Imagery Library for Intelligent Detection Systems
The Imagery Library for Intelligent Detection Systems (i-LIDS) is the government’s benchmark for video analytics (VA) systems. It has been developed by the Centre for Applied Science and Technology (CAST) in partnership with the Centre for the Protection of National Infrastructure (CPNI).
i-LIDS comprises a library of CCTV video footage based around ‘scenarios’ central to the government’s requirements. The footage accurately represents real operating conditions and potential threats.
i-LIDS currently has the following datasets available for distribution and evaluation.
Event detection scenarios:
- sterile zone
- parked vehicle
- abandoned baggage
- doorway surveillance
- new technologies
- multiple camera tracking
The datasets for these scenarios are available in training and test formats. You can find out more about the scenario definitions in the i-LIDS user guide.
For a DVD demonstrating the scenarios, please email i-LIDS.
Event detection scenarios
The datasets for the event detection scenarios each contain approximately 24 hours of footage. Each of these datasets is filmed to represent all weather conditions, time of day and scene densities expected within the scenario. The multiple camera tracking scenario datasets each contain approximately 50 hours of real world footage.
Each dataset consists of 2 or 3 camera views, referred to as stages, and is further segmented into shorter video clips of 30 to 60 minutes. The training dataset is further split into individual events. We supply each dataset with a user guide detailing the library structure, user interface and the procedure used to evaluate the systems against the relevant scenario.
See the i-LIDS user guide for information about dataset pricing.
Annual scenario evaluation
All the scenarios have a scheduled annual evaluation and manufacturers can submit their performance score to CAST to be considered for inclusion in one of the annual evaluations. See the annual timetable of evaluations.