A unified face identification and resolution scheme using cloud computing in Internet of Things
Introduction
With the fast development of computing, communication, and control technologies, the fields of IoT and Cyber–Physical Systems (CPS) have attracted much attention in the last few years [1], [2], [3]. These technologies realize the interconnection among ubiquitous things by the corresponding applications and services in both physical-space and cyber-space [4], [5], [6], [7]. In physical-space, various physical objects (e.g., persons, sensors, computers, mobile devices, and commodity) have been accessed Internet as building blocks of IoT and enable novel applications [8]. Meanwhile, a large number of cyber objects are generated in cyber-space [9], [10]. They need to rely on the identification and resolution technology to directly or indirectly communicate and cooperate with each other to reach the goals of information sensing and automatic control [11]. As an important field of IoT, identification and resolution technology has been applied various IoT scenarios. For example, entrance guard, logistics, food safety, supply chain management, Internet Finance (ITFIN), mobile payments, etc. Identification and resolution of physical object has become an important research direction for achieving object’s identity authentication, data access control, and trust establishment between object and service.
In the process of identification, various physical objects are respectively identified and associated by the corresponding identifiers. So that network and applications can control and manage these objects with identifier to implement information acquisition, processing, access control, transmission and exchange throughout both physical-space and cyber-space. In addition to physical object identification, there are communication identification and application identification. The former is used to identify the network nodes which has the communication ability, for example, IP address, E.164 number. The latter is used to identify the various application services in the application layer, for example, domain name, Uniform Resource Locator (URL). The identity resolution of physical object is a process that maps a physical object identifier to a communication identifier or application identifier or its associated information. For example, by resolving an identifier of product, we can obtain the application identifier which stores its related information or service [12]. The identification and resolution of physical object realize the mutual mapping between the individual in physical-space and the identity information or application service in cyber-space.
For identification and resolution technology, current research mainly concentrates in physical objects with identification (ID) code. Electronic product code (EPC) and ubiquitous ID (uID) are typical ID code [13], [14]. They have been widely used in various application areas, including logistics, food safety, supply chain management, commodity retail, and so on. They are accurate identifier which is comprised of numbers or alphabets with certain rules. The Object Name Service (ONS) is a typical resolution model of ID code [15]. It links EPC with its associated Physical Markup Language (PML) data file [16]. The host address on which corresponding PML file is located will be obtained by ONS. However, in some IoT scenarios, there are many objects without any available ID code. We can identify them with their properties, for example, biometric, space–time information, and other characteristics [17], [18], [19]. Compared with ID code, these identifiers are not accurate. But they can meet the demands of identification in some IoT applications. Furthermore, they have some of the particular advantages which ID code is not available.
In the future, IoT and Internet of people (IoP) will be interconnected to each other enabled by cloud [20]. More and more people are accessed to the Internet, and corresponding cyber-individuals will be generated [21]. The research on human identification and resolution has become more and more valuable for realizing identity authentication, personal information management and data access control about human.
Face is a discriminative biometric which is used to uniquely distinguish different humans. It is an inherently reliable and distinctive identification method [22], [23]. Face identification is a process that analyzes facial images, extracts special useful information such as pixel and position of feature points, size and position of eyes, nose and mouth [24], [25]. Finally, face identifier is generated by using the extracted facial feature value. The goal of face resolution is to find the application identifier stored individual’s identity information or application service by identity authentication based on face identifier.
Compared with ID code, the process of face identification and resolution is more complex, which result from the characteristics of face identifier. Some algorithms need to be performed, for example, face detection, facial image preprocessing, facial feature extraction and identifier generation, identifier matching. So the computational complexity is greater than that of ID code. Moreover, it needs more storage capacity because the size of face identifier data is relatively larger and data structure is more complex. With the increasing of IoT applications based on face identification, face image database is also growing at the same time. It causes that the time complexity and space complexity of face resolution are further increased. Especially in current big data environment, the demands for computation power and storage capacity will be greater [26], [27].
Currently, most of the identification and resolution applications are implemented in the specific and independent IoT scenarios. Simple identification and resolution service model is usually adopted. It is difficult to achieve interoperability between different industries and different platforms. Users need to frequently verify and resolve identity in different applications, which result in inconvenient operation, waste of resource, and increasing risk of privacy disclosure [28]. Furthermore, the future of IoT is moving toward the direction of ubiquity, that is ubiquitous Internet of Things (ubiquitous IoT) [29], [30]. It is the integration of multiple independent IoT applications and realizes the interconnections and cooperation among ubiquitous things, as well as pervasive management and access control of IoT resources [31]. This situation facilitates the requirement which design a unified and efficient face identification and resolution scheme to share face services for cross-industry and cross-platform IoT applications.
