Global approaches to solving recognition problems of noisy images, 1989
Thompson, Charles
1980-1989
An important problem in the area of pattern recognition is automatic detection of certain pre-assigned elements of an image distorted by noise. In this research, a global ap proach will be used. One such approach is to use an optimal smoothing algorithm which depends on efficient dynamic programming computational techniques. The basic purpose of this research is to make this dynamic programming process efficient in terms of storage requirement and computational effort. Our goal, using the objective function, is to find an optimal order of optimization and then design an effi cient computational technique. Two global techniques will be presented in this paper. Included is a graph-searching technique and the above men tioned technique using dynamic programming. Emphasis will be on the development of an algorithm using dynamic program ming. I wish to express ray deepest appreciation and sincere gratitude to those who have contributed their time, energy, and support to make this study possible. Thanks are especially due Dr. Warsi, Nazir A. , my thesis advisor. His instruction, suggestions, and patience were essential to the completion of this thesi s. Further thanks are also due Nasa Langley for providing financial, technical, and general support to help make this study possible. Special thanks are offered to Mr. Micheal Goode, Technical Monitor, for his technical support and to Dr. Samuel E. Massenberg, University Affairs Officer, for his general support.
text
application/pdf
1989-07-01
thesis
Master of Science (MS)
Clark Atlanta University
School of Arts and Sciences, Mathematical and Computer Sciences
Warsi, Nazir A.
Georgia--Atlanta
http://hdl.handle.net/20.500.12322/cau.td:1989_thompson_charles