Error-based autofocus system using image feedback in a liquid-filled diaphragm lens

• Published in: Optical Engineering Vol. 48; no. 12; pp. 123602 - 123609
• Main Authors: Abdullah, Soran J, Ratnam, Mani Maran, Samad, Zahurin
• Format: Journal Article
• Language: English
• Published: 01.12.2009
• Subjects: Algorithms; autofocus; Control systems; Diaphragms; Fluid dynamics; Fluid flow; fluid lens; Fluids; Lenses; liquid-filled lens
• ISSN: 0091-3286; 1560-2303
• DOI: 10.1117/1.3275450

A liquid-filled diaphragm (LFD) lens system that uses image feedback for automatic focus control has been designed and developed. The edge slope width ( ) of the pixel intensity profile across a binary target was selected as the focus measure parameter. An algorithm was developed to achieve single- and dual-direction autofocus control. An improved autofocus method, which is based on the error between the expected value for a focused image and the value for the current (unfocused) image, was introduced to improve the performance of the system. An empirical equation of the focus measure error was used to predict the number of autofocus operation steps required to approach a near-focus region. A stepper motor was used for actuating a syringe-driven pump mechanism that injects or withdraws fluid into (or out of) the fluid lens chamber. The lens diaphragm was made of polydimethylsiloxane (PDMS) polymer that covers the fluid chamber. A monochrome CCD camera was attached to the LFD lens to capture live images of the target. The autofocus experiments carried out using the new differential error-based algorithm proved the viability of the algorithm in determining the near-focus region. A maximum reduction of time operation was also recorded to be in comparison with the normal autofocus algorithm.