At present, video content monitoring is mostly done by manual viewing of the footage. However, prolonged manual monitoring may affect the attention span of the security personnel, resulting in an increased level of security risk. Furthermore, while conventional CCTV system is useful for retrieving evidence, it seldom provides instantaneous behaviour analysis or early warning.

With the help of computer vision and video processing technologies, (computerised) video content analysis (VCA) could increase the efficiency and effectiveness of video monitoring. It could provide on-the-spot behaviour analysis and early warning as deemed necessary. It would not only lessen the physical workload required of security personnel, but would also prove a more reliable and accurate monitoring and alert system. Automated monitoring could help in identifying, tracking, analysing and evaluating targets in the fight against crime and terrorism.

This project aims to set out the selection criteria for VCA technologies. It should start with the evaluation of the video content analytic products of various brands in the market with respect to their product characteristics and uniqueness. Then a trial platform to accommodate the various software and hardware developers should be developed, with different scenarios designed to evaluate their capabilities and performance with reference to international standards and requirements. The selection criteria for VCA should then be established.

12 months

Development of selection criteria for video content analysis technologies