STUDY OF COMPUTER GRAPHICS AND SIGNAL CLASSIFICATION APPLICATIONS.
Final technical rept. 17 Mar 69-17 Apr 70,
GENERAL ELECTRIC CO SYRACUSE N Y ELECTRONICS LAB
Pagination or Media Count:
The purpose of the study was to investigate and recommend techniques that use an interactive computer graphics system to design signal classification procedures for signal classification problems. Many of these signal classification problems have no solutions yet, and a model does not exist from which a theoretical solution can be derived. These signal classification problems are characterized by the fact that the classification information present in the signals must be extracted from a limited number of design signals and some limited knowldege concerning the environment from which the signals were obtained. Given very little a priori information concerning these signal classification problems, it is necessary to directly employ the inventive capability of the human pattern recognizer as well as any applicable mathematical or statistical techniques. In this report, various aspects of signal analysis and classification are investigated and techniques are recommended for use in a computer graphics facility to design signal classification systems. Author
- Computer Hardware