This thesis presents analysis of tomographic imaging from realistically simulated measurements of pulsed photothermal radiometry (PPTR). PPTR involves measurement of transient changes in mid-infrared (IR) emission from a sample surface after irradiation with a short light pulse. From a radiometric record obtained with a fast IR camera, light-induced temperature field inside the sample can be reconstructed in three dimensions (3D). Image reconstruction is performed with our new optimized implementation of the ''$\nu$-method'' minimization algorithm. Analysis of the tomographic imaging is based on calculations from simulated measurements with added pink noise, which realistically imitate an actual mid-IR detection system. Binning of the signal data points offers the possibility of reducing the computational time, as well as increasing the signal to noise ratio of the input signal. Statistical analysis of the image reconstruction quality has shown that progressive quadratic binning significantly reduces computational cost without adversely affecting the outcome. High temporal resolution of the radiometric records opens the possibility of detecting absorbers with dimensions below the resolution of the camera. Simulations suggest that superresolution imaging of absorbers with 10 micrometre diameter is possible with a fast IR camera with resolution of 30 micrometre.
|
