OBJECTS RECOGNITION ON DYNAMIC BACKGROUND USING TWO-DIMENSIONAL PREDICTION MODEL WITH QUADRATIC NONLINEARITY
Abstract
Method of objects recognition on dynamic background is researched in given work. Stochastic, linear and nonlinear prediction models are used for modelling of dynamic background as a signal that is changed in time and space. The best quality of object definition is received with help of simplified nonlinear model as a sum of linear and quadratic signal components. Influences of model order, supporting area size and threshold value on the signal object selection are researched.Downloads
-
PDF
Downloads: 100
Abstract views: 168
How to Cite
[1]
R. Kvietnyy and O. Bunyak, “OBJECTS RECOGNITION ON DYNAMIC BACKGROUND USING TWO-DIMENSIONAL PREDICTION MODEL WITH QUADRATIC NONLINEARITY”, Works of VNTU, no. 1, Dec. 2011.
Issue
Section
Automatics and Information Measuring Facilities