Implementation of modified control point image registration method.

M. L. Shahid; G. Raja

No Detail Found!

Implementation of modified control point image registration method.

Author(s)

M. L. Shahid Electronic Engineering Department, University of Engineering and Technology, Taxila, Chakwal Campus, Chakwal, Pakistan

G. Raja Electrical Engineering Department, University of Engineering and Technology, Taxila, Pakistan

Abstract

Control Point Image Registration (CPIR) method is semi-automatic image registration technique in which control points are selected and matched manually. CPIR method is best suited for images that have distinct features; however, it needs highly skillful expert to select and match control points. This paper describes a modified CPIR method using fine tuning to register images which makes the control point selection and matching nearly independent of expert’s skills. Fine tuning is achieved by applying normalized cross correlation which selects an 11x11 window around the input image control point and a 21x21 template across the reference image control point. The results of modified CPIR method are analyzed and it is found that modified method is more suitable and has low spatial dispersion values as compared to CPIR method.

Publication Details

Page(s) 53-60

DOI DOI not available

Published Journal: Nucleus, Volume: 50, Issue: 1, Year: 2013

Request the full text from PASTIC