Point Spread Function Estimation for Neutron Images

The neutron images produced by small power reactor like TRIGA MARK II PUSPATI research reactor (RTP) are inherently blurred due to low length-to-diameter ratio (L/D) at the neutron radiography facility. Hence, restoration is commonly undertaken in order to improve the visual quality of these image. However, point spread function (PSF) which is responsible for blurring is usually unknown, thus making the restoration a difficult process. Assuming the target is radially symmetrical for a given size and shape, the PSF can still be estimated with acceptable degree of accuracy. In this paper, an edge spread function (ESF) is used for PSF estimation, while the restoration is achieved by means of non-blind Richardson-Lucy (RL) deconvolution. Experiments show that the algorithm improves the overall quality of the image both visually and in terms of peak noise ratio (PSNR).