Patent US 7,409,653

Patent №

US 7,409,653

Granted

Owner

Lab

AI components

1

vision

Assignment

None on record

Dataset

AIPD

2023_r1 edition

Application

10946667

A plurality of images, including a first image and a second image having a higher resolution than the first image, are aligned by generating an oversampled cross correlation image that corresponds to relative displacements of the first and second images, and, based on the oversampled cross correlation image, determining an offset value that corresponds to a misalignment of the first and second images. The first and second images are aligned to a precision greater than the resolution of the first image, based on the determined offset value. Enhanced results are achieved by performing another iteration of generating an oversampled cross correlation image and determining an offset value for the first and second images. Generating the oversampled cross correlation image may involve generating a cross correlation image that corresponds to relative displacements of the first and second images, and oversampling the cross correlation image to generate the oversampled cross correlation image.

VisionG06T 7/001G06T 7/32G06V 10/24G06T 2207/30148

AI classification

Vision0.64
Evolutionary computation0.02
Natural language0.00
AI hardware0.00
Speech0.00
Knowledge representation0.00
Planning0.00
Machine learning0.00
© 2026 NYSGPT2525 LLC