Virtual Metrology of Visualizing Copper Microstructure Featured with Computer Vision and Artificial Neural Network

Virtual metrology of visualizing copper microstructure featured with computer vision and artificial neural network is reported. The focused ion beam images of copper microstructures are modeled. The input is the process variables of fabricating thin copper film and output is the microstructure image corresponding to the process variables. Computer vision and artificial neural network are used to generate model parameters. Images are generated using the model. Nucleation, grains, grain boundaries and twins are all visualized. The features of nucleation and grain growth are quantitatively analyzed in the images. The results of the predicted images support that the inference of microstructure images is proven. The technique is adaptive to most imaging systems and one of the exploratory applications of virtual metrology.