Integrated Machine Vision (IMV)
Computational VisionGroup, University of Reading.
Paolo Remagnino, Tom Grove, Anthony Worall, Steve Maybank and Keith Baker.
Dept. of Computer Science, University of Leeds.
Richard Morris and David Hogg.
This page describes the Integrated Traffic and Pedestrian Model-Based Vision System developed in the Reading and Leeds computer vision groups as part of the EPSRC's IMV program (EPSRC Grant GR/K46620).
The project combines two model based tracking systems:
(For further information about the project please contact Professor Keith Baker K.D.Baker@rdg.ac.uk )
The integrated vision system builds a dynamic 3D model of the scene. This is illustrated in the image below. A short video illustrating the integrated tracker can be dowloaded from here as either 1.8 MB MPEG-1 or 9.4 MB AVI. Using the system it is possible to locate vehicles and pedestrians within images. The eventual aim is to automatically detect events that may have security or safety implications. This information could then be used to assist a human operator.
The Vision system provides a natural language based interpretation of the evolving scene. An image illustrating the annotation is available here. Another page, describing results obtained at Leeds, can be found here.