A Hierarchical Approach to Target Recognition and Tracking. Summary of Results for the Period April 1, 1989-November 30, 1989
Abstract
This research is aimed at studying a hierarchial target extraction, identification and tracking system based on passive sensors, that could be completely integrated with other battlefield resources. Involved with this is the study of a hierarchial structure for the mutually beneficial interconnection of multiple algorithms operating on several hierarchial levels. Together these simple algorithms would cooperate in the solution of a complex problem beyond the capability of any one algorithm. Substantial progress at very modest cost (30,000 dollars) has been made in developing a passive hierarchial target identification and tracking system. A battlefield simulation capable of generating simulated images is under active development. With this simulation it is now possible to simulate images of a dynamic battlefield so that image processing and tracking algorithms can be studied. A new tracker for ground vehicles using position, attitude and terrain data has been specified. Artificial intelligence is being incorporated in two ways. First an intelligent predictor is being formulated. Second the high level reasoning module designed to use AI techniques to adjust and tune the various competing lower level modules is under development.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 07, 1990
- Accession Number
- ADA219226
Entities
People
- Dominick Andrisani Ii
- M. F. Tenorio
Organizations
- Purdue University