The advantages of cloud computing technology in providing powerful computational and storage capacity and unified cloud service access can appropriately solve the above mentioned problems [32]. It has many advantages in practical application, which mainly include resource pooling, virtualization, broad network access, high reliability, high scalability, elasticity (dynamic provisioning), service oriented architecture (SOA) [33], [34]. These characteristics motivate the utilization of cloud computing technology to store and process face identifier data. Therefore, we design a face identification and resolution scheme based on cloud computing in this paper. This scheme concentrates computation power and storage capacity in cloud platform. It can solve the problems of computation and storage caused by the increasing of applications and users in IoT. Furthermore, we design a parallel resolution mechanism to give full play to the processing capacity of cloud computing and improve the efficiency of face resolution. In this scheme, the face identifier generation model, matching algorithm and service interface are unified. So various cross-industry and cross-platform applications can conveniently access the face identification and resolution service.
In this paper, we focus on identification and resolution based on face, as well as cloud computing-based resolution framework. The main contributions are as follows:
- (1)
Face identification and resolution system model is presented to implement the face identifier generation and identifier matching. It can effectively identify an individual and realize the identity resolution with face identifier in IoT application.
- (2)
Cloud computing-based face resolution scheme is proposed to resolve face image, control personal data access and obtain identity information service. It makes full use of the advantages of cloud computing to effectively meet the demands of computation power and storage capacity. Furthermore, it provides a unified face identification and resolution service platform for cross-industry and cross-platform IoT applications.
- (3)
The parallel resolution mechanism is proposed to improve the efficiency of face resolution.
The remainder of this paper is organized as follows. Section 2 reviews the related work on face identification and resolution. Section 3 presents the face identification and resolution system model. Section 4 proposes the cloud computing-based face resolution scheme. Section 5 presents the experiment and performance evaluation for proposed scheme. Section 6 draws a conclusion.
Section snippets
Related work
The field of identification and resolution of ID code has been researched many years. Identification and resolution based on object’s properties becomes more and more concerned. The development of face recognition and computer vision technology greatly promotes the improvement of face identification accuracy. With the increasing number of IoT applications based on face identification, the capacity and efficiency of the process also need to be improved.
Haghighat et al. [35] presented CloudID
Face identification and resolution system model
In this section, we present the cloud computing-based face identification and resolution system model. We implement the face identifier generation and matching. In our another work, we has presented the face identifier generation algorithm based on Local Binary Patterns (LBP) and face identifier matching algorithm based on Euclidean distance in detail [44]. So we focus on the cloud computing-based system model in this paper.
Cloud computing-based face resolution scheme
In this section, we introduce the cloud computing-based face resolution framework. Because the operations in face identification phase need also to be performed in resolution process, we only focus on the process of face resolution here. The core of face resolution is the face identifiers matching. We present a parallel matching strategy to improve the efficiency of face resolution.
Experiment and performance evaluation
In this section, we implement a prototype system to demonstrate the practical feasibility of cloud computing-based face identification and resolution scheme proposed above. The experimental results are presented and the performance of scheme is evaluated and discussed from multiple aspects.
Conclusion and future work
In this paper, we have proposed a cloud computing-based face identification and resolution scheme for IoT applications to implement high-security identification, identity authentication, personal information management, and data access control about human. Face in physical-space is converted into face identifier in cyber-space, which is used as a reliable identification technology for human. The constituent parts and workflow of face identification and resolution system model have been
Acknowledgments
This work was funded by the National Natural Science Foundation of China (Grant No. 61471035 and Grant No. 61672131), the Fundamental Research Funds for Central Universities (Grant No. 06105031). The authors are grateful to the Caltech Computational Vision Group of California Institute of Technology for providing Caltech face database, the Center for Signal and Image Processing of Georgia Institute of Technology for providing Georgia Tech face database, the BioID for providing BioID face
Pengfei Hu received the B.E. degree from the School of Computer Science, Zhengzhou University of Aeronautics, China, in 2012. He is currently working toward the Ph.D. degree from the School of Computer and Communication Engineering, University of Science and Technology Beijing, China. He focuses on the objects modeling in cyber–physical space convergence and Internet of Things. His research interests include Internet of Things, identification and resolution of physical objects, and
References (48)
- et al.
The Internet of things: A survey
Comput. Netw.
(2010) - et al.
A smart system to manage the context evolution in the Cultural Heritage domain
Comput. Electr. Eng.
(2016) - et al.
A grid-based joint routing and charging algorithm for industrial wireless rechargeable sensor networks
Comput. Netw. Int. J. Comput. Telecommun. Network.
(2016) - et al.
ERGID: An efficient routing protocol for emergency response Internet of things
J. Netw. Comput. Appl.
(2016) - et al.
Cybermatics: Cyber–physical–social–thinking hyperspace based science and technology
Future Gener. Comput. Syst.
(2016) - et al.
Heterogeneous ad hoc networks: Architectures, advances and challenges
Ad Hoc Networks
(2017) - et al.
Physimetric identification (Physi-ID)—Applying biometric concept in physical object identification
Comput. Ind.
(2011) - et al.
A review of biometric technology along with trends and prospects
Pattern Recognit.
(2014) - et al.
Characteristic trade-offs in designing large-scale biometric-based identity management systems
J. Netw. Comput. Appl.
(2010) - et al.
