Sensing Fusion for Event Based Monitoring System (ART/361CP)

Sensing Fusion for Event Based Monitoring System (ART/361CP)

Sensing Fusion for Event Based Monitoring System (ART/361CP)
ART/361CP
Platform
16 / 05 / 2023 - 15 / 11 / 2024
7,095.500

Mr Kenny Kin-lung CHAN

The Neighbourhood Advice-Action Council
Alpha Vision Technology Limited


Developing a human space sensing fusion management system for human behaviour and environment understanding can enhance the operational efficiency of a wide range of organizations. In this project, the edge AI sensing fusion system is carefully designed to add privacy preserve sensing that effectively balances organization demand for continuous video analytics. The smart sensing fusion system is particularly helpful for the special need caring centres and elderly caring centres. Currently, such caring centres need huge manpower for redundant tasks such as, manually taking and maintaining attendance records, manually checking the temperature and alerting to admin when an abnormally high temperature is detected, manually investigating hours and hours of closed-circuit television (CCTV) footage and so on. ASTRI’s edge AI sensing fusion system effectively addresses these redundant works. ASTRI’s smart edge AI sensing fusion system can easily be integrated with existing CCTV infrastructure. This simple addon fusion system can make any surveillance infrastructure a smart sensing system. The system is designed for both public and private space • Public Space: CMOS Image Sensor (CIS) and thermal sensing fusion (e.g. in class room, training room) • Private Space: non-vision privacy preserves sensing (e.g. in toilet, bedroom) This project provides an effective solution to automatically detect predefined irregular and abnormal behavior and provide instant alerts to the management with the short incident footage. The main goal is to continuously observe and analyze human activities to instantly detect any critical events such as falls down, fighting. These kinds of situation require a quick response like calling ambulance or police. The incident can be automatically tracked under multiple cameras, and the heat map can be generated to get a quick insight into the situation on site. Multiple camera tracking aims to track persons of interest across multi-cameras and is an important element of intelligent surveillance. The system automatically maintains the attendance record with the temperature measurement. For attendance, the face will be recognized with and without a mask and the anti-spoofing prevents any false facial recognition by using mediums like photo, video, 2D/3D masks. The system also measures the temperature and provides an instant alert to the admin whenever an abnormally high temperature is detected. The automatic periodic temperature check has a significant impact on the public health system, especially during the pandemics like COVID-19.