Application of object detection and tracking techniques from unmanned aerial vehicles

Aerial surveillance is very important in military and civil applications. Trespassing and illegal border infringement by unauthorized people is a huge predicament against the United States border security force and the Department of Homeland Security. To clinch protection to the citizens of the United States, various approaches involving science and techniques have been schemed through the years. The fault lies mainly in the video surveillance systems that are meant to monitor trespassers and illegal activities. There has been a tremendous increase in the installment of more and more surveillance cameras in sensitive areas such as banks, stock-markets, train stations, airports, freeways and borders. It becomes insurmountable to warranty suspicious behavior or vigilant movement monitoring by human operators for long periods of time due to the massive amount of data involved. Video feeds are usually archived for forensic purposes in the event of some apprehensive activity. In order to assist the human operators intelligent visual surveillance is being developed. The visual surveillance system requires fast and robust methods of detecting and tracking moving objects.;In this thesis, several methods for detecting and tracking objects from Unmanned Aerial Vehicles (UAV's) are investigated. The video surveillance process is accomplished using various methods, and one of the methods is mounting a camera on a watch tower or using an unmanned vehicle for border patrol purposes. Image sensors mounted on Unmanned Aerial Vehicles capture these images. It is imperative that the hardware involved in the process remains uncomplicated, and this research focuses to find a way to do so. In this research, diverse techniques to detect and track moving objects from an aerial platform are discussed. Moving objects were detected using an adaptive background subtraction technique. The detected objects were tracked using Continuously Adaptive Mean-Shift tracking, Lucas-Kanade optical flow tracking and Kalman filter based techniques. The simulation results show the efficiency of these algorithms to detect and track moving objects in the video sequences acquired by the UAV.