Retrieval of crystal defect structures from HREM images by simulated evolution II. Experimental image evaluation

The method of structure retrieval by iterative digital image mathcing is examined with an image of a metal-ceramic interface recorded with atomic resolution microscopy. The global optimization technique of simulated evolution, described in the first part of this paper-series, is used here in two ways: first, the imaging condition is recovered by evaluating patterns of perfect crystalline material. Subsequently, the precision of locating atomic columns is examined. Random error sources as well as systematic errors from insecure knowledge of the imaging condition are included. Special attention is given to the dependency of the precision on different elements and structures.