View planning for verification using machine vision

Two different but related issues regarding view planning are: (a) how should the object be best viewed (i.e. orientation with respect to the sensor); (b) how many views are required to completely verify the object? The first step in the solution process is to determine the appropriate criteria for view planning. In the method presented here, surface area, number of surfaces and pixel resolution, have been selected as the criteria. From the inspection planning point of view, optimal view sequencing is needed to inspect a required object completely. Using two parameters (the grouped visible regions and the admissibility of the best viewpoint), a greedy heuristic search algorithm with a logarithmic ratio bound is applied to find a minimal view sequence