Image analysis – Histogram processing – For setting a threshold
Patent
1982-11-17
1984-12-25
Boudreau, Leo H.
Image analysis
Histogram processing
For setting a threshold
382 14, 382 36, G06K 900
Patent
active
044908474
ABSTRACT:
In image recognition systems of the type employing a frame store having groups (N-tuples) of locations mapped to storage forming discriminators trained, or capable of being trained, to respond to different images, a problem arises in that either a large number of separate discriminator stores are required or recognition times are long. In the present invention the discriminator storage comprises p discrete stores having address terminals divided into two groups: one group directly mapped to respective portions of the frame store and the other group being connected in parallel and used to count through a number of N-tuple groups (each of size p) and/or a number of discriminators. Such a storage arrangement allows many different combinations of number of discriminator stores and recognition time, making possible an optimum combination for a particular application.
REFERENCES:
patent: 3581281 (1971-05-01), Martin et al.
patent: 3609686 (1971-09-01), Savory et al.
Aleksander and Stonham, "Guide to Pattern Rec. Using Random-Access Memories", Computers and Digital Tech., vol. 2, No. 1, Feb. 1979, pp. 29-39.
Fairhurst and Kormilo, "Some Economic Considerations in the Design of an Optimal N-Tuple Pattern Classifier", Digital Processes, 3 (1977), pp. 321-329.
Fairhurst and Stonham, "A Class. System for Alpha-Numeric Characters Based on Learning Network Techniques", Digital Processes, 2 (1976), pp. 321-339.
Aleksander Igor
Stonham Thomas J.
Wilkie Bruce A.
Boudreau Leo H.
National Research Development Corporation
LandOfFree
Recognition apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Recognition apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Recognition apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-1156684