Version 1
: Received: 1 June 2023 / Approved: 2 June 2023 / Online: 2 June 2023 (07:27:50 CEST)
How to cite:
Kim, H.; Shin, J.; Park, S.; Paik, J. Intelligent Video Surveillance System with Abnormal Behavior Recognition and Metadata Retrieval. Preprints2023, 2023060147. https://doi.org/10.20944/preprints202306.0147.v1
Kim, H.; Shin, J.; Park, S.; Paik, J. Intelligent Video Surveillance System with Abnormal Behavior Recognition and Metadata Retrieval. Preprints 2023, 2023060147. https://doi.org/10.20944/preprints202306.0147.v1
Kim, H.; Shin, J.; Park, S.; Paik, J. Intelligent Video Surveillance System with Abnormal Behavior Recognition and Metadata Retrieval. Preprints2023, 2023060147. https://doi.org/10.20944/preprints202306.0147.v1
APA Style
Kim, H., Shin, J., Park, S., & Paik, J. (2023). Intelligent Video Surveillance System with Abnormal Behavior Recognition and Metadata Retrieval. Preprints. https://doi.org/10.20944/preprints202306.0147.v1
Chicago/Turabian Style
Kim, H., Seokmok Park and Joonki Paik. 2023 "Intelligent Video Surveillance System with Abnormal Behavior Recognition and Metadata Retrieval" Preprints. https://doi.org/10.20944/preprints202306.0147.v1
Abstract
Huge-scale video surveillance systems have become essential in crime prevention and situation recording. Traditional surveillance systems relied on human monitoring of video streams, which often led to errors and difficulties in understanding events. Furthermore, locating specific scenes within recorded videos required extensive human investigation. To overcome these challenges of inefficiency, inconvenience, and potential risks, we propose an intelligent analysis scheme that utilizes abnormal behavior recognition and metadata retrieval algorithms to replace human monitoring. The proposed method consists of three stages: i) metadata generation through object detection and tracking, ii) abnormal behavior recognition, and iii) SQL-based metadata retrieval. By incorporating specific information such as object color and aspect ratio, our technique enhances the usability of retrieval. Moreover, our abnormal behavior recognition module demonstrates robust classification capabilities for activities such as pushing, violence, falling, and crossing barriers. The proposed method can be seamlessly deployed on both edge cameras and analysis servers, making it adaptable to various surveillance setups. This approach revolutionizes the traditional surveillance paradigm, enabling more efficient, reliable, and secure video monitoring and analysis.
Keywords
Metadata generation; abnormal recognition; metadata retrieval; intelligent surveillance system
Subject
Engineering, Electrical and Electronic Engineering
Copyright:
This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The commenter has declared there is no conflict of interests.
Comment:
The paper proposes a technique to extract and store the necessary features from objects for surveillance systems, as well as to retrieve the stored information.The features detected from objects include positional information, color information of the objects, and class of object behavior.However, each of these features utilizes methods that have been developed quite a long time ago.The object tracking method used to detect the positional information of objects, called distractor-aware tracker (DATs), was proposed in 2015.The method used to detect the color information of objects, probabilistic latent semantic analysis (PLSA), was proposed in 2009.The method used for classifying object behavior appears to mimic the Two-stream Inflated 3D ConvNets (I3D) approach, which was also proposed in 2018.The methods used to detect the features of object in the paper are quite outdated, and there don't seem to be any improvements suggested.There is also no trace of contribution or consideration for new methods in terms of the storage and retrieval techniques used for feature detection.Significant improvements are needed in the method of detecting object features.Or The author seems to be in need of contemplating data storage and retrieval methods.
The commenter has declared there is no conflict of interests.
Comment:
In this paper, the authors claim to have developed a system to save and retrieve events from surveillance cameras.
In order to save the event, the position, color, and behavior information of the object was used.
To detect these features, the authors used previously developed distractor-aware trackers (DATs), probabilistic latent semantic analysis (PLSA), and a Two-stream Inflated 3D ConvNets (I3D) approach.
Training a model developed by someone else on a different data set is not a new method.
I think that what seems to be the author's contribution in this paper is simply the ability to save event and retrieve saved event information.
And these function is nothing new, already in use by many companies.
Commenter:
The commenter has declared there is no conflict of interests.
Commenter:
The commenter has declared there is no conflict of interests.
In order to save the event, the position, color, and behavior information of the object was used.
To detect these features, the authors used previously developed distractor-aware trackers (DATs), probabilistic latent semantic analysis (PLSA), and a Two-stream Inflated 3D ConvNets (I3D) approach.
Training a model developed by someone else on a different data set is not a new method.
I think that what seems to be the author's contribution in this paper is simply the ability to save event and retrieve saved event information.
And these function is nothing new, already in use by many companies.
What is the novelty in this paper?