| Abstract: |
Smuggling and stealing activities often causes enormous monetary loss in supply chain security. While a common smuggling or stealing method is to hide the contraband under clothing, the current security operation lacks an effective pre-screening mechanism to detect suspicious person from a crowd of passengers. Although an emerging technology - passive millimeter wave is capable of detecting concealed object hidden under clothing on a single target, it fails to be used in crowd scanning for pre-screening purpose. This project aims to investigate the integrated use of (1) optical camera, (2) depth camera, (3) thermal camera and (4) PMMW scanning device to provide more comprehensive information for crowd scanning analysis. The deficiency of PMMW on crowd scanning can therefore be remedied by the innovative use of multi-sensor system and machine intelligence. This project is pertained to applied research as it leverages the existing and emerging vision sensing technologies to develop an integrated system while addressing the industrial needs of crowd scanning of body-packed contraband. The deliverables of this project include a sensor fusion model integrating various vision sensors for underclothing concealment detection, a site surveying tool for setting up and calibrating multiple sensors and an alert system prototype to demonstrate the automated detection of concealed objects. |