Review and analysis of networking challenges in cloud computing
J. Netw. Comput. Appl.
(2016)
CloudID: Trustworthy cloud-based and cross-enterprise biometric identification
Expert Syst. Appl.
Tagging and classifying facial images in cloud environments based on KNN using MapReduce
Optik-Int. J. Light Electron Opt.
A comparative study of texture measures with classification based on feature distributions
Pattern Recognit.
A study on application-aware scheduling in wireless networks
IEEE Trans. Mob. Comput.
Cross-layer optimized routing in wireless sensor networks with duty cycle and energy harvesting
Wirel. Commun. Mob. Comput.
Social recommendation service for cultural heritage
Personal Ubiquitous Comput.
Cyber-physical-social-thinking space based science and technology framework for the Internet of things
Sci. China Inf. Sci.
Identity management directions in future Internet
IEEE Commun. Mag.
SHARDIS: A privacy-enhanced discovery service for RFID-based product information
IEEE Trans. Ind. Inf.
Ubiquitous ID: standards for ubiquitous computing and the Internet of things
IEEE Pervasive Comput.
Object naming service model based on two-layer P2P structure in Internet of things
Comput. Eng.
Measuring the similarity of PML documents with RFID-based sensors
Int. J. Ad Hoc Ubiquitous Comput.
nID-based Internet of things and its application in airport aviation risk management
Chin. J. Electron.
Cited by (52)
Security and privacy issues of physical objects in the IoT: Challenges and opportunities
2021, Digital Communications and NetworksCitation Excerpt :For human beings, the common biometric proof may be the fingerprint, iris, face, etc. Pengfei Hu et al. put forward a face identification resolution framework based on the cloud [24] and fog [25] computing respectively for physical objects connecting to the IoT or being mapped from the physical space to cyberspace. At the same time, they also gave the corresponding security solution for the biometric data to preserve the object’s privacy [25].
An identity management scheme for cloud computing: Review, challenges, and future directions
2023, Journal of Intelligent and Fuzzy SystemsCREATING A METHOD FOR IDENTIFYING OBJECTS IN IMAGE WITH THEIR DIFFERENT ORIENTATIONS
2023, International Journal on Technical and Physical Problems of EngineeringINTELLIGENT SECURITY SYSTEM BASED ON BIOMETRIC FACE RECOGNITION
2023, Journal of Theoretical and Applied Information Technology
Pengfei Hu received the B.E. degree from the School of Computer Science, Zhengzhou University of Aeronautics, China, in 2012. He is currently working toward the Ph.D. degree from the School of Computer and Communication Engineering, University of Science and Technology Beijing, China. He focuses on the objects modeling in cyber–physical space convergence and Internet of Things. His research interests include Internet of Things, identification and resolution of physical objects, and cyber–physical modeling.
Huansheng Ning received the B.S. degree from Anhui University in 1996 and the Ph.D. degree from Beihang University in 2001. He is a professor in the School of Computer and Communication Engineering, University of Science and Technology Beijing, China. He is the founder of Cyberspace and Cybermatics and Cyberspace International Science and Technology Cooperation Base. He is the Co-Chair of IEEE Systems, Man, and Cybernetics Society Technical Committee on Cybermatics. His current research interests include Internet of Things, Cybermatics, electromagnetic sensing and computing. He is a senior member of the IEEE.
Tie Qiu received Ph.D. and M.Sc. from Dalian University of Technology (DUT), in 2012 and 2005, respectively. He is currently Associate Professor at School of Software, Dalian University of Technology, China. He has authored/co-authored 7 books, over 50 scientific papers in international journals and conference proceedings. He is a senior member of China Computer Federation (CCF) and a Senior Member of IEEE and ACM. His research interests cover Embedded System Architecture, Internet of Things, Wireless and Mobile Communications.
Yue Xu received the B.E. degree from the School of Computer and Communication Engineering, University of Science and Technology Beijing, China, in 2016. She is currently working toward the M.S. degree from the School of Computer and Communication Engineering, University of Science and Technology Beijing, China. Her research interests include Internet of Things and the ontology modeling.
Xiong Luo received the Ph.D. degree from Central South University, China, in 2004. From 2005 to 2006, he was with the Department of Computer Science and Technology, Tsinghua University, China, as a Postdoctoral Fellow. From 2012 to 2013, he was with the School of Electrical, Computer and Energy Engineering, Arizona State University, USA, as a Visiting Scholar. He currently works as a Professor in the School of Computer and Communication Engineering, University of Science and Technology Beijing, China. His current research interests include machine learning, cyber–physical systems, and computational intelligence. He has published extensively in his areas of interest.
Arun Kumar Sangaiah received the Ph.D. degree from the School of Computer Science and Engineering, VIT University, Vellore, India. He is presently working as an Associate Professor in School of Computing Science and Engineering, VIT University, India. He has author of more than 100 publications in different journals and conference of National and International repute. He is member of international advisory board of IJIIT (IGI publisher) and editorial board member of IJIES, IJHPS etc. He is an active member of Compute Society of India. His research interests include Software Engineering, Wireless Networks, Bio-Informatics, and Embedded Systems.