| 摘要 |
A method and apparatus is provided for efficient computing of perspective transformation motion tracking parameters in a sequence of images. More specifically, a set of initial positions for N points defined as (xi yi) for i=1 to N corresponding to at least one input image in a sequence of images are defined. The initial positions are preferably chosen based on image features having good edge and corner definitions. A set of mapping parameters are then generated using a matrix processor to solve linearized equations that define a set of final positions. The mapping parameters are then applied to the initial positions to define the set of final positions.
|
