Robust autofocusing for whole slide scanning microscopy

Autofocusing is one of the key issues in automatic microscopy, which is widely used in various biomedical applications such as bacterial and cell morphology. Existing autofocusing methods are sensitive to image content density and thus the global maximum may be drowned in noises especially when the images have sparse texture. Another drawback of the existing methods is that the focus range is always limited to a well-defined scope which is not in conformity with the fact where wide range autofocusing is required. In this paper, we propose a robust autofocusing algorithm to overcome these problems. We use three strategies to make the algorithm robust. First, a focus measure enhancement method was proposed to improve the steepness of the focus curve. Second, based on the analysis of the focus curve profile, we developed a focal plane search method with unrestricted search range which is critical for wide range autofocusing. Third, a continuous autofocusing model was established for guiding whole slide scanning. Furthermore, a fully automated whole slide scanning microscopy system has been implemented and the performance of the proposed method was demonstrated on this system. The experimental results show that the proposed method outperforms existing methods by improving both the success rate and the speed of autofocusing for various specimens with different image content densities.