Algorithm for Recovery of Image Quality with Defocusing or Bluring a Technical Cause on Video Surveillance

Abstract

Processing of video data on security threats based on video surveillance, development and improvement of methods and algorithms for obtaining evidence-based information are always relevant. This article discusses the problem of improving and restoring the image quality of frames in video data due to technical reasons of additive and multiplicative noise, including "defocused" or "smeared" defects. An improvement of the random inverse convolution algorithm to reduce "smeared" defects by performing calculations of the Fouret spectra with the forward and inverse convolution operation using a tunable inverse filter is investigated.