Research on SUSAN Based Auto-focusing Algorithm for Optical Microscope Application

As a long-standing topic in the literature, autofocusing is a key issue for optical microscope based automated systems. In this paper, the existing auto-focusing algorithms are summarized firstly, then a new SUSAN (smallest univalue segment assimilating nucleus) based image sharpness function is proposed. The core problem of the proposed algorithm is selecting a suitable similarity function. In experiments, the proposed image sharpness function with different 8 similarity functions is analyzed and evaluated. Based on the evaluation results, a coarse/fine two grades global maximum searching strategy is designed to realize fast auto-focusing with proposed image sharpness function. At last, auto-focusing experiments verify the effectivity of the proposed algorithm