Artificial Neural Networks Framework for Detection of Defects in 3D-Printed Fiber Reinforcement Composites

One of the major challenges in applying tomography methods for detecting defects composite materials is large image datasets generated during imaging, which require significant effort detection damage. Machine-learning (ML) a training dataset and can be efficient processing defect detection. Methods need to developed images train ML algorithms, focus present work. An additive manufactured fiber reinforced material imaged using micro-CT scan generate an set The microstructures are processed binarized statistical features (BSIF) method compression without compromising desired information about defects. result shows that convolutional neural network model has mean square error 0.001 orientation prediction, scheme been based on predictions obtained from models.