Optimization of OpenCV based spot identification method for surface plasmon resonance imaging

Abstract

In this work, we focus on the OpenCV based microarray recognition method for Surface Plasmon Resonance Imaging (SPRi), proposing the hit-ratio of global light pixels and coverage of the potential spots in a microarray as the criteria for identification evaluation in SPRi data. We optimized the design of the ellipse fitting strategy by analyzing the impact of different parameters in the method. After optimization of the parameters, the accuracy of microarray recognition was successfully increased to over 90%. This work not only contributes to reducing errors in microarray signal extraction and improving signal processing quality but also has significant implications for applying computer graphic technology in high-throughput biochemical analysis.