Journal ArticleDOI
Video structural description technology for the new generation video surveillance systems
TLDR
A semantic based model named video structural description (VSD) for representing and organizing the content in videos is proposed, which uses the predefined ontologies including concepts and their semantic relations to represent the contents in videos.Abstract:
The increasing need of video based applications issues the importance of parsing and organizing the content in videos. However, the accurate understanding and managing video contents at the semantic level is still insufficient. The semantic gap between low level features and high level semantics cannot be bridged by manual or semi-automatic methods. In this paper, a semantic based model named video structural description (VSD) for representing and organizing the content in videos is proposed. Video structural description aims at parsing video content into the text information, which uses spatiotemporal segmentation, feature selection, object recognition, and semantic web technology. The proposed model uses the predefined ontologies including concepts and their semantic relations to represent the contents in videos. The defined ontologies can be used to retrieve and organize videos unambiguously. In addition, besides the defined ontologies, the semantic relations between the videos are mined. The video resources are linked and organized by their related semantic relations.read more
Citations
More filters
Journal ArticleDOI
Big forensic data reduction: digital forensic images and electronic evidence
TL;DR: The proposed data reduction process is intended to provide a capability to rapidly process data and gain an understanding of the information and/or locate key evidence or intelligence in a timely manner.
Journal ArticleDOI
Deep data analyzing algorithm based on scale space theory
Yiwei Zhu,Kun Gao +1 more
TL;DR: The definition and partition of data scale as well as the relationship of multiscale data set between the upper scale and lower scale based on concept hierarchy theory are given.
Journal ArticleDOI
Video structured description technology based intelligence analysis of surveillance videos for public security applications
Zheng Xu,Chuanping Hu,Lin Mei +2 more
TL;DR: A semantic based model named video structural description (VSD) for representing and organizing the content in videos is introduced firstly and the applications of VSD technology on public security from surveillance videos are given.
Journal ArticleDOI
Monitoring drivers' sleepy status at night based on machine vision
TL;DR: A night monitoring system for real-time fatiguedriving detection, which makes up for the deficiencies of fatigue driving detection technology at night, and proposes a new eye-detection algorithm that combines a Gabor filter with template matching to locate the position of the corners of the eye.
Proceedings ArticleDOI
VidCEP: Complex Event Processing Framework to Detect Spatiotemporal Patterns in Video Streams
Piyush Yadav,Edward Curry +1 more
TL;DR: This work focuses on a CEP framework where users can define high-level expressive queries over videos to detect a range of spatiotemporal video event patterns and proposes VidCEP, an in-memory, on the fly, near real-time complex event matching framework for video streams.
References
More filters
Proceedings ArticleDOI
ImageNet: A large-scale hierarchical image database
TL;DR: A new database called “ImageNet” is introduced, a large-scale ontology of images built upon the backbone of the WordNet structure, much larger in scale and diversity and much more accurate than the current image datasets.
Journal ArticleDOI
Object Detection with Discriminatively Trained Part-Based Models
TL;DR: An object detection system based on mixtures of multiscale deformable part models that is able to represent highly variable object classes and achieves state-of-the-art results in the PASCAL object detection challenges is described.
Proceedings ArticleDOI
A discriminatively trained, multiscale, deformable part model
TL;DR: A discriminatively trained, multiscale, deformable part model for object detection, which achieves a two-fold improvement in average precision over the best performance in the 2006 PASCAL person detection challenge and outperforms the best results in the 2007 challenge in ten out of twenty categories.