[IEE EC3 News] Lecture reminder - Tomorrow!

[Bill Pechey]

Dear Colleague,

You are being sent this reminder message because either you gave your
email address at one of our previous lectures or you signed up on the
web site associated with this email list:

http://lists.ourshack.com/mailman/listinfo/iee-ec3-news

I have appended the details of our next lecture and hope to see you
there.

You can download a poster from the following IET web page to display
in your place of work:

https://communities.theiet.org/files/14569

Best wishes,

Bill Pechey
IET Thames Valley Specialised Section

++++++++++

Activity monitoring on surveillance systems: Understanding Human
behaviour from video

Luis Patino, University of Reading

Thursday 26 Oct 2017 19.00 for 19.30

Palmer Building, University of Reading RG6 6UR

Surveillance systems are developed to detect threats on camera streams
and to send appropriate alarms. Detecting threatening actions from
people comes down to a behaviour recognition task. Depending on the
application domain, such systems may specialise in the detection of
some targeted events: ‘detection of a person in a forbidden or
sensitive area’, ‘person loitering’, ‘people falling’ are some
examples. Recently special attention has been placed on social
interaction detection between people in the monitored space. Detecting
people meeting or splitting from groups is thus of highest interest to
understand when a group is forming and potentially detecting a threat
such as a fight or an attack to a person.

This talk will review some the most common approaches for the
detection of individual and grouped behaviours. The talk will
particularly introduce a semantic approach recently developed at the
University of Reading. The proposed approach works by translating
people behaviour from trajectory information into semantic terms.
Having available a semantic model of the targeted behaviour, the event
detection is performed in the semantic domain. The model is learnt
employing a soft-computing clustering algorithm that combines
trajectory information and motion semantic terms. The talk will
present results obtained on different EU projects at which the
University of Reading has taken part for surveillance analysis.

This event is free of charge and open to all, with refreshments at
19.00, lecture at 19:30.

Contact: Prof Richard Mitchell
rjmitchell at theiet.org
http://www.theiet.org/ec3

For directions please visit http://www.reading.ac.uk/about/find/about-findindex.aspx