|Project Title:||Video Analytics for Resource Management|
|Hosting Institution:||The Chinese University of Hong Kong (CUHK)|
|Abstract:||In large transit facility like airport, warehouses, and distribution centers, a set of movable resources is circulating within the facility to help transport or temporarily store items going through it. For example, trolleys in an airport’s baggage reclaim hall will be fetched by passengers to carry their luggage to their choices of transportation leaving the airport. Such trolleys may be placed indoor and outdoor within the perimeter of the airport. With the fact that the CCTV visual coverage in a large facility increases over time, it is logical and economical to leverage existing CCTV infrastructure to help monitor and manage movable resources. In addition, queueing conditions in transit facility may affect or be affected by the availability of other resources. Queue performance monitoring is vital to ensure smooth transition. This project aims at using video content analytic techniques to help detect specific resources in order to achieve appropriate level of resource allocation throughout transit facility.
The project will carry out a series of pilot studies in the Hong Kong International Airport (HKIA) to measure the detection and counting capabilities of the R&D results. The first pilot study will be carried out in a relatively constant environment in Baggage Reclaim Hall to monitor number of available trolleys in trolley stations. Along with the detection module, a real-time alert system will be developed to take live CCTV feed and provide replenishment notice or alert to HKIA management and service providers to refill trolleys to appropriate trolley stations. The second pilot study will be carried out at the curbside trolley stations outside the Departure Hall where the environmental condition varies due to direct sun light at different time of the day or season as well as shadows from moving double-decker buses. The third pilot study will be at the trolley stations in the airside Departure West Hall (boarding gate areas). This pilot study is to demonstrate the extensibility of trainable object detector to detect and count trolleys of another form factor and shape under variable environmental conditions. The fourth pilot study is to demonstrate the capability to collect waiting time statistics in makeshift passenger queues in the Departure Hall.
The video-based technologies developed in this project are applicable vertically to airports around the world, or horizontally to other CCTV-ready facilities like shopping malls and exhibition centers with indoor, outdoor, and mixed environment to monitor allocations of resources inside premises.
|Project Coordinator:||Prof Chun-hung CHENG|
|Approved Funding Amount:||HK$5.26M|
|Project Period:||22 Sep 2015 - 21 Sep 2017|