Canon Kabushiki Kaisha

Daniel John Wedge

Ashley John Partis

David Kenji See

Peter Jan Pakulski

Disclosed herein are a method, system, and computer program product for determining a correspondence between a first object () tracked in a first field of view and a second object tracked () in a second field of view. The method determines a first area () in the first field of view, based on the location and size of the first object (). The method utilizes a predetermined area relationship between the first area () in the first field of view and at least one area () in the second field of view to determine a second area () in the second field of view. In one embodiment, the method determines the second area () in the second field of view by comparing predetermined area relationships between the first area () and any areas () in the second field to determine a best match. The method determines a correspondence between the first object () and the second object (), based on a comparison between a first object signature associated with the first object () and a second object signature associated with the second object ().

Location-based signature selection for multi-camera object tracking

Disclosed herein are a computer-implemented method and a camera system for determining a current spatial representation for a detection in a current frame of an image sequence. The method derives an expected spatial representation () for the detection based on at least one previous frame, generates a spatial representation () of the detection, and extends the spatial representation () to obtain an extended spatial representation (), based on the expected spatial representation (). The method determines a similarity measure between the extended spatial representation () and the expected spatial representation (), and then determines the current spatial representation for the detection based on the similarity measure.

Video object fragmentation detection and management

Disclosed herein are a method and system for appearance-invariant tracking of an object in an image sequence. A track is associated with the image sequence, wherein the track has an associated track signature comprising at least one mode. The method detects the object in a frame of the image sequence (). A representative signature is associated with the detected object. The method determines a spatial difference measure between the track and the detected object, and determines, for each mode of the track signature, a visual difference () between the mode of the track signature and the representative signature to obtain a lowest determined visual difference (). The method then utilises the spatial difference measure and the lowest determined visual difference to perform at least one of the following steps of: (i) associating the detected object with the track (), and (ii) adding a new mode to the track signature (), based on the representative signature.

Multi-modal object signature

Disclosed herein are a computer-implemented method and a camera system for determining a current spatial representation for a detection in a current frame of an image sequence. The method derives an expected spatial representation () for the detection based on at least one previous frame, generates a spatial representation () of the detection, and extends the spatial representation () to obtain an extended spatial representation (), based on the expected spatial representation (). The method determines a similarity measure between the extended spatial representation () and the expected spatial representation (), and then determines the current spatial representation for the detection to based on the similarity measure.

Growing community of inventors

St. Leonards, Australia

Daniel John Wedge