Blind-Deconvolution-Based Terahertz Image Restoration Algorithm

Terahertz waves, with their unique penetrability and safety, hold significant potential in the field of non-destructive testing. However, the imaging process is susceptible to diffraction limits, noise, and system distortions, leading to degraded image quality. Traditional image restoration methods rely on known degradation models, which are difficult to apply to complex realworld degradation scenarios. This paper introduces blind image restoration technology, which simultaneously estimates the degradation model and the original image, effectively enhancing image resolution and fidelity. Experimental results indicate that this algorithm outperforms traditional methods in terms of noise suppression and image clarity, providing a higher-quality image restoration solution for terahertz non-destructive testing.