Image Restoration for Blurred License Plates Extracted from Traffic Video Surveillance using Lucy Richardson Algorithm
Ronnier Franz H. Torres, Ramon G. Garcia · 2022
Vehicle related accidents happens at urban areas, this includes, national roads, municipal roads, provincial roads and crowded boulevard streets and most of the captured vehicle license plate by the CCTV that is involve in a hit and run incident was blurred and not clear enough to identify and trace the perpetrators. Several studies have presented to restore a blurred image by performing different deconvolution techniques. Most of this researches focuses only on removing the blur effect on a non-moving object and mostly in a situation where there is an enough light in the surrounding and the blur parameter was either added or setup intentionally or user-defined. In this paper, the study aims to restore a blurred image specifically on a license plate of a moving test vehicle using Lucy-Richardson Algorithm from a real video footage of a surveillance camera subjected in different light condition. In addition to that, the vehicle will also be subjected to different speeds to know its effect on its restoration. The study will significantly test whether the deconvolution algorithm can restore a real motion blurred license plate image exposed at different lighting conditions test in different car speeds.