Real Time Abnormal Activity Recognition in Academic Environments

  IJETT-book-cover  International Journal of Engineering Trends and Technology (IJETT)          
© 2017 by IJETT Journal
Volume-47 Number-8
Year of Publication : 2017
Authors : Muthusathiya .E, Usha Kingsly Devi.K
DOI :  10.14445/22315381/IJETT-V47P271


Muthusathiya .E, Usha Kingsly Devi.K "Real Time Abnormal Activity Recognition in Academic Environments", International Journal of Engineering Trends and Technology (IJETT), V47(8),430-436 May 2017. ISSN:2231-5381. published by seventh sense research group

Abnormal activity detection plays a crucial role in an academic environment. A novel framework has been proposed for the real-time abnormal activity recognition in an academic environment. To develop an abnormal activity recognition system the work is divided into three modules: software module, embedded main board module, communication module. For effective identification of an activity, a real time video can be taken and processed using root of sum of square method in which different window size is used to record magnitude of pixel intensity. Histogram of Gradient used for performing feature extraction. The human activity is classified into two groups: normal activities and abnormal activities based on the Random forest approach. Finally, a software-based simulation work using MATLAB is performed and the results of the conducted experiments show the excellent abnormal activity recognition in an academic environment with no human intervention. When a student involves in such abnormal activity inside the academic environment it can be intimated to the management and staff with a message using GSM modem. The results reveal that the proposed system has achieved an accuracy of 93 % and it performs well than the other existing systems.


[1] Junejo IN, Xiaochun C, Foroosh H (2007) Auto configuration of a Dynamic Nonoverlapping Camera Network. Syst Man Cyber B Cybern IEEE Trans 37:803–816
[2] Duque D, Santos H, Cortez P (2007) Prediction of abnormal behaviors for intelligent video surveillance systems. In: IEEE symposium on computational intelligence and data mining, 2007. CIDM 2007, pp 362–367
[3] Chris Stauffer (2008) Adaptive background mixture models for real-time tracking
[4] Maddalena L, Petrosino A (2008) A self-organizing approach to background subtraction for visual surveillance applications. Image Process IEEE Trans 17:1168–1177
[5] Elarbi-Boudihir M, Al-Shalfan KA (2012) Intelligent Video Surveillance System Architecture for Abnormal Activity Detection. In: The international J on informatics and applications (ICIA2012), pp 102–111
[6] Ahad, M.A.R., Tan, J.K., Kim, H., Ishikawa, S.: Motion history image: its variants and applications. Mach. Vision Appl. 23(2), 255–281 (2012)
[7] Yannick Benezeth(2012) Abnormality Detection Using Low-Level Co-occurring EventsPattern Recognition Letters, Elsevier, 32 (3), pp.423-431.
[8] Yahya Chetouani(2013) Use of a Multi-Layer Perceptron Artificial Neural Network (MLP-ANN) for the Dynamic Behavior Modelling pp. 267-278.
[9] Javan RM, Levine MD (2013) An on-line, real-time learning method for detecting anomalies in videos using spatio-temporal compositions. Comput Vis Image Underst 117:1436–1452
[10] Robertas damasecivius (2016) Human Activity Recognition in AAL Environments Using Random Projections. Computational and Mathematical Methods in Medicine

Root Of Sum Of Square, Random Forest Approach, HOG , PIC controller.