Publication Type:

Conference Paper

Source:

2017 International Conference on Intelligent Computing, Instrumentation and Control Technologies, ICICICT 2017, Institute of Electrical and Electronics Engineers Inc., Volume 2018-January, p.1330-1333 (2018)

ISBN:

9781509061068

URL:

https://www.scopus.com/inward/record.uri?eid=2-s2.0-85049411743&doi=10.1109%2fICICICT1.2017.8342762&partnerID=40&md5=010dd5d255bac2e83b3f9193da2e3cc5

Keywords:

Abnormal event detections, Activity recognition, Comparative evaluations, Decision trees, Detection accuracy, Human actions, Image recognition, Intelligent computing, Random forests

Abstract:

As the crime rates in the ATMs are increasing, a security system which detects abnormal events is the need of the hour. Several classifiers such as Random Forest, SVM and KNN are used for recognizing human actions. This paper intends to compare the effectiveness of all these methods for abnormal event detection in ATMs. Feature Extraction is done by HOG technique for all three classifiers. Based on the experimental results, it has been found that Random forest, with a detection accuracy of 96.4 %, is the most effective one. © 2017 IEEE.

Notes:

cited By 0; Conference of 2017 International Conference on Intelligent Computing, Instrumentation and Control Technologies, ICICICT 2017 ; Conference Date: 6 July 2017 Through 7 July 2017; Conference Code:136091

Cite this Research Publication

V. Ashokan and Murthy, O. V. R., “Comparative evaluation of classifiers for abnormal event detection in ATMs”, in 2017 International Conference on Intelligent Computing, Instrumentation and Control Technologies, ICICICT 2017, 2018, vol. 2018-January, pp. 1330-1333.