Extraction of the line drawing of 3-dimensional objects by sequential illumination from several directions

摘要

The conversion of a three-dimensional physical scene to a line-drawing is made more reliable by using information obtained by illuminating the scene, sequentially, from several different directions. By applying two-dimensional logical operations to the separately obtained pictures, one can eliminate many of the bad effects of poor edge contrast (due to too small difference of illumination of two planes) and of shadows. The method gives better results on scenes containing many overlapping polyhedra than are obtained by conventional one-picture edge-finding methods. The basic operations are OR, which extracts lines contained in either drawing and AND, which finds lines contained in both.The project is part of the programming for the ETL-ROBOT, an intelligent hand-eye automaton being developed at the Electrotechnical Laboratory.