Description
Long-wavelength optical measurement has attracted significant attention in non-destructive inspection as a non-contact, non-invasive method. Research utilizing carbon nanotube (CNT) film sensors combines their ultra-wide bandwidth with high optical absorbance. This unique combination enables observers to evaluate optical properties across a range of wavelengths. Consequently, the evaluation process facilitates the precise identification of target shapes and constitutive materials. Previous studies using CNT film devices primarily relied on conventional transmissive optical systems. Those systems typically detect transmitted light during the direct irradiation of the target object. This transmissive approach fails to distinguish between light absorption and surface reflection. This limitation significantly restricts the variety of identifiable materials during the measurement process. Therefore, this study presents advanced non-destructive measurements utilizing a reflective optical system. This reflective setup simultaneously achieves accurate material identification and precise shape reconstruction. For shape reconstruction, the proposed system employs a contour reconstruction method utilizing surface angle estimation. This analytical method geometrically calculates the precise surface angle of the target object. The calculation utilizes the device-to-subject distance and the specific coordinates of the light-receiving elements. The estimation process then correlates this calculated surface angle with the corresponding scanning distance. The system records this scanning distance exactly at the response peak of each sensor element. This crucial correlation allows the analytical system to determine lengths and estimate the overall shape. Furthermore, the photodetector responses clearly distinguish the materials on the surface of the subject. This distinction effectively separates reflective materials from light-absorbing materials covering the target object. The experimental results demonstrate accurate shape estimation for a metal cylinder under an opaque obstruction. The measurement system achieved this precise shape estimation with a radius error strictly below 5%. Finally, this approach successfully reconstructs shapes and identifies materials. The validation process utilizes multiple light sources in the near-, mid-, and far-infrared regions.
| I am the presenting author | Yes |
|---|