Automatic Compensation for Defects of Laser Reflective Patterns in Optics-Based Auto-Focusing Microscopes

Abstract

Since the optomechatronic system is characterized by a high energy density and high response speed, system performance can be substantially improved by integrating optical components with electromechanical systems. For micro-processing, conventional machining has been commonly replaced with laser processing. Auto-focusing plays a crucial role in processing, so developing an autofocus (AF) system with high precision and speed is vital. AF modules integrated with the knife edge method typically use a reflective semicircular laser image to calculate the defocus distance for the optical instrument and evaluate the accuracy. On the stage of an optical instrument, a reflective image is typically not a standard semicircular segment because of the different reflection characteristics of objects or the positioning error of the devices. In this study, a novel method is proposed to automatically compensate for these defects and errors. In addition to the compensation part, the improved adaptive anisotropic diffusion filter and wavelet-based edge detection are both proposed for pattern smoothing and enhancement, and pattern edge detection. For reflective patterns obtained using an image sensor in an optics-based AF module, the proposed method, compared with six state-of-the-art approaches in terms of several metrics, such as compensation accuracy, sensitivity, specificity, Dice coefficient, mean absolute distance, Hausdorff distance, and speed / computation time, enables and facilitates accurately reconstructing the complete semicircular segment. Experimental results show that the proposed method yields more satisfactory, stable and accurate results in comparison with six other approaches in performance analysis and statistical evaluation.