Single-Pixel THz imaging using mechanical SLM and compressive sensing

We propose a single-pixel terahertz (THz) imaging system at a frequency of 100 GHz using a mechanical spatial light modulator (SLM). In this method, the THz beam passes through the object and masks and then is received by a single-pixel THz detector. The masks donot allow the THz beam to pass through all points of the target, so for each mask, an incomplete image of the target is focused in thedetector by a lens. To recover the image with N pixels, N different measurements are required by the detector. Masks, using their unique pattern, allow some THz beams to pass through and radiate to the object and not allow someTHz beams to pass through. To create a complete image of the object, we need to masks as many as the number of pixels, which we achieved mechanically. To increase the imaging speed, we use fewer masks, so the number of measurements will be less than the number of pixels, which leads to the formation of an incomplete image per mask. In order to form a complete image of the object and compensate for the number of measurements per mask, we use a compressive sensing (CS) algorithm. Applying the CS algorithm, the object image with dimensions of 10 cm and a resolution of 20 × 20 pixels is reconstructed. The reconstruction results of the object image indicate the superiority of TV norm method compared to L1-magic and SL0 methods in terms of RMSE criteria.