Parallel Thermal Wave Imaging Using a Vector Lock-In Video Technique

Abstract

The appearance of the IR video camera has extended the wavelength range of the visible video camera to the thermal IR range (3–12 µm), thus providing a powerful tool to researchers in thermal wave imaging. However, imaging in the thermal IR range has its special handicaps, not shared by its visible counterpart. Most objects in conventional photography reflect rather than emit light of their own. As a result one often has the freedom to choose the intensity, direction and color of illumination to accentuate the aspects of the object to be photographed. In thermal IR imaging the situation is very different in that nearly all objects emit thermal radiation of their own, in addition to reflecting radiation of other objects. What is recorded in a thermograph is always a mixture of emitted and reflected radiation, some of which even comes from components of the camera itself, including lenses and their supporting structures. This problem is particularly severe in the 8–12 µm range, because it corresponds to the peak of blackbody radiation at room temperature. It is this same range of wavelength that is most relevent in non-destructive evaluation. In conventional scanned thermal wave imaging applications this problem is overcome by the use of a lock-in analyzer synchronized to the source of the thermal wave. Without the lock-in technique, the IR video camera is capable of observing only very slow thermal phenomena[1], despite the fact that the intrinsic band width of the camera is very broad. This limitation offsets the main advantage of the IR video camera, namely its high data-acquisition rate. In this paper we report on instrumentation development which combines the lock-in technique with the IR video camera. With this technique the information of each pixel of an image is handled in the manner of a lock-in analyzer, while the object is illuminated (i.e., heated) or stimulated (e.g., joule heating) with a signal which is synchronous with the reference signal of the lock-in detection. This way the unsynchronous background radiation is rejected and the signal-to-noise ratio is enhanced.