Toward a Remote-Controlled Weapon-Equipped Camera Surveillance System

Author(s):  
Robert Bisewski ◽  
Pradeep K. Atrey
2004 ◽  
Vol 7 (4) ◽  
pp. 430-446 ◽  
Author(s):  
James Black ◽  
Dimitrios Makris ◽  
Tim Ellis

2014 ◽  
Vol 3 (2) ◽  
pp. 95-112 ◽  
Author(s):  
Sergio Orts-Escolano ◽  
Jose Garcia-Rodriguez ◽  
Vicente Morell ◽  
Miguel Cazorla ◽  
Jorge Azorin ◽  
...  

2013 ◽  
Vol 16 (11) ◽  
pp. 1272-1280
Author(s):  
Viet Quoc Nguyen ◽  
HoSeok Kang ◽  
Sun-Tae Chung ◽  
Seongwon Cho

2021 ◽  
Vol 11 (19) ◽  
pp. 9197
Author(s):  
Muhammad Tahir ◽  
Saeed Anwar

Person Re-Identification is an essential task in computer vision, particularly in surveillance applications. The aim is to identify a person based on an input image from surveillance photographs in various scenarios. Most Person re-ID techniques utilize Convolutional Neural Networks (CNNs); however, Vision Transformers are replacing pure CNNs for various computer vision tasks such as object recognition, classification, etc. The vision transformers contain information about local regions of the image. The current techniques take this advantage to improve the accuracy of the tasks underhand. We propose to use the vision transformers in conjunction with vanilla CNN models to investigate the true strength of transformers in person re-identification. We employ three backbones with different combinations of vision transformers on two benchmark datasets. The overall performance of the backbones increased, showing the importance of vision transformers. We provide ablation studies and show the importance of various components of the vision transformers in re-identification tasks.


Author(s):  
Sonal Beniwal ◽  
Usha Saini ◽  
Puneet Garg ◽  
Rakesh Kumar Joon

This paper is proposing an IoT-based camera surveillance system. The objective of research is to detect suspicious activities by camera automatically and take decision by comparing current frame to previous frame. Major motivation behind research work is to enhance the performance of IoT-based system by integration of edge detection mechanism. Research is making use of numerous cameras, canny edge detection-based compression module, picture database, picture comparator. Canny edge detection has been used to minimize size of graphical content to enhancing the performance system. Simulation of output of this work is made in MATLAB simulation tool. Moreover, MATLAB has been used to give comparative analysis among IoT-based camera surveillance system and traditional system. Such system requires less space, and it takes less time to inform regarding any suspicious activities.


2007 ◽  
Vol 127 (6) ◽  
pp. 837-843 ◽  
Author(s):  
Kenji Iwata ◽  
Yutaka Satoh ◽  
Ikushi Yoda ◽  
Katsuhiko Sakaue

Sign in / Sign up

Export Citation Format

Share Document