IoT based smart surveillance monitoring by using model-based human action recognition design

In recent years, with the rapid development in IP camera usage, massive video surveillance data are produced at an unprecedented speed. IP camera is one of the most important elements in In Internet of things (IoT) and smart cities, in this era. IoT Based Smart Surveillance provides resources over t...

Full description

Saved in:

Bibliographic Details
Published in:	2021 5th International Conference on Internet of Things and Applications (IoT) pp. 1 - 6
Main Authors:	Ebrahimy, Amir Reza, NaghshNilchi, Ahmad Reza, Monadjemi, Amir Hassan, SaeidEhsani, Mohmmad
Format:	Conference Proceeding
Language:	English
Published:	IEEE 19-05-2021
Subjects:	Annotations Cameras Computational modeling Feature extraction Gaussian mixture model Human Action Recognition Manuals Smart cities Smart Surveillance System Unsupervised Feature Extraction Video surveillance
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	In recent years, with the rapid development in IP camera usage, massive video surveillance data are produced at an unprecedented speed. IP camera is one of the most important elements in In Internet of things (IoT) and smart cities, in this era. IoT Based Smart Surveillance provides resources over the Internet and allows a plethora of applications to be deployed to provide services for different applications. Traditional solutions to deal with the big video data would require a large amount of computing and storage resources. one of the major bottleneck being faced the computational cost of spatio-temporal deep neural networks making them run as fast as their 2D counterparts while preserving accuracy on video recognition benchmarks. To this end, we present the new model-based human action recognition design. In this method we use fixed dimensional representation for actions in video clips of varying lengths to decrees computational complexes. This representation, simplify clustering and comparing between videos. In addition, the methods based on CNN usually need manual annotation at the beginning and the meaning of characteristics is an abstract wonder and subjective phenomenon and hence a manual annotation of attributes is highly inconsistent [1]. Anyhow our proposed smart surveillance system is unsupervised and don't need inconsistent manual annotation of video clips attributes. Many researchers believe smart surveillance system is a critical component of smart cities and can be a solid foundation for future applications in intelligent surveillance systems focuses on human action Recognition.
DOI:	10.1109/IoT52625.2021.9469601