Real-time imaging systems' combination of methods to achieve automatic target recognition
| Autor: | Elmer F. Williams, Alan Z. Pezzulich, Carlos G. Maraviglia |
|---|---|
| Rok vydání: | 1998 |
| Předmět: |
Computer science
business.industry Real-time computing Image processing Machine learning computer.software_genre Asset (computer security) Target acquisition Expert system Weapon system Automatic target recognition Feature (computer vision) Artificial intelligence business computer Real-time operating system |
| Zdroj: | SPIE Proceedings. |
| ISSN: | 0277-786X |
| DOI: | 10.1117/12.300069 |
| Popis: | Using a combination of strategies real time imaging weapons systems are achieving their goals of detecting their intended targets. The demands of acquiring a target in a cluttered environment in a timely manner with a high degree of confidence demands compromise be made as to having a truly automatic system. A combination of techniques such as dedicated image processing hardware, real time operating systems, mixes of algorithmic methods, and multi-sensor detectors are a forbearance of the unleashed potential of future weapons system and their incorporation in truly autonomous target acquisition. Elements such as position information, sensor gain controls, way marks for mid course correction, and augmentation with different imaging spectrums as well as future capabilities such as neural net expert systems and decision processors over seeing a fusion matrix architecture may be considered tools for a weapon system's achievement of its ultimate goal. Currently, acquiring a target in a cluttered environment in a timely manner with a high degree of confidence demands compromises be made as to having a truly automatic system. It is now necessary to include a human in the track decision loop, a system feature that may be long lived. Automatic Track Recognition will still be the desired goal in future systems due to the variability of military missions and desirability of an expendable asset. Furthermore, with the increasing incorporation of multi-sensor information into the track decision the human element's real time contribution must be carefully engineered.© (1998) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|