Proceedings ArticleDOI
Surveillance Video Synopsis
VK Choudhary,Anil Kumar Tiwari +1 more
- pp 207-212
Reads0
Chats0
TLDR
A fundamental goal of any video summarization or synopsis technique with reference to a surveillance video is to reduce the Spatio-temporal redundancy, which is removed by detecting the frames having low activity and then deleting those frames.Abstract:
Video is a powerful tool to show various activities but generally we use still images to show a condensed video, which is problematic in viewing and comprehending. Thus, there is a need for a summarized surveillance video. A fundamental goal of any video summarization or synopsis technique with reference to a surveillance video is to reduce the Spatio-temporal redundancy. The activity in any surveillance video is very less as compared to the total length of the video. The spatial redundancy is removed by showing two activities that happened in different frames at different spatial locations in a single frame. Temporal redundancy is removed by detecting the frames having low activity and then deleting those frames. We then generate a stroboscopic video, which traces path of the extracted object. Lastly, we introduce a media Player, which indexes the video synopsis to the original video demonstrating how the video synopsis can be used as an effective tool.read more
Citations
More filters
Journal ArticleDOI
Optimization method for trajectory combination in surveillance video synopsis based on genetic algorithm
TL;DR: The method utilizes the temporal combination ways of trajectories to deal with the optimization problem of motion trajectory combination under the idea of genetic algorithm (GA).
Proceedings ArticleDOI
Multilevel Framework for Summarization of Surveillance Videos
TL;DR: This paper presents a mechanism to select salient blocks which can contribute in the video summarization by propagating motion information from frame level to segment level and further to block level and presents the experimental results on surveillance videos.
Proceedings ArticleDOI
Single Camera Surveillance Video Synopsis: A Review and Taxonomy
Subhankar Ghatak,Suvendu Rup +1 more
TL;DR: A comprehensive, categorical review based on single-camera surveillance video synopsis methodologies with a focus to develop an insight into the current research trends and compared and projected in several ways for future research.
Book ChapterDOI
A Parallel Non-Linear Surveillance Video Synopsis System with Operator Eye-Gaze Input
Ulas Vural,Yusuf Sinan Akgul +1 more
Journal ArticleDOI
Detecting Criminal Activities of Surveillance Videos using Deep Learning
TL;DR: A novel approach dealing with Convolutional Neural Networks using Deep Learning was used to sample the priority information from the surveillance videos and the results of the CNN model effectively were able to extract suspicious activity frames from a long video and thus extract suspicious frames and create a video from it.
References
More filters
Proceedings Article
A density-based algorithm for discovering clusters a density-based algorithm for discovering clusters in large spatial databases with noise
TL;DR: In this paper, a density-based notion of clusters is proposed to discover clusters of arbitrary shape, which can be used for class identification in large spatial databases and is shown to be more efficient than the well-known algorithm CLAR-ANS.
Proceedings Article
A density-based algorithm for discovering clusters in large spatial Databases with Noise
TL;DR: DBSCAN, a new clustering algorithm relying on a density-based notion of clusters which is designed to discover clusters of arbitrary shape, is presented which requires only one input parameter and supports the user in determining an appropriate value for it.
Journal ArticleDOI
An integrated scheme for automated video abstraction based on unsupervised cluster-validity analysis
Alan Hanjalic,Hong-Jiang Zhang +1 more
TL;DR: A novel method for generating key frames and previews for an arbitrary video sequence by first applying multiple partitional clustering to all frames of a video sequence and then selecting the most suitable clustering option(s) using an unsupervised procedure for cluster-validity analysis.
Proceedings ArticleDOI
Making a Long Video Short: Dynamic Video Synopsis
TL;DR: This work presents dynamic video synopsis, where most of the activity in the video is condensed by simultaneously showing several actions, even when they originally occurred at different times.
Book ChapterDOI
Background cut
TL;DR: Zhang et al. as mentioned in this paper proposed a real-time foreground layer extraction algorithm based on background contrast attenuation, which adaptively attenuates the contrasts in the background while preserving the contrasts across foreground/background boundaries.