AUTOMATED HUMAN DETECTION AND TRACKING FOR SURVEILLANCE APPLICATIONS

Abstract
Safeguarding a country’s border is very vital to its nation. A nation’s security is determined by the kind of security imposed at the border. Proper surveillance system needs to be imposed at border because of numerous reasons which include drug smuggling, illegal immigrants crossing the borders and last but not the least, terrorist intrusion. To protect our border from these activities we need to monitor, filter and detect if there is any motion or any kind of activity that might lead to intrusion in our borders. For this we need 24×7, surveillance and hence an automated surveillance system needs to be developed. The existing systems are not efficient enough to detect threats if the intruders are using some kind of camouflage or cover. To overcome these drawbacks of the existing systems we intend to use thermal imaging cameras. The focus is to use thermal imaging cameras to capture the images for detection and tracking of the human/object, so as to analyze and predict their behaviour and intentions. The use of thermal imaging unlike the optical camera will help the system to perform detection of suspects which are hidden behind an obstacle for e.g. a bush. The thermal imaging will perform the detection at night with the same efficiency as in the daytime. The surveillance system is designed in order to achieve the objective of safeguarding the border, thus satisfying the aim of detecting threats and minimizing false alarms.

Authors
Avaneesh Tripathi, Milind Pathak, Amritha Sharma, Parth Vijay, S A Jain
MIT Academy of Engineering, India

Keywords
Image Processing, Thermal Imagers, Machine Learning, Border Surveillance, Neural Networks
Published By :
ICTACT
Published In :
ICTACT Journal on Image and Video Processing
( Volume: 9 , Issue: 4 )
Date of Publication :
May 2019

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.