Paper Key : IRJ************229
Author: R. Jagadish Vijay
Date Published: 03 Mar 2023
Abstract
In this paper, a method for quick and concurrent video processing on Map Reduce-based clusters like Apache Hadoop is proposed. The technique can handle massive amounts of video data and processing time can be drastically decreased by using clusters. The mechanism for transferring conventional video processing algorithms created for a single computer to the suggested system is revealed, along with other technical aspects of doing video analysis on clusters. Algorithms for face detection, motion detection, and tracking have been put into practice on clusters as case studies. Performance tests on a six-machine Apache Hadoop cluster demonstrate that the system is able to lower the two implemented algorithmsoperating times to under 25% of those of a single computer. The applications of the system include smart city video surveillance, services provided by video sites and satellite image processing
DOI LINK : 10.56726/IRJMETS33990 https://www.doi.org/10.56726/IRJMETS33990