Publications by authors named "Rosen K Rachev"

Article Synopsis
  • Machine learning has the potential to improve defect characterization in nondestructive evaluation (NDE), but lacks sufficient real defect training data.
  • A hybrid finite element and ray-based simulation approach is proposed to train a convolutional neural network (CNN) for crack characterization in inline pipe inspections.
  • The CNN significantly outperforms the traditional 6-dB drop method, showing much lower average absolute errors in predicting crack length and angle, demonstrating the effectiveness of deep learning in NDE applications.
View Article and Find Full Text PDF
Article Synopsis
  • Plane wave imaging (PWI) is used for high-resolution ultrasonic inspections, but existing methods require knowledge of component geometry for effective imaging.
  • A new technique called PWI adapted in postprocessing (PWAPP) allows for imaging without the need for prior geometric information, using captured data to reconstruct components and adapt the imaging process.
  • PWAPP demonstrated superior results in terms of signal-to-noise ratio (SNR) compared to conventional PWI across various experiments, proving effective even with fewer transmission firings.
View Article and Find Full Text PDF