Page:Advanced Automation for Space Missions.djvu/9

From Wikisource
Jump to navigation Jump to search
This page needs to be proofread.

electromagnetic (visual, infrared, X-ray, microwave), aural, tactile, chemical; the possibilities are virtually unlimited.


Figure 1.1.—Overview of NASA/ASEE 1980 Summer Study on Advanced Automation for Space Missions.


The basic problem in perception is to extract from a large amount of sensed data some feature or characteristic that permits object identification. If viewed scenes can contain only two possible object classes, say, round or square, then the problem of deciding which is present may be relatively simple. But if thousands of characteristics are important in the scene, the task of creating a perceptual model of sufficient richness to permit unambiguous identification may be formidable indeed.


1.1.3 Natural Languages


One of the most difficult problems in the evolution of the digital computer has been communication between machine and human operator. The operator would like to use an everyday language—a natural language—to gain