Informatics news
MVA Most Influential Paper of the Decade awarded to Dr Luc Berthouze
By: Laura Tovey
Last updated: Friday, 8 March 2013
The Machine Vision Applications organization has awarded Dr Luc Berthouze (Senior Lecturer in Informatics) the Most Influential Paper of the Decade Award, for his paper "Adaptive background estimation: Computing a pixel-wise learning rate from local confidence and global correlation values" (co-authored with Michael Pic and Takio Kurita), which was presented at the MVA 2002 conference.
The paper describes a method of processing video that can be used to distinguish moving objects, such as people, from a static background. The technique can be used in a wide wide of applications, ranging from security surveillance and traffic monitoring to medical and space imaging. The approach is robust to motion within the background, and situations in which the foreground and background share similar colours, there are frequent changes in lighting conditions, or there is noise in the camera image.
Dr Berthouze has been invited to receive the award at the 13th IAPR Conference on Machine Vision Applications (MVA 2013) in Kyoto, Japan.