Smart Surveillance Analysis System Using Hadoop
Deepali Babanrao Maske, Mehzabin Shaikh, Ketki Keskar, Nikita Vikrant Chavan, Priyanka Mahangare · International journal of advance research, ideas and innovations in technology · 2017
Security and time play very important role in everyone's life. So it is required to develop a system which will make an individual feel protected. In case if any changes are detected then it can be analyzed in a minimum time. Up to some extent, environmental changes are also detected. The first task of multi-camera surveillance is Automation object detection analysis. Changes in the CCTV recording are detected using OpenCV and they are represented by using particular colors. On single machine it will take ample of time to analyze the CCTV video recording, this drawback is overcome by using Hadoop. Hadoop has a feature of centralized server processing and distributed processing. The analysis is done using algorithms. Color histogram, Mean square difference and SSIM index are the algorithms used for frame comparison. An algorithm with the accurate result is preferred. Experimental results show a reduction in analysis time.