|
Autofocus optical imaging systems are widely used in industrial inspection, medical diagnosis, drone photography, machine vision and other fields. Image-based autofocus algorithms typically consist of two key steps: image sharpness evaluation and search strategy. In this study, we propose an advanced image-based autofocus algorithm specifically designed for industrial image measurement. In order to improve the accuracy of the clarity assessment, we improve the existing method by introducing Gaussian filtering and threshold controlled Laplace operators. This improvement effectively reduces the effects of noise and light intensity changes, improving the reliability of clarity measurements. In addition, we propose a novel distance adjustment strategy combining coarse and fine tuning as part of the search strategy. This strategy reduces the interference of local peaks, allowing the algorithm to accurately identify the best image focus. The proposed image-based autofocus algorithm has several advantages, e.g. high focusing accuracy, high repeatability, and stability under light intensity changes. These advantages make it ideal for industrial image measurement applications. To verify the effectiveness of our method, we build an experiment setup in a lab using ADLINK PCI-9114-DG board and a custom-designed lens. The results show that the sensitivity and performance of the autofocus system meet certain requirements. In summary, the image-based autofocus algorithm that we developed enables accurate and reliable focusing. It overcomes some challenges such as noise and light intensity changes to ensures optimal image focus, and improves overall image quality. The successful implementation of our autofocus system can benefit industrial inspection, medical diagnostics and many other applications.
|
