PSIVT15 Workshop on Video Surveillance(VSWS 2015)

摘要

These contributions have been presented at the First Workshop on Video Surveillance (VSWS2015), held in conjunction with the7th Pacific Rim Symposium on Video and Image Technology (PSIVT 2015), in Auckland, New Zealand, on November 24, 2015. Surveillance can function both as a deterrent to help prevent crime as well as an investigative tool to aid us in identifying the actors and causes of incidents. Surveillance systems record multi-modal evidence and help carry out intelligent analytics to obtain a multi-perspective understanding. It can also alert security staff about security breaches, potentially dangerous events and hazardous situations. These security related events are captured in the content of surveillance video as low-level semantic signals, which bridges the gap between the physical world and the semantic cyberspace. If a picture is worth more than a thousand words, an event is perhaps worth thousands of pictures. Thus, the multi-modal surveillance content can convey invaluable security-related information. Without proper automated means of analyzing this information, it is extremely hard to manually process this big surveillance data for detecting events of interest.