Unusual Activities Classification In Video Using Deep Learning Approach


Tay, Nian Chi (2020) Unusual Activities Classification In Video Using Deep Learning Approach. Masters thesis, Multimedia University.

Full text not available from this repository.


Intelligent visual surveillance has received an increasing research attention due to the need for greater security in public areas like airports, shopping malls, railway stations, and crowded parks. According to a survey by the British Security Industry Association, more than five million CCTV cameras have been installed around main cities in the United Kingdom. The widespread use of CCTV cameras leads to huge amount of data that need to be analysed for safety purpose. Analysis of these video footages is limited through manual inspection. This study proposes to perform intelligent detection of abnormal behaviour in video scenes. Efforts are made to design a system that can learn what a normal behaviour is, and is able to distinguish between a normal and an abnormal behaviour in a given context. Sometimes, a normal behaviour in a context may be abnormal in the others. For example, a running action is normal in a sports event but is considered abnormal in a shopping mall. The specific focuses of this research are (1) Two-person interactions, e.g. fighting, and (2) crowdbased interactions, e.g. people running in a crowd. The proposed study is useful for effective public monitoring and early prevention of incidents like accidents, crime and terrorism. The research also contributes towards new designed solutions of robust visual surveillance system for public security in smart cities.

Item Type: Thesis (Masters)
Additional Information: Call No.: QA76.87 .T39 2020
Uncontrolled Keywords: Neural networks (Computer science)
Subjects: Q Science > QA Mathematics > QA71-90 Instruments and machines > QA75-76.95 Calculating machines
Divisions: Faculty of Information Science and Technology (FIST)
Depositing User: Ms Nurul Iqtiani Ahmad
Date Deposited: 21 Sep 2020 07:12
Last Modified: 21 Sep 2020 07:23
URII: http://shdl.mmu.edu.my/id/eprint/7743


Downloads per month over past year

View ItemEdit (login required)