Suspicious Activity Detection and Tracking through Unmanned Aerial Vehicle Using Deep Learning Techniques
Madala Gayathri · International Journal of Advanced Trends in Computer Science and Engineering · 2020
This paper focuses on crime detection that is extensively used in real world visual monitoring applications like video surveillance.Crime prevention is widely implemented in some countries, through police force and in many cases, private policing methods such as private security and home defense.Implementation of such private and public security could not completely eradicate such awful crimes.Therefore, there's a requirement for credible and effective surveillance in order to diminish the brutal and skeevy crimes.The reason for occurrence of many offences is late communication and unestablishment of authentic security or surveillance system .It's tragic that more than 20000 people lost their lives in Hit-and-run cases in India alone.Many notorious criminals who deserve dreadful punishment for their recklessness fly the coop due to the lack of evidence.Many other instances can be drawn out where there's a requirement for an organized system of surveillance like ATM's, shopping malls, etc.To stop such awful offences there's a need for genuine security system.The idea of an Intelligent Unmanned Aerial Vehicle (UAV) proposed in this paper, is inspired from various sources [1], that is capable of monitoring its surroundings constantly for suspicious or illegal activities.Whenever it encounters a suspicious activity, it automatically captures and processes the scene, sends out an alert to the administrator and waits for his/her command.The main intention is to constantly monitor the surroundings using a bird shaped surveillance device, with very minimal human intervention.