SVAS: Surveillance Video Analysis System

Kardas, Karani; Cicekli, Nihan Kesim

doi:10.1016/j.eswa.2017.07.051

Cited by 31 publications

(13 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the proposed method works at microscopic level. An Interval-Based Spatio-Temporal Model (IBSTM) have been proposed in (Kardas and Cicekli, 2017) in order to detect untoward events in a video. However, the proposed method is a microscopic event model that cannot deal with macroscopic events such as crowd flow.…”

Section: Computer Vision-based Methodsmentioning

confidence: 99%

Estimation of linear motion in dense crowd videos using Langevin model

Behera

Dogra

Bandyopadhyay

et al. 2020

Expert Systems with Applications

View full text Add to dashboard Cite

Crowd gatherings at social and cultural events are increasing in leaps and bounds with the increase in population. Surveillance through computer vision and expert decision making systems can help to understand the crowd phenomena at large gatherings. Understanding crowd phenomena can be helpful in early identification of unwanted incidents and their prevention. Motion flow is one of the important crowd phenomena that can be instrumental in describing the crowd behavior. Flows can be useful in understanding instabilities in the crowd. However, extracting motion flows is a challenging task due to randomness in crowd movement and limitations of the sensing device. Moreover, low-level features as compared to existing popular crowd segmentation methods. The evaluation of the proposed model has been carried out on publicly available datasets as well as using our dataset. It has been observed that the proposed method is able to estimate and segment the linear flows in the dense crowd with better accuracy as compared to state-of-the-art techniques with substantial decrease in the computational overhead.

show abstract

Section: Computer Vision-based Methodsmentioning

confidence: 99%

Estimation of linear motion in dense crowd videos using Langevin model

Behera

Dogra

Bandyopadhyay

et al. 2020

Expert Systems with Applications

View full text Add to dashboard Cite

show abstract

“…It is capable of describing patterns of knowledge via specific representations including Elements of Context Representation (ECR), Action representation (AR), and General Descriptors of Context (GCD) with corresponding reasoning rules. However, existing approaches [4], [16], [17], both knowledge-based and rule-based schemes through semantic representations, require heavy manual modeling work.…”

Section: B Related Work On Event Analysismentioning

confidence: 99%

Multi-Event Modeling and Recognition Using Extended Petri Nets

Qiu

Wang

et al. 2020

IEEE Access

View full text Add to dashboard Cite

This paper addresses the event modeling and recognition problem in video surveillance systems using the system net on the Petri Net (PN) formalism. The single event on the foundation of prior knowledge is first modeled via arranging event schemes from the design document. Then, finite sequential runs (FSRs) determined by semantic features in training clips drive elements of the proposed single event model. Finally, the multi-event model is built automatically from single event models via a proposed integration method using multi-level features (including high-level semantic features and low-level features like numerical characters of individual trajectories). We provide a novel solution to event conflict issue through an extended system net where the resultant event type is determined by the decision tree technique. The comparison between the proposed methodology and other approaches in the literature is reported via experiments on an acknowledged public dataset. INDEX TERMS Event recognition, high-level video event modeling, video surveillance system, Petri nets, video analysis application.

show abstract

“…Nowadays, surveillance systems contribute vitally to public security. The development of artificial intelligence, especially artificial intelligence for computer vision [1], has made it easier to analyze the resulting videos [2,3]. Several studies have recently addressed the problem of event detection in video surveillance [4] which requires the ability to identify and localize specified spatiotemporal patterns.…”

Section: Introductionmentioning

confidence: 99%

“…(1) We introduce a race dataset of Vietnamese people collected from a social network and published for academic use. (2) We propose an efficient framework including three modules for information collection (IC), face detection and preprocessing (FD&P), and RR. (3) For the RR module, we propose two independent models: an RR model using a CNN (RR-CNN) and a fine-tuning model based on VGG (RR-VGG).…”

Section: Introductionmentioning

confidence: 99%

Race Recognition Using Deep Convolutional Neural Networks

Nguyen

2018

Symmetry

View full text Add to dashboard Cite

Race recognition (RR), which has many applications such as in surveillance systems, image/video understanding, analysis, etc., is a difficult problem to solve completely. To contribute towards solving that problem, this article investigates using a deep learning model. An efficient Race Recognition Framework (RRF) is proposed that includes information collector (IC), face detection and preprocessing (FD&P), and RR modules. For the RR module, this study proposes two independent models. The first model is RR using a deep convolutional neural network (CNN) (the RR-CNN model). The second model (the RR-VGG model) is a fine-tuning model for RR based on VGG, the famous trained model for object recognition. In order to examine the performance of our proposed framework, we perform an experiment on our dataset named VNFaces, composed specifically of images collected from Facebook pages of Vietnamese people, to compare the accuracy between RR-CNN and RR-VGG. The experimental results show that for the VNFaces dataset, the RR-VGG model with augmented input images yields the best accuracy at 88.87% while RR-CNN, an independent and lightweight model, yields 88.64% accuracy. The extension experiments conducted prove that our proposed models could be applied to other race dataset problems such as Japanese, Chinese, or Brazilian with over 90% accuracy; the fine-tuning RR-VGG model achieved the best accuracy and is recommended for most scenarios.

show abstract

SVAS: Surveillance Video Analysis System

Cited by 31 publications

References 26 publications

Estimation of linear motion in dense crowd videos using Langevin model

Estimation of linear motion in dense crowd videos using Langevin model

Multi-Event Modeling and Recognition Using Extended Petri Nets

Race Recognition Using Deep Convolutional Neural Networks

Contact Info

Product

Resources

About