Thick pinhole imaging system is an effective tool for diagnosing intense pulsed radiation sources, but its spatial resolution decreases due to limitations in field of view and energy flux. To address this limitation, image super-resolution reconstruction algorithms are commonly employed to enhance the system's spatial resolution. However, existing algorithms face two main issues: incomplete or inaccurate point spread function (PSF) of the imaging system and susceptibility of image reconstruction algorithms to noise. This paper extensively studies the model and parameters of the neutron thick pinhole imaging system and proposes a PSF calculation method based on the characteristics of each component. Additionally, we present a novel image super-resolution reconstruction algorithm for the neutron thick pinhole imaging system, utilizing image denoising priors based on the half-quadratic splitting method. The results indicate that, compared to the traditional Richardson-Lucy algorithm, the reconstructed images generated by this algorithm exhibit fewer artifacts and sharper boundaries. Furthermore, this algorithm can surpass the spatial resolution limitations imposed by the imaging system.

Thick pinhole imaging,Single image super-resolution,Image denoiser prior,Half quadratic splitting