Computer assisted microscopy image analysis for high throughput statistical analysis and 3D structure determination in nanotechnology

As nano-micro hybrid structures start to approach biological complexity, two problems demand computational assistance: Determination of three dimensional (3D) structure suffers from ambiguity due to the 2D nature of images, and insight into the complex growth mechanisms of the structures as well as optimization of emergent properties in their multidimensional parameter spaces, both need accurate statistics, for example in order to obtain the precise shape of size distributions. Computer image recognition is now cheap, simple to customize, and can integrate with on the spot simulation and statistical analysis. We introduce these issues generally (guiding philosophy) and with examples from scanning and tunneling electron microscopy (SEM/TEM), pedagogically showing how relatively little effort can result in crucial progress in diverse fields such as connectivity analysis of particle networks for sensors, nanowire strength analysis, metallic thin film twinning dislocation analysis, and precise determination of shapes in the transformation of nano shapes during calcination.