 |
|
|
|
|
|
|
|
|
|
|
|
|
 |
|
|||||||||||
 |
 |
 |
|
|||||||||
 |
|
|||||||||||
 |
 |
 |
|
|||||||||
 |
|
|||||||||||
 |
 |
 |
|
|||||||||
 |
|
|||||||||||
 |
 |
|
||||||||||
 |
|
|||||||||||
 |
 |
 |
|
|||||||||
 |
|
|||||||||||
 |
 |
 |
|
|||||||||
 |
|
|||||||||||
 |
 |
 |
|
|||||||||
 |
|
|||||||||||
 |
|
|||||||||||
 |
|
|||||||||||
Applications of Simulated Annealing Minimization Algorithm to Conduct Unsupervised Classification of Remotely Sensed Data
Principal Investigator: Siamak Khorram
Co-Principal Investigator: Xiaolong Dai
Project Scientist: Hui YuanObjective and Introduction
Unsupervised classification has been applied to substantial applications in remote sensing. However, the convergence of classical unsupervised algorithms is initialization-dependant and limited by many different undesirable local minima. In this research, we propose to develop algorithms and applications to use the Simulated Annealing optimization algorithm in unsupervised classifier for solving this problem. Global optimal classification based on Simulated Annealing is expected to reduce the uncertainties and improve the efficiency and classification accuracy for land cover classification.Deliverables
Simulated Annealing-based unsupervised algorithm for land cover classification coded in C++.
A final report about the methods and algorithms developed, experimental results comparing algorithms, result analysis and discussion for future research on this subject.
Presentation of research results at seminars and professional scientific symposium and publications in peer-reviewed journals.A component of the 1999 Cray Grant Program
Contact the Webmaster Contents Copyright © 2000 NC State University