A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
A blockchain-based edge collaborative detection scheme for construction internet of things
Abstract For smart construction, it is increasingly necessary to develop innovative methods to automatically detect whether construction workers are wearing safety helmets or not. With the development of Internet of Things and object detection in deep learning, it is possible to enable object detection in end devices. In this paper, we propose a coordinated recognition scheme based on YOLOv3 to automatically detect safety helmets at construction sites. Our scheme provides coordinated recognition from multiple devices to obtain comprehensive detection results with multiple viewpoints, which improves the detection accuracy. Besides, to strengthen the security and reliability of detection devices, we propose a peer-to-peer cooperation scheme which is based on smart contract over blockchain. Our scheme ensures that only trusted devices can initiate or participate coordinated recognition tasks. All access and detection records are stored in the blockchain, which is auditable and traceable. Moreover, we propose a fair video sharing mechanism which encourages trusted devices to actively participate in the coordinated recognition tasks. Only participated nodes can access the shared detection videos from others, which forms an alliance of fair sharing with long term interest. The experimental results and analysis justify that our scheme performs well in terms of security and detection accuracy at few expense of a certain startup delay. Our framework can be employed for detecting other objects related to risks in smart construction.
Highlights In this paper, we propose a coordinated recognition scheme over multiple edge nodes, and we select automatically detecting safety helmets at construction sites based on YOLOv3 as a typical case study. We propose a coordinated recognition scheme over multiple IoT devices. The accuracy of detection will be improved by expanding the visible coverage and exchanging detection information, which can overcome the limitations of single device such as shelter, target omission, line of sight, blur video and foggy weather. We implement blockchain and smart contract experiments to achieve secure and robust coordinated interaction between multiple devices, which empower identity authentication to ensure that only trusted nodes can initiate and participate in the coordinated recognition task. All access and operation records are stored in the blockchain, which can be traceable and undeniable. Besides, our scheme provides a cross-domain cooperation mechanism for multiple edge nodes without the delay caused by video uploading to cloud. A detection video sharing mechanism is proposed to encourage trusted nodes to actively participate in the coordinated recognition task. This mechanism ensures that only the nodes that make contributions to the coordinated detection task can obtain permission to access the shared videos as their reward.
A blockchain-based edge collaborative detection scheme for construction internet of things
Abstract For smart construction, it is increasingly necessary to develop innovative methods to automatically detect whether construction workers are wearing safety helmets or not. With the development of Internet of Things and object detection in deep learning, it is possible to enable object detection in end devices. In this paper, we propose a coordinated recognition scheme based on YOLOv3 to automatically detect safety helmets at construction sites. Our scheme provides coordinated recognition from multiple devices to obtain comprehensive detection results with multiple viewpoints, which improves the detection accuracy. Besides, to strengthen the security and reliability of detection devices, we propose a peer-to-peer cooperation scheme which is based on smart contract over blockchain. Our scheme ensures that only trusted devices can initiate or participate coordinated recognition tasks. All access and detection records are stored in the blockchain, which is auditable and traceable. Moreover, we propose a fair video sharing mechanism which encourages trusted devices to actively participate in the coordinated recognition tasks. Only participated nodes can access the shared detection videos from others, which forms an alliance of fair sharing with long term interest. The experimental results and analysis justify that our scheme performs well in terms of security and detection accuracy at few expense of a certain startup delay. Our framework can be employed for detecting other objects related to risks in smart construction.
Highlights In this paper, we propose a coordinated recognition scheme over multiple edge nodes, and we select automatically detecting safety helmets at construction sites based on YOLOv3 as a typical case study. We propose a coordinated recognition scheme over multiple IoT devices. The accuracy of detection will be improved by expanding the visible coverage and exchanging detection information, which can overcome the limitations of single device such as shelter, target omission, line of sight, blur video and foggy weather. We implement blockchain and smart contract experiments to achieve secure and robust coordinated interaction between multiple devices, which empower identity authentication to ensure that only trusted nodes can initiate and participate in the coordinated recognition task. All access and operation records are stored in the blockchain, which can be traceable and undeniable. Besides, our scheme provides a cross-domain cooperation mechanism for multiple edge nodes without the delay caused by video uploading to cloud. A detection video sharing mechanism is proposed to encourage trusted nodes to actively participate in the coordinated recognition task. This mechanism ensures that only the nodes that make contributions to the coordinated detection task can obtain permission to access the shared videos as their reward.
A blockchain-based edge collaborative detection scheme for construction internet of things
Xiong, Feng (author) / Xu, Cheng (author) / Ren, Wei (author) / Zheng, Rongyue (author) / Gong, Peisong (author) / Ren, Yi (author)
2021-11-23
Article (Journal)
Electronic Resource
English
| BASE | 2022
Internet-based collaborative decision-making system for construction
| Tema Archive | 2004