Human Centred Computer Vision Tool (HCV) – LESTCROWD Tool Presentation Cards #6
by Matteo Mauri
Download the LETSCROWD Toolcards. The HCV tool provides 3 functionalities aimed at supporting LEA operators and forensic investigators in the use of video surveillance systems for monitoring and investigation activities during and after mass gathering events.
The HCV tool comprises the following modules:
1. Appearance-based person re-identification (video), to search for an individual of interest in several recorded videos, based on clothing appearance, using an image as a query.
2. Attribute-based people search (video), to search in several recorded videos for individuals who match an attribute profile of clothing appearance (e.g., colours), gender, carried items.
3. Crowd countin and density estimation (video), for real-time estimation of crowd size and detection of related anomalous behaviours, e.g.: overcrowding with respect to an user-defined threshold, sudden increase or decrease of crowd size.
WHO IS THIS TOOL FOR?
LEA operators and forensic investigators using video surveillance systems for monitoring and investigation activities during mass gathering events.
WHEN SHOULD THE TOOL BE USED?
– During event execution, to monitor the size of a crowd in one or more regions of interest of the event venue.
– After a mass gathering event, in forensic investigations on crimes or incidents, to search for suspect individuals in the videos recorded by video surveillance systems.
WHAT IS THE ADDED VALUE OR BENEFITS FOR LEAs OPERATORS AND OTHER STAKEHOLDERS?
– Reducing LEA operators’ burden in real-time crowd monitoring tasks during event execution.
– Reducing forensic investigators’ effort and time to search for suspect individuals in recorded videos during post-event investigations of crimes and incidents.
The Human Centred Computer Vision Toolcard is available below.
See all the LETSCROWD toolcards.
08 Nov 2019 - news