Unnatural/Abnormal Behaviour Detection System
Author(s):
Arunabh Ashish, Martand Joshi, Lakshya Meena, S.H. Patil
Keywords:
Machine Learning, OpenCV, sklearn, skimage, computer vision SVM, Logistic Regression, k- nearest neighbor
Abstract
The contents of this study primarily focus on various data mining approaches that are useful in predicting unnatural/abnormal behaviors. This paper presents a novel method of utilizing observed history for detecting abnormal behaviors in surveillance applications. An unsupervised algorithm is proposed to detect abnormal behaviors and re-train itself in real-time. Motion vectors of objects are estimated using the optical flow method. This method has been evaluated under both indoor and outdoor surveillance scenarios. It demonstrates promising results that this detection procedure is able to discover abnormal behaviors and adapt to changes in the behavioral patterns incrementally. In this study, we have selected a video clip and used Online-Convert to convert it into multiple frames of 10 frames per second. This dataset is used as the training dataset whereas the video clip is used for the testing phase. We have used various classifier methods in order to improve accuracy, which is then summarized further. Support Vector Machine, k-nearest neighbour, and Logistic Regression are the methods in question. The models used have performed equally or even better than other models. This research offers a development in which fundamental prefixes such as movement, gesture, speed, neighbour density and others are used to determine any kind of abnormal behaviour. Our aim ahead is to improve the system and implement it in public sectors using various equipment and models.
Article Details
Unique Paper ID: 155952

Publication Volume & Issue: Volume 9, Issue 2

Page(s): 643 - 649
Article Preview & Download


Share This Article

Join our RMS



Conference Alert

NCSEM 2024

National Conference on Sustainable Engineering and Management - 2024

Last Date: 15th March 2024

Latest Publication

Call For Paper

Volume 11 Issue 1

Last Date for paper submitting for Latest Issue is 25 June 2024

About Us

IJIRT.org enables door in research by providing high quality research articles in open access market.

Send us any query related to your research on editor@ijirt.org

Social Media

Google Verified Reviews

Contact Details

Telephone:6351679790
Email: editor@ijirt.org
Website: ijirt.org

Policies

Important Links

Browse Rsearch papers