R-PCNN method to rapidly detect objects on THz images in human body security checks

Hong Xiao; Rongyue Zhang; Hao Wang; Feng Zhu; Cheng Zhang; Hongning Dai; Yubin Zhou

doi:10.1109/SmartWorld.2018.00300

R-PCNN method to rapidly detect objects on THz images in human body security checks

Hong Xiao, Rongyue Zhang, Hao Wang, Feng Zhu, Cheng Zhang, Hongning Dai, Yubin Zhou

Department of Computer Science

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

7 Citations (Scopus)

Abstract

Terahertz human body security images have low resolution and a low signal-to-noise ratio. In the traditional method, image segmentation, positioning, and identification are applied to detect objects carried by humans in the THz images. However, it is difficult to satisfy the requirements of detection accuracy and speed with this approach. The current research presents a faster detection framework (R-PCNN) combining the preprocessing and structure optimization of Faster R-CNN. The experiment results show that this method can effectively improve the accuracy and speed of object detection in human body THz images. A detection accuracy of 84.5% can be achieved in dense flow scenes, with an average detection time of less than 20 milliseconds for each image.

Original language	English
Title of host publication	2018 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI)
Editors	Frederic Loulergue, Guojun Wang, Md Zakirul Alam Bhuiyan, Xiaoxing Ma, Peng Li, Manuel Roveri, Qi Han, Lei Chen
Publisher	IEEE
Pages	1777-1782
Number of pages	6
ISBN (Electronic)	9781538693803
ISBN (Print)	9781538693803
DOIs	https://doi.org/10.1109/SmartWorld.2018.00300
Publication status	Published - 6 Dec 2018
Event	4th IEEE SmartWorld, 15th IEEE International Conference on Ubiquitous Intelligence and Computing, Advanced and Trusted Computing, Scalable Computing and Communications, Cloud and Big Data Computing, Internet of People and Smart City Innovations, SmartWorld/UIC/ATC/ScalCom/CBDCom/IoP/SCI 2018 - Guangzhou, China Duration: 7 Oct 2018 → 11 Oct 2018

Conference

Conference	4th IEEE SmartWorld, 15th IEEE International Conference on Ubiquitous Intelligence and Computing, Advanced and Trusted Computing, Scalable Computing and Communications, Cloud and Big Data Computing, Internet of People and Smart City Innovations, SmartWorld/UIC/ATC/ScalCom/CBDCom/IoP/SCI 2018
Country/Territory	China
City	Guangzhou
Period	7/10/18 → 11/10/18

Scopus Subject Areas

Computer Networks and Communications
Computer Science Applications
Information Systems and Management
Artificial Intelligence

User-Defined Keywords

Faster R-CNN
Human body security check
Image enhancement
Object detection
Terahertz image

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1109/SmartWorld.2018.00300

Cite this

Xiao, H., Zhang, R., Wang, H., Zhu, F., Zhang, C., Dai, H., & Zhou, Y. (2018). R-PCNN method to rapidly detect objects on THz images in human body security checks. In F. Loulergue, G. Wang, M. Z. A. Bhuiyan, X. Ma, P. Li, M. Roveri, Q. Han, & L. Chen (Eds.), 2018 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI) (pp. 1777-1782). Article 8560278 IEEE. https://doi.org/10.1109/SmartWorld.2018.00300

Xiao, Hong ; Zhang, Rongyue ; Wang, Hao et al. / R-PCNN method to rapidly detect objects on THz images in human body security checks. 2018 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI). editor / Frederic Loulergue ; Guojun Wang ; Md Zakirul Alam Bhuiyan ; Xiaoxing Ma ; Peng Li ; Manuel Roveri ; Qi Han ; Lei Chen. IEEE, 2018. pp. 1777-1782

@inproceedings{0840b24b9a91439e991341cab0e91d60,

title = "R-PCNN method to rapidly detect objects on THz images in human body security checks",

abstract = "Terahertz human body security images have low resolution and a low signal-to-noise ratio. In the traditional method, image segmentation, positioning, and identification are applied to detect objects carried by humans in the THz images. However, it is difficult to satisfy the requirements of detection accuracy and speed with this approach. The current research presents a faster detection framework (R-PCNN) combining the preprocessing and structure optimization of Faster R-CNN. The experiment results show that this method can effectively improve the accuracy and speed of object detection in human body THz images. A detection accuracy of 84.5% can be achieved in dense flow scenes, with an average detection time of less than 20 milliseconds for each image.",

keywords = "Faster R-CNN, Human body security check, Image enhancement, Object detection, Terahertz image",

author = "Hong Xiao and Rongyue Zhang and Hao Wang and Feng Zhu and Cheng Zhang and Hongning Dai and Yubin Zhou",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE. Applied science and technology research and development in Guangdong province special fund project: development of urban rail transit security inspection equipment based on terahertz imaging technology (NO.2015B090923004)National natural science foundation project: research on self-organizing reinforcement strategy of IOT network in intelligent manufacturing(NO. 61672170); 4th IEEE SmartWorld, 15th IEEE International Conference on Ubiquitous Intelligence and Computing, Advanced and Trusted Computing, Scalable Computing and Communications, Cloud and Big Data Computing, Internet of People and Smart City Innovations, SmartWorld/UIC/ATC/ScalCom/CBDCom/IoP/SCI 2018 ; Conference date: 07-10-2018 Through 11-10-2018",

