Improvement of Subaperture Phase-Stitching Algorithm by Autofocus

碩士 === 國立中央大學 === 照明與顯示科技研究所 === 106 === With the development of technology, optical components such as mirrors, various types of lenses, prisms, beam-splitters, etc. are widely used in various products and equipment, and the quality of optical components directly affects the quality of an optical s...

Full description

Bibliographic Details
Main Authors: Wen-Ching Wang, 王文慶
Other Authors: Chien-Cheng Kuo
Format: Others
Language:zh-TW
Published: 2018
Online Access:http://ndltd.ncl.edu.tw/handle/4tce48
Description
Summary:碩士 === 國立中央大學 === 照明與顯示科技研究所 === 106 === With the development of technology, optical components such as mirrors, various types of lenses, prisms, beam-splitters, etc. are widely used in various products and equipment, and the quality of optical components directly affects the quality of an optical system. Therefore, the technology for testing optical components is quite important. In this paper, scanning white light interference technology is used to scan and record the intensity of the interference signal, and the interference intensity is converted into phase information through mathematical calculation. Then, the subaperture phase stitching technology is used to optical. The surface of the component is restored. In the process, in order to measure the phase information of multiple regions, the position of the component to be tested needs to be moved, and this step will cause the interference signal to disappear, which requires manual adjustment, but the adjustment process cannot be accurate. The adjustment may cause errors or even loss of phase information restored after scanning. Therefore, this study introduces a passive autofocus algorithm, based on the mathematical operation of edge detection, with the electronic platform of vertical movement, records the image intensity, calculates the focus value, and automatically found. After comparing the data of the lens LA-1708, the curvature radius error calculated by manual focus is 2.79%, and the error of the radius of curvature calculated by autofocus reduced to 1.45%, which can effectively reduce the error caused by artificial adjustment.