Autofocusing for Automated Microscopic Evaluation of Blood Smear and Pap Smear

摘要

Autofocusing is a fundamental procedure towards automated microscopic evaluation of blood smear and pap smear samples for clinical diagnosis. This paper presents comparison results of 16 focus algorithms based on a total of 8000 bright-field images from 10 blood smear and pap smear samples. A ranking methodology adapted from our previously proposed ranking system is used for thoroughly evaluating the performance of the selected 16 focus algorithms. Experimental results demonstrate that the variance algorithm provides the best overall performance, which will be selected for our future implementation of an automated microscopic system for computer-assisted blood smear and pap smear evaluation. Together with our previously reported findings, we hypothesize that the variance algorithm or the Normalized variance algorithm is the optimal focus algorithm for all non- fluorescence microscopy applications