Reducing Data Redundancy and Analysing Video Using Deep Learning

Abstract

Video surveillance has for ages been used to track protection delicate areas such as banks, department stores and highways and crowded people places and boundaries. Traditionally the video sparks are processed on the web by human operators and also the increase in many cameras from the now utilized surveillance technologies overload the storage devices with substantial quantities of data and make it rather tricky for your human operators to correctly track the videos. Inside this paper, we offer a remedy to eliminate redundant data from the surveillance procedures using a movement detection algorithm. In addition to that, we now apply thing detection about videos to form them that they can be easily processed from the human operators.