. We show in details how this computation is formulated and give an analysis on the error bounds of the computation steps showing that the method is stable. Computation results on real equidistance range contours are also shown.