year = "2018",

month = dec,

day = "6",

doi = "10.1109/SmartWorld.2018.00300",

language = "English",

isbn = "9781538693803",

pages = "1777--1782",

editor = "Frederic Loulergue and Guojun Wang and Bhuiyan, {Md Zakirul Alam} and Xiaoxing Ma and Peng Li and Manuel Roveri and Qi Han and Lei Chen",

booktitle = "2018 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI)",

publisher = "IEEE",

address = "United States",

}

Xiao, H, Zhang, R, Wang, H, Zhu, F, Zhang, C, Dai, H & Zhou, Y 2018, R-PCNN method to rapidly detect objects on THz images in human body security checks. in F Loulergue, G Wang, MZA Bhuiyan, X Ma, P Li, M Roveri, Q Han & L Chen (eds), 2018 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI)., 8560278, IEEE, pp. 1777-1782, 4th IEEE SmartWorld, 15th IEEE International Conference on Ubiquitous Intelligence and Computing, Advanced and Trusted Computing, Scalable Computing and Communications, Cloud and Big Data Computing, Internet of People and Smart City Innovations, SmartWorld/UIC/ATC/ScalCom/CBDCom/IoP/SCI 2018, Guangzhou, China, 7/10/18. https://doi.org/10.1109/SmartWorld.2018.00300

R-PCNN method to rapidly detect objects on THz images in human body security checks. / Xiao, Hong; Zhang, Rongyue; Wang, Hao et al.
2018 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI). ed. / Frederic Loulergue; Guojun Wang; Md Zakirul Alam Bhuiyan; Xiaoxing Ma; Peng Li; Manuel Roveri; Qi Han; Lei Chen. IEEE, 2018. p. 1777-1782 8560278.

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - R-PCNN method to rapidly detect objects on THz images in human body security checks

AU - Xiao, Hong

AU - Zhang, Rongyue

AU - Wang, Hao

AU - Zhu, Feng

AU - Zhang, Cheng

AU - Dai, Hongning

AU - Zhou, Yubin

N1 - Publisher Copyright: © 2018 IEEE. Applied science and technology research and development in Guangdong province special fund project: development of urban rail transit security inspection equipment based on terahertz imaging technology (NO.2015B090923004)National natural science foundation project: research on self-organizing reinforcement strategy of IOT network in intelligent manufacturing(NO. 61672170)

PY - 2018/12/6

Y1 - 2018/12/6

N2 - Terahertz human body security images have low resolution and a low signal-to-noise ratio. In the traditional method, image segmentation, positioning, and identification are applied to detect objects carried by humans in the THz images. However, it is difficult to satisfy the requirements of detection accuracy and speed with this approach. The current research presents a faster detection framework (R-PCNN) combining the preprocessing and structure optimization of Faster R-CNN. The experiment results show that this method can effectively improve the accuracy and speed of object detection in human body THz images. A detection accuracy of 84.5% can be achieved in dense flow scenes, with an average detection time of less than 20 milliseconds for each image.

AB - Terahertz human body security images have low resolution and a low signal-to-noise ratio. In the traditional method, image segmentation, positioning, and identification are applied to detect objects carried by humans in the THz images. However, it is difficult to satisfy the requirements of detection accuracy and speed with this approach. The current research presents a faster detection framework (R-PCNN) combining the preprocessing and structure optimization of Faster R-CNN. The experiment results show that this method can effectively improve the accuracy and speed of object detection in human body THz images. A detection accuracy of 84.5% can be achieved in dense flow scenes, with an average detection time of less than 20 milliseconds for each image.

KW - Faster R-CNN

KW - Human body security check

KW - Image enhancement

KW - Object detection

KW - Terahertz image

UR - http://www.scopus.com/inward/record.url?scp=85060302725&partnerID=8YFLogxK

UR - https://ieeexplore.ieee.org/document/8560278

U2 - 10.1109/SmartWorld.2018.00300

DO - 10.1109/SmartWorld.2018.00300

M3 - Conference proceeding

AN - SCOPUS:85060302725

SN - 9781538693803

SP - 1777

EP - 1782

BT - 2018 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI)

A2 - Loulergue, Frederic

A2 - Wang, Guojun

A2 - Bhuiyan, Md Zakirul Alam

A2 - Ma, Xiaoxing

A2 - Li, Peng

A2 - Roveri, Manuel

A2 - Han, Qi

A2 - Chen, Lei

PB - IEEE

T2 - 4th IEEE SmartWorld, 15th IEEE International Conference on Ubiquitous Intelligence and Computing, Advanced and Trusted Computing, Scalable Computing and Communications, Cloud and Big Data Computing, Internet of People and Smart City Innovations, SmartWorld/UIC/ATC/ScalCom/CBDCom/IoP/SCI 2018

Y2 - 7 October 2018 through 11 October 2018

ER -

Xiao H, Zhang R, Wang H, Zhu F, Zhang C, Dai H et al. R-PCNN method to rapidly detect objects on THz images in human body security checks. In Loulergue F, Wang G, Bhuiyan MZA, Ma X, Li P, Roveri M, Han Q, Chen L, editors, 2018 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI). IEEE. 2018. p. 1777-1782. 8560278 doi: 10.1109/SmartWorld.2018.00300

R-PCNN method to rapidly detect objects on THz images in human body security checks

Abstract

Conference

Scopus Subject Areas

User-Defined Keywords

UN SDGs

Access to Document

Other files and links

Cite this