REGISTRATION FOR DATA FUSION BY EXACT MAXIMUM LIKELIHOOD METHOD

PDF

Authors
  1. Zhou, Y.
  2. Yip, P.
Corporate Authors
McMaster Univ, Hamilton ONT (CAN) Communications Research Lab;Defence Research Establishment Ottawa, Ottawa ONT (CAN)
Abstract
The combination of data and information from multiple sources or sensors constitutes one of the most important problems in signal processing to-day. This process of data fusion has been defined as one of dealing with the association, correlation and combination of data. In this report, an exact maximum likelihood method is used to solve the problem of registration errors in data fusion. Registration is defined as the co-ordinate conversion of multiple source data. Error-free registration is a pre-requisite in the process of data fusion. In the exact maximum likelihood method (EML), a likelihood function is constructed based on the errors of co-ordinate transformation of the local sensor locations to a common system plane. Optimization is then carried out in a recursive two-step Gauss-Newton type procedure, which produces the correct system biases for proper registration. Simulation results show that EML is more efficient when compared with conventional methods of registration.
Keywords
Maximum Likelihood Method;Data registration;Registration error
Report Number
CRL-323;DREO-CR-96-745 — Contract Report (Final)
Date of publication
31 Mar 1996
Number of Pages
43
DSTKIM No
97-00581
CANDIS No
500711
Format(s):
Hardcopy;Document Image stored on Optical Disk

Permanent link

Document 1 of 1

Date modified: