Non-uniformity Correction Algorithm for THz Array Detectors in High-Resolution Imaging Applications

One of the major challenges in THz imaging using array detectors is the responsivity non-uniformities across different pixels. This is caused by different pixels responding differently and non-linearly to the same incident field. Unfortunately, such responsivity non-uniformities cannot be corrected by the traditional amplitude normalization method, which only accounts for linear responsivity. Therefore, we propose a pixel calibration scheme that corrects a non-linear responsivity of individual pixels. This calibration scheme is particularly useful in a multi-frame sub-pixel imaging technique, which assumes that all pixels have the same characteristics. Using this imaging technique would allow us to achieve an image pixel size smaller than a detector pixel size, in much faster acquisition time compared with most existing systems employing a single-pixel detector. We demonstrate the performance of our pixel calibration approach through THz imaging of various objects by the multi-frame sub-pixel imaging technique. For a square aluminum piece as a test sample, total variation (TV) in the THz image is reduced by at least 12% after applying the pixel calibration. Finally, we present example uses of our proposed technique in moisture distribution imaging. The proposed technique holds promises towards rapid THz high-resolution imaging in non-destructive testing (NDT) and quality control (QC) applications.