This study examines the measurement of film thickness, curvature, and defects on the surface or inside of an optical element using a highly accurate and efficient method. This is essential to ensure their quality and performance. Existing methods are unable to simultaneously extract the three types of information: thickness, curvature, and defects. Spectral-domain optical coherence tomography (SD-OCT), a non-invasive imaging technique with imaging depths down to the millimeter scale, provides the possibility of detecting the optical element components' parameters. In this paper, we propose an error correction model for compensating delay differences in A-scan, field curvature, and aberration to improve the accuracy of system fitting measurements using SD-OCT. During data processing, we use the histogram-equalized gray stretching (IAH-GS) method to deal with strong reflections in the thin film layers inside the optics using individual A-scan averages. In addition, we propose a window threshold cutoff algorithm to accurately identify defects and boundaries in OCT images. Finally, the system is capable of rapidly detecting the thickness and curvature of film layers in optical elements with a maximum measurement depth of 4.508 mm, a diameter of 15 × 15 mm, a resolution of 5.69 microns, and a sampling rate of 70 kHz. Measurements were performed on different standard optical elements to verify the accuracy and reliability of the proposed method. To the best of our knowledge, this is the first time that thickness, curvature, and defects of an optical film have been measured simultaneously, with a thickness measurement accuracy of 1.924 µm, and with a difference between the calibrated and nominal curvature measurements consistently within 1%. We believe that this research will greatly advance the use of OCT technology in the testing of optical thin films, thereby improving productivity and product quality.
Read full abstract