Automated probe-mark analysis for advanced probe technology characterization

This article describes a software tool for automated probe-mark analysis that minimizes the chance of human subjectivity and error. Researchers have developed a Python-based software tool for automated probe-mark analysis for micro-bumps to characterize the quality of probing of large-array fine-pitch micro-bumps in the context of prebond testing of 3D-SICs. In this article, researchers compare manual and automatic measurements. The results show that the automatic approach yields comparable results with the measured offsets we have gathered by manual measurements, but at a much faster speed. The algorithms in the software tool are deterministic, such that different runs on the same data will give identical results, while the results of the manual approach were often influenced by human subjectivity. The researchers also show their automated approach to be robust in the sense that the tool handles top-view as well as tilted-view images, images with varying SEM contrast settings, and multiple probe marks.