Self-calibration of a noisy multiple-sensor system with genetic algorithms

Document Type

Conference Proceeding

Publication Date

1-1-1996

Abstract

This paper explores an image processing application of optimization techniques which entails interpreting noisy sensor data. The application is a generalization of image correlation; we attempt to find the optimal gruence which matches two overlapping gray-scale images corrupted with noise. Both taboo search and genetic algorithms are used to find the parameters which match the two images. A genetic algorithm approach using an elitist reproduction scheme is found to provide significantly superior results. The presentation includes a graphic presentation of the paths taken by tabu search and genetic algorithms when trying to find the best possible match between two corrupted images.

Publication Source (Journal or Book title)

Proceedings of SPIE - The International Society for Optical Engineering

First Page

20

Last Page

30

This document is currently not available here.

Share

COinS