Modified algorithm for defocused image restoration

This study develops a modified inverse filter algorithm that corrects defocus aberrations through optimization of the Zernike defocus coefficient. By strategically adjusting the Zernike-based point-spread function (PSF), our method achieves precise compensation for small defocus in optical systems. Through simulations and actual tests using a custom-designed infrared lens, we demonstrate that this approach achieves a 2.25-fold improvement in contrast ratio at 25 lp/mm compared to the uncorrected original images, with the improvement being more pronounced in high-spatial-frequency regions. Additionally, our method significantly reduces computation time compared to conventional iterative algorithms, such as blind deconvolution, making it highly efficient for practical applications. Future work will focus on further optimizing this method for more general cases and different optical systems.