Patent US 3,984,671

Patent №

US 3,984,671

Granted

Owner

Lab

AI components

2

ml · vision

Assignment

None on record

Dataset

AIPD

2023_r1 edition

Application

05502138

A method of producing single-class and multi-class composite classification maps from multispectral data is provided. The multispectral data is transformed into a binary matrix format which is then encoded on an optical medium such as photographic film. The encoded data is holographically correlated with coded patterns representing selected spectral signatures to produce signal-class classification maps. Several single-class maps are optically superimposed to produce multi-class composite classification maps.

Machine learningVisionG06E 1/04G06V 10/895Y10S 359/90

AI classification

Vision1.00
Machine learning0.53
Knowledge representation0.16
Speech0.01
AI hardware0.01
Natural language0.00
Evolutionary computation0.00
Planning0.00
© 2026 NYSGPT2525 LLC