Dichroism mapping at terahertz frequencies via polarization-sensitive computational imaging

An efficient terahertz (THz) single pixel computational imaging system is implemented to spatially map the dichroism of an anisotropic fiber distribution in wood. Current terahertz non-destructive evaluation systems have yet to be adopted in industry due to expensive and fragile components or long acquisition times. The demonstrated terahertz single pixel imager involves durable and inexpensive components and simple optical alignment with the ability to efficiently acquire polarization-resolved terahertz images of wood products-vital for wood grain inspection. Polarizing S-cyclic Hadamard mask patterns with dimensions of 17 × 19 are used to spatially encode the 0.29 THz imaging beam. A spruce sample is imaged for azimuthal rotations, providing a comprehensive dichroism analysis. The determined dichroism is within the accuracy of traditional terahertz time-domain spectroscopy systems while avoiding drawbacks of terahertz inspection systems.