Multiple Objects Tracking in Surveillance System Using Big Data Analytics

Usha, M Naga (2017) Multiple Objects Tracking in Surveillance System Using Big Data Analytics. MTech thesis.

[img]PDF (Fulltext is restricted upto 31.01.2020)
Restricted to Repository staff only



Video surveillance is an active research areas in the field of computer vision. It helps in monitoring the behaviors, activities and changing information of people through CCTV cameras. It’s applications are in the field of , Military, Law enforcement and criminal investigation, Traffic congestion and pattern recognition etc. There have been many methods and algorithms developed for detecting and tracking the moving objects. A content-based image retrieval (CBIR) system helps the users (even if they are unfamiliar with the database) retrieve relevant images based on their contents. Object tracking is a challenging problem and the tracking is performed in high level applications which require location and object in every frame. Big Data Analytics is defined as process of gathering, arranging, examining the huge data sets for discovering various patterns and other important information. It helps the user to uncover the hidden patterns There were so many forms of data here we are working with video data, as we know videos grow faster in size as they are simply nothing but group of images. Video analytics algorithms is implemented to analyze video, it helps the user to search particular video when it is required. Video tracking is defined as the process of locating a moving object (or multiple objects) over time using a camera. The main objective of video tracking is to associate target objects in the consecutive video frames and the automatic detection and motion-based tracking of moving objects in a video is performed by stationary camera. Here, for detecting and tracking moving objects we use Background subtraction algorithm using Gaussian mixture models, morphological operations for removing noise and Kalman filter to predict the motion of each track. Finally by using big data analytics approach we map the images across multiple videos collected from different regions and then we analytically process on Hadoop cluster by applying single Gaussian background subtraction algorithm to MapReduce framework.

Item Type:Thesis (MTech)
Uncontrolled Keywords:Object Tracking; Surveillance; Big Data; Hadoop; Video Analytics
Subjects:Engineering and Technology > Computer and Information Science > Data Mining
Engineering and Technology > Computer and Information Science > Image Processing
Divisions: Engineering and Technology > Department of Computer Science
ID Code:9076
Deposited By:Mr. Kshirod Das
Deposited On:03 May 2018 12:33
Last Modified:03 May 2018 12:33
Supervisor(s):Korra, Sathya Babu

Repository Staff Only: item control page