Urban authorities are focusing their efforts on reinforcing methods to secure and enhance public safety by preventing crimes, protecting properties and assets. Public video surveillance is becoming ubiquitously deployed as a valuable tool but denotes specific drawbacks as the absence of effective data processing, and the lack of advanced machine intelligence features such as a DROP-oriented (Distinctive Regions Of interest or Patterns) retrieval architecture. DROP may include color regions, tattoos, logos or any other distinctive features that appear in a given "incident" images and currently starts to replace traditional analysis techniques such as face detection or color/motion information. The concept derives from real police video-based forensics investigations. Many times clear face identification of the suspect is not possible, therefore DROP could be used to track and identify relevant target instances in the entire surveillance system. To cope with the aforementioned limitations, the proposed technology out-coming from the SPOTTER research project will be capable of automatically finding the occurrence of a DROP instance by running specialized algorithms embedded directly on the IP video cameras, and provide the results as efficient as possible to the human operator.
The SPOTTER project is funded under research grant PN-III-P2-2.1-PED-2016-1065, agreement 30PED/2017 (http://uefiscdi.gov.ro/) and is spanning over 18 months (January 2017 to June 2018).