Material prediction from confocal images of lasered samples
Microscopy and Microanalysis
Effective use of laser for fine-machining of material requires fine-tuning of laser-machining parameters. Based on the machining requirement and the composition of the material that needs to be ablated, proper lasering/scanning parameters must be practiced to achieve satisfactory results. Nevertheless, oftentimes, a priori accurate information about the material composition of sample of interest is not at hand and thus the material composition must be inferred during the laser-machining
... Non-trial-and-error existing methods that could be used for this purpose include energy dispersive spectroscopy (EDS) and laser induced breakdown spectroscopy (LIBS). The complexities associated with integrating such techniques with laser machining often acts as a prohibitive factor on the way of using them. Herein, we report on the development of a new technique that can predict material composition while laser machining is taking place using confocal images that have been obtained from the surface of lasered samples together with a knowledge of the lasering parameters. A multilayer fully connected neural network was trained, using a training data set, to predict the material composition of samples, within a set of unseen data, that have undergone laser machining, followed by confocal imaging. Note that, although lasering must start before material composition can be detectedwhich is also the case for LIBSthe amount of lasering that is needed for this purpose is minimal.