Authors
Nour Ahmed , Mariam Hesham, Samiha Hesham, Sandra Fares, Nour El-Hoda Hisham, Eng. Lobna Shahen and Dr. Islam Tharwat
Publishing Date
December 28, 2020
Abstract
Surveillance systems are of vital importance for the development of smart cities. These systems can be considered vision organs of such cities. It is expected that a huge amount of data (Big Data) will be generated in smart cities. Therefore, to ensure the safety of its citizens, it is important to provide an efficient and real-time analysis of these data to get real-time responses, when catastrophic events occur. Accordingly, transmitting this massive data to the cloud, to be processed, is relatively slow. Therefore, the purpose of this project is to implement a fog/edge computing-based surveillance system to offer real-time data processing. When surveillance videos capture an incident, the data get transferred to the edge for processing. Moreover, a rapid response is then provided to properly handle the occasion. Furthermore, despite tackling scalability obstacles, the system should handle privacy-sensitive data to overcome the privacy challenges in smart cities.
1.1 Purpose of this document
The purpose of this document is to present a detailed description of how to use the edge-fog model to handle criminal cases and reduce them. It will figure both functional and non-functional requirements of the system in addition to the interfaces, GUI, and constraints that will cover the system. Moreover, this document will define both stakeholders and developers of the system.
1.2 Scope of this document
The scope of this document is to show the basic outlines of our system requirements in order to understand our system for any further updates later on and the issues we may face. Furthermore, the objectives that we will reach in our system. Requirements outlined in this document are subject to be changed.
1.3 System Overview
The system aims to prevent criminal acts in smart cities by fast prediction and detection of those incidents, using a framework composed of three computing layers. First, the cameras located at the potential crime scenes act as the edge nodes; Light-weighted algorithms are deployed on them for any possible failure in the device and human motion detection. Once it detects abnormal motion, the data moves to the second layer, which is the fog nodes layer, where the processing is completed and alerts are generated to the authorities, with possibilities to track and detect the kidnapper’s face too. Then transmits the data to the third layer, the cloud layer where data backup and system updates are done. The system must ensure safety and privacy measures for the citizens while using and transferring the captured data throughout the day.
1.4 System Scope
• Fog/Edge computing-based surveillance system to monitoring the whole city 24/7.
• Efficiently detect any abnormal actions caused by human behavior and response immediately.
• Privacy assurance by blur people’s faces to secure their exposed identity in the video.
• Scalability system to handle such massive data generated from surveillance cameras and sensors.
• Efficiently detection for any failure that impacts the usefulness of the video streams.
• Dispatch crime details (i.e., video and location of the event, criminal’s face) and giving alert to the nearest competent authority.
• Applying a prediction function to predict possible abnormal events.
• Providing efficient real-time tracking for the criminal in case of detecting a massive crime.
• Face recognition for a criminal when an event occurs in the far distance through high resolution video.