Correction of thermal airflow distortion in warpage measurements of microelectronic packaging structures via deep learning-based digital image correlation.

ABSTRACT: The projected speckle-based three-dimensional digital image correlation method (3D-DIC) is being increasingly used in the reliability measurement of microelectronic packaging structures because of its noninvasive nature, high precision, and low cost. However, during the measurement of the thermal reliability of packaging structures, the thermal airflow generated by heating introduces distortions in the images captured by the DIC measurement system, impacting the accuracy and reliability of noncontact measurements. To address this challenge, a thermal airflow distortion correction model based on the transformer attention mechanism is proposed specifically for the measurement of thermal warpage in microelectronic packaging structures. This model avoids the oversmoothing issue associated with

SUBMITTER: Gao Y 

PROVIDER: S-EPMC11347651 | biostudies-literature | 2024 Aug

REPOSITORIES: biostudies-literature