Certificate from Supervisor

Abandoned object detection is one of the most practically useful areas of computer vision due to its application in automated video surveillance systems for the detection of suspicious activities that might endanger public safety, especially in crowded places like airports, railway stations, shopping malls, movie theatres and the like. An abandoned object is defined as one that has been lying stationary at a certain place with no apparent human attendance for an extended period of time. Such objects are usually inconspicuous commonplace objects that people often carry around including backpacks, suitcases and boxes. Detection of abandoned objects is of prime importance in uncovering and forestalling terrorist activities since it is a reasonable supposition that an abandoned object, if left behind on purpose, may be hiding dangerous items like explosives. The present work is an attempt to create a flexible and modular framework that can be used to experiment with several different methods for each stage of the overall task of detecting abandoned and removed objects in a video stream. Several existing methods have been implemented for each of these stages and integrated into the system in a way that makes it possible to switch between them in real time. This enables the user to observe and compare the performance of different methods of solving the same problem and choose the one most suited to any given scenario. The system has also been designed to allow new methods to be added for any stage of the system with minimum programming effort.