Normalized Cross Correlation based Solution for Automated Optical Inspection during Boundary Scan Testing

This paper proposes a Normalized Cross Correlation (NCC) based solution for integrating automated visual inspection with Boundary Scan Test (BST, IEEE Std. 1149.1). The proposed approach can serve as a guide when implementing an industrial BST station and it emerged as a necessity in the context of proliferation of such tests. BST procedures may include activation/deactivation of signaling LEDs mounted on the printed circuit boards (PCBs). Successful testing of such elements often relies on human visual inspection and manual confirmation, for example by key/button pressing. The aim of our work is to provide a low-cost alternative. By using XJTAG tools, NI LabVIEW, and MATLAB, we demonstrate the practical implications of visual inspection by means of NCC in an automated test scenario. Finally, a report on different integration strategies applied with the purpose of maximizing Device Under Test (DUT) coverage and reducing test design effort and costs, is discussed.