Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/375240
Title: | Semantic Video Interpretation For Surveillance Using Machine Learning Technique |
Researcher: | SAXENA, PARUL |
Guide(s): | JADON, R. S. |
Keywords: | Computer Science Computer Science Software Engineering Engineering and Technology |
University: | Rajiv Gandhi Proudyogiki Vishwavidyalaya |
Completed Date: | 2020 |
Abstract: | Video Surveillance refers to an automated monitoring process that involves data newlineacquisition, analysis, and interpretation for understanding objects and their behavior. newlineIt addresses real-time observation of people and vehicles within a busy environment, newlineleading to a description of their actions and interactions. Automated surveillance newlinesystems are mostly used for military, law enforcement, and commercial applications. newlineThe technical issues include moving object detection and tracking, object newlineclassification, crowd monitoring, human motion analysis, and activity understanding, newlinetouching on many of the core topics of computer vision, pattern analysis and artificial newlineintelligence. Sensors of different types and characteristics in surface-based or aerialbased newlineplatforms are used for the acquisition of data of large areas sometimes covering newlineseveral square miles. Our aim in this thesis has been to develop a vision based newlinesurveillance system using semantic interpretation of objects and scenes in the video. newlineWe have assumed stationary camera, recording contiguous frames. Most of the newlineobjects are either stationary of having uniform motion. We have developed soft newlinecomputing techniques for charactering the objects like people and vehicle in the scene newlineand further to detect anomalies and abnormalities in terms of these semantic objects. newlineIn the current years such automated surveillance is getting significant as more and newlinemore surveillance video is made available due to enhancements in recording and newlinestorage technologies. We use morphological erosion and dilation method to reduce the newlinenoises, then detects objects from surveillance video using background subtraction, newlinespatio-temporal frame differencing and optical flow computation. The interpretation newlineof these low level features is done using Neural network based systems. newlineCompared with the traditional camera surveillance, intelligent monitoring analysis newlinetechnology require relatively low cost and price of hardware. |
Pagination: | 10.MB |
URI: | http://hdl.handle.net/10603/375240 |
Appears in Departments: | Department of Computer Applications |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf | Attached File | 156.83 kB | Adobe PDF | View/Open |
02_declaration.pdf | 160.63 kB | Adobe PDF | View/Open | |
03_certificate.pdf | 214.12 kB | Adobe PDF | View/Open | |
04_acknowledgement.pdf | 64.43 kB | Adobe PDF | View/Open | |
05_contents.pdf | 86.52 kB | Adobe PDF | View/Open | |
06_list of graph and table.pdf | 93.24 kB | Adobe PDF | View/Open | |
07_chapter 1.pdf | 222.67 kB | Adobe PDF | View/Open | |
08_chapter 2.pdf | 217.99 kB | Adobe PDF | View/Open | |
09_chapter 3.pdf | 395.3 kB | Adobe PDF | View/Open | |
10_chapter 4.pdf | 479.83 kB | Adobe PDF | View/Open | |
10_chapter 5 a.pdf | 5.09 MB | Adobe PDF | View/Open | |
10_chapter 6 b.pdf | 2.68 MB | Adobe PDF | View/Open | |
10_chapter 7 c.pdf | 64.6 kB | Adobe PDF | View/Open | |
11_references.pdf | 124.12 kB | Adobe PDF | View/Open | |
80_recommendation.pdf | 44.68 kB | Adobe PDF | View/Open | |
abstract.pdf | 44.68 kB | Adobe PDF | View/Open | |
preliminary page.pdf | 156.83 kB | Adobe PDF | View/Open |
Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
Altmetric Badge: